Capital and Imperialism: Theory, History, and the Present

by Utsa Patnaik

The Marxian and the Keynesian Systems

  On the basic issue of a wealth demand for money and the possibility of generalized ex ante overproduction (or what Keynes called “involuntary unemployment”), there is much in common between the Marxian and the Keynesian systems. Indeed, one can say without exaggeration that almost seventy years before Keynes’s General Theory, the basic conclusions of that opus had been anticipated by Marx, although, having recognized the possibility of generalized overproduction, Marx did not theorize about where the economy would settle in such a situation. That is, he lacked a theory of income determination based on the “multiplier,” which is what Richard Kahn and Keynes provided in the 1930s.

  But Marxists saw deficient aggregate demand only as a cyclical phenomenon that would reverse itself, so that on average, the system would function at a certain level of capacity utilization, on the basis of which the input coefficients that go into the determination of labor values could be worked out. This was an error, because once the possibility of a deficiency of aggregate demand is admitted, then an isolated capitalist system acquires a “knife-edge” property and gets pushed toward a stationary state or simple reproduction (a zero trend situation), a possibility seen only by Rosa Luxemburg (1913) and later theorized by Kalecki.2 For a long time, Marxian economics remained handicapped due to not following up on Marx’s insights into the theory of aggregate demand.

  This similarity between the Marxian and Keynesian systems may appear surprising to many because they have very different notions of money. Indeed Nicholas Kaldor has argued that though monetarism does not hold in a credit money world, since such a world is characterized by money supply “endogeneity” (where money supply adjusts to the demand for money), it is likely to characterize a commodity money world since money supply there is exogenously given.3 But the existence of a hoard, which signifies a wealth demand for money even in a commodity money world, can have exactly the same implications, by way of negating monetarism and Say’s Law, as occur in a credit money world.

  However, one fundamental difference between the Marxian and the Keynesian systems exists, even in this terrain. Whereas in Keynes an increase in money wages gives rise to an increase in prices, leaving real wages unchanged, begging the question why trade unions existed at all, in Marx, as in Ricardo, a rise in money wages increases real wages at the expense of profits. In a famous pamphlet, Wages, Prices and Profits, which is the text of his speech at a meeting of the International Working Men’s Association, Marx had argued against Citizen Weston, a follower of John Stuart Mill, who had suggested on the basis of Mill’s “Wages Fund” theory that workers could not raise their real wages, and that the gains of some could only come through losses of others. But this was not the case, Marx said, contending that all workers could gain through trade union action at the expense of the capitalists. This was probably the first explicit formulation of what later came to be called the “factor-price frontier,” that is, the inverse relationship between the real wage rate and the rate of profit. What was striking, however, was that this formulation was made for a world that could, by Marx’s own recognition, be characterized by “involuntary unemployment.”

  In a world with “involuntary unemployment,” since output is demand-constrained, a rise in real wages could simply raise output—the rise in real wages leads to an increased demand for output, which, in turn, raises the quantity produced without altering the magnitude of profits and hence the rate of profit. A downward-sloping “factor-price frontier” can be postulated here only in the following sense: if real wages increase at a given output, then the rate of profit will fall. Or put differently, a rise in real wages would lower the “given-output rate of profit,” and a rise in money wages, since it necessarily raises real wages, would also do so. In interpreting Marx’s “factor-price frontier,” which shows an inverse relationship between money wages and the rate of profit, we are not asking the question what happens if money wages rise in an economy; we are asking the question what happens if money wages rise in an economy with given output?

  Let us now pull together the threads of Marx’s argument with regard to money. Marx’s view of the characteristics of a money-using capitalist economy can be summed up in three propositions:

  • Proposition 1: Money is held in a capitalist economy not just for circulating commodities but also as a form of wealth, so that the money held on average over any period may far exceed what is needed for circulation.

