SOCIAL RELATIONS AND THE KEYNESIAN MULTIPLIER

Massimo De Angelis

(University of East London)

[Version with no footnotes]

February 1997

ABSTRACT

This paper offers a reading of the simple Keynesian multiplier through Marx's lenses. The objective is to make explicit, within the Keynesian multiplier, what for a Marxian perspective should be made explicit in any discourse about the capitalist economy, namely the role of the social relations of production. Since in given historical time social relations are expressed in institutional forms, this paper also explicates the basic institutional requirements at the basis of the Keynesian discourse, requirements without which the Keynesian multiplier could not be operational.

1. Introduction.

It is some time now that Keynesian analytical tools and policy targets have lost the status of orthodoxy in economic thought. Perhaps because of this, many of the critics of neoliberal thinking are drawn back to the Keynesian framework of analysis in the attempt to reformulate alternative proposals to face the economic problems of our age in the new conditions provided by the global economy. This is understandable, since for students of economics Keynes represents one of the most influential reference points in the theoretical and policy battle against economic liberalism. Another influential reference point of critical thought is of course the one provided by Marx. However, the terminology and mode of discourse of this paradigm seem so alien to Keynesian constructs, that seldom there have been attempts to understand Keynes in terms of Marx. The result is that Marxist scholars have either dismissed Keynesian analysis as "bourgeois" and ideologically biased, or have melted their Marxist frame of reference within the Keynesian paradigm, producing in this operation a more "socially concerned" form of Keynesianism. Both these cases leave Marxists scholars with no independent interpretative framework with which to assess the meaning of Keynesianism and the policies formulated within its framework, in terms of basic categories having strategic relevance within the Marxist discourse.

To help to fill this gap, in this paper I offer a reading of the simple Keynesian multiplier through Marx's lenses. My objective is to make explicit, within the Keynesian multiplier, what for a Marxian perspective should be made explicit in any discourse about the capitalist economy, namely the role of the social relations of production. To avoid any misunderstandings, I must here make explicit what this paper is not about. This paper is not about Marx being right and Keynes wrong or viceversa. Nor is it about comparing and contrasting different theories, or functional relations, or causality direction in the two approaches. This paper is simply about a translation, and its implications; that is, the insights we can gain through it, in order to make explicit what is implicit and evaluate the consequences. This operation of course requires a translator, a mapping device that enables us to read analytical categories in ways that are different from those originally intended and which became common in the community of professional economists. Thus, in this paper I will try to make the basic analytical tool of Keynesianism the fiscal multiplier intelligible and understandable in terms of Marx's categories. As Marx's main concern is the exposition of social relations, here my focus is to search for the role played by social relations in the Keynesian multiplier. However, since in given historical time social relations are expressed in institutional forms, I also explicate the basic institutional requirements at the basis of the Keynesian discourse, requirements without which the Keynesian multiplier could not be operational.

This paper is thus structured. After having drawn the object of this paper against the background of the broad Keynesian discourse (section 2), in section 3 I briefly outline the basic tenets of my interpretative framework. In section 4 I dissect the income determination model and express the fiscal multiplier in terms of a gap between productivity and wage rate. In section 5 I briefly comment on the relation between this reinterpreted multiplier and two Marxian categories: the rate of exploitation and the rate of profit. In section 6 and 7 I show how the basic analytical tool of the Keynesian paradigm, the fiscal multiplier, embeds particular assumptions regarding the relation between productivity and wages, and the extension of working time. The predictable multiplier effect of aggregate demand over national income, and therefore employment, can be conceived only to the extent that this relation and this extension are believed to be stable. This assumption, I argue, is not only a distribution assumption, but, following a Marxian interpretation, it entails the pattern of class relations at the point of production. In light of the previous discussion in section 8 I also briefly discuss some important policy implications.

