I    Digital Labour: Class, Digital Divide and Gender

The Underpinnings of Class in the Digital Age: Living, Labour and Value

［导读］2 在互联网研究中，对于如何界定数字劳工的讨论愈发激烈，这些讨论强调“工作”和“玩耍”、“生产”和“消费”之间越来越模糊的界限——尤其是“无酬劳动”是否也创造剩余价值，是否存在剥削和异化？在作者看来，对这些问题的回答，需要我们重新审视劳动价值理论（the labour theory of value）。

劳动价值理论是马克思在对资本主义作为一种社会关系分析过程中使用的核心概念，其连接了从劳动中产生的剩余价值、劳动以及工人对生存的需要——离开了劳动，资本就无法积累、资本主义也就无法运作。但是，什么是劳动？什么样的劳动才能创造剩余价值？如何界定“必要劳动”？作者指出，仅仅从是否具有生产性的角度来界定劳动，往往遮蔽了这样的事实：存在着大量创造价值的“无酬劳动”，但它们并不改善劳动者的生计。“数字劳动”中普遍存在“无酬劳动”；同时，“数字劳工”在整个工人阶级中的比重不断增加，且他们中的一些收入微薄、穷困潦倒。因此，对于研究者而言，理解数字劳工在全球资本主义中的作用、数字劳工的组成，以及数字劳工所表达出的阶级忠诚变得格外重要。

从劳动中产生的剩余价值：在资本主义中，租金、贸易和通过商品的生产而创造的剩余价值是资本家获得利润的三种主要途径。在对新媒体企业的分析中，我们不难发现：（1）租金：来源于广告代理商支付给社交媒体或者搜索引擎公司大量的金钱用以将这些广告传递给用户；用户为信息付费，如使用在线数据库；在线服务平台向服务提供方、使用者，或者广告商收取费用，如eBay；（2）贸易：来源于在买卖过程中赚取差价，如线上对他人知识产权的占有和销售、使用他人免费翻译的网站内容来盈利等；（3）商品生产过程中创造的剩余价值：“数字劳工”无处不在，如果根据马克思在《政治经济学批判大纲》（Grundrisse）中强调的将“生产性的劳动”界定为使产品出现在市场上的一系列过程，那么现代企业中的诸多功能都具有“生产性”，包括市场营销、物流管理、流通、运输、客服、零售和批发、快递等等——也就是说，从出工厂大门到最后到消费者手中的整个价值循环链都具有生产性，而这些过程中的劳动都创造剩余价值。

劳动：“无酬劳动”是一个重要的概念。第一种“无酬劳动”指代那些独立于市场、在家庭中创造使用价值的劳动，如家务劳动等“体力劳动”和记住生日、安排聚会等“非体力劳动”，而后者的大部分存在于今天虚拟的网络世界中，如在线社交网络活动就属于此类范畴。第二种“无酬劳动”指代“消费工作”，消费者承担了市场中原本由“雇佣劳动”所从事的商品生产的分配过程中的部分劳动，这些劳动往往具有生产性。第三种“无酬劳动”指代“创意劳动”，包括写博客、在网上发布照片、音乐和视频等，这些不具有生产性的劳动往往产生“社会使用价值”。第四种“无酬劳动”指代日趋流行的“无酬实习”和“志愿者劳动”。

工人对生存的需要：这里涉及对于“必要劳动时间”的讨论。虽然工人通常以独立个体的形式进入劳动力市场，但他们往往和其他人一同生活，所以对工人必要劳动时间的计算并非易事。这种计算因养老金、福利待遇和减税政策，以及其他形式的补助变得更为困难。对于网络上“无酬的数字劳工”来说，他们维持生计的收入来源可能是多样的：来自父母的经济支持，来自养老金或社会福利，或来自日常工作的工资等。但是有一点十分明确：离开了经济来源（不管是上述哪一种形式），这种“无酬劳动”就无法继续。

在资本高度集中，信息传播技术如此发达，并影响劳动在时间和空间的分工，进而使个人的工作和生活紧紧粘连在一起的今天，我们是否都是没有区别的劳动力，为没有区别的资本创造没有区别的价值呢？答案必然是否定的。资本主义是一种社会关系，而这种社会关系取决于“工人的认同”（workers’ consent），因为工人在商品的生产过程中发挥着特殊的作用。这种认同，对于工人之间的共同利益诉求和工人之间的联合异常重要。仅仅因为“无酬劳动”对“雇佣劳动”产生的冲击，而将“无酬劳工”视为“害群之马”，这一想法过于简单——它忽视了催生“无酬劳动”的结构性要素，以及从更广义的现实来看，在“无酬劳动”和“雇佣劳动”中都存在剥削——不管它们以何种形式出现。因此从这个意义上来说，工人阶级是否具有挑战资本的能力成为了劳动价值理论隐含的内核。

最后，作者格外小心地勾勒了未来的阶级结构。为资本家企业所雇佣、从事商品生产的劳工仍然是发展最为迅速的群体。虽然他们中并非所有的人都拥有稳定的工作，但毋庸置疑，他们都从事“生产性劳动”，创造剩余价值。不同地区、不同职业和不同社会身份的工人可能认为他们之间没有任何的相似之处，反而存在着竞争性的利益关系，如何将他们联合起来，分享共同的阶级意识是一个值得探讨的问题；同时，管理的、职业的和技术的工人更多地自我认同与雇主成为联盟而非其他工人阶级，他们的困惑应该如何解决？在资本主义社会关系之外，还存在着许多“工人阶级”：如从事非常小规模商品生产、租赁和贸易者，公共服务机构的工作者，以及大量的“无酬劳工”。揭示全球价值链的复杂性以及将劳动过程的研究和它联系起来是一件劳心劳力的事情，但这对于我们更好地、一起行动起来改变现存的体系，以及开始思考哪些是可以替代的可能性却是亟需的。

As Marxism has segued in and out of vogue, there is hardly a Marxian concept that has not at some time been questioned as anachronistic, in the light of the transformations in economic and political conditions that have occurred over the last century and a half. The current renewal of interest in Marx’s ideas is no exception. It is indeed no easy task to apply theoretical concepts developed in the mid nineteenth century to a world where capitalism has penetrated every region and every aspect of life, where the pace of technological change is so rapid that labour processes are obsolescent within months of being introduced and where the division of labour is so intricate that no single worker has any chance of grasping it in its full complexity. Divisions between manual and non-manual work dissolve and are reconstituted, the boundaries between production, distribution and consumption melt away, and, whilst some paid work morphs into unpaid work, new jobs and new economic activities are generated from areas of life which were traditionally seen as beyond the scope of any market. In the suck and blow of commodification, the abstract becomes concrete and the concrete abstract, casting doubt on conceptual categories that formerly seemed self-evident. It may seem that we need new definitions of the most basic concepts used by Karl Marx, including ‘class’, ‘commodity’ and ‘labour’.

One current idea that has attracted considerable support, especially among the young, is the notion that the idea of a working class defined by its direct relationship to production is outmoded. Since all aspects of life, such arguments go, have been drawn into the scope of the capitalist cash nexus in some way, all those who are not actually part of the capitalist class must be regarded as part of an undifferentiated ‘multitude’. In Michael Hardt and Antonio Negri’s formulation, this ‘multitude’ takes the place of a working class, while according to Guy Standing, a ‘precariat’ constitutes a new class in and for itself alongside the traditional proletariat.^［1］ Standing does not attempt to locate this ‘precariat’ with any precision in relation to capitalist production processes. However, many of the followers of Hardt and Negri have engaged in elaborate attempts to do so in relation to the ‘multitude’. Two questions in particular, have puzzled them: what sorts of commodities are being produced by members of this multitude?^［2］ And how does the value produced by this labour accrue to capital?

