Take two branches of business with the same length of working day — say ten hours — cotton spinning and locomotive building. In one branch, a fixed amount of finished product comes off the line every day or every week: cotton yarn. In the other, the work process has to be repeated for maybe three months before there is one finished product: a locomotive. In the first case the product is discrete — the same work starts fresh each day or week. In the second, the process is continuous: it stretches over a long run of daily work processes, and only their connection, their unbroken continuity, yields a finished product after a much longer stretch. The length of the daily work process is the same in both cases. But there is a big difference in how long it takes to complete the act of production — how long the repeated work processes must run before the product is finished, sent to market as a commodity, and turned from productive capital into commodity-capital. This has nothing to do with the difference between fixed and circulating capital. The difference just described would exist even if both branches used exactly the same proportions of fixed and circulating capital.
These differences in how long the act of production takes show up not only between different branches of production but within the same branch, depending on how big the product is. An ordinary house is built faster than a large factory and so needs fewer connected work processes. Building a locomotive takes three months; building an ironclad warship takes one year or more. Growing grain takes almost a year; raising cattle takes several years; growing timber can take anywhere from 12 to 100 years. A country road might be built in a few months where a railway takes years; an ordinary carpet might take a week, a Gobelins tapestry years. The differences in how long the act of production takes are endlessly varied.
A difference in how long the act of production takes must obviously produce a difference in how fast the capital turns over, given an equal outlay of capital — that is, in the length of time for which a given capital stays advanced. Suppose the machine-spinning mill and the locomotive factory employ equally large capitals; the split between constant and variable capital is the same in both, and so is the split between the fixed and the fluid parts of the capital; and finally the working day is the same length, split the same way between necessary labour and surplus labour. And to clear away everything that comes from the circulation process and has nothing to do with this comparison, let's assume both the yarn and the locomotive are made to order and paid for on delivery of the finished product. At the end of the week, on delivery of the finished yarn, the spinning manufacturer (we set aside surplus-value here) gets back the circulating capital he laid out, plus the wear of the fixed capital that is now sitting in the value of the yarn. He can start the same cycle over again with the same capital — his turnover is complete. The locomotive manufacturer, by contrast, has to lay out fresh capital in wages and raw material week after week for the whole three months. Only after three months, when the locomotive is delivered, does the circulating capital he has been laying out bit by bit — for one and the same act of production, making one and the same commodity — come back in a form that lets him start the cycle again. Only then, too, does he get back the wear on his machinery over those three months. One manufacturer's outlay is for one week. The other's is the weekly outlay multiplied by twelve. Everything else being equal, one of them has to have twelve times as much circulating capital on hand as the other.
That the two weekly outlays happen to be equal doesn't matter here. Whatever the size of the advanced capital, in one case it is advanced for only a week, in the other for twelve weeks, before it can be put to work again — either repeating the same operation or starting a different one.
The difference in how fast the turnover goes — in how long an individual capital must stay advanced before the same capital-value can serve again for a new process of work or valorization — comes from the following:
Suppose building the locomotive, or some other machine, costs 100 working days. For the workers employed in spinning and in machine-building alike, these 100 working days are a discontinuous, discrete quantity: on our assumption, 100 separate, successive ten-hour work processes. But looked at from the side of the product — the machine — those same 100 working days form a continuous quantity: a working day of 1,000 working hours, one single connected act of production. Only with respect to the product does it look this way, not to the worker, whose days stay ten hours long and separate as ever. A working day built this way, out of a run of more or less numerous connected working days, is what I call a working period. When we speak of the working day, we mean the length of working time during which the worker has to spend his labour-power, has to work, each day. When we speak instead of the working period, we mean the number of connected working days that a given branch of business needs to deliver one finished product. Here, the product of each single working day is only a partial product, carried a little further each day, and it only takes its finished shape — becomes a finished use-value — at the end of the longer or shorter working period.
So interruptions and disturbances in social production — from a crisis, say — hit products that are discrete in nature very differently than they hit products that need a long, connected period to make. If a certain amount of yarn or coal is produced today, in the discrete case tomorrow's production of yarn or coal simply doesn't happen — that's all. Not so with ships, buildings, railways. There it isn't just the labour that stops; a connected act of production is broken off. If the work isn't carried on, the means of production and the labour already used up in it have been spent for nothing. And even if work starts up again, deterioration will always have set in during the interruption.
