Under capitalism, the only thing worse than being exploited is not being exploited. Since the beginnings of the wage-labour economy, wageless life has been a calamity for those dispossessed of land, tools and means of subsistence. Expelled from work, the wageless also became invisible to science: political economy, as Marx noted in the earliest formulations of his critique of the discipline, ‘does not recognize the unemployed worker’: ‘The rascal, swindler, beggar, the unemployed, the starving, wretched and criminal workingman—these are figures who do not exist for political economy but only for other eyes, those of the doctor, the judge, the gravedigger, and bum-bailiff, etc; such figures are spectres outside its domain.’  Karl Marx and Frederick Engels, Collected Works, New York 1975ff (hereafter mecw), Volume 3, p. 284. This essay was originally written as part of the Yale Working Group on Globalization and Culture. I would like to thank the other members for their suggestions and criticisms, and Achille Mbembe for his response to an earlier text at the Wits Institute for Social and Economic Research, University of the Witwatersrand, South Africa, 22 February 2006. These days, Marxism—more often seen as an example of political economy than as its critique—and other labour-based analyses face the same objection. Understandings built upon wage labour cannot, we are told, account for the reality lived by the most numerous and wretched of the world’s population: those without wages, those indeed without even the hope of wages. Bare life, wasted life, disposable life, precarious life, superfluous life: these are among the terms used to describe the inhabitants of a planet of slums. It is not the child in the sweatshop that is our most characteristic figure, but the child in the streets, alternately predator and prey.
In face of this situation, neither of the classic Marxist designations for the wageless—the reserve army of labour or the lumpenproletariat—seems adequate. For some, only a theory of citizenship and exclusion from it, or rights and their absence, can capture this reality: to speak of labour is to speak of the already enfranchised. Others have turned to a biopolitics or necropolitics of bare existence. Neither of these alternatives is persuasive. Though the struggle for social and cultural inclusion as well as political citizenship is vital in a world of sans-papiers, too often the theoretical battles over citizenship and human rights remain caught in fantasies of sovereignty. On the other hand, the rhetoric of life and death sometimes has a false immediacy, seeing a state of exception or emergency in what is unfortunately a state of normality. To speak repeatedly of bare life and superfluous life can lead us to imagine that there really are disposable people, not simply that they are disposable in the eyes of state and market.
Moreover, bare life is not without practical activity. A critical account of living and making a living under capitalist imperatives must, I believe, begin not from the accumulation of capital but from its other side, the accumulation of labour. They are, dialectically, the same: as Marx put it, ‘Accumulation of capital is therefore multiplication of the proletariat.’  Karl Marx, Capital, vol. 1, Harmondsworth 1976, p. 764. But to approach the issue from the point of view of capital is, as Hegel and Marx might put it, one-sided. A number of contemporary critics of political economy have noted this imbalance. Michael Lebowitz argues that Marx’s book on capital was meant to be accompanied by one on wage labour; in The Limits to Capital, David Harvey describes ‘Marx’s rather surprising failure to undertake any systematic study of the processes governing the production and reproduction of labour power itself’ as ‘one of the most serious of all the gaps in Marx’s own theory’.  Michael Lebowitz, Beyond Capital: Marx’s Political Economy of the Working Class, New York 2003; David Harvey, The Limits to Capital, Chicago 1982, p. 163.
In what follows, I will suggest that we need a similar reversal regarding wage labour. Wageless life has almost always been seen as a situation of lack, the space of exclusion: the unemployed, the informal. I do not claim to solve this semantic problem: my own working vocabulary—the wageless—is a parallel construction. However, I want to insist that we decentre wage labour in our conception of life under capitalism. The fetishism of the wage may well be the source of capitalist ideologies of freedom and equality, but the employment contract is not the founding moment. For capitalism begins not with the offer of work, but with the imperative to earn a living. Dispossession and expropriation, followed by the enforcement of money taxes and rent: such is the idyll of ‘free labour’. In those rare moments of modern emancipation, the freed people—from slavery, serfdom and other forms of coerced labour—have never chosen to be wage labourers. There may be a ‘propensity to truck, barter and exchange one thing for another’, as Adam Smith put it, but there is clearly no propensity to get a job.
Rather than seeing the bread-winning factory worker as the productive base on which a reproductive superstructure is erected, imagine the dispossessed proletarian household as a wageless base of subsistence labour—the ‘women’s work’ of cooking, cleaning and caring—which supports a superstructure of migrant wage seekers who are ambassadors, or perhaps hostages, to the wage economy. These migrations may be short in distance and in interval—the daily streetcars or buses from tenement to factory, apartment block to office, that will come to be called ‘commuting’—or they may be extended to the yearly proletarian globe-hopping of seasonal workers by steamship, railroad and automobile, as well as the radical separation of airborne migration linked by years of remittances and phone calls. Unemployment precedes employment, and the informal economy precedes the formal, both historically and conceptually. We must insist that ‘proletarian’ is not a synonym for ‘wage labourer’ but for dispossession, expropriation and radical dependence on the market. You don’t need a job to be a proletarian: wageless life, not wage labour, is the starting point in understanding the free market.