In these debates, particular attention has been paid to the value created online by ‘virtual’ or ‘digital’ labour. In the field which is becoming known as ‘Internet studies’, there have recently been energetic discussions about ‘digital labour’ and how it should be conceptualised.^［3］ These debates have addressed the increasingly blurred boundaries between ‘work’ and ‘play’ (encapsulated in the term ‘playbour’^［4］) and between production and consumption (‘prosumption’^［5］ and ‘co-creation’^［6］); discussed the problematic category of ‘free labour’^［7］ and questioned whether such labour, paid or unpaid, can be regarded as producing surplus value and whether it is ‘exploitative’ or ‘alienated’. With the exception of Andrew Ross, few of these authors have drawn parallels with other forms of labour carried out offline. Yet, many of the questions they raise apply much more generally to labour under capitalism. These debates thus provide a useful starting point for investigating the labour theory of value itself, and how – or, some would wonder, even if – it can be applied in twenty-first century conditions.

This essay argues that it is still possible to apply Marx’s theory in current conditions, to define what is, or is not, a commodity, to identify the point of production of such commodities, whether material or immaterial, and to define the global working class in relation to these production processes. In order to do so, however, it is necessary to re-examine the labour theory of value in all its dimensions. I pay particular attention to ‘digital’ or ‘virtual’ labour not only because it is currently attracting so much attention, but also because online labour is particularly difficult to conceptualise. It is thus a fertile source of cases against which to test more general hypotheses. If a theory can apply here, then it should be more generally applicable. The aim of doing this is to enable a mapping of the working class across the whole economy by applying the theory more broadly (as Marx did). This is an important task, in my view, because without a clear sense of which workers are engaged directly in the antagonistic relation to capital that characterises commodity production, and without identifying where that point of production is located, it is impossible to identify strategies that will enable labour to confront capital where it is possible to exercise some power to shape the future in its own interests.

LABOUR AND CAPITALISM

The labour theory of value is the knot at the heart of Marx’s conceptualisation of capitalism as a social relationship. It integrally links three things: workers’ need for subsistence, their labour, and the surplus value expropriated from the results of that labour, without which capital cannot be accumulated or capitalism perpetuated. The expropriation of labour is the act of violence at the heart of this relationship. It is the worker’s labour time which constitutes the bone which is fought over in this relationship, so an understanding of how and under what circumstances this expropriation takes place is critical to an understanding both of capitalism as a system and of which workers can be said to belong to the working class. The knot cannot be undone: each rope is essential to holding the system together. Nevertheless it seems necessary to examine it, strand by strand, so that we can grasp how it is put together, what tightens it, and what enables new threads to be drawn in or existing ones more elaborately entangled.

In its basic form, the argument is remarkably simple: the worker, obliged to do so in order to subsist, works a given number of hours for the capitalist, producing a certain value as a result. Some of this value is essential to cover the cost of subsistence, and the hours worked to produce this value (‘necessary labour time’) are (usually) reimbursed. The remainder (‘surplus value’) is appropriated by the capitalist to distribute as profit and invest in new means of production. On close examination, however, just about every element of this simple story turns out to be open to question. What, exactly, is ‘labour’? And, more particularly, what labour is productive of surplus value? How is ‘subsistence’ to be defined? Does it include only what the individual worker needs to keep going, or does it also include what is required for the sustenance of his or her entire household? If we cannot define subsistence precisely, how can we possibly calculate necessary labour time? And, just because all value within capitalism ultimately derives from the results of human labour applied to the earth’s raw materials, does this mean that all value that accrues to individual capitalists is necessarily surplus value?

The current debates around ‘digital labour’ skim past some of these questions and oversimplify others. This essay will not attempt to rewrite Marx’s entire theory, as that would be both hopelessly over-ambitious and misguided. Rather, it will take some of the questions raised in these debates about digital labour as starting points for examining the factors that will have to be taken into account in any modern elaboration of Marx’s theory, an elaboration that will, in my view, be an essential precondition for understanding the new class formations that are emerging in the twenty-first century in all their complex and contradictory dimensions. I will do this by attempting to unravel the three strands – living (or subsistence), labour, and value – in order to categorise their separate components. I will do this in reverse order, reflecting the priorities of current debates in this field. These concepts are all well used and difficult to re-employ without bringing along a large freight of associated meanings, both intended and unintended. So it is perhaps useful to begin with two explanatory notes.

The first concerns the terminology. In advanced capitalist societies, not only is the division of labour extremely complex but so, too, is the distribution of wealth. Workers’ subsistence is achieved not only as a direct result of waged labour but also via redistribution through the financial system (in the form of credit, private insurance and pension schemes, etc.) and through the state (in a monetary form through tax and social security systems, and in kind by means of state-provided services). In such a context, the direct connection between labour and value can be obscured. It is common for analysts to follow Marx in classifying labour as ‘productive’ or ‘unproductive’. The approach I adopt in this essay draws on insights from feminism and makes a slightly different distinction. This is a distinction between labour which is productive for capitalism as a whole (which can be termed ‘reproductive’) and labour which is directly productive for individual capitalists (which, for lack of a better term, I have named ‘directly productive’). I draw a further distinction between work that is paid and work that is unpaid. I argue that (dependent though it is on other forms of labour for its reproduction) the quintessential form of labour that characterises capitalism is labour that both produces value for capital and produces the income that is necessary for the worker’s survival. This is work whose very performance contains within itself the contestation of labour time between worker and capitalist at whose heart lies the wrench of expropriation, the experience of which Marx described as ‘alienation’ (a term which has, unfortunately, become so contaminated with other meanings that it can no longer be used with the precision with which Marx employed it). And this is therefore the work that lies at the centre of the accumulation process. The workplace is not, of course, the only place that labour confronts capital. But, because capital cannot be accumulated without workers’ consent, it is the site at which labour has the greatest potential power to wrest concessions from capital (without resorting to bloodshed).

Table 1 Labour: A Schematic Typology

The term ‘waged labour’ encompasses work which Marx would have designated as both productive and unproductive. It also excludes various forms of labour (piecework, freelance work, etc.) paid in non-wage forms, which contribute directly both to capital accumulation and workers’ subsistence. Defining labour only in terms of whether it is productive or not, in Marx’s sense, ignores the reality that (as will be discussed below) there is a considerable amount of unpaid labour which produces value directly for capital without contributing to the worker’s subsistence. Conversely, of course, there is paid labour which contributes to subsistence without creating value directly for capital. After spending some time considering a range of alternatives (including ‘contested productive labour’, ‘alienated productive labour’, ‘directly productive labour’ and ‘productive waged labour’), I have, for the purposes of this essay, decided to use a shorthand term to distinguish this form of labour from other forms of productive and waged labour. Drawing on the metaphor I have used to describe the labour theory of value, I therefore refer to it below as labour which is ‘inside the knot’ (Quadrant C in Table 1).

Labour ‘inside the knot’, in this definition, is labour carried out directly for a capitalist employer by a worker who is dependent on this labour for subsistence and is therefore a front-line adversary in the struggle between capital and labour over how much labour time should be exchanged for how much money. This may seem like a somewhat narrow definition. It is indeed the sort of definition that was much criticised in the 1960s and 1970s for excluding large groups of workers who often saw themselves as part of the working class, including public sector workers and some service workers, whose relationship to production was indirect. In using it here, I am not arguing that such workers are not productive. On the contrary, many of the tasks they perform are essential for the reproduction of labour. However, these workers’ exposure to the coercive logic of capitalism may be somewhat mitigated, either because they are working under older forms of employment (for instance as domestic servants or as petty commodity producers), or because they are employed by the state to provide as yet uncommodified services.

These forms of labour still exist, of course, but, as I have argued elsewhere, in the current wave of commodification, these forms of work are diminishing and the workers who carried them out are rapidly being drawn ‘inside the knot’.^［8］ In other words, the commodification of public services has produced a major shift of labour from Quadrant A in the diagram above to Quadrant C.

This is not the only movement that is occurring.^［9］ The more general commodification of consumer goods and services has also involved large shifts from Quadrant B to Quadrant D, transforming the nature of some unpaid work from the direct production of use values for household members to the purchasing of commodities in the market, involving a direct relationship with capitalist production and distribution activities. In a further twist, there has also been a shift of labour from Quadrant C to Quadrant D as capitalist production and distribution companies have reduced their labour costs, increasing the exploitation of their paid workers by externalising more and more tasks onto consumers who have to carry them out as unpaid self-service activities. In a parallel process, austerity measures are also leading to a shift of activities from Quadrant A to Quadrant B, which in turn puts more pressure on the further shift from B to C. Thus, whilst labour ‘inside the knot’ constitutes a sub-set of all labour, it is a sub-set that is rapidly expanding to become the overwhelming majority of paid labour.