  • Proposition 2: For any given level of capacity utilization of the fixed capital stock, there is a unique money value of the aggregate output produced, which is independent of the total money supply.

  • Proposition 3: For any given level of capacity utilization, a rise in the money-wage rate lowers the maximum realizable rate of profit.

  Although the Keynesian system accepts the first two propositions, the Marxian system accepts all three, and therein lies its uniqueness.

  The difference between the two systems, which relates only to the third of these propositions, would be attributed immediately to the fact that Marx is dealing with a commodity money world. In it, the value of the money commodity in terms of non-money commodities is fixed independently, whereas Keynes is dealing with a credit money world where there is no such fixity.

  Although this is certainly true, the question to ask of the Marxian system is: If money wages increase, why should prices not increase? What is the mechanism through which the fixity or the relative fixity of the value of money is maintained when money wages increase?

  To say that money consists, say, of gold, and that the value of gold in terms of other commodities is determined by their relative labor costs (either immediately or in some refracted fashion) is not enough. This amounts to asserting that any increase in money wages ipso facto results in a rise in real wages (because the value of money in terms of commodities is fixed). The question here is: Why does the value of money in terms of commodities not change when the money wage rate rises?

  Looked at differently, suppose the money-wage rate rises, and suppose the capitalists put up their prices proportionately, so that the profit margin and hence the profit rate do not fall. What is there to prevent such a denouement in the Marxian schema, while such precisely is what is supposed to happen in the Keynesian-Kaleckian schema (though the manner in which it is supposed to happen is not identical for Keynes and Kalecki)?

  The immediate answer might be that since Marx was talking about “free competition,” raising prices by capitalists in view of the money wage increase should be ruled out because that is not the way “free competition” is supposed to operate. This answer, however, is unconvincing for two reasons. First, it would amount to saying that trade unions can raise real wages under free competition but not under oligopoly (where there is markup pricing), which then raises the same question as before: Why is there so much hullaballoo over trade unions under oligopolistic capitalism? Second, a focus on the nature of competition obscures that the answer is supposed to lie in the nature of money. Indeed, that would have been Marx’s own answer, emphasizing that the money he was talking about was commodity money. Hence, dragging in the distinction between “free competition” and oligopoly/monopoly is inapposite here.

  So, if in a commodity money world there is a rise in money wages and the capitalists mark up their prices equiproportionally, then what is there to prevent it? There is no point saying that they cannot, given the fixity of the value of money, since the whole point is to investigate how this fixity in the value of money is sustained.

  Oddly, Marx does not say much on how exactly this fixity in the value of money is actually maintained in the economy. An implicit answer in Wages, Prices and Profits goes as follows: If the economy is not a gold-producing one but imports its gold against a certain bundle of other commodities, there is no reason whatsoever why the gold exporters should accept any smaller bundle of commodities in exchange for what they supply just because money wages have gone up inside the gold-importing economy. T
he fixity of the value of money in the face of money-wage changes arises, in other words, because the terms of exchange between its gold imports and commodity exports remain unchanged when money wages change. This argument is plausible, though it does not cover the case where gold is domestically produced.

  The Marxist tradition has not been concerned with this question at all.4 Typically, the tendency has simply been to take this fixity for granted and argue on its basis that a rise in money wages must raise real wages. But this, as we have maintained, does not amount to proving the point.

  Fixity of the Value of Money

  Although Marx did not specify how exactly the fixity in the value of money is sustained, we can adduce a possible mechanism for it. Suppose everyone in the economy believes that this fixity will be maintained, because it actually has been maintained in the past. Then if money wages rise by 10 percent, say, and the money prices are also marked up by 10 percent, the value of money falls by 10 percent. Since economic agents believe that the usual value of money will be maintained, they would expect that the value of money would rise by 10 percent from the level to which it has currently fallen; that is, they would expect commodity prices to fall by 10 percent. There would, therefore, be a reduction in the demand for commodities, as people postpone purchases and reduce commodity stocks in order to increase money holdings. This will have the actual effect of pushing commodity prices back to their old level, and ensuring that the increase in money wages actually results in a corresponding rise in real wages, and hence a fall in the “given-output rate of profit.”