Emergence of unemployment
In this essay, I want to explore the lineaments of wageless life over the past century by offering a genealogy of two key representations which not only name and seek to regulate it, but draw a dramatic line between its conceptions in capitalism’s imperial metropoles and its periphery: the figures of unemployment and the informal sector. The former was the founding trope of twentieth-century social democracy, invented in the midst of the great economic crises which gripped the industrial capitalisms of the North Atlantic and reverberated across their colonial territories. It displaced a host of earlier conceptions of the poor, the idle and the dangerous, and became a central part of state and popular discourse through the next century, particularly during the moments of mass unemployment: the Great Depression of the 1930s and the Great Recession of the 1970s. On the other hand, the term ‘informal sector’ was coined in the early 1970s to reckon with the mass of wageless life in the newly independent Third World, which seemed to escape the categories of employment and unemployment alike. It too displaced earlier conceptions—perhaps most notably that of the lumpenproletariat figured by Frantz Fanon’s Wretched of the Earth—and continues to be part of official and unofficial discourse.
An older institutional history might say that the welfare state was created in response to unemployment: the spectre of the unemployed returns with every depression and recession, as illustrators and photographers try to represent the absence of work in icons ranging from Victorian cartoonist Tom Merry’s ‘The Meeting of the Unemployed’ to Dorothea Lange’s ‘White Angel Breadline’. But a more recent biopolitical history suggests that the emerging social state invented unemployment in the process of normalizing and regulating the market in labour.  Biopolitical readings of unemployment are in a way the product of the intellectual upheaval triggered by the third wave of mass unemployment; two landmark texts both date from 1986: Robert Salais, L’invention du chômage: histoire et transformations d’une catégorie en France des années 1890 aux années 1980, Paris 1986, and Alexander Keyssar, Out of Work: The First Century of Unemployment in Massachusetts, Cambridge 1986. See also Christian Topalov, Naissance du chômeur: 1880–1910, Paris 1994. A more recent study that draws on this work is William Walters, Unemployment and Government: Genealogies of the Social, Cambridge 2000. The word itself emerged just when the phenomenon became the object of state knowledge production in the long economic downturn of the 1880s and 1890s. The term was first used in English in 1887, when the chief of Massachusetts’s Bureau of Labor Statistics, Carroll D. Wright, attempted to count the unemployed, triggering a statistical practice that became central to the modern state, and by the following decade was in common use. The earliest theoretical treatment, the 1895 article ‘The Meaning and Measure of “Unemployment”’ by the liberal economist J. A. Hobson (best known for his influential analysis of imperialism), set the agenda for a century of debate: how does one define and measure it? The French word for unemployed, chômeur, dates from the same era, and the German equivalent, Arbeitslosigkeit, was rarely used before the 1890s. Indeed, as John Garraty, the author of the still-standard Unemployment in History, points out, Marx himself did not use the expression. In Capital, as well as in the passage from the 1844 manuscripts quoted earlier, Marx writes of die Unbeschäftigen—the not-busy, the unoccupied in one English translation—rather than die Arbeitslosen, the contemporary term for the unemployed.  John Garraty, Unemployment in History, New York 1978, pp. 109, 4; J. A. Hobson, ‘The Meaning and Measure of “Unemployment”’, Contemporary Review 67, March 1895.
The modern notion of unemployment depended on the normalization of employment, the intricate process by which participation in labour markets is made ordinary. As employers make rules, workers insist on customary practices, while courts, legislatures and factory inspectors set standards. ‘The creation of a normal working day [ein Normalarbeitstag]’, Marx argued, ‘is, therefore, the product of a protracted civil war, more or less dissembled, between the capitalist class and the working class.’ Indeed, he insisted that: ‘in place of the pompous catalogue of the “inalienable rights of man” comes the modest Magna Carta of a legally limited working day.’  Marx, Capital, pp. 303, 307.
Normalizing employment made possible the normalization of unemployment in at least three senses. First, to be unemployed was to lose one’s usual employment—and indeed the first forms of unemployment protection came from trade unions that tried to maintain the going wage rate by offering members out-of-work benefits. In his discussion of unemployment and government William Walters proposes that ‘the status of “out-of-work” was actually invented by trade unionism’. The second form of normalization arose as the wageless began to meet and march as the unemployed. The canonical starting point is the famous February 1886 London riot. A Tory-led Fair Trade League had called a meeting of the unemployed in Trafalgar Square that attracted 20,000 jobless building- and dock-workers; when the Social Democratic Federation led part of the crowd down Pall Mall, windows were smashed, shops were looted and London, according to The Times, was in a panic. Similar demonstrations continued and grew through 1887, culminating that November in Bloody Sunday, the protest against coercion in Ireland, in which police attacked demonstrators and three were killed.  Walters, Unemployment, p. 18. See also Gareth Stedman Jones, Outcast London, New York 1984, pp. 291–6. In his letters Engels was very critical of the sdf’s ‘bunkum about social revolution’. His characterization of their procession as merely comprised of ‘idlers, police spies and rogues’ is one of the classic passages on the lumpenproletariat; mecw 47, pp. 407, 408.