My second cautionary note concerns the danger of extrapolation from a typology of labour to a typology of workers, and hence to a class typology. Whilst part of my aim is to classify different forms of labour in their relation both to capital accumulation and workers’ subsistence, I do not intend in so doing to produce a classification of workers that can be read off in any simple manner from this typology. Most workers engage in several different kinds of labour, paid and unpaid, both simultaneously and over the course of their lives, crossing these simple categories. Even more importantly, most workers live in households where different kinds of labour are carried out by different household members, some of whom, at any given time, may be unemployed. Whether or not members of such households perceive themselves, or can be perceived by others, as belonging to the working class is a large question. In my conclusion, I will attempt to sketch out some of the ways that it might be possible to map working classes in the twenty-first century drawing on this analysis. But the analysis of labour constitutes only a small first step in that larger process and this exercise is necessarily speculative.

‘DIGITAL LABOUR’ IN A MATERIAL WORLD

Before embarking on this analysis, it is worth noting that digital labour cannot be regarded as a discrete form of labour, separated hermetically from the rest of the economy. As I argued in these pages, the existence of a separately visible sphere of non-manual labour is not evidence of a new ‘knowledge-based’, ‘immaterial’, or ‘weightless’ realm of economic activity.^［10］ It is simply an expression of the growing complexity of the division of labour, with a fragmentation of activities into separate tasks, both ‘mental’ and ‘manual’, increasingly capable of being dispersed geographically and contractually to different workers who may be barely aware of each other’s existence. This is a continuing process, with each task subject to further divisions between more creative and/or controlling functions on the one hand, and more routine, repetitive ones on the other.

Furthermore, whilst there has clearly been an enormous expansion in non-manual work, both routine and deskilled and otherwise, it remains a minority of all labour. As I have argued previously, the growing visibility of apparently dematerialised labour, dependent on information and communications technologies to observers in developed economies has sometimes served to obscure the reality that this ‘virtual’ activity is dependent on a highly material basis of physical infrastructure and manufactured commodities, most of which are produced out of their sight, in the mines of Africa and Latin America, the sweatshops of China and other places in the developing world. Without the generation of power, cables, satellites, computers, switches, mobile phones and thousands of other material products, the extraction of the raw materials that make up these commodities, the launching of satellites into space to carry their signals, the construction of the buildings in which they are designed and assembled and from which they are marketed, and the manufacture and operation of the vehicles in which they are distributed, the Internet could not be accessed by anyone.

Whilst 20 per cent of the world’s 100 largest transnational corporations are now service companies, it should not be forgotten that 80 per cent are not.^［11］ And, according to UNCTAD, in 2012 it was manufacturing companies that were expanding their foreign investment the fastest.^［12］ The physical production of material commodities is still capitalism’s preferred method for generating profit; it is still growing; and it seems likely to continue to employ the largest proportion of the world’s workforce. There is, moreover, a continuum between tasks that mainly involve the exercise of physical strength or dexterity and those that involve mental agility, engagement or concentration. There are few jobs that do not require workers to bring their own knowledge, judgement and intelligence to the task in hand, and even fewer that do not involve some physical activity, even if this just entails speaking, listening, watching a screen or tapping keys.

That said, a large and growing proportion of the workforce is involved in performing ‘digital labour’ whose products are intangible, much of it low-paid and menial. And many members of this workforce are descended from or cohabiting with workers who would by any definition be assigned to the working class. It is therefore important to understand what role their labour plays in global capitalism, what the composition of this workforce is, how it is changing and what class allegiances these workers might express.

VALUE

Put simply, it could be said that there are three main ways that enterprises generate profit under capitalism, the first two of which also existed under other systems. These are rent, trade and the generation of surplus value through commodity production. Because it is the paradigmatic form of value generation under capitalism, it is commodity production that receives the most attention from Marxian analysts. If value is observably being generated from some activity, the tendency is to search for the commodity at its source. If a commodity cannot easily be identified, or if it does not appear to be produced by extracting surplus value from paid workers, then it is sometimes concluded that this means that Marx’s labour theory of value does not apply and is either outmoded or in need of adaptation. However, before leaping to the conclusion that entirely new theories are needed to explain online activities, it is worth analysing them in relation to traditional forms of value generation to see whether they fit these categories.

Rent

The ways in which commercially-mediated online activities seem to encroach indiscriminately on work, leisure, consumption and personal relationships draws attention to the extent to which capitalist relations have spread into all aspects of life, or as Marx put it, ‘in the modern world, personal relations flow purely out of relations of production and exchange’, encouraging broad-ranging speculation about how the monetisation of online exchanges can be understood and theorised.^［13］ The starting point for many of the current discussions about the value that is generated on the Internet is the indisputable reality that online companies like Google and Facebook are hugely profitable. If they are making profits, it is then argued, this must be because some commodity is being produced, which in turn begs the question of what precisely these commodities might be and whose labour is producing them. In the case of Google and Facebook, the main source of income is revenues from advertising, which can be targeted with great precision as a result of the ever-more sophisticated analysis of data generated by users. Here, Dallas Smythe’s concept of the ‘audience commodity’^［14］ has been seized on by a number of commentators, including Christian Fuchs.^［15］ Originally developed as part of a Marxian attempt to understand the economics of advertising in commercial radio and TV, this concept portrays the media audience as the commodity which is sold to advertisers to generate revenue: ‘Because audience power is produced, sold, purchased and consumed, it commands a price and is a commodity.’^［16］ Fuchs applies this logic to the Internet: ‘the productive labour time that is exploited by capital ... involves ... all of the time that is spent online by the users’. He goes on to say that ‘the rate of exploitation converges towards infinity if workers are unpaid. They are infinitely exploited’. Other contributors to the digital labour debate suggest that ‘reputation’^［17］ or even life itself (produced by ‘bio-labour’)^［18］ have become commodities.

Whilst Smythe’s concept has undoubtedly opened up useful insights into the nature of the mass media, it has also led to much confusion. The underlying assumption among Smythe’s followers seems to be that the term ‘commodity’ can be used to refer to anything that can be bought and sold. There is a certain circular logic operating here. Since Marx declares that ‘commodities are nothing but crystallised labour’ and that ‘a good only has a value because labour is objectified or materialised in it’, then it must follow, according to this logic, that anything described as a commodity must be the result of productive labour.^［19］ But how useful is such a broad conception of the term?

It seems to me that in order to understand the distinctive nature of the commodity form under capitalism, a somewhat different definition needs to be used. I have defined commodities elsewhere as ‘standardised products or services for sale in a market whose sale will generate profits that increase in proportion to the scale of production’ (all else being equal).^［20］ This definition singles out capitalist commodities as fundamentally different from those produced under other systems. A traditional carpenter making chairs and selling them directly to the public makes more or less the same profit on each chair. The capitalist who opens a factory and employs workers to mass-produce chairs has to make an investment in machinery, buildings and so on and will not make a profit on the first chair, but the more chairs that are produced in that factory, the greater will be the profit on any given one. This gives the chairs produced in the factory a fundamentally different character from those produced individually by a single artisan in relation to their value. There are a number of services, including intangible ones (such as insurance policies or software programs) which have the same character as commodities. It is the social relations under which they are produced (the coerced labour of waged workers, under the control of the capitalist) that gives them this character.^［21］ Such a definition of commodity inverts the logic of Smythe’s followers. It takes as its starting point the nature of the capitalist-labour relationship rather than the fact that something is being sold.