Throughout the whole working period, the portion of value that the fixed capital gives off to the product each day piles up, layer on layer, until the product is finished. And here the difference between fixed and circulating capital shows its practical importance. Fixed capital is advanced to the production process for a long stretch of time; it doesn't need to be renewed before that stretch — maybe several years — is up.
Whether the steam engine gives off its value to the yarn — the product of a discrete work process — piece by piece each day, or gives it off to a locomotive — the product of one continuous act of production — over three months, makes no difference at all to the outlay needed to buy the steam engine in the first place. In one case its value flows back in small doses, say weekly; in the other, in bigger lumps, say every three months. But either way, the steam engine itself is only renewed after maybe 20 years. As long as each of these paybacks — where its value flows back piece by piece through the sale of the product — is shorter than the engine's own working life, the same steam engine goes on functioning through several working periods.
It's different with the circulating parts of the advanced capital. The labour-power bought for this week is used up during this week and has taken shape in the product. It has to be paid at the end of the week. And this outlay in labour-power repeats every week for the three months — spending this part of the capital in one week doesn't let the capitalist cover next week's purchase of labour. Fresh, additional capital has to be laid out every week to pay for labour-power, and — leaving all credit arrangements aside — the capitalist has to be able to advance three months' worth of wages, even though he pays it out only in weekly doses. The same goes for the other part of the circulating capital, the raw and auxiliary materials. Layer of labour piles on layer of labour on the product. Not only is the value of the labour-power used up added onto the product — surplus-value is added too, continuously, all through the work process. But it's added onto a product that isn't finished yet, that doesn't yet have the shape of a finished commodity — so it can't circulate yet. The same holds for the capital-value transferred from the raw and auxiliary materials onto the product, layer by layer.
Depending on whether the working period required — by the product's own particular nature, or by the use-effect to be achieved — is longer or shorter, there's a constant need for extra outlay of circulating capital — wages, raw and auxiliary materials — none of which is sitting in a form that could circulate and so serve to start the same operation over again. Instead, each part gets locked up, one after another, as a piece of the product taking shape within the sphere of production — tied up in the form of productive capital. But turnover time is equal to the sum of the capital's production time and its circulation time. So lengthening the production time cuts the speed of turnover just as much as lengthening the circulation time would. In the case in front of us, though, there are two things to note:
First: the longer stay in the sphere of production. The capital advanced, say, in the first week — in labour, raw material, and so on — along with the portions of value the fixed capital has given off to the product, stay locked into the sphere of production for the whole three-month term. Built into a product that is still taking shape, still unfinished, they cannot enter circulation as a commodity.
Second: since the working period the act of production needs lasts three months — really forming just one connected work process — a new dose of circulating capital has to keep being added to the previous ones, week after week. So the mass of capital advanced in this piling-up way grows with the length of the working period.
We've assumed that the spinning mill and the machine factory have equally large capitals invested, that these capitals are split in the same proportions between constant and variable capital, and likewise between fixed and circulating, and that the working days are the same length — in short, that everything is the same except the length of the working period. In the first week, the outlay is equally large for both. But the spinner's product can be sold, and the proceeds used to buy new labour-power and new raw materials — production carries on at the same scale. The machine-builder, by contrast, can only turn the circulating capital he spent in the first week back into money after three months, once his product is finished — and only then can he operate with it again. So, first, there's a difference in how fast the same amount of advanced capital comes back. But second: during the three months, an equally large productive capital is at work in the spinning mill and in machine-building — yet the size of the capital outlay is quite different for the spinner and the machine-builder. In one case, the same capital renews itself quickly and so the same operation can be repeated straight away; in the other, it renews itself only slowly, relatively speaking, so fresh amounts of capital keep having to be added to the old ones until the time it's renewed comes round. So it isn't just the length of time it takes for a given portion of capital to renew itself — the length of the advance — that differs; the mass of capital that has to be advanced differs too (even though the capital applied per day or per week is the same), depending on how long the work process runs. This is worth holding onto, because the length of the advance can grow — as it may in cases we'll look at in the next chapter — without the mass of capital that has to be advanced growing in proportion to that length of time. The capital has to be advanced for longer, and a bigger quantity of capital ends up tied up in the form of productive capital.