Finally, unemployment was integrated into the work of turn-of-the-century theorists such as Hobson and William Beveridge, who argued that it was not a matter of individual depravity or idleness but was a normal and unavoidable aspect of industrial society. ‘Personal causes, no doubt, explain in a large measure who are the individuals that shall represent the 10 per cent “unemployed”’, Hobson argued, ‘but they are in no true sense even contributory causes of “unemployment”’. These analyses built on the earlier notion that capitalism created a reserve army of labour, a concept often taken to be distinctively Marxist since it appears in Capital’s discussion of capitalism’s relative surplus population. However, Marx was simply adopting the rhetoric of the British labour movement. Radicals, particularly the Chartists and Fourierist associationists, imagined the new factory workers as great industrial armies, and this common trope led the Chartist leader Bronterre O’Brien to write of a reserve army of labour in the Northern Star in 1839. The young Engels picked up that image in The Condition of the Working Class in England in 1844, and Marx would invoke the metaphor occasionally, distinguishing between the active and reserve armies of the working class. By the end of the nineteenth century, it was part of the commonsense understanding of unemployment: by 1911, even the Massachusetts Bureau of Statistics of Labor could conclude that, ‘however prosperous conditions may be, there is always a “reserve army” of the unemployed’.  Hobson quoted in Walters, Unemployment, p. 32. See also Stedman Jones, Languages of Class, Cambridge 1983, p. 159. Massachusetts Bureau quoted in Keyssar, Out of Work, p. 72.
Risk and relief
This normalization of unemployment was the basis for the great social-democratic techniques that sought to contain the spectre of wageless life. The first moment was characterized by an initial conceptualization of unemployment as an insurable risk, an accident like illness, fire, theft or death. This was the basis of Britain’s National Insurance Act of 1911, the first government programme of its kind. Imitating Bismarck’s regime of welfare provision, the Asquith government created a state-managed fund to insure workers against unemployment. However, the logic of insurance fails in cases of collective disaster, when there are too many accidents all at once. And thus it was the mass unemployment during the Great Depression of the 1930s that made clear the limits of such safety-nets. A new generation of unemployed movements emerged, usually led by young communist militants, such as the Comités des Chômeurs in France or the Unemployed Councils in the United States, where a third of the population was out of work. The most celebrated street processions and eviction protests were in these industrial heartlands—the 1930 Wall Street riot, the Ford hunger march two years after, the Lille-to-Paris hunger march in late 1933—but there were similar demonstrations in the colonies as well, such as the 1933 hunger march in Jamaica.
The subsequent Keynesian reconceptualization of unemployment as an economic indicator subject to national macroeconomic fine-tuning became the basis for the post-wwii welfare states, which imagined a full-employment economy. For two decades, it appeared as if mass unemployment was a thing of the past. However, the Great Recession of the 1970s in Europe and North America marked the return of the spectre of wageless life, now under the sign of redundancy—the permanent shuttering of plants as entire regions underwent an Industrial Counterrevolution. A new wave of movements arose, particularly in France in the winter of 1997–98. As in the 1930s, deindustrialization is often understood to be a First World phenomenon, but, as we will see, it took place in rust belts around the globe like Ahmedabad, the Manchester of India.
But for some theorists, deindustrialization marked the end of unemployment as a political and conceptual tool. Among those arguing that we had reached the end of work was Ulrich Beck, the German theorist of neoliberalism’s risk society, who pointed to the shift from a ‘uniform system of lifelong full-time work organized in a single industrial location, with the radical alternative of unemployment, to a risk-fraught system of flexible, pluralized, decentralized underemployment, which, however, will possibly no longer raise the problem of . . . being completely without a paid job.’  Ulrich Beck, Risk Society, London 1992, p. 143. Neoliberal economists insisted that involuntary joblessness did not even exist; unemployment was either a choice for the marginal utility of leisure, or a temporary blockage of the labour market caused by high wages made too sticky by union monopoly and the state’s minimum wage.
It is also worth noting the great weakness of the social-democratic normalization of employment and unemployment. It constituted a normal subject: the wage earner. As a result, much of capitalism’s multitude was unrecognizable to a labour movement that had been reconstituted by state apparatuses into an employment movement, the agent of wage-earners divided into collective-bargaining units. Across society, there were many who lived outside typical employment and unemployment—women working in their own households, deindustrialized and disinvested communities without wages, those subjected to racial codes, even wage-earners in officially unrecognized industries and workplaces (in the us, for example, domestic, agricultural and academic workers not covered by the National Labor Relations Board). As a generation of feminist critics of the welfare state argued, this led to a gendered and unjust bifurcation of social security. Working-class households and neighbourhoods were divided between the independent, characteristically male, subjects of social insurance, and the dependent, characteristically female, subjects of social relief. One arm of the state apparatus insured and secured the normative male breadwinner against the risk of involuntary unemployment; another arm tested the ways and means of women raising children, before doling out a stigmatized relief. If the social-democratic conception of unemployment broke with the nineteenth-century rhetoric of the Poor Laws by understanding it as systemic rather than individual, as a waste of social labour rather than a malingering of the idle and dissolute, it also drew a stark and ideological line across the working multitude.