If they do not derive from the sale of commodities, how can we understand the profits made by online social networking or search engine companies? There is an alternative explanation, and it is one that has long antecedents in the offline world: they derive from rent. A simple historical example of a similar way of generating income could be provided by a street market where the rent charged for a stall-space is higher in areas where the most customers (or the richest customers) will pass by. Bricks-and-mortar examples can be found in New York’s Fifth Avenue, London’s Oxford Street, or any other street with a large and lucrative footfall: the more well-trafficked the site, the higher the rent. For well over a century, properties that border busy highways have been able to make money by renting space for billboards. Don’t these online companies simply follow the same model, albeit with sites that are virtual rather than paved and rather more sophisticated means of identifying the most lucrative customers and gaining intelligence about their desires? The value that accrues to the social networking and search engine sites does indeed ultimately derive from surplus value produced by labour. But this is the labour of the workers who produced the commodities that are advertised on these sites, not the labour of the people who use the sites.^［22］

Some participants in the digital labour debate, such as Adam Arvidsson and Eleanor Colleoni, dispute Fuchs’s notion that social media users are producing surplus value.^［23］ They, too, argue that the value that is generated can be more properly regarded as rent. However, they use the term ‘rent’ to refer to the value that accrues to financial investors in these companies. But in this respect, they do not say what it is that makes online companies different from any other companies that are quoted on stock exchanges and attract financial investments. In attempting to classify what, precisely, it is that generates the value that attracts such investors, they develop an explanation whereby ‘social media platforms like Facebook function as channels by means of which affective investments on the part of the multitude can be translated into objectified forms of abstract affect that support financial valuations’. They further argue that such companies gain their share of ‘socially produced surplus value’ through ‘the ability to attract affective investments ... from the multitude or the global public’.^［24］ This somewhat convoluted model sidesteps the rather more prosaic question of who is paying whom for what in order to generate the return on investment for the shareholders. It can, in my opinion, be rather simply answered by saying that it is the advertisers (producers of commodities for sale) who are paying the social media or search engine companies for the opportunity to advertise to their users. This is not to deny that social media sites do not incidentally also facilitate other forms of labour which could be regarded as more directly productive. These will be discussed below.

There are, of course, a number of ways that value is generated online other than through the use of search engine or social media sites. There are many other online activities that rely on rent for the generation of income. These include a variety of other sites that rely on advertising revenue, but also sites that charge rents to their users for access to information (such as online databases), sites from which copyright music or videos can be downloaded (such as iTunes), companies that sell software licenses online, and online games for which subscriptions have to be bought (on the same principle as software licenses).

Other sites can be regarded as essentially online equivalents of offline businesses that generate income from rent. These include online marketplaces (such as eBay), dating sites (such as eHarmony or Match.com), online employment agencies which match freelance workers to employers (such as oDesk or Elance), price comparison sites, online travel booking or accommodation finding sites (such as Opodo or Expedia) or various forms of peer-to-peer services allowing people to find bed-and-breakfast accommodation (such as Airbnb) or car shares (such as Lyft). The connection with offline businesses is often evident here. For example, one of the largest of the online peer-to-peer car rental services, RelayRides, was launched with funding from GM Ventures (the investment arm of General Motors) in 2011 and has now been acquired by Zipcar, which in turn was acquired by Avis in January 2013.^［25］

Whatever the specific mix of sources of revenue, most of the profit of such enterprises comes from some combination of charging usage or commission fees to service providers and/or service users and/or advertisers – in other words, rent. It is interesting to note that some of these sites seem to be enabling the development of new forms of petty commodity production and rentier activity or allowing older forms to survive offline. Etsy, for example, makes it possible for individuals to sell craft products in the online equivalent of a crafts market. Airbnb lets them make an income from renting out rooms in their homes for bed and breakfast (taking a percentage of the cost). Peer-to-peer car rental services enable people to provide taxi services or charge others to borrow their cars.

Trade

Trade involves acquiring something at one price (including stealing it) and selling it a higher price, making a profit in the process. Some forms of stealing, such as the appropriation of other people’s intellectual property, may take place online. These include the reselling of captured images or music or the plagiarism of text for sale or some more elaborate forms of theft which are currently emerging, such as the exploitation of the unpaid labour of language learners to obtain free translation of web content by the website Duolingo.com, or reCAPTCHA’ s reuse of users’ attempts to decode distorted images of letters and numbers (required for security on many sites ‘to ensure that you are not a robot’) that cannot be recognised by automatic optical scanning systems.^［26］

However, there are also a very large number of companies that sell online (Amazon being probably the most famous) in a manner that replicates offline commercial trade. Indeed, many established merchants now buy and sell both online and offline. Although there may be some blurring of traditional boundaries between the distribution activities of manufacturers, wholesalers and retailers, and some labour processes may be rather different, there is nothing mysterious about how value is generated by such companies. The scale of many of these companies, and the fact that they have had to put extensive infrastructure in place for processing payments internationally, has meant that some of them have been able to diversify into rental activities which have in turn created the basis for new forms of commodity production, discussed in the next section.

Commodity Production

This brings us to the final category: value which is generated from the production of commodities. Here, the analyst seeking to isolate the role of digital labour in value creation is faced with considerable challenges. The spread of computing across most sectors of the economy, combined with the near-universal use of telecommunications, means that there are few economic activities that do not involve some element of digital labour, whether they take place in farms, factories, warehouses, offices, shops, homes or on moving vehicles. Furthermore, these activities are linked with each other in complex chains which cross the boundaries between firms, sectors, regions and countries. Tracing the connection of any given activity back to its origins, or forward to the final commodity to whose production it has contributed, is no easy task. Nevertheless, it is by no means impossible. One useful approach here is to analyse economic activities in functional terms.^［27］

The functions of research and development and design, for instance, clearly make direct inputs to the development of new commodities (or the adaptation of older ones). Much of the labour involved in these activities nowadays comes into the category of digital labour in that it involves computer-based tools and/or is delivered in digital form to the workers who will take it forward to the production stage. The same goes for activities whose purpose is to develop content for books, films, CDs or other cultural products. Here, some activities may be more directly ‘digital’ than others: actors or musicians, for instance, may be performing in a manner that is ‘live’, but if the end result is going to be incorporated into a reproducible commodity, then their functional relation to capital is the same as that of fellow workers sitting at screens or mixing desks.^［28］ Digital labour is also involved in a variety of ways in production processes, whether this involves the operation of digitally-controlled tools, the maintenance of software, the generation of immaterial products or the supervision of other workers engaged in these processes.

When it comes to ‘service’ activities, it is useful – though increasingly difficult – to make a general distinction between those that contribute directly to production (such as cleaning the factory floor or servicing the machines); those that contribute to the maintenance or management of the workforce (such as processing payroll data or staff recruitment or training); those that contribute to the more general management of the enterprise (including financial management); those that are involved in activities connected with purchasing, sales and marketing; and those that are involved in distribution. All of these categories include activities that are carried out online and/or using a combination of information and communications technologies. They are, however, becoming more and more difficult to tell apart, for several interconnected reasons.

The first of these is the increasingly generic nature of many labour processes. Workers inputting numerical data on a keyboard, for instance, may be doing it for a bank, a government department or a manufacturing company, for purposes entirely unknown to them. Call centre operators may be using standard scripts to deal with sales, customer services, debt collection, government enquiries, fund raising or a variety of other functions, cutting across any neat classification scheme that would allow them to be sorted into different categories by function. Software engineers may be working on the development of new products, or the maintenance of existing ones.

Closely linked with this form of standardisation is the growing propensity of such activities to be outsourced, often to companies that bundle together a number of different functions for different clients into clusters of activities carried out in shared service centres. The possibility for these and other services to be carried out online has further blurred the distinction between services provided to businesses and those provided directly to final customers. If everyone can order goods online, to be delivered to the door from a central warehouse, then the distinction between ‘wholesale’ and ‘retail’ becomes an artificial one. Similarly, there is a growing range of standardised immaterial products, ranging from software licenses to bank accounts to insurance policies that can be sold as readily to individuals as to companies.

The existence of online platforms through which labour can be coordinated has led to the development of an extreme form of subdivision of tasks, sometimes known as ‘micro-labour’, ‘crowd work’^［29］ or ‘crowd-sourcing’.^［30］ These include ‘pay-per-click’ work whereby workers are paid by commercial companies to ‘like’ their Facebook posts or blog entries, or platforms like Amazon’s Mechanical Turk, whose users are paid a few cents to perform a variety of very small tasks, so fragmented that they are very unlikely to understand what relation any given task has to the final commodity to which it contributes.