In the less developed stages of capitalist production, undertakings that need a long working period — a lot of capital laid out for a long stretch — are not run as capitalist businesses at all, especially when they only make sense on a large scale. Roads, canals, and the like get built at the expense of the community or the state instead (in earlier times mostly using forced labour, as far as the actual work went). Or, where a product does need a long working period to make, only a small part of it gets financed out of the capitalist's own funds. Take house-building: the private person the house is being built for pays the builder in instalments, portion by portion, so they are really paying for the house piece by piece, as it gets built. In the developed capitalist era, though, things differ. Huge amounts of capital sit concentrated in individual hands; alongside individual capitalists there is now the associated capitalist — the joint-stock company; and credit is fully developed. A capitalist builder now only builds to order for private individuals as an exception. His business is building rows of houses and whole city districts for the market — the same way some capitalists make a business of building railways as contractors.
Testimony from a builder before the 1857 Bank Committee shows what capitalist production did to house-building in London. He said that in his youth, houses were mostly built to order, with the buyer paying the builder in instalments as certain stages of the building were finished. Very little was built speculatively — builders mainly did that just to keep their workers steadily employed and together. Over the last forty years all this changed. Almost nothing is built to order any more. Someone who needs a new house picks one out from among the houses already built speculatively, or still being built. The builder no longer works for a customer but for the market — like any other manufacturer, he has to have finished goods ready in the market. Where a builder might once have had three or four houses under speculative construction at a time, he now has to buy an extensive plot of land (or, as they'd put it on the continent, lease it, usually for ninety-nine years), put up as many as 100 or 200 houses on it, and take on an undertaking twenty to fifty times bigger than his own wealth. He raises the funds by taking out mortgages, and the money is handed to him as the building of the individual houses goes forward. If a crisis then comes along and the instalment payments stop coming in, the whole undertaking usually collapses — at best the houses stay unfinished until better times, at worst they go under the hammer and are sold off at half price. No builder can get anywhere today without building speculatively, and on a large scale at that. The profit from the building itself is tiny; his real gain comes from the rise in ground rent, from choosing and exploiting building land skilfully. It is this kind of speculation, anticipating the demand for houses, that has built almost the whole of Belgravia and Tyburnia, and the countless thousands of villas around London. (Condensed from Report from the Select Committee on Bank Acts, Part I, 1857, Evidence, questions 5413-5418, 5435-5436.)
Work with a considerably long working period and on a large scale only falls fully into the hands of capitalist production once the concentration of capital is already very substantial — and once the credit system has developed far enough to offer the capitalist a convenient way to advance somebody else's capital instead of his own, and so risk somebody else's capital instead of his own. But it goes without saying that whether the capital advanced to production belongs to the person using it or not makes no difference at all to how fast it turns over, or to the turnover time.
The same things that let a single day's labour produce more — cooperation, division of labour, the use of machinery — also shorten the working period for connected acts of production. Machinery shortens the time it takes to build houses or bridges; the reaping machine and the threshing machine shorten the working period needed to turn ripened corn into a finished commodity. Better shipbuilding, by adding speed, shortens the turnover time of capital laid out in shipping. But these improvements, which shorten the working period and so shorten the time for which circulating capital has to be advanced, mostly come bundled with a bigger outlay of fixed capital. On the other hand, in certain branches the working period can be shortened simply by extending cooperation: a railway gets finished sooner because large armies of workers are put to work on it, attacking the job from many points in space at once. Here the turnover time is shortened by growth in the capital advanced — more means of production and more labour-power have to be brought together under the capitalist's command.
So the shortening of the working period usually goes together with an increase in the capital that has to be advanced for that shorter time — the shorter the time it is advanced for, the larger the mass in which it has to be advanced. This is worth remembering: quite apart from how much social capital exists overall, what matters is the degree to which the means of production and of subsistence — or the power to dispose of them — are scattered among many hands or gathered into the hands of individual capitalists; in other words, how far the concentration of capitals has already gone. Wherever credit helps bring about, speeds up, and increases the concentration of capital in one hand, it helps to shorten the working period, and with it the turnover time.
In branches of production where the working period — whether continuous or broken up — is fixed by definite natural conditions, none of the means described above can shorten it.