Favelas and bidonvilles
If unemployment dominated the imagination of the capitalist states of the West, it was not to be the governing concept in the development discourse of the post-colonial states. Here the spectre of wageless life in the sprawling shanty towns and favelas of Asia, Africa and Latin America overwhelmed any clear divide between employed and unemployed. Wageless life was not a temporary accident that might be insured against, nor a macroeconomic failure of aggregate demand; it appeared to be the main mode of existence in a separate, almost autonomous, economy.
The idea of the informal sector emerged following two decades of extraordinary Third World migration to cities, in which the urban working population doubled between 1950 and 1970. Colonial and settler-colonial regimes, as well as the plantation economies of the Americas, had restricted and even criminalized migration to the city; thus many mid-century revolts were based on the insurgency of peasants and rural agricultural workers. But in the wake of national liberation, ‘the poor’, as Mike Davis put it, ‘eagerly asserted their “right to the city”, even if that meant only a hovel on its periphery’.  Mike Davis, Planet of Slums, London 2006, p. 55. New forms of livelihood and struggle emerged out of the great squatter cities of the 1950s, and even before the development economists and sociologists had named the informal sector, filmmakers represented the wageless life of the new shanty towns in films that became paradigmatic for the rest of the century: Marcel Camus’s Black Orpheus (1959), which launched the first World Music—bossa nova—out of a mythic romanticization of Rio’s favelas during carnival; and Gillo Pontecorvo’s Battle of Algiers (1966), which lastingly portrayed the anti-colonial Algerian revolution not as the peasant war it was but through the epic metonymy of the defeated urban insurrection of 1956–57.
The first great theoretical engagement with this new form of wageless life also came out of a reflection on the Algerian revolution: Frantz Fanon’s revival of the nineteenth-century Marxist word ‘lumpenproletariat’ in The Wretched of the Earth. Coined by Marx in the 1840s as one of a family of terms—the lumpenproletariat, the mob, i lazzaroni, la bohème, the poor whites—it characterized the class formations of Second Empire Paris, Risorgimento Naples, Victorian London and the slave states of North America. In most cases, Marx even used the original language to suggest the historical specificity of these formations rather than the theoretical standing of the concept. For him, such expressions had two key connotations: on the one hand, of an unproductive and parasitic layer of society, a social scum or refuse made up of those who preyed upon others; on the other hand, of a fraction of the poor that was usually allied with the forces of order—as in the account of Louis Napoleon’s recruitment of the lumpenproletariat in The Eighteenth Brumaire, or Marx’s analysis of the slaveholders’ alliance with poor whites in the us South.
In these formulations, Marx had two antagonists. First, he was combating the established view that the entire working class was a dangerous and immoral element. He drew a line between the proletariat and the lumpenproletariat to defend the moral character of the former. Second, he was challenging those—particularly his great anarchist ally and adversary Bakunin—who argued that criminals and thieves were a revolutionary political force.  See Hal Draper, Karl Marx’s Theory of Revolution, vol. 2, New York 1978, ch. 15 and appendix G: ‘On the Origin of the Term Lumpenproletariat’. By the mid-twentieth century, the concept of the lumpenproletariat had pretty much disappeared from socialist and Marxist discourse. However, its reinvention in The Wretched of the Earth to describe the entirely new urban populations of the Third World made it one of the key stakes in the theoretical debates of the 1960s and 1970s. The discussion of the lumpenproletariat comes primarily in the book’s second essay, ‘Spontaneity: Its Strength and Weakness’, in which Fanon delineates the contradictions of the anti-colonial coalition, as urban nationalist militants turn to the peasant masses. He makes three powerful and controversial claims. The first is a sociological one about the emergence of a new dispossessed population, the people of les bidonvilles: ‘Abandoning the countryside . . . the landless peasants, now a lumpenproletariat, are driven into the towns, crammed into shanty towns and endeavour to infiltrate the ports and cities, the creations of colonial domination’; ‘These men, forced off the family land by the growing population in the countryside and by colonial expropriation, circle the towns tirelessly, hoping that one day or another they will be let in.’ Fanon resorts to biological metaphors: ‘The shanty town is the consecration of the colonized’s biological decision to invade the enemy citadels at all costs, and, if need be, by the most underground channels.’ It is an ‘irreversible rot’, a ‘gangrene eating into the heart of colonial domination’. ‘However hard [this lumpenproletariat] is kicked or stoned it continues to gnaw at the roots of the tree like a pack of rats.’  Frantz Fanon, The Wretched of the Earth, New York 2004, pp. 66, 81.
Secondly, Fanon, like Marx, argues that this lumpenproletariat is readily manipulated by the repressive forces of colonial order—if it is not ‘organized by the insurrection, it will join the colonialist troops as mercenaries’—and gives examples from Madagascar, Algeria, Angola and the Congo. Thirdly, and most famously, against the accepted wisdom of both nationalist and communist movements, he insists that
it is among these masses, in the people of the shanty towns and in the lumpenproletariat that the insurrection will find its urban spearhead. The lumpenproletariat, this cohort of starving men, divorced from tribe and clan, constitutes one of the most spontaneously and radically revolutionary forces of a colonized people . . . These jobless, these species of subhumans, redeem themselves in their own eyes and before history.  Fanon, Wretched of the Earth, pp. 87, 81–2.