If such activities, however dispersed, are carried out by paid workers, in the employ of enterprises set up to make a profit, then they can unproblematically be assigned to the category of work that directly produces surplus value for capital – labour ‘inside the knot’. However, as the borderlines between production, distribution and consumption become increasingly fuzzy and the same activity can be carried out interchangeably by paid and unpaid workers, this simple position needs some modification. Marx was somewhat ambivalent about distribution labour, regarding transport workers as productive but not retail workers. However, at one point in the Grundrisse, he asserted that the whole process of bringing a product to market should be regarded as productive labour: ‘Economically considered, the spatial condition, the bringing of the product to the market, belongs to the production process itself. The product is really finished only when it is on the market’.^［31］ Following this logic, a wide range of functions to be found in a modern corporation can be assigned to this directly productive category, including marketing, logistics management, distribution, transport, customer service, retail and wholesale sales (whether online or offline) and delivery – in short, the whole value chain from factory gate (or software development site) to the final consumer should be regarded as productive labour. But what happens when the customer’s unpaid labour is substituted for that of the productive waged worker? What if, for instance, you go and fetch a purchase yourself from the store or warehouse? Or design your own product, selecting a unique combination of standard features from a website? And what, exactly, is the difference between booking your own holiday via a website, keying in your own data, and doing so over the phone to a (paid) call centre operator who keys it in on your behalf? In the latter case, the labour falls comfortably into what is traditionally regarded as the ‘productive’ category. But what about the former? In my view, all these activities should be regarded as productive. However, only those carried out by paid workers fall ‘inside the knot’ whereby their relationship to capital is both direct and, actually or potentially, contested.

LABOUR

Any attempt to categorise different forms of labour has to begin by confronting the extraordinarily difficult question of what labour actually is. The word itself covers a vast spectrum of meanings from the physical exertion of giving birth at one extreme to formal participation in employment, or the political representation of people who do so, at the other. If we take it to refer to activities which are actually or potentially reimbursed by wages in a ‘labour market’, then we have to include a large range of activities which most people carry out without pay, including sex, caring for children, cooking, cleaning, gardening, singing, making people laugh and holding forth on topics that interest us.

If we apply a more subjective filter and try to exclude activities that are carried out for pleasure, then we are confronted with the awkward reality that the same activity may be experienced as a chore or a joy under differing circumstances and, furthermore, that some activities, paid or unpaid, may be both onerous and enjoyable simultaneously. The baby, for instance, may give you a beaming smile whilst its smelly diaper is being changed; a truck driver’s long lonely journey may suddenly bestow a heart-stoppingly beautiful glimpse of landscape; hard physical work in harsh surroundings may engender a camaraderie between workers that leaves a warm glow long after the muscle ache has subsided; solving a tricky problem may release a sudden gush of satisfaction, even if the problem is not one’s own.

Another dimension that might help to distinguish between ‘labour’ and ‘pleasure’ is whether or not the activity is carried out voluntarily or by coercion, under the direction of another person or organisation. Here again what seems a simple distinction becomes remarkably difficult to apply in practice. One difficulty results from the historically determined ways in which such things as gender roles, concepts of duty or even caste-based divisions of labour are internalised, rendering patterns of power and coercion invisible to all parties and, indeed, giving many acts of service the subjective quality of freely offered gifts of love even when objective analysis might suggest that they involve the exploitation of one person’s labour by another. Coercion may also be exercised in more indirect ways. An addicted gambler, for instance, may perceive his or her compulsion as internally generated, not recognising the societal pressures that impel it. The same could, perhaps, be said of many of the online activities that people spend so much time on, including online gaming and interacting with others on social media sites. It is perhaps some inkling of these social pressures that leads so many commentators in the digital media debates to insist that these unpaid activities are a form of ‘free’ labour.^［32］

Unpaid labour is not, of course, a new phenomenon. It has however received only rather fitful attention from Marxian scholars, except as a kind of vestigial repository of pre-capitalist social relations from which waged labour later emerged. Apart from debates about slavery among historians, most of the attention paid to unpaid labour until recently was in the context of what could loosely be called ‘reproductive labour’, in particular in feminist debates during the 1970s. In these discussions, the main question raised was whether unpaid domestic labour or ‘housework’ could be regarded as producing surplus value because without it capitalism could not exist. The reproduction of the workforce depended crucially, it was argued, on unpaid labour in the home, not only for bringing up the next generation of workers, but also to provide the nutrition, cleaning and bodily maintenance services that allow the current workforce to perform effectively in the labour market. In 1976, Batya Weinbaum and Amy Bridges published a ground-breaking article in which they argued that, under monopoly capital conditions, much of this labour did not only involve producing services in the home but also consuming commodities produced in the market.^［33］ The concept of ‘consumption work’, in which unpaid labour is substituted for what was formerly the paid labour of distribution workers, is one that I developed further in the late 1970s and, I argue here, is relevant for understanding some of the new forms of unpaid labour that take place both on and offline.^［34］ Drawing on some of this work, I propose here a somewhat rough-and-ready typology of unpaid labour in the hope that it can provide a starting point for a categorisation that will bring some clarity to these debates.

The first category is the labour that is carried out independently of the market to produce use values in the home, the category of labour located in Quadrant B in the diagram above. It is ‘unproductive’ in the sense that it produces no direct value for capital in the form of surplus value from somebody’s direct labour, but ‘reproductive’ in the sense that it is necessary for the reproduction of the workforce. It includes many of the tasks traditionally carried out in subsistence agriculture and housework. If someone is employed to do this kind of work by the direct user of the service (e.g. a domestic servant, nanny, cleaner or gardener) that worker is, in Marx’s opinion, an unproductive worker, although if they are employed via a capitalist intermediary (e.g. a commercial childcare, cleaning or gardening company), then they move into the category of productive worker (in terms of the diagram above, from Quadrant A to Quadrant C).^［35］ However, we are concerned here with unpaid labour. To the extent that maintaining the emotional health of a family and sustaining the social networks in which it is embedded is a necessary part of ensuring the survival of a household, then a range of non-physical activities can be included in this category, including such seemingly trivial tasks as remembering birthdays, writing letters of condolence or arranging social get-togethers which help to produce and reproduce the solidaristic bonds that may be necessary for survival in times of crisis. It also includes acquiring the skills and habits that enable someone to be employable. Even courtship can be regarded as a necessary prelude to this family maintenance project. Many of these activities are carried out online these days; thus, at least a part of online social networking activity could be assigned to this category (represented by Quadrant B). Whether or not the person carrying out this labour is exposed to advertising in the process of carrying it out is as incidental to the productivity of the labour as whether or not they might pass a billboard on the way to visit a sick grandmother or be exposed to cinema commercials whilst on a date.