2. Social Relations and Kenesianism

As the philosopher of science Thomas Kuhn reminds us, scientific revolutions happen only to the extent they are able to shift a paradigm, to make a large majority of the scientific community accept the tenets of the new paradigm. With this comes new "problems available for scientific scrutiny" and the standards by which the profession determines "what should count as an admissible problem or as a legitimate-problem-solution" (Kuhn, 1970, p. 6). Thus, to the extent a Keynesian revolution occurred and became institutionalised practice in the universities and think-tanks of the world in the post-war era, it had to define a new problem around which a new economics could unfold, viz. full employment. It also had to define a means for the solution of the problem, viz. growth, and a set of instruments through which growth could be managed and achieved, viz. monetary and, especially, fiscal policies. It is this triad ÄÄ problem, means, and instruments ÄÄ that define the Keynesian revolution as the world experienced it, independently of idiosyncrasies that distinguish it from Keynes' original message. It may be true that there are major differences between Keynes and post-war mainstream Keynesianism in terms of the modelling of time and the use of the concept of equilibrium. It may be true that for the sake of the correct analytical classification of doctrines, we should add the adjective "bastard" in front of the noun "Keynesian" to characterise post-war economic orthodoxy (Robinson, 1962). Yet, the shift in the central problem of economic theory and policy and the active engagement of the government and the state apparatus in the managing of the economy legitimately entitle the new economic strategies to a revolutionary status. The triad defined above, and especially the nature of the problem, constitutes my working definition of Keynesianism, merely understood as orthodoxy, as the "conventional wisdom" of mainstream paradigm of post-war economic community, which influenced both the teaching, research and the policies of Western governments.

Yet, there is a fourth element in the post-war Keynesian revolution that has not found much explicit attention in standard textbooks, although it constituted a corner stone upon which the entire post-war Keynesian building was founded. It is the relation between wages and productivity, a relation that not only defines profit margins and therefore, ultimately, the entire raison d'être and motivation of capitalist production, but also uncovers the political and social dimension behind the veil of pure economics. Kalecki (1943), in his seminal paper Political Aspects of Full Employment, puts the finger upon this relation when discussing the question of the maintenance of full employment.

Indeed, under a regime of permanent full employment, the `sack' would cease to play its role as a disciplinary measure. The social position of the boss would be undermined, and the self-assurance and class consciousness of the working class would grow. Strikes for wage increases and improvements in conditions of work would create political tensions. It is true that profits would be higher under a regime of full employment than they are on average under laissez-faire; and even the rise in wages rates resulting from the stronger bargaining power of the workers is less likely to reduce profits than to increase prices, and thus affect only the rentier interest. But `discipline in the factories' and `political stability' are more appreciated than profits by business leaders. Their class instinct tells them that lasting full employment is unsound from their point of view, and that unemployment is an integral part of the `normal' capitalist system (Kalecki, 1943, p. 351).

It is arguable whether full employment is the only condition for political instability in a capitalist system. Indeed, as some have argued (Armstrong, Glyn, Harrison, 1984; De Angelis, 1995) the post-war government "pledge" for full employment originated out of the social turmoil after the first world war, the Russian Revolution, the strike waves of the 1930s and 1940s. Yet, Kalecki addresses here a crucial issue. His argument explores the likely effect of full employment on the relation between productivity ÄÄ that is, "discipline in the factory" (Glyn, 1995) ÄÄ and wages. He thus shifts the focus from the mere distribution effect of full employment to its effect on the dynamic of social relations. "Discipline in the factory" and "political stability" are closely related and they are both "more appreciated than profits by business leaders" because it is on their terms that the activity of profit-making itself, and therefore the social existence of profit makers, rather than a mere quantum of profit, is defined.

There is some indication that Keynes, for whom Kalecki's article was "exceedingly good" and "very acute" (Osiatynski, 1990, p. 573), looked at the stability of the relation between productivity and wages with a hopeful eye. As early as 1926 in his polemic with the cotton industrialists of Lancashire, Keynes pointed to the disastrous effect of cuts in money-wages in a situation of pressing international competition and technical inefficiency (Keynes, 1926a). Instead of attacking "directly" the money-wages, the reduction in the cost of production instead should have been pursued "indirectly" through the restructuring of the productive processes, through a social restructuring of the working day which implied capital concentration, technological efficiency of the industry and better and more rational use of the labour power (Gobbini, 1972, p. 56). But rationalisation of production and restructuring of the working day cannot be implemented without raising problems of discipline in the factories. Keynes thus seems to find an answer to this problem by looking back to the experience of "war socialism", the "social pact" between capital and unions during W.W.I:

War experience in the organisation of socialised production has left some near observers optimistically anxious to repeat it in peace conditions. War socialism unquestionably achieved a production of wealth on a scale far greater than we ever knew in peace, for though the goods and services delivered were immediate and fruitless extinction, none the less they were wealth (Keynes, 1926b, p. 286).