Birth of informality
Fanon’s appropriation of the nineteenth-century term fuelled political debates throughout the 1960s. Virtually all the pioneering studies of labour in the Third World addressed his formulation: Pierre Bourdieu on work and workers in Algeria; Ken Post on the Jamaican labour uprisings of the 1930s; Charles van Onselen on everyday life on the Witswatersrand. Development economists and sociologists struggled to name the new reality that Fanon identified. In his landmark history of the economic development of the Third World, Paul Bairoch argued that ‘concepts of unemployment and underemployment as they have been formulated in the West cannot be applied . . . except in a very crude and approximate way.’  Paul Bairoch, The Economic Development of the Third World since 1900, Berkeley 1975, p. 165. Working in a social-democratic tradition, the Jamaican economist W. Arthur Lewis developed an influential model of the colonial ‘dual economy’ in the early 1950s. By the mid-1960s, the Argentine Marxist economist José Nun’s concept of the marginal mass had provoked an important debate.
The phrase that came to dominate official discourse—the ‘informal sector’—was coined in the early 1970s by a British development economist, Keith Hart, who was studying the communities of Frafra migrants from northern Ghana living in the Nima shanty town on the northern outskirts of the old city of Accra. ‘A very large part of the urban labour force is not touched by wage employment’, Hart wrote. He went on to outline the forms of ‘self-employment’ that made up the means of livelihood of Nima slum-dwellers: ‘the distinction between formal and informal income opportunities is based essentially on that between wage-earning and self-employment.’ The term was quickly adopted by the International Labour Organization in a 1972 study of employment in Kenya. Twenty years later the ilo had developed standards for the statistical measurement of the informal sector, and there were distinct debates not only in Anglophone Africa, but also in South Asia and Latin America. The ‘informal sector’ became the master trope for representing wageless life in cities around the world. According to the ilo ‘informal employment comprises one half to three-quarters of non-agricultural employment in developing countries’: 48 per cent in North Africa, 51 per cent in Latin America, 65 per cent in Asia and 72 per cent in sub-Saharan Africa. Moreover, ‘three types of non-standard and atypical work—self-employment, part-time work, and temporary work—comprise 30 per cent of overall employment in 15 European countries and 25 per cent of total employment in the United States.’ By the end of the century, the informal economy (as it had been renamed) had been made visible not only in Accra and Nairobi but in Los Angeles and Moscow.  Keith Hart, ‘Informal Income Opportunities and Urban Employment in Ghana’, Journal of Modern African Studies, vol. 11, no. 1, March 1973, pp. 62, 68; Paul E. Bangasser, The ilo and the Informal Sector, ilo Employment Paper 2000/9, p. 10; and ilo, Women and Men in the Informal Economy, Geneva 2002, p. 7.
In his essay on Accra, Hart launched the debate about informal wageless life that has continued ever since: ‘It is generally understood that growing residual underemployment and unemployment in the cities of developing countries is “a bad thing”. But why should this be so? In what way precisely does this phenomenon constitute a problem?’ His question might be seen as the beginning of the normalization of the informal economy. Earlier models of the dual economy had treated it as the ‘bad’ legacy of colonialism’s incomplete modernization, a transitional moment on the way to formal employment and unemployment. These states had inherited colonial labour apparatuses that had tried to discipline and regularize casual work. And, indeed, the mid-century era of import-substitution industrialization did see the growth of formal-sector employment in Latin America and even in some parts of Asia and Africa; the emergence of new armies of organized industrial workers gave rise to the great labour uprisings of South Africa, Brazil and South Korea. However, by the 1970s the growth of such jobs had stalled, and the discourse that named the informal sector saw it as a normal—indeed under neoliberalism, expanding—sphere of economic activity, part of the logic of post-colonial capitalist accumulation.  Hart, ‘Informal Income Opportunities’, p. 81. See also Alejandro Portes and Kelly Hoffman, ‘Latin American Class Structures: Their Composition and Change during the Neoliberal Era’, Latin American Research Review, vol. 38, no. 1, February 2003.
Just as the definition of unemployment in the late nineteenth century had depended on a new understanding of the economy, so the discovery of the informal sector depended on a sense of the state’s formal wage-labour apparatuses, which set minimum wages and maximum hours and provided unemployment insurance and social security. It was not the size of the enterprise that characterized the informal sector, nor the form of the labour process, but its relation to the state. The central issue then becomes the strength or weakness of the state: for some, informal economies develop when states regulate too much, driving economic activity to an underground, unregulated, untaxed world; for others, they are a product of weak or failed states, unable to provide social protections to their citizens and enforce rules or collect taxes. Neoliberal critics of state regulation have tended to celebrate the entrepreneurial gusto of the informal sector, its micro-enterprises that need only micro-credit to thrive. Defenders of social democratic welfare states have advocated the formalization of the informal: the extension of social protections and representation in unions.