The second category of unpaid labour is what I have referred to above as ‘consumption work’ (Quadrant D). This involves the consumer taking on tasks in the market that were previously carried out by paid workers as part of the distribution processes of commodity production. Since these tasks are necessary to the distribution of these commodities, and increase the profits of the commodity-producing companies by eliminating forms of labour that were formerly paid for, there are strong arguments for categorising this kind of work as ‘productive’, even when it is unpaid. However, because it does not generate income directly for the worker, it has to be treated differently from paid labour in relation to its contribution to subsistence, a topic to which I will return below. It is in other words ‘outside the knot’. As already noted, increasing amounts of consumption work are carried out online, with the Internet having opened up a range of new ways of externalising labour over distance.^［36］

The third category involves creative work. Here, Marx made his position clear:

Milton, for example ... was an unproductive worker. In contrast to this, the writer who delivers hackwork for his publisher is a productive worker. Milton produced Paradise Lost in the way that a silkworm produces silk, as the expression of his own nature. Later on he sold the product for £5, and to that extent became a dealer in a commodity. ... A singer who sings like a bird is an unproductive worker. If she sells her singing for money, she is to that extent a wage labourer or a commodity dealer. But the same singer, when engaged by an entrepreneur who has her sing in order to make money, is a productive worker, for she directly produces capital.^［37］

According to this conception, to the extent that it is carried out for the purposes of self-expression, unpaid artistic work, such as blogging or posting one’s photographs, music or videos on the Internet comes straightforwardly into Marx’s category of ‘unproductive’ labour (which I would prefer to regard as unpaid reproductive labour, producing social use values). If the product of this labour is subsequently sold, or stolen, to become the basis of a commodity, then this does not change that status. It is only if the worker is hired to do the work for a wage that it becomes productive labour in Marx’s sense of the term (i.e. it moves from Quadrant B to Quadrant C). As Ross has pointed out, many artistic workers may oscillate between these forms: ‘Creatives have been facing this kind of choice since the eighteenth century when the onset of commercial culture markets offered them the choice of eking out a living with the scribblers on Pope’s Grub Street or of building a name-recognition relationship with the fickle public’.^［38］ The fact that the same person does both kinds of work does not, however, invalidate the distinction between them. Creative work thus has to be seen as straddling a number of different positions in the labour market, including self-employment, paid employment and petty commodity production, leading, very often, to contradictory identities for creative workers.^［39］

The same logic applies even in the much-discussed case of the ‘free labour’ that built the Internet, much of which was designed by idealistic software developers who donated their labour for nothing in the belief that they were creating a common benefit for humankind (in other words, they were producing social use value without pay, placing them in Quadrant B). As Marx said, ‘labour with the same content can be both productive and unproductive’.^［40］ In this case, it seems that although the results of their labour were appropriated by capital to incorporate into new commodities, their original unpaid labour cannot be regarded as productive in the sense of producing surplus value for capital under coercive conditions (i.e. it is not ‘inside the knot’). Rather, the value that was produced from it should more properly be put into the category of trade, which, as I noted above, also includes theft.

A fourth – but overlapping – form of unpaid labour, which is increasingly discussed, is the widespread use of unpaid internship or ‘voluntary’ labour.^［41］ This, too, seems to have precedents in various forms of apprenticeship labour, such as the production of ‘show pieces’ to impress potential employers. Situated ambiguously between education and self-promotion, it is undoubtedly used in highly exploitative ways by employers as a direct substitute for paid work. Sometimes, direct coercion is involved to oblige the worker to undertake unpaid ‘work placements’, for instance by state job search agencies which threaten the withdrawal of unemployment benefit from those who refuse to take them. Nevertheless, like the unpaid consumption labour already discussed, whilst clearly contributing value to commodity production, this form of labour plays no part in generating present income for the worker and must therefore be regarded as ‘outside the knot’, even if it is producing value indirectly for the unpaid worker in the form of ‘employability’. It is clear that in order to make sense of the relationship of unpaid labour to capital, we have to take into account the third rope in the knot that constitutes the labour theory of value: the worker’s subsistence, or ‘living’.

LIVING

The question of how the worker pays for the cost of subsistence is surprisingly absent from most of the debates about ‘free’ digital labour. Perhaps because they themselves often have secure academic jobs, the majority of the authors who have contributed to these discussions fail to ask how those dedicated workers who built the Internet with their free labour actually made a living. Nor, among those who advocate a ‘Creative Commons’ on the Internet, to which all authors are supposed to donate their work for free, is it ever made clear how these authors are supposed to pay their rent and provide for their families?

Yet, the labour theory of value cannot be operationalised without this information. In order to know how much surplus value is generated, and how, from any given unit of labour, we need to know the cost of that worker’s reproduction, and how much of his or her working time is the ‘necessary labour time’ required to sustain life. Only then can we see how much of the remainder is left over to be appropriated as surplus value and begin to formulate demands for its redistribution. This is not, of course, a mechanical calculation. It is perfectly possible for workers to be employed below the cost of subsistence. What does the employer care if they die, if there are plenty more where they came from? Equally, it is possible for well-organised groups of workers with scarce skills to punch above their weight and claim back from capital a higher wage than that required for bare survival – even one which allows them to employ other workers as servants. Nevertheless, capitalism as a system, in Marx’s model, requires a working class that is compelled to sell its labour in order to survive, just as it requires capitalists who are able to employ that labour to produce commodities whose collective value on the market exceeds the total wages of the workforce required to produce them. And it is the direct experience of being obliged to contest ownership of their labour time with the employer that produces the alienation likely to lead to class consciousness. The question of ‘necessary labour time’ cannot therefore be ducked.

But even in Marx, this is quite a problematic concept. One reason for this is that although workers normally enter the labour market as separate individuals, their subsistence takes place in households where several people may co-habit.^［42］ Because these households vary considerably in size and composition and in the number of members who engage in paid work, the same wage may have to stretch to cover the subsistence of varying numbers of people. Marx and Engels discuss the ‘natural’ (sic) division of labour in the family, which they regard as a form of ‘latent slavery’ that can even be regarded as the origin of all property.^［43］ From their premise that women and children are the property of the male head of household, it is possible for them to conclude that, when women and children enter the workforce, ‘Formerly, the sale and purchase of labour-power was a relation between free persons; now, minors or children are bought; the worker now sells wife and child – he becomes a slave-dealer’.^［44］

In the twenty-first century, when women make up nearly half the workforce in most developed countries and only a minority are economically inactive, such an explanation will not suffice. Every worker who enters employment needs to be separately accounted for as an individual with his or her own cost of subsistence to be raised. The fact that people co-habit with other workers can, however, mean that this ‘necessary labour time’ should be regarded as producing a fraction, rather than the whole, of any individual’s cost of subsistence. Or, in other words, that the concept of a ‘family wage’ is redundant in most circumstances. A number of other factors have also intervened to make it difficult to identify a simple correspondence between what a person earns and what it costs them to survive, at least in situations where that person is co-habiting with, or responsible for, economic dependents. These complicating factors include societal transfers in the form of pensions, welfare benefits or tax credits, intergenerational transfers within families, remittances from migrants working abroad and other forms of subsidy for some (or drains on the resources of others). Tax credits, the favoured neoliberal model of social transfer, have played a particularly pernicious role in disguising not only the extent to which many jobs pay wages that are well below subsistence level, but also in concealing from public awareness the reality that a large and growing proportion of social benefit payments go not to unemployed ‘scroungers’ but to workers in employment.^［45］ Such transfers could thus be seen as having played an important role in blunting class-consciousness and diverting workers’ energy away from direct conflict with their employers.

Despite very real difficulties of precise calculation, it is possible to analyse the income of any given individual in any given household and produce some estimate of how this is generated. In the case of ‘free labour’ on the Internet, it is likely that a number of different income sources may be involved. Some of this labour may be contributed by people who are economically dependent on their parents, some by people drawing pensions or receiving some other form of welfare benefit, some by people with regular salaries from jobs that leave them with enough leisure time to blog, surf the net or write Wikipedia entries; some might be done by people (such as freelance journalists, consultants, or academics) whose jobs require them to engage in self-promotion. And others might be being supported from rents, gambling, the proceeds of trade, crime or other activities. What is clear, however, is that these unpaid contributors could not engage in this unpaid activity without some kind of subsidy from somewhere. Otherwise, how would they eat? Arguments that postulates the production of surplus value at a societal level from their labour seem untenable. Such arguments could also be seen as playing a similar role to societal financial transfers in diverting workers’ attention away from confronting the employers directly expropriating their labour towards expressing their anger and sense of exploitation towards abstract targets (such as ‘globalisation’). In failing to organise at the point of production, they give away their strongest weapon: the power to withdraw their labour.

CLASS CONFIGURATION IN THE TWENTY-FIRST CENTURY

We live in a society where capital is highly concentrated, with most commodity production carried out by companies whose fates are largely shaped by financial investors. The commodities that they produce, whether material or immaterial, are made available to us in a global marketplace, delivered through complex value chains in whose operation our own unpaid labour as consumers is increasingly implicated. Information and communications technologies have so affected the spatial and temporal division of labour that for many of us, the boundaries between work and private life are inextricable muddled and few relationships are unmediated by them. In such a situation, are not the kinds of distinctions made in this essay not ridiculously nit-picking? Should we not just accept that all of us are, in some way or another, part of a huge undifferentiated workforce, producing undifferentiated value for an undifferentiated capital?