Almost twenty years later, in 1943, Keynes maintains that the task of keeping "efficiency wages" (wages per unit of output) stable is a political rather than an economic problem and in 1944 he acknowledges the problem of restraining real wages in presence of full employment and collective bargaining (Winch, 1989, p. 107; Glyn, 1995, p. 37).

Yet, in the General Theory, Keynes explicitly accepts the second postulate of classical theory, that wages are equal to the marginal product of labour, thus linking any increase in employment to a reduction in real wages. Later in How to Pay for the War, Keynes states again his caution in handling the issue of wages: "I have not attempted to deal directly with the problem of wages. It is wiser, I expect, to deal with it indirectly" (Keynes, 1940, p. 55). Here again Keynes prefers to appeal to the indirect economic mechanism of money illusion to keep wages in check, rather than explicitly dealing with the overt political issue of bargaining. As Dillard notes, "Keynes always dealt with the wage problem indirectly and never developed anything that could properly be called a theory of wages. He strenuously objected to reductions in money-wage rates during depression periods, but on the other hand he did not advocate higher wage rates" (Dillard, 1984, p. 320).

Post-war Keynesian orthodoxy shows a similar reluctance to expressively handle the question of wages and profitability within its main theoretical framework. One of the main analytical tools used by post-war Keynesian orthodoxy, the IS-LM model, presupposes the second postulate of classical theory. Wages and productivity are regarded as two hidden endogenous variables determined by the position of the demand led economy on a particular point of the aggregate supply and aggregate demand curve.

Yet, another of the main analytical tools used by post-war macroeconomics offers us the opportunity to bring at the forefront what is generally left behind the scenes. The Keynesian fiscal multiplier offers this opportunity. Social relations here are social relations of production and are captured by the spread between productivity and wage rate, that is the profit per hour.

3. Searching for the embedded assumption in the Keynesian system.

Since Keynes' General Theory, the consumption function has become one of the pillars of modern economics. Every textbook of economics starts with this basic concept as the building block of the entire edifice of macroeconomic theory. Its importance resides in the definition of the fiscal multiplier and in the mirrored relation to the savings and consumption functions. These are concepts which assume a strategic meaning as far as the explanation and regulation of economic activity are concerned. Therefore the proposal of policies which attempt to shape the growth rate of an economy and, in the old days of Keynesian orthodoxy, the achievement of full employment, are based on a clear understanding of the quantitative side of these expressions.

Yet, the concepts building on the definition of the consumption function, accept a basic interpretation of reality which is in line with the methodological individualism of neo-classical economics. "Aggregate" consumption is by its very nature a concept which lumps together "individual" consumption decisions by simply adding them up independently of the different social role that different consumption activities may fulfil. Of course, the role of consumption is, from the perspective of human needs, consumption of use values for the satisfaction of human needs, a tautology this that indeed serve as an intuitive starting point of any textbook of economics. But if we want to concede the Marxian claim that the relations among human beings and, indeed, what human beings are, is socially and historically determined, then individual and aggregate consumption must present themselves in terms of the social role they fulfil.

Post-Keynesian writers have attempted to do justice to this Marxian insight starting with the works of Kalecki (1943), Kaldor (1956) and Pasinetti (1962). In these works classes were introduced to the analytical structures of models in order to discuss the role of income distribution between classes on the level of economic activity. Although this was of course an important innovation within the context of an economic discourse which was dominated by the conception of the economy as formed by the collection of atomised and classless abstract economic agents, to a certain extent their effort represented a return to the roots of pre-Marxian classical political economy. This because the classes that were introduced were sociological classes rather than social, in that classes only were determined in terms of an intuitive criterion arising from the observation of everyday life, the distribution of income, and not in terms of the social relations of production, which are more hidden and require us to dig through the surface of the phenomenal world. In Marx's framework profits and wages, the income of capitalist and workers, are the form, acquired in everyday life, of surplus value and variable capital. The last two categories are defined not simply in terms of a distributive criteria, but in terms of the social role they acquire in the reproduction of the social relations of the capitalist process of production. Taking into consideration "capital as a whole," profit is defined not in terms of capitalists' unlimited wants in search of satisfaction, but as surplus value that is, in terms of the unlimited drive of profit-making. Furthermore, in this fashion reinvested profit implies the reproduction of the set of class relation of productions defining the capitalist system. Through the insight of the analysis of surplus value therefore, profit is not seen as the income of the capitalists, but as the result and precondition of the maintenance of an historically defined system of production. In this sense, the specific social agents which in real actual historical time and circumstances are responsible for the implementation of this profit making activity, are secondary as far as the general analysis is concerned. Also in the case of wages, the theoretical priority of social relations of production over income distribution is evident. As variable capital, wage is not the income of labour, but the value of labour power, meaning that it corresponds to what is necessary, in given social, cultural and historical circumstances, to reproduce workers "and their race" for another day, month, or year of work within the given set of capitalist social relations of production.