  It may be felt that with this mechanism it would become difficult for any change to occur in the market prices of commodities, such as would arise as a prelude to involuntary unemployment. After all, before any quantity adjustment occurs, there is likely to be some fall in prices in general owing to ex ante overproduction; but if every fall gives rise to an expectation of a price increase and to a larger demand for commodities, then there can scarcely be a deficiency of aggregate demand (or fluctuations in the level of aggregate demand).

  We can, however, salvage Marx’s theoretical vision from this logical problem by adopting one of two possible routes. One, which can be borrowed from Kalecki, does not visualize any actual divergence of market prices from the “equilibrium” (or markup) prices. Variations in demand directly affect inventories and through them capacity utilization without causing any price changes. At the same time, any rise in money wages, if it is fully passed on as a general price rise, is doomed to failure, since everyone will believe that money prices will come back to their earlier level in the future, no matter what their current level may be. This route entails what some would call a “fix-price” economy. Though a far cry from Marx’s own formulation, it can be one way of ensuring that his various propositions on money hold together.

  The other route is to assume that economic agents can distinguish between “cost-induced” changes in the price level and “demand-induced” changes in the price level. Where the money commodity is domestically produced and thus affected by a rise in money wages, like all other commodities, they would expect money prices of non-money commodities to remain more or less unchanged even when money wages rise, and this would prevent any actual rise in money prices of non-money commodities. Money-wage increases in this case will lead to real wage increases. But aggregate demand–induced changes in the price level affect only the non-money commodities vis-à-vis the money commodity. The expectation then would be that the change in money prices would continue for some time, instead of reversing itself. And this would cause output adjustment, or involuntary unemployment.

  We can thus adduce possible ways of defending Marx’s position on money in the context of an isolated capitalist sector. But even these, no matter whether they are persuasive, are not enough. The irreconcilability between a closed capitalist sector and the three propositions advanced by Marx, which are valid propositions in a money-using economy, becomes obvious when we consider paper money, as Marx himself had done.

  A Problem with Paper Money

  It would appear at first sight that the Marxian system is immune to one important strand of the general criticism we have been making here of the various theoretical systems in economics. Our general criticism has been that the isolated capitalist economy, consisting only of workers and capitalists, with the state not directly intervening in any significant way in economic life, which is the usual conceptual representation of it in the economic literature, is incompatible with a money-using economy. This is so for a number of reasons, one of which is that two properties of a money-using economy cannot both be satisfied in an isolated capitalist sector. First, a rise in money wages must lead to a reduction in the “given-output rate of profit,” otherwise we cannot understand the reality of class struggle and the employers’ persistent efforts to roll back trade union rights. And second, the economy cannot avoid the possibility of involuntary unemployment.5

  The Marxian system appears to be immune to this particular strand of our criticism. It seems that there can both be involuntary unemployment, and at any level of such unemployment if there is a rise in money wages enforced by the workers, then it would lead not to a rise in the price level but to a reduction in the rate of profit. In other words, with an invariance of money-price to money-wage changes at every level of unemployment, the system can settle at alternative levels of unemployment. (And, typically, lower levels of unemployment are associated with lower “given-output rates of profit.” In such a system, it is clear why employers are dead-opposed to trade unions.

  But there is still a problem here. To go back to our previous discussion, if money wages rise and prices rise pari passu, then, with people expecting that the value of the money commodity will come back to the old level, there will be an actual reduction in the demand for commodities to bring the price level back to the original position and reduce the given-output rate of profit. But if money consists not just of the money commodity but also of paper currency convertible into it, then we are dealing with two sets of expectations, not one, namely the expectation of the value of the money commodity vis-à-vis the non-money commodities, and the expectation of the value of paper currency vis-à-vis the money commodity. Even if the paper currency is statutorily convertible, people still may not always be confident that it will remain so when the crunch comes.