Organizing in Ahmedabad
At the same time as development economists like Hart were discovering the informal sector, the first major organization of informal-sector workers took shape. In 1972, an activist in the Gandhian Textile Labour Association, Ela Bhatt, began to bring together the women head loaders and street vendors of the Gujarat mill town of Ahmedabad into a union, the Self-Employed Women’s Association. She had been assigned to survey families affected by the closure of two major textile mills.
While the men were busy agitating to reopen the mills . . . it was the women who were earning money and feeding the family. They sold fruits and vegetables in the streets; stitched in their homes at piece-rate for middlemen; worked as labourers in wholesale commodity markets, loading and unloading merchandise; or collected recyclable refuse from city streets . . . jobs without definitions. I learned for the first time what it meant to be self-employed. None of the labour laws applied to them; my legal training was of no use in their case.
‘Ironically’, she recalls three decades later, ‘I first glimpsed the vastness of the informal sector while working for the formal sector.’  Ela Bhatt, We Are Poor but So Many: The Story of Self-Employed Women in India, Oxford 2006, p. 89.
Over the next thirty years, sewa became a cluster of three types of membership-based organizations of the poor: first, a union—by 2004, the largest primary union in India—of a variety of informal trades—rag pickers, home-based chindi and garment stitchers, bidi rollers, vegetable vendors—bargaining with buyers, contractors and municipal authorities over piece-rates and pavement space; second, a coalition of dozens of producer co-operatives, producing shirt fabrics, recycling waste paper and cleaning offices; and third, several institutions of mutual assistance and protection, including a sewa bank and health cooperatives, organized around midwives who were themselves part of the informal sector.
A key part of its history has been a struggle over representation. ‘When someone asks me what the most difficult part of sewa’s journey has been’, Bhatt writes,
I can answer without hesitation: removing conceptual blocks. Some of our biggest battles have been over contesting preset ideas and attitudes of officials, bureaucrats, experts and academics. Definitions are part of that battle. The Registrar of Trade Unions would not consider us ‘workers’; hence we could not register as a ‘trade union’. The hard-working chindi workers, embroiderers, cart pullers, rag pickers, midwives and forest-produce gatherers can contribute to the nation’s gross domestic product, but heaven forbid that they be acknowledged as workers! Without an employer, you cannot be classified as a worker, and since you are not a worker, you cannot form a trade union. Our struggle to be recognized as a national trade union continues.  Bhatt, We Are Poor, p. 17–8.
sewa rejected the rhetoric of the informal sector that dominated official discourse: ‘dividing the economy into formal and informal sectors is artificial’, Bhatt argues, ‘it may make analysis easier, or facilitate administration, but it ultimately perpetuates poverty’: ‘to lump such a vast workforce into categories viewed as “marginal”, “informal”, “unorganized”, “peripheral”, “atypical”, or “the black economy” seemed absurd to me. Marginal and peripheral to what, I asked . . . In my eyes, they were simply “self-employed”.’ Indeed the women street vendors who were among the first to build sewa called themselves traders.  Bhatt, We Are Poor, pp. 18, 10, 11.
This rhetoric of self-employment drew on the ideologies of the Gandhian wing of Indian trade unionism from which sewa emerged, and it has been adopted by other organizations of wageless workers, notably the Durban-based South African Self-Employed Women’s Union founded in the mid-1990s. However, in retrospect, it seems to have been a nominal place-holder, as sewa took as one of its key tasks the representation of a world of wageless work which was invisible to the labour apparatuses of the state. When sewa organized the women who stitched chindi—fabric scraps discarded by textile mills—into khols (quilt covers) in the late 1970s, they began by depicting them, in spite of their scepticism:
in order to better understand the problems of chindi workers, we decided to conduct a survey in the seven poles, or streets, where most of the khols were stitched. Karimaben [one of the militant workers] had no patience for a survey. She complained, ‘We all know exactly what the problem is. Let me tell you that I spend more on a khol than I earn from making it’.
Nonetheless, sewa insisted on ‘proceeding methodically and conducting a survey’, reporting the findings to the chindi workers, and using them to fight for an increase in piece-rates both to khol traders and Labour Department officials. Surveys, Bhatt argues, ‘have served sewa well over the years. They help us gain a thorough understanding of the issues before taking any action, and the process helps us identify potential leaders in the community.’
Bhatt, We Are Poor, p. 63. These studies have given a much more complex view of the world of the self-employed. By 2004, sewa’s research had divided its members into more than eighty occupations in four main categories: street vendors and hawkers, home-based producers, labourers and service providers, and rural producers.
Martha Alter Chen, Self-Employed Women: A Profile of sewa’s Membership, Ahmedabad 2006, p. 12. Table 1 shows the growth of each of these categories since the 1970s: notice how the most visible group—the street vendors who make up about two per cent of urban India—were a major part of the early sewa, before dropping off proportionally.
After beginning in the cities, the organization of rural producers and agricultural labourers took off in the 1990s. Two-thirds of their members are not so much self-employed as what Jan Breman has called ‘wage hunters and gatherers’, casual labourers and service providers who work for others in the intricate disguises of contracted and piece-rate jobs.