The primary role of social relations of production in Marx leads to the interpretation of economic categories that arise from everyday practice and their integration in economic theory, not as wrong categories, but as fetishised representations of these social relations of production. This Marx's insight, which I have discussed at length elsewhere (De Angelis, 1996), enables us to move beyond the often sterile Marxist criticism of orthodox economic theory which labels it simply as wrong and ideologically biased. Because if there is a correspondence between fetishised categories and social relations of production, the task is to investigate how this correspondence is played out. In other words, if we recognise the standard economic categories as fetishes, the role of critical theory is not to dismiss them as such, but to investigate what lies behind the fetish. This task is made relatively easy by the fact that due to their fetish-like character, due to the fact that the fetish originates from the actual relations of production, the categories used by the economic discourse may want to speak for themselves. We therefore do not need to proceed like Kalecki, Kaldor and Pasinetti by bringing from the outside the assumption that there are classes, but start from the categories themselves and investigate the extent to which they hide social relations of production. Once we have done this, we can proceed to investigate the class meaning of the quantitative changes of these categories and understand them in terms of various strategic options.

4. The dissection of the simple income determination model.

Textbook macroeconomics starts with the so called income determination model. This is built starting from the identity of national income which, in a closed economy, looks like this:

1. Y = C + I + G.

For reason of simplicity, let assume there is no public expenditure, so identity 1. becomes

2. Y = C + I

Once we interpret consumption in terms of the simple consumption function 3., in which A is autonomous consumption and b is the marginal propensity to consume, identity 2 turns into equation 4.

3. C = A + bY,

1

4. Y = (1/1-b) (A + I).

Assuming that autonomous consumption is zero, 4. turns into

1

5. Y = (1/1-b) * I.

The fraction 1/(1-b) is the Keynesian multiplier, the most important strategic variable underling post-war Keynesian modelling. A stable marginal propensity to consume b (and a correspondent stable marginal propensity to save 1-b) allows, ceteris paribus, a stable multiplier, and therefore a predictable effect of a change in aggregate demand (say, I) on the level of income and therefore employment. One could argue that a stable multiplier is the key for the successful management of the economy by the government. Coddington (1983) has pointed out the stability of income and expenditure flows as a basic assumption of what he calls "hydraulic Keynesianism". Here I want to reinterpret the Keynesian multiplier and its stability criterion as a strategic variable rather than as a result. This is done by reinterpreting the denominator of the Keynesian multiplier in terms of the spread between productivity and wages. I will do so by rewriting the standard national income identity with few modifications. National income Y can be expressed as:

6. Y = PRD*(L),

in which PRD is labour productivity (y/L), and L = N*h, where N = number of people employed, h = average hours worked (daily, weekly, etc.). Equation 5. can therefore be expressed as

7. PRD(L) = (1/1-b) * I,

or

8. L = (1/(PRD - PRD*b)* I.

In general terms, equation 8. gives us the amount of labour put in motion by a given level of investment. In order to understand this relation however, we have to investigate the meaning of the denominator of 8, and in particular, of the product between productivity and the propensity to consume.

Let us start with the unit of measurement. The propensity to consume is a pure number, with no dimension, as it relates the change in consumption (in money units) to the change in income (in the same money units). As productivity is expressed in terms of money units per hour of labour, the product between productivity and the marginal propensity to consume can be interpreted as the proportion of the hourly product that goes to consumption, that is, consumption per hour of work.

