Meanwhile, the wave of industrialisation that absorbed those who had been pushed out of agriculture came up against its own limits in the 1970s. Since then, the major capitalist countries have seen an unprecedented decline in their levels of industrial employment. Over the past three decades, manufacturing employment fell 50 percent as a percentage of total employment in these countries. Even newly “industrialising” countries like South Korea and Taiwan saw their relative levels of industrial employment decline in the past two decades. At the same time the numbers of both low-paid service-workers and slum-dwellers working in the informal sector have expanded as the only remaining options for those who have become superfluous to the needs of shrinking industries. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

The “general law of capitalist accumulation”, with its clear implications for the interpretation of Capital, has been overlooked in our own time because under the name of the “immiseration thesis” it was taken up and abandoned many times over in the course of the 20th century. It was held that Marx’s prediction of rising unemployment, and thus the increasing immiseration of the working population, has been contradicted by the history of capitalism: after Marx’s death, the industrial working class both grew in size and saw its living standards rise. Yet quite apart from the fact that these tendencies are often over-generalised, more recently their apparent reversal has made the immiseration thesis seem more plausible. The last 30 years have witnessed a global stagnation in the relative number of industrial workers. A low-wage service sector has made up the difference in the high GDP countries alongside an unparalleled explosion of slum-dwellers and informal workers in the low GDP countries. So is the immiseration thesis correct after all? That is the wrong question. The question is: under what conditions does it apply? Marx wrote about the growth of consolidated surplus populations in 1867. Yet the tendency he described — by which newer industries, because of their higher degree of automation, absorb proportionally less of the capital and labour thrown off by the mechanisation of older industries — did not play out as he had envisaged. As we can see from the graph on the facing page, Marx’s view was correct, in his own time, for the UK: the rising industries of the early second Industrial Revolution — such as chemicals, railways, telegraph etc. — were not able to compensate for declining employment in the industries of the first Industrial Revolution. The result was a steady fall in the rate of growth of manufacturing employment, which looked set to become an absolute decline sometime in the early 20th century. What Marx did not foresee, and what actually occurred in the 1890s, was the emergence of new industries that were simultaneously labour and capital absorbent, and which were able to put off the decline for more than half a century. The growth of these new industries, principally cars and consumer durables, depended on two 20th century developments: the increasing role of the state in economic management, and the transformation of consumer services into consumer goods. The emergent industries of which Marx wrote in the 1860s — gas-works, telegraphy, and railways, (we would add only electrification) — were already in his time beginning to be made available to consumers. Yet the consumer services generated from these technologies — initially reserved for the enjoyment of a wealthy elite — were secondary to the services they provided within the internal, planned economy of industrial firms. Railways emerged as a labour-saving innovation within mining, which was subsequently extended to other industries. It became a service offered to consumers only after extensive national-rail infrastructures had been developed by state-supported cartels. Even as costs fell and mechanised transportation via rail became available to more and more people, as a consumer service it preserved many of the features of its initial employment as a “process innovation” within industry. National railways, carrying passengers in addition to freight, absorbed large amounts of capital and labour in their construction but were subsequently relatively automated processes requiring less capital and labour for their upkeep. The advent of the automobile industry, subsidised by state funding of roads, eventually transformed the consumer service of mechanised transport into a good that could be purchased for individual consumption. This segmentation and replication of the product — the transformation of a labour-saving process innovation into a capital-and-labour absorbing “product innovation”— meant that this industry was able to absorb more capital and labour as its market expanded. A similar story can be told of the shift from telegraphy to telephones, and from electronic manufacture to consumer electronics. In each case, a collectively consumed service — often emerging from an intermediary service within industry — was transformed into a series of individually purchasable commodities, opening up new markets, which in turn became mass markets as costs fell and production increased. This provided the basis for the “mass consumerism” of the 20th century, for these new industries were able to simultaneously absorb large amounts of capital and labour, even as productivity increases reduced relative costs of production, such that more and more peasants became workers, and more and more workers were given stable employment. Yet, as the unprecedented state deficit-spending which supported this process indicates, there is no inherent tendency to capital that allows for the continual generation of product innovations to balance out its labour-saving process innovations. On the contrary, product innovations themselves often serve as process innovations, such that the solution only worsens the initial problem. When the car and consumer durables industries began to throw off capital and labour in the 1960s and 70s, new lines like microelectronics were not able to absorb the excess, even decades later. These innovations, like those of the 2nd industrial revolution described above, emerged from specific process innovations within industry and the military, and have only recently been transformed into a diversity of consumer products. The difficulty in this shift, from the perspective of generating new employment, is not merely the difficulty of policing a market in software — it is that new goods generated by microelectronics industries have absorbed tendentially diminished quantities of capital and labour. Indeed computers not only have rapidly decreasing labour requirements themselves (the microchips industry, restricted to only a few factories world-wide, is incredibly mechanised), they also tend to reduce labour requirements across all lines by rapidly increasing the level of automation. Thus rather than reviving a stagnant industrial sector and restoring expanded reproduction — in line with Schumpeter’s predictions — the rise of the computer industry has contributed to deindustrialisation and a diminished scale of accumulation — in line with Marx’s. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