'The phrase "quicker turnover" cannot be applied to corn harvests, since only one turnover a year is possible. As for livestock, we need only ask: how is one to speed up the turnover of two- and three-year-old sheep, and four- and five-year-old oxen?' (W. Walter Good, Political, Agricultural, and Commercial Fallacies, London 1866, p. 325.)
The need to have money in hand earlier — to pay fixed obligations like taxes or ground rent, for instance — gets around this by having livestock sold and slaughtered before it has reached the age that suits it economically, greatly to the harm of agriculture. In the end this also pushes up the price of meat.
'The people who used to breed cattle mainly to stock the pastures of the Midland counties [inland counties] in summer and the stalls of the eastern counties in winter... have been brought so low by the swings and falls in corn prices that they are glad to profit from the high price of butter and cheese. They bring the butter to market every week to cover their running costs; against the cheese they take advances from a factor, who collects it as soon as it can be moved, and who of course sets his own price for it. For this reason, and because farming is run on the principles of political economy, the calves that used to be sent south from the dairy districts to be reared are now sacrificed wholesale, often when they are only eight or ten days old, in the slaughterhouses of Birmingham, Manchester, Liverpool, and other big towns nearby. If malt were not taxed, the tenant farmers would not only have made more profit and so been able to keep their young stock until it grew older and heavier — malt could also have been used instead of milk to rear calves, by people who keep no cows — and the present alarming shortage of young stock would largely have been avoided. If these small farmers are now advised to rear their calves, they say: we know very well that rearing them on milk would pay, but first we would have to lay out money, and we cannot; and second we would have to wait a long time to get our money back, whereas in dairying we get it back at once.' (Ibid., pp. 11, 12.)
If the lengthening of the turnover has such effects even on smaller English tenant farmers, it is easy to see what disruption it must cause among the small farmers of the continent.
The longer the working period — and so the longer the stretch of time before the commodity is finished and ready to circulate — the more the portion of value that fixed capital gives off to the product, layer by layer, piles up, and the more the return of that portion of value is delayed. But this delay does not cause any fresh outlay of fixed capital. The machine goes on working in the production process whether the money that replaces its wear flows back slowly or quickly. Circulating capital is different. Not only must capital be tied up for longer, the longer the working period; fresh capital must also constantly be advanced for wages and for raw and auxiliary materials. So a delayed return affects the two kinds of capital differently. Fixed capital goes on working whether the return is slow or fast. Circulating capital, though, stops being able to function once the return is delayed, if it is stuck in the form of an unsold or unfinished product that cannot yet be sold, and there is no extra capital on hand to replace it in kind.
'While the peasant starves, his cattle thrive. There had been fairly good rain, and the grass stood thick. The Indian peasant will starve beside a fat ox. The rules of superstition may look cruel toward the individual, but they preserve society: keeping working cattle alive keeps agriculture going, and with it the sources of future subsistence and wealth. It may sound hard and sad, but it is so — in India a man is easier to replace than an ox.' (Return, East India. Madras and Orissa Famine, no. 4, p. 44.)
Compare this with the sentence from the Manava-Dharma-Sastra, Chapter X, §62:
'Giving up one's life without reward, to preserve a priest or a cow... can secure the bliss of these low-born tribes.'
Naturally it is impossible to deliver a five-year-old animal before five years are up. But within certain limits, it is possible to change the way animals are handled so that they are ready for their purpose sooner. Bakewell in particular achieved this. English sheep used to be — like French sheep still in 1855 — not ready for slaughter before their fourth or fifth year. Under Bakewell's system, even a one-year-old sheep can already be fattened, and in every case it is fully grown by the end of its second year. Through careful selective breeding, Bakewell, the tenant of Dishley Grange, reduced the sheep's bone skeleton to the minimum needed for it to survive. His sheep were called the New Leicesters.
'The breeder can now bring three sheep to market in the time it used to take him to finish one, and with a broader, rounder, larger development of the parts that give the most meat. Almost the whole of their weight is pure meat.' (Lavergne, The Rural Economy of England etc., 1855, p. 20.)
The methods that shorten the working period apply in very different degrees in different branches of industry, and they do not even out the differences in how long the various working periods are. To stick with our example: using new machine tools may shorten, in absolute terms, the working period needed to build a locomotive. But if improved processes in spinning increase the finished product delivered daily or weekly at a far faster rate, then the working period in machine manufacture has still grown longer relative to spinning — even though it shortened in itself.