Jan Breman, Wage Hunters and Gatherers: Search for Work in the Urban and Rural Economy of South Gujarat, Delhi 1994. A more specific breakdown in 2004 (Table 2, below) shows not only the variety of informal trades—from vegetable vendors, waste pickers to head loaders—but the overwhelming numbers of agricultural labourers.
Thus, organizations of workers in the so-called informal sector have mapped their world less by its relation to a formal state-regulated economy than by its workplaces, particularly the street and the home. When sewa pioneered transnational alliances of informal-workers associations in the 1990s, they did so by creating StreetNet and HomeNet. Increasingly, the two key representations of informalized workers in both official discourse and popular culture are the street vendor and the home-based worker.
Wandering the market
What can we conclude from this genealogy of representations of wageless life? It seems clear that neither of the great twentieth-century terms—unemployment and the informal sector—remain adequate, not least because of their segregation to specific zones of the capitalist world system; even the scholarly literatures on them barely speak to each other. This sense of conceptual exhaustion also applies to their traditional Marxist analogues: the socialist adoption of Marx’s ‘industrial reserve army’, and the anti-colonial adoption of Fanon’s re-figuring of the lumpenproletariat. But what alternatives do we have?
As I suggested earlier, two types of metaphor seem to dominate our contemporary imagination. The first points to the insecurity of many kinds of contemporary work: we speak of casualization, informalization and the proliferation of temporary and precarious jobs. In 1999, the ilo—long a site of struggle over forms of representation of work, its 1996 convention on home-based work the product of a protracted battle led in part by sewa—tried to cut across the formal–informal divide by characterizing such work as vulnerable, against which they called for decent work. This demand is both a retreat—a recognition that formal labour regulation does not touch the majority—and an advance—an argument for social protections and labour rights for the vulnerable. In the face of the many still-pompous invocations of inalienable human rights, one might note that we still await the modest Magna Carta of decent work.
A second metaphor goes further, suggesting that we have passed a historical watershed, the end of work as we have known it. Work, we are told, has lost its centrality to life; wageless life is workless, wasted life. Noting the dramatic break in popular discourse between the rhetoric of unemployment and that of redundancy, Zygmunt Bauman writes that ‘“redundancy” shares its semantic space with “rejects”, “wastrels”, “garbage”, “refuse”—with waste. The destination of the unemployed, of the “reserve army of labour”, was to be called back into active service. The destination of waste is the waste-yard, the rubbish heap’. ‘The production of “human waste”, or more correctly wasted humans . . . is an inevitable outcome of modernization’; ‘refugees, asylum seekers, immigrants’ are ‘the waste products of globalization’.  Zygmunt Bauman, Wasted Lives, Cambridge 2004, pp. 12, 5, 66.
Bauman’s apocalyptic denunciation of our culture of waste is powerful, but it misses the mark for two reasons. First, in its overly glib linking of material waste and human waste, it repeats one of the oldest tropes regarding the wageless—that they are akin to garbage, rubbish. Such metaphors run throughout this literature: early on Hobson characterized unemployment as waste; Marx was not immune, referring to the lumpenproletariat as refuse in The Eighteenth Brumaire. And indeed there is a connection: for those without wages have long worked as scavengers. As I noted earlier, not only are waste pickers a significant part of sewa, but many of their trades, like the chindi stitchers, were built out of the by-products of the textile industry. In March 2008, the first international conference of waste-pickers’ organizations was held in Bogotá.
That globalization produces redundancy would be better understood not through the deceptively concrete image of wasted lives, but through Marx’s two dialectically related concepts: the relative surplus population and the virtual pauper. The one is from Capital; the other from the Grundrisse. In the key chapter on ‘The General Law of Capitalist Accumulation’ in Capital, Marx views the problem from the vantage point of capital: ‘it is capitalist accumulation itself that constantly produces, and produces indeed in direct relation with its own energy and extent, a relatively redundant working population, i.e. a population which is superfluous to capital’s average requirements for its own valorization, and is therefore a surplus population.’ He continues: ‘this is a law of population peculiar to the capitalist mode of production; and in fact every particular historical mode of production has its own special laws of population’. Indeed, ‘the relative surplus population exists in all kinds of forms. Every worker belongs to it during the time when he is only partially employed or wholly unemployed.’ The industrial reserve army is thus merely one of these forms; in fact, as might be expected, Marx’s specific examples of the relative surplus population are the most dated part of his analysis.  Marx, Capital, pp. 782,783–4, 794.
The fundamental metaphor in Marx’s account is that of opposing forces: it is not as if there are two kinds of workers, employed and unemployed, or two sectors of the economy, formal and informal; rather, there is a process in which ‘greater attraction of workers by capital is accompanied by their greater repulsion . . . the workers are sometimes repelled, sometimes attracted again in greater masses’. The ‘higher the productivity of labour, the greater is the pressure of the workers on the means of employment, the more precarious therefore becomes the condition for their existence, namely the sale of their own labour-power’. Intriguingly, almost the entire contemporary vocabulary—redundant, superfluous, precarious—can be found in this chapter.  Marx, Capital, pp. 783, 794, 798.