Deindustrialisation began in the US, where the share of manufacturing employment started falling in the 1960s before dropping absolutely in the 80s, but this trend was soon generalized to most other high-GDP countries, and even to countries and regions that are seen as “industrializing”. [In no country (with the exception of the UK) did deindustrialisation involve a decline in real industrial output. In 1999, manufacturing still accounted for 46 percent of total US profits, but only 14 percent of the labour force.] The explosive growth of a low-wage service-sector partially offset the decline in manufacturing employment. However, services have proven incapable of replacing manufacturing as the basis of a new round of expanded reproduction. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

After 1973, the situation changed. Markets for manufactures were becoming saturated, and it was increasingly the case that a few countries could provide the manufactures for all of the world (one chinese firm currently supplies over half the world's microwaves). Thus the resulting crisis of the capital-labour relation, which is to say, a combined crisis of over-production and under-consumption, signalled by a global fall in the rate of profit and issuing in a multiplication of forms of unemployment and precarious employment. As the capital-labour accord snapped, having always been based on healthy rates of growth worldwide, wages stagnated. Capital in all countries became even more dependent on international trade, but from now on, capitals in some countries would expand only at the expense of those in others. Though they had not yet caught up to the high-GDP countries, the low-GDP countries took part in the same international crisis. Structural Adjustment Programs only accelerated their transition to a new, unstable international framework. Deindustrialisation, or at least the stagnation of industrial employment, set in almost universally among industrialising countries in the 1980s and 90s. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

It should thus be clear that de-industrialization is not caused by the industrialization of the “third world”. Most of the world’s industrial working-class now lives outside the “first world”, but so does most of the world’s population. The low-GDP countries have absolutely more workers in industry, but not relative to their populations. Relative industrial employment is falling even as agricultural employment collapses. Just as de-industrialisation in the high-GDP countries entails both the exit from manufacturing and the failure of services to take its place, so also the explosive growth of slums in the low-GDP countries entails both the exit from the countryside and the failure of industry to absorb the rural surplus. Whereas the World Bank used to suggest that the growing surplus populations throughout the world were a mere transitional element, they are now forced to admit the permanence of this condition. More than a billion people today eke out a terrible existence via an endless migration between urban and rural slums, searching for temporary and casual work wherever they can find it. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

In the 19th century when England was the workshop of the world, 95 percent of that world were peasants. Today, when the vast majority of the world’s population depends on global markets for their survival, the ability of one country to produce for all the others spells ruin, both for those who must be kept impoverished in order to maintain export prices, and for the vast multitudes whose labour is no longer necessary, but who, equally, can no longer rely on their own resources to survive. In this context the remainder of the world’s peasantry can no longer act as a weapon of modernisation, i.e. as a pool of both labour and consumer demand that can be drawn on in order to accelerate the pace of industrialisation. It becomes a pure surplus. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