It is at this juncture that we can bring in Marx's framework of interpretation. First, along with Marx (1858; 1867), the wage is the form taken up by the more substantive category of the value of labour power. The latter is defined in terms of the labour necessary for the reproduction of the commodity labour power, and it presents itself in its objective form as the bundle of commodities which are necessary for this purpose. For this reason, even if workers save as individuals, which is a starting point of methodological individualism of mainstream economics, they do not do so when considered as a class. The rationale of this interpretation is that for workers' standpoint, savings represent their postponed consumption, and not a means for enrichment, it does not represent the basis for an "unceasing movement of profit-making" (Marx, 1867, p. 254). Thus, even if workers formally save, this saving serves, in the hand of today's capitalists, as part of advance in capital for production. Since for Marx what is important is capital as a social relation of production, and accumulation as accumulation of a social relation, it is irrelevant whether some of the financial resources for investment are formally "owned" by workers. To the extent they finance investment which is controlled by "capitalists", these resources acts as "capital" vis-à-vis the workers. Thus, at any given point in time, whatever workers consume represents in aggregate what is necessary to reproduce them as labour power, which takes the form of wage. In this light, the product between labour productivity and propensity to consume, that we have interpreted as consumption per hour, can be interpreted as a proxy for the social wage rate.

Second, this is of course a proxy, because in the calculation of the propensity to consume b, also the consumption of capitalists is taken into account. However, since the proportion of property income over total national income is very small, in considering pb as the social wage rate, we may be only slightly overestimating it. Thus we can safely regard the product between labour productivity and the marginal propensity to consume as representing a proxy for a social wage rate, understood in Marx's terms, as a money representation of the labour necessary for the reproduction of labour power as a class per unit of hour worked.

It is therefore possible to rewrite 8. as

9. L = (1/PRD - w) * I.

in which w = PRD*b I call the social wage rate, the difference PRD - w is the profit per hour, and the ratio 1/(PRD - w) I call the social multiplier. By social wage rate I understand the wage rate that in a given time period is necessary to reproduce society as a whole for another round of capitalist accumulation. By social multiplier I understand the Keynesian fiscal multiplier understood in terms of the spread between productivity and social wage rate.

Formally, as the standard fiscal multiplier defined the increase in output for a given change in aggregate demand, the social multiplier, being a simple transformation of the fiscal multiplier, defines the increase in employment for a given change in aggregate demand. While in the case of the fiscal multiplier the size of this depended on society's propensity to consume (and therefore to save), the social multiplier depends on the spread between productivity and social wage, that is the profit per hour worked. In the case of the fiscal multiplier, an exogenous change in aggregate demand leads to a change in the same direction in output (and therefore employment) which in turn induces a change in consumption (that is, demand) which affects output, and so on. The net final change in output (and employment) which depends on both initial and induced changes in aggregate demand is captured by the size of the multiplier. In the case of the social multiplier, an initial increase in aggregate demand leads to a change in the same direction in total hours worked in the economy (and therefore output and, for a given working time, employment) which in turn induces an change in consumption (that is, demand) which affects L, and so on. The net final change in total hours of labour worked (and therefore output and employment) which depends on both initial and induced changes in aggregate demand is captured by the size of the social multiplier.

It is clear that the higher is the spread between labour productivity and wage rate, that is the higher is the profit per hour worked, the lower is the social multiplier, as in the traditional case, the higher the propensity to save the lower is the fiscal multiplier. Thus, a given initial increase in aggregate demand dI, will induce an initial increase in total labour hours of dL = dY/p which then will induce an increase in demand of wdL (=bdY) and so on. So, ceteris paribus , the higher is p the lower is the initial increase in employment following an increase in aggregate demand, and the higher is w, the higher is the induced increase in employment. Thus net effect in a given time depends on the spread between labour productivity and social wage.

5. The Keynesian multiplier and Marxian categories: rate of surplus value and rate of profit.

It is useful to spend some space on the meaning, in Marxian terms, of the spread PRD - w, the profit per hour, and of the social multiplier. The profit per hour has a direct link with Marx's category of the rate of surplus value, or rate of exploitation, or balance between surplus and necessary labour. This is defined as s/v, where s is surplus value (profit) and v is variable capital (wages). Dividing both numerator and denominator by L, the rate of exploitation s' is

s/L = PRD - w

10. s' = (PRD - w)/ w

Thus, for a given wage rate, if the profit per hour increases, the rate of surplus value increases. If wage rate and productivity grow at the same rate, as in the case for example of the productivity deals of the Keynesian period, then the profit per hour also grows at the same rate, thus leaving the balance between surplus and necessary labour (rate of exploitation) constant. If productivity grows more than the wage rate, the profit per hour increases more than the wage rate, thus the balance between surplus and necessary labour increases.