Take two branches of business with the same length of working day — say ten hours — cotton spinning and locomotive building. In one branch, a fixed amount of finished product comes off the line every day or every week: cotton yarn. In the other, the work process has to be repeated for maybe three months before there is one finished product: a locomotive. In the first case the product is discrete — the same work starts fresh each day or week. In the second, the process is continuous: it stretches over a long run of daily work processes, and only their connection, their unbroken continuity, yields a finished product after a much longer stretch. The length of the daily work process is the same in both cases. But there is a big difference in how long it takes to complete the act of production — how long the repeated work processes must run before the product is finished, sent to market as a commodity, and turned from productive capital into commodity-capital. This has nothing to do with the difference between fixed and circulating capital. The difference just described would exist even if both branches used exactly the same proportions of fixed and circulating capital.
These differences in how long the act of production takes show up not only between different branches of production but within the same branch, depending on how big the product is. An ordinary house is built faster than a large factory and so needs fewer connected work processes. Building a locomotive takes three months; building an ironclad warship takes one year or more. Growing grain takes almost a year; raising cattle takes several years; growing timber can take anywhere from 12 to 100 years. A country road might be built in a few months where a railway takes years; an ordinary carpet might take a week, a Gobelins tapestry years. The differences in how long the act of production takes are endlessly varied.
A difference in how long the act of production takes must obviously produce a difference in how fast the capital turns over, given an equal outlay of capital — that is, in the length of time for which a given capital stays advanced. Suppose the machine-spinning mill and the locomotive factory employ equally large capitals; the split between constant and variable capital is the same in both, and so is the split between the fixed and the fluid parts of the capital; and finally the working day is the same length, split the same way between necessary labour and surplus labour. And to clear away everything that comes from the circulation process and has nothing to do with this comparison, let's assume both the yarn and the locomotive are made to order and paid for on delivery of the finished product. At the end of the week, on delivery of the finished yarn, the spinning manufacturer (we set aside surplus-value here) gets back the circulating capital he laid out, plus the wear of the fixed capital that is now sitting in the value of the yarn. He can start the same cycle over again with the same capital — his turnover is complete. The locomotive manufacturer, by contrast, has to lay out fresh capital in wages and raw material week after week for the whole three months. Only after three months, when the locomotive is delivered, does the circulating capital he has been laying out bit by bit — for one and the same act of production, making one and the same commodity — come back in a form that lets him start the cycle again. Only then, too, does he get back the wear on his machinery over those three months. One manufacturer's outlay is for one week. The other's is the weekly outlay multiplied by twelve. Everything else being equal, one of them has to have twelve times as much circulating capital on hand as the other.
That the two weekly outlays happen to be equal doesn't matter here. Whatever the size of the advanced capital, in one case it is advanced for only a week, in the other for twelve weeks, before it can be put to work again — either repeating the same operation or starting a different one.
The difference in how fast the turnover goes — in how long an individual capital must stay advanced before the same capital-value can serve again for a new process of work or valorization — comes from the following:
Suppose building the locomotive, or some other machine, costs 100 working days. For the workers employed in spinning and in machine-building alike, these 100 working days are a discontinuous, discrete quantity: on our assumption, 100 separate, successive ten-hour work processes. But looked at from the side of the product — the machine — those same 100 working days form a continuous quantity: a working day of 1,000 working hours, one single connected act of production. Only with respect to the product does it look this way, not to the worker, whose days stay ten hours long and separate as ever. A working day built this way, out of a run of more or less numerous connected working days, is what I call a working period. When we speak of the working day, we mean the length of working time during which the worker has to spend his labour-power, has to work, each day. When we speak instead of the working period, we mean the number of connected working days that a given branch of business needs to deliver one finished product. Here, the product of each single working day is only a partial product, carried a little further each day, and it only takes its finished shape — becomes a finished use-value — at the end of the longer or shorter working period.
So interruptions and disturbances in social production — from a crisis, say — hit products that are discrete in nature very differently than they hit products that need a long, connected period to make. If a certain amount of yarn or coal is produced today, in the discrete case tomorrow's production of yarn or coal simply doesn't happen — that's all. Not so with ships, buildings, railways. There it isn't just the labour that stops; a connected act of production is broken off. If the work isn't carried on, the means of production and the labour already used up in it have been spent for nothing. And even if work starts up again, deterioration will always have set in during the interruption.