I argue that we should not. Capitalism is a social relationship in which workers play specific roles in relation to the production of specific commodities. This relationship relies crucially on workers’ consent. If we cannot understand this relationship in its specificity, we cannot identify the critical points in the processes of production and distribution where workers’ agency can be implemented to some effect. And if we cannot identify these, workers cannot understand their powers to consent to, or refuse, the specific deal that is on offer to them. This prevents them from actively renegotiating the terms of the deal – their only option for improving their situation. Neither, without this knowledge, can we see which groups of workers have interests in common, how these common interests might become mutually visible or how their labour may be interconnected?

Each of the different forms of unpaid labour described above has an impact on paid labour, opening up the potential for tensions and fissures within the working class. Interns, working for nothing to make themselves employable, erode the bargaining position of paid workers in the same roles. Carrying out unpaid consumption work affects service workers by reducing overall employment levels and intensifying work through the introduction of new forms of standardisation and Taylorisation, leading to deteriorating working conditions. Writing Wikipedia entries, blogging or posting video clips or photographs online without payment threatens the livelihoods of journalists, researchers or other creative workers who lack a subsidy from an academic salary or other source and rely on their creative work to provide an income. In many cases, the same people occupy several of these paid and unpaid roles in different capacities. Even more commonly, different members of the same household may be doing so. To regard unpaid workers as scabs who are undermining paid workers is of course much too simplistic, ignoring the imperatives that propel these behaviours and the broader reality that exploitation takes place in all of them, albeit in different forms. But an analysis that equates a common exploitation with an identical role in the generation of surplus value, and collapses all these separate positions into a common collective identity as a ‘multitude’ makes it impossible to identify the point of production: the point where workers have the power to challenge capital; the centre of the knot.

Starting from a detailed analysis of how value chains are structured, it is possible to begin to sketch out the lineaments of the class configuration that might confront us in coming years. However, this exercise has to be embarked on with extreme caution because, as noted earlier, many of us are engaged simultaneously or consecutively in a number of different forms of labour, with different relations to capital, or live in households where multiple forms of labour take place.

Leaving aside the rural populations that still subsist, at least in part, from their own direct labour on the land, in this emerging labour landscape, the largest, and by far the most rapidly-growing group is that of workers ‘inside the knot’: those who are employed by capitalist enterprises producing commodities, both material and immaterial. Many of these have been sucked into directly capitalist labour relations comparatively recently, coming to this work as migrants from the countryside or from other countries, being transferred from public sector employment, or recruited from a previous existence in petty commodity production. Not all of these workers have the status of permanent employees, with many paid by piece rates or employed on a casual or temporary basis. They are, nevertheless, productive workers, directly producing surplus value. However, the ways in which their labour processes connect with each other are not obvious.

A product like a smart phone contains within it the results of the labour of miners, assembly-line workers, chemical workers, designers, engineers, call centre workers, invoice clerks, cleaners and many more. Scattered in different countries, with different occupational and social identities, these workers may not perceive themselves as having anything whatsoever in common. Indeed, they may believe their interests to be directly opposed to each other. If and when they organise themselves, this might be on the basis of skill, occupation or the company they work for, but it might also be on the basis of a shared regional, linguistic or cultural identity, a shared political history or a response to a shared form of discrimination. What forms of solidarity or shared consciousness might emerge from these forms of organisation is an open question.

Another open question is the extent to which managerial, professional and technical workers within these value chains will identify with other workers rather than aligning themselves with the employer. These are volatile groups, made up of people who, in the accelerating speed of technological change and economic restructuring, find many of their labour processes undergoing standardisation and deskilling even whilst new opportunities to become managers are emerging. On the one hand, their employers want to nurture them as sources of innovation; on the other, they want to cheapen their labour and drive up their productivity. Caught between these two contradictory imperatives, these intermediate workers may be put into a position where they have to decide whether to continue to internalise management priorities and take the pain, to leave, to look for individualistic solutions or to throw in their lot with other workers and resist.

Alongside, and overlapping with, this explosively growing body of workers ‘inside the knot’ of capitalism are other groups less directly involved in capitalist social relations. These include people patching together a living out of petty commodity production, small-scale rent or trade, a class which Marx assumed would die out but which appears to have been given a new lease of life by the Internet, although it is doubtful whether such sources of income can ever supply a sustainable livelihood for more than a minority of the population. In many cases, it seems likely that this way of earning a living, often cobbled together from several different kinds of economic activity, is a transitional one, adopted by people who have been displaced from the formal labour market, or have not yet managed to enter it. It is not new. Working-class biographies have always thrown up many examples of people making ends meet by taking in lodgers, child-minding, pet-breeding or making small items for sale. But it cannot be taken for granted that all such people will necessarily identify their interests with those of workers ‘inside the knot’.

Groups that are ‘outside the knot’ also include people involved in paid reproduction work: public sector workers working in the increasingly rare fields of service provision that remain uncommodified; domestic servants; and other forms of service work that are not directly involved in the market (such as work in the voluntary sector). Their work is, of course, necessary for the reproduction of capitalism, but it is ‘outside the knot’ according to my earlier definition. Again, these groups cover a diverse range of social identities and may not perceive themselves as having interests in common, either with each other or with workers ‘inside the knot’.

Added to these are large numbers of people who are not paid workers but who nevertheless also produce value, either in the form of reproduction, such as unpaid childcare or housework, or (externalised) production, in the form of consumption work. Many of these will be women, and their unpaid status may place them into relations of dependency on paid workers or on the state. History has given us many examples of reproduction workers throwing in their lot with the production workers to whom their lives are linked (for instance in the organisation of miners’ wives in the UK coalminers’ strike in the 1980s) and of consumption workers acting in solidarity with production workers, for instance in the consumer-based Clean Clothes Campaign which organises petitions and boycotts to improve working conditions for garment workers.^［46］

These are broad categories and a much more detailed mapping of the composition of these groups and their interrelationships with each other will be necessary to predict the class configuration that will confront us globally in the twenty-first century. Tedious though it may be to unravel the complexities of global value chains and position our labour processes in relation to them, this seems to be an absolutely necessary task if we are to learn how this system might be changed, act collectively to change it, and start to imagine what alternatives might be possible.

Notes

［1］ M. Hardt and A. Negri, Multitude: War and Democracy in the Age of Empire, New York: Penguin, 2004; G. Standing, Precariat: The New Dangerous Class, London and New York: Bloomsbury, 2011.

［2］ Hardt and Negri, Multitude; T. Terranova, ‘Free Labor: Producing Culture for the Digital Economy’, Social Text, 18(2), 2000, pp. 33-58.

［3］ See for instance, M. Andrejevic, ‘Exploiting YouTube: Contradictions of User-Generated Labor’, in P. Snickers and P. Vonderau, eds., The YouTube Reader, Stockholm: National Library of Sweden, 2009; A. Arvidsson and E. Colleoni, ‘Value in Informational Capitalism and on the Internet’, The Information Society, 28(3), 2012, pp. 135-50; J. Banks and S. Humphreys, ‘The Labor of User Co-Creators’, Convergence, 14(4), 2008, pp. 401-18; C. Fuchs, ‘Labor in Informational Capitalism and on the Internet’, The Information Society, 26(3), 2010, pp. 179-96; C. Fuchs, ‘With or Without Marx? With or Without Capitalism? A Rejoinder to Adam Arvidsson and Eleanor Colleoni’, Triple C, 10(2), 2012, pp. 633-45; D. Hesmondhalgh, ‘User-Generated Content, Free Labour and the Cultural Industries’, Ephemera, 10(3/4), 2011, pp. 267-84; A. Ross, ‘On the Digital Labour Question’, in T. Scholz, ed., The Internet as Playground and Factory, New York: Routledge, 2012; and Terranova, ‘Free Labor,’ in Scholz, Internet as Playground and Factory.

［4］ J. Kücklich, ‘Precarious Playbour: Modders and the Digital Games Industry’, The Fibreculture Journal, Issue 5, 2005.