It must be pointed out that from the point of view of capital there is a minimum level of productivity growth which is compatible with a non declining profit rate. The latter is in fact defined as

11. r = s /(C + v)

in which r is the rate of profit, s is the total amount of surplus value (profit), C is the price value of fixed and circulating capital invested (assuming one year turnover), and v is variable capital (wages). Dividing both numerator and denominator by L, 11. can be rewritten as

p - w

12. r = (PRD - w)/(k + w)

in which k is the capital-labour ratio. Thus, if productivity deals are institutionalised such that productivity grows in line with the wage rate, and if increases in productivity are obtained through increases in the capital labour ratio k, then from 12, in order to maintain a non declining profit rate, dr/dt = 0, it is required that d(PRD - w)/dt = d(k + w)/dt. Thus, the higher the increase in the capital/labour ratio following technical change and increase in wages, the higher must be the increase in the spread between productivity and wages, the profit per worker, in order to maintain the same degree of profitability.

6. Implicit assumptions and implications of the dissected income determination model I.

6.1. Assumptions

The transformed multiplier makes clear that the validity of the standard simple Keynesian argument that an increase in aggregate demand will accompany an increase in national income (and therefore employment) depends on the following two crucial assumptions.

1. The multiplier is, in the very short term, given, that is on the assumption of a given spread between labour productivity and wage rate, and, for a given wage rate, a given rate of exploitation in the economy. In Marxian terms, and indeed also in Kaleckian terms, this means that what is assumed as given is a particular balance of forces between classes at the point of production (p) and in the labour market (w).

2. In a growing economy the social multiplier is stable and predictable. It is only to the extent that this stability condition holds that the Keynesian multiplier can be used as an analytical tool for the design and implementation of growth strategies. However, the dissected multiplier also tells us that the effectiveness of these strategies depends on the actual stability and predictability of the social multiplier. This means that Keynesian policies rely heavily on institutions able to guarantee a social deal among representatives of workers and capitalists, and that these representatives are actually able to implement and deliver the predicaments of these deals. These deals are obviously essentially productivity deals, in that they entail growth in labour productivity and in the wage rate. As the growth in productivity depends on the introduction of technical change and therefore affects labour process, organisations, and rhythms and patterns of work, the Keynesian multiplier presupposes the ability to govern not only income distribution among classes, but also the very core of class relations at the point of production.

Furthermore, in a context of growing productivity and productivity deals that link increases in wages to increases in productivity, the social multiplier tends to fall. This makes the employment effect of expansions of aggregate demand smaller and smaller, requiring larger and larger investment outlays to maintain a given level of employment growth.

6.2. Implications

The assumptions of given, stable and predictable class relations thus forms the core of Keynesian income determination theory through movements of aggregate demand. The exposition of these hidden assumptions makes it relatively simple to envisage the effects of movements in productivity and wage rate on the employment impact of investment (EII) within the framework of this dissected Keynesian multiplier.

1. Technical change and process innovation, together with increases in labour intensity, increase labour productivity and thus reduce EII.

2. More relaxed working life, that is, the reduction of labour intensity which by no means implies the reduction of technological change reduces labour productivity and increases EII.

3. EII is positively related to movements in wage rates, increasing when the latter increase and viceversa.

.

7. Implicit assumptions and implications of the dissected income determination model II.

7.1. Assumption

Another crucial assumption of the standard income determination model is a given extension of working time. Equation 9. can be rewritten as

1

N x h = (1/ (PRD - w)) * I, or

13. N = (1/h*(PRD - w)) * I.

where h is the average working time, and N is the number of people employed. Equation 10 shows that for a fixed working time and predictable growth of p - w, investments can have a predictable effect on employment N. We can thus see that the stability condition that we have encountered in relation to the spread between productivity and wages is now applied to working time, as this has an effect on the social multiplier. As the productivity deals were the institutional framework regulating the post-war spread between productivity and wages, so one of the most remarkable stylised facts of post-war accumulation has been the ending of the secular downward trend of working time. This trend occurred through successive series of "shocks" brought about by cycles of struggles for the reduction of working time (Roediger and Foner, 1989). The stability in the social multiplier required by the post-war Keynesian strategies led to the devising of an institutional environment able to prevent the occurrence of these "shocks".

7.2. Implication

One clear implication of the introduction of working time in the model is the negative relation between working time and EII. A lower h means, in principle, a higher multiplier and therefore a higher employment effect of a given change in aggregate demand.

8. Policy implications: strategic options to increase employment.

We are now in a position to explore alternative strategic options aimed at employment generation within the framework of the dissected Keynesian multiplier. Employment can be generated in the following ways:

1. Classical Keynesian strategy. This relies on a given and a stable balance between productivity and wages and the expansion of aggregate demand. It thus relies on productivity deals that are operational.