Throughout the whole working period, the portion of value that the fixed capital gives off to the product each day piles up, layer on layer, until the product is finished. And here the difference between fixed and circulating capital shows its practical importance. Fixed capital is advanced to the production process for a long stretch of time; it doesn't need to be renewed before that stretch — maybe several years — is up.
Whether the steam engine gives off its value to the yarn — the product of a discrete work process — piece by piece each day, or gives it off to a locomotive — the product of one continuous act of production — over three months, makes no difference at all to the outlay needed to buy the steam engine in the first place. In one case its value flows back in small doses, say weekly; in the other, in bigger lumps, say every three months. But either way, the steam engine itself is only renewed after maybe 20 years. As long as each of these paybacks — where its value flows back piece by piece through the sale of the product — is shorter than the engine's own working life, the same steam engine goes on functioning through several working periods.
It's different with the circulating parts of the advanced capital. The labour-power bought for this week is used up during this week and has taken shape in the product. It has to be paid at the end of the week. And this outlay in labour-power repeats every week for the three months — spending this part of the capital in one week doesn't let the capitalist cover next week's purchase of labour. Fresh, additional capital has to be laid out every week to pay for labour-power, and — leaving all credit arrangements aside — the capitalist has to be able to advance three months' worth of wages, even though he pays it out only in weekly doses. The same goes for the other part of the circulating capital, the raw and auxiliary materials. Layer of labour piles on layer of labour on the product. Not only is the value of the labour-power used up added onto the product — surplus-value is added too, continuously, all through the work process. But it's added onto a product that isn't finished yet, that doesn't yet have the shape of a finished commodity — so it can't circulate yet. The same holds for the capital-value transferred from the raw and auxiliary materials onto the product, layer by layer.
Depending on whether the working period required — by the product's own particular nature, or by the use-effect to be achieved — is longer or shorter, there's a constant need for extra outlay of circulating capital — wages, raw and auxiliary materials — none of which is sitting in a form that could circulate and so serve to start the same operation over again. Instead, each part gets locked up, one after another, as a piece of the product taking shape within the sphere of production — tied up in the form of productive capital. But turnover time is equal to the sum of the capital's production time and its circulation time. So lengthening the production time cuts the speed of turnover just as much as lengthening the circulation time would. In the case in front of us, though, there are two things to note:
First: the longer stay in the sphere of production. The capital advanced, say, in the first week — in labour, raw material, and so on — along with the portions of value the fixed capital has given off to the product, stay locked into the sphere of production for the whole three-month term. Built into a product that is still taking shape, still unfinished, they cannot enter circulation as a commodity.
Second: since the working period the act of production needs lasts three months — really forming just one connected work process — a new dose of circulating capital has to keep being added to the previous ones, week after week. So the mass of capital advanced in this piling-up way grows with the length of the working period.
We've assumed that the spinning mill and the machine factory have equally large capitals invested, that these capitals are split in the same proportions between constant and variable capital, and likewise between fixed and circulating, and that the working days are the same length — in short, that everything is the same except the length of the working period. In the first week, the outlay is equally large for both. But the spinner's product can be sold, and the proceeds used to buy new labour-power and new raw materials — production carries on at the same scale. The machine-builder, by contrast, can only turn the circulating capital he spent in the first week back into money after three months, once his product is finished — and only then can he operate with it again. So, first, there's a difference in how fast the same amount of advanced capital comes back. But second: during the three months, an equally large productive capital is at work in the spinning mill and in machine-building — yet the size of the capital outlay is quite different for the spinner and the machine-builder. In one case, the same capital renews itself quickly and so the same operation can be repeated straight away; in the other, it renews itself only slowly, relatively speaking, so fresh amounts of capital keep having to be added to the old ones until the time it's renewed comes round. So it isn't just the length of time it takes for a given portion of capital to renew itself — the length of the advance — that differs; the mass of capital that has to be advanced differs too (even though the capital applied per day or per week is the same), depending on how long the work process runs. This is worth holding onto, because the length of the advance can grow — as it may in cases we'll look at in the next chapter — without the mass of capital that has to be advanced growing in proportion to that length of time. The capital has to be advanced for longer, and a bigger quantity of capital ends up tied up in the form of productive capital.