［5］ Alvin Toffler coined this term in his 1980 book The Third Wave, published by Bantam Books. It has since been taken up by a number of other writers working in a Marxist framework, including Christian Fuchs and Ed Comer.

［6］ Banks and Humphreys, ‘The Labour of User Co-creators’, using a term derived from C.K. Prahalad and V. Ramaswamy, ‘Co-Opting Customer Competence’, Harvard Business Review, (January/February), 2000.

［7］ A term coined by Tiziana Terranova in her influential article, ‘Free Labor’.

［8］ U. Huws, ‘Crisis as Capitalist Opportunity: The New Accumulation through Public Service Commodification’, Socialist Register 2012, Pontypool: Merlin, 2011, pp. 64-84.

［9］ U. Huws, ‘Domestic Technology: Liberator or Enslaver?’, in U. Huws, The Making of a Cybertariat: Virtual Work in a Real World, New York: Monthly Review Press, 2003, pp. 35-41.

［10］ U. Huws, ‘Material World: The Myth of the Weightless Economy’, Socialist Register, 1999, pp. 29-56.

［11］ UNCTAD, World Investment Report, Geneva, 2008.

［12］ According to UNCTAD, 60 per cent of manufacturing TNCs were planning to increase their FDI in the next year, compared with 45 per cent of firms in the primary sector and 43 per cent of those in services. See World Investment Report, 2012, p. 19.

［13］ Karl Marx, Grundrisse, ‘Chapter on Money’ Part II, available at http://www. marxists.org.

［14］ D. W. Smythe, ‘Communications: Blindspot of Western Marxism’, Canadian Journal of Political and Social Theory, 1(3), 1977, pp. 1-27.

［15］ C. Fuchs, ‘Dallas Smythe Today – The Audience Commodity, the Digital Labour Debate, Marxist Political Economy and Critical Theory. Prolegomena to a Digital Labour Theory of Value’, Triple C, 10(2), 2012, pp. 692-740.

［16］ D. W. Smythe, ‘On the Audience Commodity and its Work’, in M.G. Duncan and D.M. Kellner, eds., Media and Cultural Studies, Malden, MA: Blackwell, 1981, p. 233.

［17］ A. Hearn, ‘Structuring Feeling: Web 2.0, Online Ranking and Rating, and the Digital “Reputation” Economy’, Ephemera, 10(3/4), 2010, pp. 421-38.

［18］ C. Morini and A. Fumagalli, ‘Life Put to Work: Towards a Life Theory of Value’, Ephemera, 10(3/4), 2010, pp. 234-52.

［19］ K. Marx, Capital, Chapter 1, available at http://www.marxists.org.

［20］ Huws, Making of a Cybertariat, p. 17.

［21］ This point is made a little differently in a discussion of the distinction between productive and unproductive labour by Marx in Capital, Chapter 4.

［22］ Except in some special circumstances, such as when workers are paid to go on Facebook and click ‘like’ on commercial websites in the ‘pay per click’ model. But here they are not employed by Facebook but by companies linked to these commercial websites which have some commodity to sell, so they should more accurately be regarded as belonging to the value chain of these commodity-producing companies.

［23］ Arvidsson and Colleoni, ‘Value in Informational Capitalism’.

［24］ Arvidsson and Colleoni, unpublished manuscript.

［25］ ‘All Eyes on the Sharing Economy’, The Economist, 9 March 2013.

［26］ I am indebted to Kaire Holts for drawing my attention to this explanation of the business model of reCAPTCHA by its originator, who also founded Duolingo, available at http://www.willhambly.com. See also the related video, available at http://www.inmyinnovation.com.

［27］ I have discussed the concept of the ‘business function’ and its relation to Marxist analysis in several publications. See for instance, U. Huws, ‘The Restructuring of Global Value Chains and the Creation of a Cybertariat’, in Christopher May, ed., Global Corporate Power: (Re)integrating Companies into International Political Economy (International Political Economy Yearbook Volume 15), Lynne Rienner Publishers, 2006, pp. 65-84; and U. Huws, ‘The Emergence of EMERGENCE: The Challenge of Designing Research on the New International Division of Labour’, Work Organisation, Labour and Globalisation, 1(2), 2007, pp. 20-35.

［28］ I have analysed the relationship of creative labour to capital elsewhere. See, for instance, U. Huws, ‘Expression and Expropriation: The Dialectics of Autonomy and Control in Creative Labour’, Ephemera, Volume 10(3/4), 2010.

［29］ Kittur et al., ‘The Future of Crowd Work’, 2013, available at http://hci. stanford.edu.

［30］ K. Holts, ‘Towards a Taxonomy of Virtual Work’, Hertfordshire Business School Working Paper, 2013.

［31］ Marx, Grundrisse, Notebook V. It should be noted that this interpretation of this passage is disputed. Marx is often considered to be making a special exception of transport workers (perhaps because they were a group with strong potential trade union organisation – a potential that was more-than-realised in the twentieth century when transport workers played a key role in industrial action). It is my view that his argument applies equally to other forms of labour involved in getting products to market, many of which were inconceivable at the time when he was writing.

［32］ Terranova, ‘Free Labor’.

［33］ B. Weinbaum and A. Bridges, ‘The Other Side of the Paycheck: Monopoly Capital and the Structure of Consumption’, Monthly Review, 28(3), 1976.

［34］ See for instance, Huws, ‘Domestic Technology’.

［35］ See K. Marx, Economic Manuscripts, Chapter 4, available at http://www. marxists.org.

［36］ I use the term ‘externalising’ here to refer to the ways in which employers increase the productivity of paid staff by transferring some or all of their unpaid tasks to unpaid consumers in the form of self-service, whether through the operation of machines such as ATMs or self-service supermarket or online activities such as booking tickets, filling in tax returns or ordering goods.

［37］ Marx, Economic Manuscripts, Chapter 2.

［38］ A. Ross, ‘In Search of the Lost Paycheck’, in Scholz, Internet as Playground and Factory, p. 15.

［39］ I have anatomised these in greater detail in Huws, ‘Expression and Expropriation’, pp. 504-21.

［40］ Marx, Economic Manuscripts, ‘Productive and Unproductive Labour’.

［41］ See, for instance, R. Perlin, Intern Nation: How to Earn Nothing and Learn Little in the Brave New Economy, London: Verso, 2011.

［42］ I have written more extensively about this in U. Huws, ‘The Reproduction of Difference: Gender and the Global Division of Labour’, Work Organisation, Labour and Globalisation, 6(1), 2012, pp. 1-10.

［43］ K. Marx, ‘Division of Labour and Forms of Property – Tribal, Ancient, Feudal’, Part 1, A, The German Ideology, 1845, available at http://www.marxists.org.

［44］ F. Engels, On Marx’s Capital, Moscow: Progress Publishers, 1956 [1877], p. 89.

［45］ For more on this, see my blog post on ‘Hunger in a Supermarketocracy’, available at http://ursulahuws.wordpress.com. In the UK, according to HM Revenue and Customs, ‘the numbers of families without children receiving Working Tax Credits-only has risen over time, almost doubling from 235,000 in April 2004 to around 455,000 in April 2009 and now at just over 580,000 in April 2012’ and ‘the numbers of families benefiting from the childcare element has consistently risen over time, from 318,000 in April 2004 to around 493,000 in April 2011’. By this date, tax credits (paid to workers in employment) already accounted for 27 per cent of all benefit spending – by far the largest single component. By comparison, Job-seekers Allowance (paid to the unemployed) accounted for only 4 per cent. In the USA, similarly, many large companies rely on government-provided benefits, such as food stamps and Medicaid, to subsidise below-subsistence wages. For instance, Wal-Mart employees are estimated to receive $2.66 billion in government assistance every year, or about $420,000 per store. See HM Revenue and Customs, Child and Working Tax Credits Statistics, Office of National Statistics, 2012; and P. Ryan, ‘Walmart: America’s Real “Welfare Queen”’, Daily Kos, 2012, available at http://www. dailykos.com.

［46］ See http://www.cleanclothes.org.