2. Alternative strategy I. Increase the social wage rate for a given productivity.

3. Alternative strategy II. Reduce productivity for a given wage rate.

4. Alternative strategy III. Reduce the working time.

5. Alternative strategy IV. Any combination between 2., 3. and 4.

The classic Keynesian strategy which entailed the practice of productivity deals presupposes of course the establishment of a suitable institutional environment. In the United States this has been obtained through the institutionalisation of unions, especially after W.W.II, the active involvement of their bureaucratic apparatuses into negotiations and practices of control of grassroots' militancy (Edwards, 1979; Glaberman,1952, 1966, 1975; Harris, 1982), as well as the establishment of wage rounds practices across the economy. The spread between productivity and wage rate can also be targeted with the practice of income policies. These have the same aim as the productivity deals, although while the latter are established at the micro level, from which they then spread throughout the economy, income policies move from the macroeconomic level. Income policies were often attempted when productivity deals failed to keep in check the spread between p and w. A reasonably stable accumulation occurs to the extent that productivity deals are successful. However, this strategic option must be discarded as the collapse of Keynesianism and the "golden age" of post-war capitalist accumulation in the late 1960s and early 1970s was brought about by the collapse of the institutional conditions supporting the stability condition (De Angelis, 1995). Furthermore, as mentioned before, even if productivity deals are guaranteed by institutions, in a regime of growth the social multiplier tends to fall, thus making necessary larger and larger levels of investment to guarantee the same level of job creation. As equation 12 shows, in a context of increasing capital-labour ratio, there are increasing pressures to increase the denominator of the social multiplier (numerator of the rate of profit) in order to operate counter-tendencies to the fall in the rate of profit. This of course further depresses the social multiplier and further reduces the employment impact of a given level of investment.

All the other strategies entail a reduction in the rate of surplus-value. This of course make all these strategies non viable from the point of view of capital accumulation and the profit motive, although they all are in principle technically possible. The level of employment may fail to increase as a consequence of any of these strategies due to the fall in the level of investment consequent to the reduction in the profit rate. Of course this situation can be reversed through a direct socialisation of investment and people's participation in investment decisions based on needs rather than profit.

9. Conclusion

In this paper I have translated the basic Keynesian income determination model into Marxist categories. The main result is that the heuristic significance of the model thus translated is not so much in terms of a proper description of reality, rather in terms of its strategic meaning. We have seen that the Keynesian model relies on the assumption of stable class relations and therefore on the existence of an institutional environment able to contain and control class relations at the point of production.

The income determination model became famous and widely used during the post-war boom. It reflected an optimistic vision of the future as plentiful and secure. Growth became the magic word which could cure all social problems and would inhibit social conflict. Yet, in the income determination model, this inhibition is presupposed as an assumption. Thus, the image of the future given to us by this model, is one of more goods, more work, more investments, and the same basic social relations of production. Progress was defined as a simple multiplication of the present. A critical look at equation 13. however, easily take us to another dimension, to another vision of the future. It is in fact possible to imagine another scenario, one based on the centrality of social needs and aspirations. These include the need for creative activity, the need for those material goods which are necessary for people's reproduction and the need for time freed from work. Not only these needs can become the basis upon which resources are directed to production of new kind, but also the yardstick upon which to measure the speed and direction of a post-capitalist strategy. Thus, essential co-ordinates of this strategy could be a systematic increase in w above p, a systematic reduction in working time, or a combination between increases in w below productivity accompanied by reduction of working time. Of course, in this context all the variables would also acquire different meaning. The "social wage rate" would immediately be recognised as the "socially necessary labour rate", while the "profit per hour" would immediately reveal itself as "surplus labour per hour worked". In the context of production directly for social needs, a higher productivity for a given "socially necessary labour rate", will give higher surplus labour, which can be used either as a means to produce more goods to satisfy more needs, and/or to reduce working time to satisfy the basic human need of time freed from external necessity. In this context, with a systematic decline in the extension of working time and/or increase in necessary labour ("w") in relation to surplus labour ("p - w"), a given change in resources invested will lead to a higher and higher increase in level of "employment." It seems to me that there is only one constraint in the way to the actualisation of this scenario, and this is outside equation 13, and is a political constraint. The systematic reduction of working time and/or increase in the wage rate over productivity would cause the reduction in the profit per hour worked, and this would be unacceptable to profit earners. Thus, at this juncture, production for profit meets production for use and the result is a big clash.