In the less developed stages of capitalist production, undertakings that need a long working period — a lot of capital laid out for a long stretch — are not run as capitalist businesses at all, especially when they only make sense on a large scale. Roads, canals, and the like get built at the expense of the community or the state instead (in earlier times mostly using forced labour, as far as the actual work went). Or, where a product does need a long working period to make, only a small part of it gets financed out of the capitalist's own funds. Take house-building: the private person the house is being built for pays the builder in instalments, portion by portion, so they are really paying for the house piece by piece, as it gets built. In the developed capitalist era, though, things differ. Huge amounts of capital sit concentrated in individual hands; alongside individual capitalists there is now the associated capitalist — the joint-stock company; and credit is fully developed. A capitalist builder now only builds to order for private individuals as an exception. His business is building rows of houses and whole city districts for the market — the same way some capitalists make a business of building railways as contractors.
Testimony from a builder before the 1857 Bank Committee shows what capitalist production did to house-building in London. He said that in his youth, houses were mostly built to order, with the buyer paying the builder in instalments as certain stages of the building were finished. Very little was built speculatively — builders mainly did that just to keep their workers steadily employed and together. Over the last forty years all this changed. Almost nothing is built to order any more. Someone who needs a new house picks one out from among the houses already built speculatively, or still being built. The builder no longer works for a customer but for the market — like any other manufacturer, he has to have finished goods ready in the market. Where a builder might once have had three or four houses under speculative construction at a time, he now has to buy an extensive plot of land (or, as they'd put it on the continent, lease it, usually for ninety-nine years), put up as many as 100 or 200 houses on it, and take on an undertaking twenty to fifty times bigger than his own wealth. He raises the funds by taking out mortgages, and the money is handed to him as the building of the individual houses goes forward. If a crisis then comes along and the instalment payments stop coming in, the whole undertaking usually collapses — at best the houses stay unfinished until better times, at worst they go under the hammer and are sold off at half price. No builder can get anywhere today without building speculatively, and on a large scale at that. The profit from the building itself is tiny; his real gain comes from the rise in ground rent, from choosing and exploiting building land skilfully. It is this kind of speculation, anticipating the demand for houses, that has built almost the whole of Belgravia and Tyburnia, and the countless thousands of villas around London. (Condensed from Report from the Select Committee on Bank Acts, Part I, 1857, Evidence, questions 5413-5418, 5435-5436.)
Work with a considerably long working period and on a large scale only falls fully into the hands of capitalist production once the concentration of capital is already very substantial — and once the credit system has developed far enough to offer the capitalist a convenient way to advance somebody else's capital instead of his own, and so risk somebody else's capital instead of his own. But it goes without saying that whether the capital advanced to production belongs to the person using it or not makes no difference at all to how fast it turns over, or to the turnover time.
The same things that let a single day's labour produce more — cooperation, division of labour, the use of machinery — also shorten the working period for connected acts of production. Machinery shortens the time it takes to build houses or bridges; the reaping machine and the threshing machine shorten the working period needed to turn ripened corn into a finished commodity. Better shipbuilding, by adding speed, shortens the turnover time of capital laid out in shipping. But these improvements, which shorten the working period and so shorten the time for which circulating capital has to be advanced, mostly come bundled with a bigger outlay of fixed capital. On the other hand, in certain branches the working period can be shortened simply by extending cooperation: a railway gets finished sooner because large armies of workers are put to work on it, attacking the job from many points in space at once. Here the turnover time is shortened by growth in the capital advanced — more means of production and more labour-power have to be brought together under the capitalist's command.
So the shortening of the working period usually goes together with an increase in the capital that has to be advanced for that shorter time — the shorter the time it is advanced for, the larger the mass in which it has to be advanced. This is worth remembering: quite apart from how much social capital exists overall, what matters is the degree to which the means of production and of subsistence — or the power to dispose of them — are scattered among many hands or gathered into the hands of individual capitalists; in other words, how far the concentration of capitals has already gone. Wherever credit helps bring about, speeds up, and increases the concentration of capital in one hand, it helps to shorten the working period, and with it the turnover time.
In branches of production where the working period — whether continuous or broken up — is fixed by definite natural conditions, none of the means described above can shorten it.