A common objection to the account we have so far provided would be to point to China as an obvious exception to this picture of global stagnation, particularly in so far as it relates to otherwise global trends of deindustrialisation and under-employment. Of course, over these years China became a global industrial powerhouse, but it did so not through opening new markets or innovating new productive techniques, but rather by massively building out its manufacturing capacity at the expense of other countries. [41] Everyone assumes that this expansion must have brought about a historic increase in the size of the Chinese industrial working class, but that is flatly false. The latest statistics show that, on balance, China did not create any new jobs in manufacturing between 1993 and 2006, with the total number of such workers hovering around 110 million people. [41]: In the 1990s Japan devolved its more labour-intensive industries onto developing countries in Asia – first to the East Asian Tigers, then to the ASEAN countries and then to China. But the absorption of industries by China has undermined the hierarchy of production within the region. […] China has not only grown on the basis of labour-intensive manufacturing. Its low wages have helped it to compete across a spectrum of industries, from textiles and toys to cars and computers. The incorporation of existing labour-saving innovations into the firms of developing countries, including China, has meant that, even with growing geographic expansion, each set of industrialising countries has achieved lower heights of industrial employment (relative to total labour force). That is to say, not only has China lost manufacturing jobs in its older industries; the new industries have absorbed tendentially less labour relative to the growth of output. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

Yet if the “general law” [Marx’s “general law of capital accumulation” leading to increased surplus population] was suspended for much of the 20th century for the reasons we have outlined above, the current growing global masses of under-employed cannot be attributed to its reassertion, at least in any simple sense. For the trajectory of surplus capital distorts the trajectory of surplus labour described by Marx, and not only in the ways that we have already described. Most importantly, surplus capital built up in international money markets over the last 30 years has masked some of the tendencies to absolute immiseration, through the growing debt of working class households. This tendency, which has kept the bottom from falling out of global aggregate demand, has equally prevented any possibility of recovery, which would be achieved only through the “slaughtering of capital values” and “setting free of labour”. For while asset-price deflation may raise the possibility of a new investment boom, the devalorisation of labour-power will, in this context, only lead to increasing levels of consumer default and further financial breakdowns. Thus it is not only its capacity to generate employment, but the sustainability of the recovery itself which remains in question today. The coming decades may see a series of blowouts, if states fail to manage global deflationary pressures, or they may see a long and slow decline. While we are not catastrophists by inclination, we would warn against those who might forget that history sometimes rushes forward unpredictably. Regardless, the catastrophe for which we wait is not something of the future, but is merely the continuation of the present along its execrable trend. We have already seen decades of rising poverty and unemployment. Those who say of the still-industrialized countries that it is not so bad, that people will soldier on — in a phrase, that the proletariat has become indifferent to its misery — will have their hypothesis tested in the years to come, as levels of debt subside and household incomes continue their downward trend. In any case, for a huge chunk of the world’s population it has become impossible to deny the abundant evidence of the catastrophe. Any question of the absorption of this surplus humanity has been put to rest. It exists now only to be managed: segregated into prisons, marginalised in ghettos and camps, disciplined by the police, and annihilated by war. Endnotes, Misery and Debt, On the Logic and History of Surplus Population and Surplus Capital, Endnotes #2 (2010), https://endnotes.org.uk/articles/misery-and-debt