BIBLIOGRAPHY

Ando, A. and F. Modigliani. 1963. The `Life Cycle' Hypothesis of Saving: Aggregate Implications and Tests. American Economic Review , vol. 53, no. 1.

Armstrong, Philip, Andrew Glyn, and John Harrison. 1984. Capitalism since World War II, . London, Fontana.

Coddington, Alan. 1983. Keynesian Economics. The Search for First Principles, London, George Allen and Unwin.

De Angelis, Massimo. 1995. Social Antagonism and Economics: a Reformulation of the Radical Critique of Keynesianism, Ph.D. Dissertation, Salt Lake City, University of Utah.

De Angelis, Massimo. 1996. Social relations, Commodity-Fetishism and Marx's Critique of Political Economy, Radical Review of Political Economics, vol. 28, no. 4, December.

Dillard, Dudley. 1984. The Economics of John Maynard Keynes. The Theory of a Monetary Economy, Westport, Greenwood Press.

Edwards, Richards. 1979. Contested Terrain, New York, Basic Books.

Glaberman, Martin. 1952. Punching Out, Detroit, Bewick Edition.

Glaberman, Martin. 1966. `Be His Payment High or Low' The American Working Class in the Sixties, Detroit, Facing Reality Publishing Committee.

Glaberman, Martin. 1975. Unions vs Workers in the Seventies: The Rise of Militancy in the Auto Industry, The Working Class and Social Change, Toronto, New Hogtown Press.

Glyn, Andrew. 1995. Social Democracy and Full Employment, New Left Review, no. 211, May/June.

Gobbini, Mauro. 1972. Lo Sciopero Generale Inglese del '26, Operai e Stato, Milano, Feltrinelli.

Osiatynski, J. (ed.). 1990. Collected Works of Michael Kalecki, vol. 1, Oxford 1990.

Harris, John H. 1982. The Right to Manage. Industrial relations Policies of American Business in the 1940s, London, The University of Wisconsin Press.

Kaldor, Nicholas. 1956. Alternative Theories of Distribution, Review of Economic Studies no. 30, vol. 23.

Kalecki, Michael. 1943. Political Aspects of Full Employment, Osiatynski (ed.), Collected Works of Michael Kalecki, vol. 1, Oxford 1990.

Keynes, John Maynard. 1926a. Industrial Reorganization: Cotton, The Collected Works of John Maynard Keynes, vol. XIX, London, Macmillan 1981.

Keynes, John Maynard. 1926b. The End of Laissez-Faire, Essays in Persuasion, The Collected Works of John Maynard Keynes, vol. IX, London, Macmillan. 1972.

Keynes, John Maynard. 1936. The General Theory of Employment, Interest, and Money, New York, Harvest/HBJ Book.

Keynes, John Maynard. 1940. How to Pay for the War, Essays in Persuasion, The Collected Works of John Maynard Keynes, vol. IX, London, Macmillan. 1972.

Kuhn, Thomas S. 1970. The Structure of Scientific Revolutions, Chicago, The University of Chicago Press.

Marx, Karl. 1858. Grundrisse, New York, Penguin. 1974.

Marx, Karl. 1867. Capital, vol. 1. New York. Penguin Books. 1976.

Noble, David. 1986. Forces of Production. A Social History of Industrial Automation. Oxford, Oxford University Press.

Panzieri, Raniero. 1961. The Capitalist Use of Machinery: Marx Versus the `Objectivists', Outlines of a Critique of Technology, Phil Slater (ed.), Atlantic Highlands, Humanity Press. 1980.

Pasinetti, Luigi. 1962. Rate of Profit and Income Distribution in Relation to the Rate of Economic Growth, Review of Economic Studies, vol. 91, no.1.

Robinson, Joan. 1962. Review of Harry Johnson's Money, Trade, and Economic Growth, Collected Economic Papers, vol. 3., Oxford. Basil Blackwell. 1979.

Schor, Juliet. 1991. Overwork American, New York, Basic Books.

Roediger, David R. and Philip S. Foner. 1989. Our Own Time. A History of American Labor and the Working Day, London, Verso.

Winch, Donald. 1989. Keynes, Keynesianism and State Intervention, P. Hall (ed.), The Political Power of Economic Ideas, Princeton.