  Though people may confidently expect that commodity prices will come down vis-à-vis the money commodity if there is a rise in prices because of a rise in money wages, they may also expect that paper currency will not maintain its value. That is, it may not remain convertible at the old rate for long. There would then be a rush to gold from both directions, from those holding commodities and from those holding paper currency. And with every flight from paper currency to gold, the confidence in its remaining convertible at the old rate would diminish, which would trigger further flight. Indeed, maintaining convertibility would actually become impossible in such a situation.

  In the Marxian system, despite commodity money and convertible paper currency, we therefore have something of the same problem that afflicted the Keynesian system. A rise in money wages, if it is passed on through higher prices, threatens the value of the part of money that consists of paper currency. On the other hand, there is no reason why capitalists should not wish to pass on the higher wages in the form of higher prices and meekly accept a lower rate of profit. And if higher money wages are passed on as higher prices, with no fall expected in these prices vis-à-vis the paper currency, which is also depreciating, then wealth held in the form of paper currency will be subject to losses. Thus, if the economy has paper currency, even though convertible to the money commodity at a fixed rate, and some wealth is held in the form of such currency, then a rise in money wages threatens the value of such currency and hence the wealth held in this form.

  Reconciling three obvious and observable phenomena, namely money as a form of wealth-holding, the possibility of involuntary unemployment, and
trade unions’ ability to raise real wages through higher money-wage bargains, all three of which characterize capitalism and are recognized within the Marxian system, seems impossible within an isolated capitalist sector.

  It follows, then, that even the Marxian system faces a problem in reconciling a conceptual representation of capitalism as an isolated sector consisting only of workers and capitalists with the reality of a money-using economy, though it is more keenly aware of this reality than other theoretical systems, including even the Keynesian one.

  A Critique of the Marxian System: Summing Up

  Notwithstanding Marx’s deep understanding of the relationship between capitalism and the surrounding pre-capitalist segments of the world economy that were forcibly dragged into its orbit, Marx’s basic concept of capitalism in Capital is of an isolated capitalist sector. This understanding gets expressed in his numerous writings on the colonial question but makes only fleeting appearances in Capital, especially Volume I, which he completed in his lifetime. As a result, the tendency has been to see Marx, like the other economic theorists, as conceptualizing a capitalist economy basically as a closed sector (with the pre-capitalist environment being simply an add-on but not in any sense essential to this core sector).

  This perception has also influenced subsequent Marxist writings on economics in an unfortunate direction. An obvious example of this is the belief widely held in Marxist circles that the process of primitive accumulation of capital occurred only in the prehistory of capitalism and that once the system was in place, subsequent accumulation occurred only on the basis of the generation of surplus value within it. The reality, however, is that primitive accumulation occurs throughout the history of capitalism, and even in a wholly explicit form, of a tax-based appropriation of surplus by the metropolis, under colonialism.6

  It is this understanding of the Marxian system, for which one cannot altogether exonerate Marx from blame, that we are critiquing. And our overall critique of the Marxian system, including what was said in earlier chapters, focuses on three points. First, since Marx rightly rejects Say’s Law and visualizes the possibility of ex ante overproduction, the capitalist sector he is examining would settle down at a state of simple reproduction. Or, to be more precise, in such a universe where aggregate demand can be deficient, if we have an investment function that takes adequate cognizance of demand, then the only stable equilibrium is one with a zero trend.7 Sustained growth, or expanded reproduction, in the capitalist sector therefore cannot be explained unless we go beyond its isolated existence. We have, in sum, a theoretical contradiction between the concept of a money-using economy (which causes ex ante overproduction) and an isolated capitalist sector if sustained growth is seen, as it must be, as one of its characteristics.