According to an old developmental narrative, agricultural employment would decline as agriculture became more productive, precipitating lots of potential new workers into towns, who would then be taken up by expanding industrial production. These developments would eventually bring every country into modernity. For orthodox Marxists, this would tendentially form a proletariat unified under the hegemony of its most “advanced” fractions in industry. But as the global peak of industrialisation recedes into history, it looks like something else is now happening. While agricultural employment has not halted its decline, those workers shed are less likely to join the ranks of the industrial working class than to enter a vast and heterogeneous service sector. At the world level, there are now twice as many workers in services as compared to industry: services account for 44 percent of global employment, while industry accounts for just 22 percent. […] Just as agriculture comes to employ a decreasing share of the workforce, since there are limits to growth in the demand for food, so too with manufacturing: there are supposedly limits to the demand for goods (apparently, there is, however, a limitless demand for services). The result, according to this perspective, is that over time, a rising demand for services will dynamically pull workers into the service sector, just as in an earlier phase workers were pulled into the industrial sector. In reality, the dynamic draw of manufacturing during industrialisation was unique to that sector. To manufacture something is to take a good — or to transform a service, such as dishwashing, into a good, such as a dishwasher — and to produce that good in a factory, according to ever-more efficient techniques. It is the resulting rise in the efficiency of production within the space of the factory that rapidly lowers costs of production in manufacturing lines. That leads, in turn, to a rapid fall in relative prices. Markets for manufactures expand, making possible a dramatic expansion of output. Concomitantly, huge masses of humanity are pulled into work in manufacturing lines. That is the key to the dynamic growth of manufacturing output and employment: the former is very rapid, and that is why, in spite of high rates of labour productivity growth, the latter expands, raising the manufacturing employment share. The same does not take place in the service sector. Services are precisely the sorts of activities that cannot be — or have not yet been — substituted by goods. In services, labour productivity tends to increase slowly if at all, and concomitantly, prices follow the same trajectory. In fact, as long as real wages are rising, the relative price of services will itself tend to increase. Since relative prices do not fall dramatically, there is no impetus for markets for services to expand rapidly. Hence, there is no dynamic tendency to dramatically expand output and thus to draw lots of labour into the sector; instead, employment in the service sector expands slowly. […] The shift from industrialisation to deindustrialisation is necessarily the shift from an economy that grows rapidly, with big booms and busts, to one that grows slowly, tending towards stagnation. In such a context, booms and busts are given merely by financial bubbles being blown up and deflated around the world by surplus capital. There is a corollary to this theory, which explains why a large portion of the surplus population ends up in the service sector, particularly in the low-wage, super-exploited section and in the informal, self-exploiting section. As service work tends to be labour-intensive, a large proportion of the final costs are made up of wages. Because real wages do not usually fall across the economy, it is difficult for service sector firms to lower their costs on a regular basis (general tendencies towards falling costs in industry and agriculture are due to increases in the efficient use of more expensive labour). This results in a relatively low level of output growth in services. But precisely for that reason, when workers are expelled from other sectors, it is possible to get much cheaper workers into services — as those discarded as surplus will usually have to accept a lower wage level. This lowers costs and allows for some expansion in demand for, and output of, services. In the service sector, there is greater room to expand the market by lowering wages. By contrast, in most manufacturing activities, wages make up only a small portion of the final cost of the product, so there is less room for manoeuvre. […] While it is difficult to identify a precise and determinate “logic” as to why some activities become really subsumed and others do not, the fact that certain activities require delicate work or direct human contact, and therefore must remain labour-intensive, is clearly key. There appears always to be a remainder of such activities, an assortment of differentiated tasks, mostly in services. Insofar as services remain services, they tend predominantly to be a source only of “absolute”, and not “relative” surplus value. This is simply another way of saying that there are limits to raising productivity. Consequently, economies that are “post-industrial” and concentrated around service work tend to be low-growth. In such conditions, it is imperative for capitalists to get as much out of their workers as possible, by increasing the duration or intensity of labour. To some extent, the prerequisite for the existence of many jobs becomes pressurised work conditions. If super-exploited sectors take up a growing share of the labour market, this also puts downward pressure on all wages, and increases insecurity, as workers lose bargaining power and bosses are emboldened to demand ever more flexibility. With this, the door is opened for a whole range of abuses to be unleashed upon the worker — sexual, emotional and psychological, as well as the stealing or retention of wages and chronic overworking. Certain positions, such as that of the low-wage service sector worker, thus appear as a kind of special category of surplus worker, akin to the informally self-employed in low-income countries (and in high income countries over the past decade or so). Low-wage service workers must become extreme self-exploiters, as well as being super-exploited, if they are to get work. Many of these jobs (deliveries, house-cleaning, supermarket baggers, and so on) can only exist because the wages of the people performing the service are a fraction of what those consuming the service are paid. Thus, the condition for finding a job in a growing service sector is often accepting a significantly lower than average wage. Endnotes, An Identical Abject-Subject?, Endnotes #4 (2015), https://endnotes.org.uk/articles/an-identical-abject-subject

Additionally, in the global economy — in which mobile flows of surplus capital discipline states — high-income states must do everything they can to prevent outright unemployment, and thus unemployment provisions, from growing too dramatically. Welfare expenditures, which are ultimately funded from tax receipts, must be kept to a minimum to avoid worrying bondholders and taxpayers. Current UK government policy, for example, is to try to eradicate, as far as possible, possibilities for unemployment as any kind of stable category, transforming welfare into workfare. As a result, in the high-income countries, many workers fall in and out of relative superfluity during their lifetime, due both to the increasing flexibilisation of the labour market and its destabilisation of categories of employment at a structural level, as well as the falling demand for labour. Endnotes, An Identical Abject-Subject?, Endnotes #4 (2015), https://endnotes.org.uk/articles/an-identical-abject-subject