In chapter 15 of Capital, vol. 1, ``Machinery and Large-Scale Industry'', Marx offers a masterly account of the impact on the working class of the shift towards mechanization and large-scale industry within the social relations of capitalism. He marshalls a wealth of technical detail on the industrial processes involved and draws on the careful social documentation of the Factory Inspectors, putting this material into a context at once theoretical, historical and moral. He shows himself in command of a large literature in political economy on the subject, much of which he dismisses scathingly as apologetics.
The main theme of the chapter is announced at the start:
John Stuart Mill says in his Principles of Political Economy: 'It is questionable if all the mechanical inventions yet made have lightened the day's toil of any human being.' That is, however, by no means the aim of the application of machinery under capitalism.... The machine is a means for producing surplus value. (p. 492)
While one might naively think that labour-saving machinery might actually save labour, in the sense of lightening the worker's toil, Marx goes on to argue that, in a cruel irony, the introduction of machinery generally gives the capitalist both the opportunity and the incentive to lengthen and intensify the working day.
Marx's general argument to this effect is powerful and, I think, valid. I will focus here on one particular aspect that I think is questionable. Let me be clear: I don't regard the questioning of this particular point as undermining Marx's general thrust.
The point at issue comes up in section 3, part (b) of the chapter, ``The Prolongation of the Working Day''. Having mentioned the physical deterioration of fixed capital equipment over time, Marx says:
But in addition to the material wear and tear, a machine also undergoes what one might call a moral depreciation. It loses exchange-value, either because machines of the same sort are being produced more cheaply than it was, or because better machines are entering into competition with it. In both cases, however young and full of life the machine may be, its value is no longer determined by the necessary labour-time actually objectified in it, but by the labour-time necessary to produce either it or the better machine. It has therefore been devalued to a greater or less extent. (p. 528)
This ``moral depreciation'' threatens the capitalist with a loss on capital account. He may purchase an expensive new piece of capital equipment, only to find that shortly thereafter its value is drastically reduced by some further innovation.
If we ignore moral depreciation, we could think of a piece of fixed capital as gradually, over the course of its normal working life, transferring its initial value to the product. Thus for instance a machine might have a value of $10,000 and a working life of five years: in that case we could think of it as transferring $2,000 to the product each year (after five years it's worthless scrap).
Moral depreciation threatens to truncate this process. In the extreme case the machine may be rendered worthless before the five years of its ``physical'' life are up. This could happen if a better machine comes on the scene, capable of producing the given output for a total cost below the variable cost of its production using the older machine, and if the cost associated with the new machine sets the market price. (Here I'm using ``standard'' rather than Marxian terminology for costs.) Short of this extreme, the owner of the old machine can still suffer a setback. The old machine is devalued, even if not to zero.
Now Marx draws out an implication: ``The shorter the period taken to reproduce [the machine's] total value, the less is the danger of moral depreciation; and the longer the working day, the shorter that period in fact is'' (p. 528). That is, the capitalist can reduce his exposure to the hazard of moral depreciation if he can ``use up'' his machine (get it to transfer its value fully to the product) more quickly. The way to use it up more quickly is to run it more hours in the day, i.e. to prolong the working day.
Marx says that this ``special incentive'' to prolong the working day is felt most acutely in the ``early days of a machine's life'', when it is most vulnerable to competition from the cascade of related innovations that often accompanies the birth of a new technology.
The aspect of Marx's argument here that I'm unsure about is his claim that moral depreciation provides a ``special incentive'' to lengthen the working day-that is, an incentive that would not be present even in the absence of moral depreciation.
Suppose a capitalist invests in a fixed capital of value K which, absent moral depreciation, will last n years and transfer Cd = 1/n of its value to the product each year, when run for 12 hours per day. Suppose that operating the equipment at this rate involves hiring a workforce for an annual wage bill (variable capital) of V and the working up of materials to an annual value of Cm. Let the rate of surplus value be 100 percent. Assume the capitalist pays wages in arrears (out of the value of the product sold) and gets his materials on trade credit, so that his total capital stock is just K.
Under these assumptions the annual value of output is
Assuming the product is sold at its value, the capitalist makes an annual profit of S = V and gains a rate of profit on his capital stock of V/K. This is sustained over n years, at the end of which the capitalist has his K back, ready to buy a new machine, plus profit.
That's the benchmark. Now suppose the capitalist moves to operating his plant 24 hours per day. What happens?
As regards the fixed capital, let's assume this change simply means it wears out twice as quickly. It lasts n/2 years, and transfers to the product 2K/n per year.
As regards materials and labour costs, the simplest assumption would be that everything just scales proportionately: in shifting to 24-hour operation the capitalist pays twice the wages and twice the materials bill for twice the amount of labour and materials, and generates twice the output. The rate of surplus value remains at 100 percent. We'll run with this for the moment but will consider alternative scenarios shortly.