If the passage in Capital tells the story from the point of view of the accumulation of capital, the parallel passage in the Grundrisse begins from the point of view of living labour: ‘It is already contained in the concept of the free labourer, that he is a pauper: a virtual pauper . . . If the capitalist has no use for his surplus labour, then the worker may not perform his necessary labour’. Marx is not arguing that all workers are or will become beggars, as in the immiseration thesis often attributed to him. Rather, this is his account of bare life: since the exchange required for the means of living—the selling of labour-power—is accidental and indifferent to their organic presence, the worker is a virtual pauper.  Marx, Grundrisse, trans. Martin Nicolaus, New York 1973, p. 604. For an intermediate draft of this passage in the 1861–63 manuscripts, see mecw 30, p. 40. Virtual paupers: this strange figure—which combines an almost lost word with one that has taken on entirely new connotations—will be my temporary resting place. In a letter written as he turned fifty, Marx wrote: ‘half a century on my shoulders and still a pauper’. A century and a half on again, the spectre of wageless life still weighs upon us.
 Karl Marx and Frederick Engels, Collected Works, New York 1975ff (hereafter mecw), Volume 3, p. 284. This essay was originally written as part of the Yale Working Group on Globalization and Culture. I would like to thank the other members for their suggestions and criticisms, and Achille Mbembe for his response to an earlier text at the Wits Institute for Social and Economic Research, University of the Witwatersrand, South Africa, 22 February 2006.
 Karl Marx, Capital, vol. 1, Harmondsworth 1976, p. 764.
 Michael Lebowitz, Beyond Capital: Marx’s Political Economy of the Working Class, New York 2003; David Harvey, The Limits to Capital, Chicago 1982, p. 163.
 Biopolitical readings of unemployment are in a way the product of the intellectual upheaval triggered by the third wave of mass unemployment; two landmark texts both date from 1986: Robert Salais, L’invention du chômage: histoire et transformations d’une catégorie en France des années 1890 aux années 1980, Paris 1986, and Alexander Keyssar, Out of Work: The First Century of Unemployment in Massachusetts, Cambridge 1986. See also Christian Topalov, Naissance du chômeur: 1880–1910, Paris 1994. A more recent study that draws on this work is William Walters, Unemployment and Government: Genealogies of the Social, Cambridge 2000.
 John Garraty, Unemployment in History, New York 1978, pp. 109, 4; J. A. Hobson, ‘The Meaning and Measure of “Unemployment”’, Contemporary Review 67, March 1895.
 Marx, Capital, pp. 303, 307.
 Walters, Unemployment, p. 18. See also Gareth Stedman Jones, Outcast London, New York 1984, pp. 291–6. In his letters Engels was very critical of the sdf’s ‘bunkum about social revolution’. His characterization of their procession as merely comprised of ‘idlers, police spies and rogues’ is one of the classic passages on the lumpenproletariat; mecw 47, pp. 407, 408.
 Hobson quoted in Walters, Unemployment, p. 32. See also Stedman Jones, Languages of Class, Cambridge 1983, p. 159. Massachusetts Bureau quoted in Keyssar, Out of Work, p. 72.
 Ulrich Beck, Risk Society, London 1992, p. 143.
 Mike Davis, Planet of Slums, London 2006, p. 55.
 See Hal Draper, Karl Marx’s Theory of Revolution, vol. 2, New York 1978, ch. 15 and appendix G: ‘On the Origin of the Term Lumpenproletariat’.
 Frantz Fanon, The Wretched of the Earth, New York 2004, pp. 66, 81.
 Fanon, Wretched of the Earth, pp. 87, 81–2.
 Paul Bairoch, The Economic Development of the Third World since 1900, Berkeley 1975, p. 165.
 Keith Hart, ‘Informal Income Opportunities and Urban Employment in Ghana’, Journal of Modern African Studies, vol. 11, no. 1, March 1973, pp. 62, 68; Paul E. Bangasser, The ilo and the Informal Sector, ilo Employment Paper 2000/9, p. 10; and ilo, Women and Men in the Informal Economy, Geneva 2002, p. 7.
 Hart, ‘Informal Income Opportunities’, p. 81. See also Alejandro Portes and Kelly Hoffman, ‘Latin American Class Structures: Their Composition and Change during the Neoliberal Era’, Latin American Research Review, vol. 38, no. 1, February 2003.
 Ela Bhatt, We Are Poor but So Many: The Story of Self-Employed Women in India, Oxford 2006, p. 89.
 Bhatt, We Are Poor, p. 17–8.
 Bhatt, We Are Poor, pp. 18, 10, 11.
 Bhatt, We Are Poor, p. 63.
 Martha Alter Chen, Self-Employed Women: A Profile of sewa’s Membership, Ahmedabad 2006, p. 12.
 Jan Breman, Wage Hunters and Gatherers: Search for Work in the Urban and Rural Economy of South Gujarat, Delhi 1994.
 Zygmunt Bauman, Wasted Lives, Cambridge 2004, pp. 12, 5, 66.
 Marx, Capital, pp. 782,783–4, 794.
 Marx, Capital, pp. 783, 794, 798.
 Marx, Grundrisse, trans. Martin Nicolaus, New York 1973, p. 604. For an intermediate draft of this passage in the 1861–63 manuscripts, see mecw 30, p. 40.