'The phrase "quicker turnover" cannot be applied to corn harvests, since only one turnover a year is possible. As for livestock, we need only ask: how is one to speed up the turnover of two- and three-year-old sheep, and four- and five-year-old oxen?' (W. Walter Good, Political, Agricultural, and Commercial Fallacies, London 1866, p. 325.)
The need to have money in hand earlier — to pay fixed obligations like taxes or ground rent, for instance — gets around this by having livestock sold and slaughtered before it has reached the age that suits it economically, greatly to the harm of agriculture. In the end this also pushes up the price of meat.
'The people who used to breed cattle mainly to stock the pastures of the Midland counties [inland counties] in summer and the stalls of the eastern counties in winter... have been brought so low by the swings and falls in corn prices that they are glad to profit from the high price of butter and cheese. They bring the butter to market every week to cover their running costs; against the cheese they take advances from a factor, who collects it as soon as it can be moved, and who of course sets his own price for it. For this reason, and because farming is run on the principles of political economy, the calves that used to be sent south from the dairy districts to be reared are now sacrificed wholesale, often when they are only eight or ten days old, in the slaughterhouses of Birmingham, Manchester, Liverpool, and other big towns nearby. If malt were not taxed, the tenant farmers would not only have made more profit and so been able to keep their young stock until it grew older and heavier — malt could also have been used instead of milk to rear calves, by people who keep no cows — and the present alarming shortage of young stock would largely have been avoided. If these small farmers are now advised to rear their calves, they say: we know very well that rearing them on milk would pay, but first we would have to lay out money, and we cannot; and second we would have to wait a long time to get our money back, whereas in dairying we get it back at once.' (Ibid., pp. 11, 12.)
If the lengthening of the turnover has such effects even on smaller English tenant farmers, it is easy to see what disruption it must cause among the small farmers of the continent.
The longer the working period — and so the longer the stretch of time before the commodity is finished and ready to circulate — the more the portion of value that fixed capital gives off to the product, layer by layer, piles up, and the more the return of that portion of value is delayed. But this delay does not cause any fresh outlay of fixed capital. The machine goes on working in the production process whether the money that replaces its wear flows back slowly or quickly. Circulating capital is different. Not only must capital be tied up for longer, the longer the working period; fresh capital must also constantly be advanced for wages and for raw and auxiliary materials. So a delayed return affects the two kinds of capital differently. Fixed capital goes on working whether the return is slow or fast. Circulating capital, though, stops being able to function once the return is delayed, if it is stuck in the form of an unsold or unfinished product that cannot yet be sold, and there is no extra capital on hand to replace it in kind.
'While the peasant starves, his cattle thrive. There had been fairly good rain, and the grass stood thick. The Indian peasant will starve beside a fat ox. The rules of superstition may look cruel toward the individual, but they preserve society: keeping working cattle alive keeps agriculture going, and with it the sources of future subsistence and wealth. It may sound hard and sad, but it is so — in India a man is easier to replace than an ox.' (Return, East India. Madras and Orissa Famine, no. 4, p. 44.)
Compare this with the sentence from the Manava-Dharma-Sastra, Chapter X, §62:
'Giving up one's life without reward, to preserve a priest or a cow... can secure the bliss of these low-born tribes.'
Naturally it is impossible to deliver a five-year-old animal before five years are up. But within certain limits, it is possible to change the way animals are handled so that they are ready for their purpose sooner. Bakewell in particular achieved this. English sheep used to be — like French sheep still in 1855 — not ready for slaughter before their fourth or fifth year. Under Bakewell's system, even a one-year-old sheep can already be fattened, and in every case it is fully grown by the end of its second year. Through careful selective breeding, Bakewell, the tenant of Dishley Grange, reduced the sheep's bone skeleton to the minimum needed for it to survive. His sheep were called the New Leicesters.
'The breeder can now bring three sheep to market in the time it used to take him to finish one, and with a broader, rounder, larger development of the parts that give the most meat. Almost the whole of their weight is pure meat.' (Lavergne, The Rural Economy of England etc., 1855, p. 20.)
The methods that shorten the working period apply in very different degrees in different branches of industry, and they do not even out the differences in how long the various working periods are. To stick with our example: using new machine tools may shorten, in absolute terms, the working period needed to build a locomotive. But if improved processes in spinning increase the finished product delivered daily or weekly at a far faster rate, then the working period in machine manufacture has still grown longer relative to spinning — even though it shortened in itself.