In that case the annual value of output becomes
Under these simplifying assumptions, therefore, the capitalist has a clear incentive to operate his fixed capital up to the maximum possible hours per day-even supposing there is no threat of moral depreciation.
How might matters change if we attempt to be a bit more realistic? Let's stay with the assumptions above regarding the depreciation of fixed capital, but drop the assumption that ``everything just scales'' with regard to labour and materials. If the capitalist wants to extend the hours of operation of his factory from 12 to 24 he'll have to do some combination of the following:
To the extent he manages to get more labour out of his existing workforce for the same daily wage, the capitalist is garnering an increase in the rate of surplus value, and hence the increase in the rate of profit will be greater than above. If, on the other hand, the hourly wage stays the same, then to a first approximation the rate of surplus value remains constant (as assumed above). But it's possible that prolongation of the typical worker's working day runs the capitalist into diminishing returns (the exhausted workers can't produce so much per hour as in the shorter day). In that case the rate of surplus value falls and the increase in the rate of profit will be less than above.
With shift work, will the capitalist have to offer a higher hourly wage to get hold of workers for the back shift? If so, the rate of surplus value falls (at the margin) and again the increase in the rate of profit shown above is an overstatement.
One other consideration should be mentioned. This is all under the assumption that the capitalist can sell as much output as he wants at a price corresponding to its value. Also, the time-phasing of production doesn't matter (stuff produced in the middle of the night is ``just as good'' as stuff produced during the day). These conditions are not necessarily met. What if the firm is very large in relation to the market, so that a doubling of its output would drive down the market price? What if it's producing a perishable commodity for a market that operates during the day only? But I think it's clear Marx has in mind competitive manufacturing firms producing a storable commodity for a large market. In that context, I'm saying, capitalists have a clear incentive to produce on a 24-hour basis unless (a) they run into severely diminishing returns to labour, or (b) they have to pay a much higher wage for shift labour, to the extent that the cost of labour-power and materials, per unit of output, exceeds the selling price of the product at the margin of the working day.
If that analysis is correct, it seems there's nothing left for moral depreciation to add to the capitalist's calculation. Given his appetite for surplus value he should already be producing to the maximum daily hours. Yes, he'd like to escape the hazard of moral depreciation, but this merely overdetermines a course of action to which he's already committed. The limit to the prolongation of the working day is the condition of zero marginal profit, and the position of that limit is governed by the degree of diminishing returns to labour and the characteristics of the market in shift labour, along with product-market conditions.
Although I'm arguing that Marx's appeal to moral depreciation doesn't really strengthen the main argument of this chapter of Capital, I'd like to clarify one point in a way that brings me back closer to Marx.
I'm not simply saying that the capitalist ``always has an incentive to make the working day as long as possible'', period, and that this in itself undermines the use Marx wants to make of moral depreciation. If we take the daily wage as given, and if diminishing returns to working hours don't actually take us into the region of negative returns, then it's obvious that capitalists have an incentive to lengthen the working day. Marx calls this absolute surplus value, and he's quite clear that capitalists will grab as much of it as they can.
My point is rather that if fixed capital is important, and if capitalists are calculating in terms of the rate of return on capital stock, then (under the conditions stated) they will have an incentive to lengthen the working time to the maximum, even if they have to pay for the extra labour at the original rate per hour or somewhat more. My point is not to do with any increase in the rate of surplus value that might be achieved by lengthening the daily working time: it turns on an increase in the mass of surplus value gained per year, as a ratio to the capital stock invested.
In other words, I'm not dissolving the point Marx wants to make in chapter 15 into the obvious generality that capitalists will take absolute surplus value whenever they can get it. I'm agreeing with him that with ``machinery and large-scale industry''-with an increased importance for fixed capital, and with an increase in the proportion of the economy represented by competitive firms producing durable commodities for large markets-there is an increasing premium on round-the-clock operation. And I have no problem with his general account of the impact this has on the workers.
To sustain Marx's point that the threat of moral depreciation makes capitalists push to lengthen the working day with a greater force than that due to the general importance of fixed capital and industrial production for large markets (as explained above), we'd need to conceive of capitalists in a particular way: not as rigorous profit-maximizers but as ``satisficers'' (in Herbert Simon's terminology). We'd have to think of them not as pursuing maximum profit in any given circumstances, but as waiting around until they're kicked in the pants by the threat of moral depreciation, before they take an action that was available to them before and would have increased their profits before. There may actually be some truth in this characterization of capitalists (at least for some places and times), but it's not the way in which Marx generally represents them.
1Pages references are to the Penguin edition, tr. Ben Fowkes.