Siddharth Kara’s Bonded Labor, written for a popular readership, is a welcome addition to the sparse literature on the persisting relations of servitude in South Asia.¹ Spanning a wide range of sectors (from agriculture to stone-breaking and construction) and almost all South Asian countries (mainly Bangladesh, India, Nepal, and Pakistan), it is an excellent documentation of the incidence of bonded labour, its economics, its oppressive conditions, and of the resounding failure of the progressive legislation that was set up – especially since 1976 in India and its neighbouring countries – to root out this contemporary practice.

The main contribution of this book is to document the extent and nature of the slave-like relations that exist in different sectors in South Asian countries.² Relations of servitude are explored in multiple spatial and sectoral locations, such as in agriculture through the kamaiya system in Nepal and the hari system in Pakistan; in shrimp production and distribution and on tea plantations in Bangladesh; in brick-making and carpet-making in Uttar Pradesh, bidi-making in West Bengal, stone-breaking in Faridabad (Haryana), and construction work in New Delhi (for the prestigious Commonwealth Games) in India; in domestic work, including kamaliri practices in Nepal and other arrangements in several other parts of South Asia; and in mining across South Asia. According to the author’s calculations, the number of bonded labourers (who are the most prevalent slave-like workers) in the world today ranges from 18 to 20.5 million, out of which South Asia accounts for over 85 per cent.³

The author details the legislation that has been enacted to eliminate this institution in India, including the The Bonded Labour System (Abolition) Act, 1976 and two landmark Supreme Court judgments, viz., People’s Union for Democratic Rights vs. Union of India and Others, 1982 and Bandhua Mukti Morcha vs. Union of India and Others, 1983. His account bears testimony to the fact that the institution of bonded labour has thrived in the Indian context despite the existence of such progressive legislation. Kara also offers his own suggestions to improve the enforcement of laws and policies in general (from both the demand and supply sides) to tackle and end this odious practice in South Asia.

The author lucidly defines the basic structure of the institution of bonded labour early on in the book, and uses this description to outline its concrete manifestations in different sectors and spaces. “Bonded labour” derives its name from the fact that the worker pledges, or bonds, his or her labour (power) to another individual (not necessarily the final employer, but also, for example, a labour contractor such as a jamadar or a sirdar) for obtaining a certain amount of money as a loan.⁴ It is therefore an instance of the worker being at one end of an inter-linked structure of markets – in this case, the labour and credit markets. Until the loan is repaid, the worker has to labour for the employer–moneylender, and, in most cases, a meagre subsistence amount (an amount that is much less than a living wage) is granted to the worker during the intervening period. The period of repayment can vary quite significantly, ranging from a few months to an entire lifetime, and sometimes may stretch across generations as well. During this period, the worker is a bonded labourer, which, according to the author, is one kind of slave labour.

The author provides several reasons to justify his use of the term “slaves” to refer to workers in the period of their bonds or contracts. I list six broad reasons below, not necessarily in the way the author articulates them, nor as an exhaustive list of what he offers. First, even if, on occasion, there appears to be a voluntary beginning to their status as bonded labourers, there is a whole host of coercive economic circumstances (e.g., daily consumption needs, wedding expenses or the death expenditure for a family member) that may lead to the initial borrowing and the pledging of their labour. Second, during the period of contract, the workers are frequently treated as sub-human, in the sense that various humiliations and violations (both physical and symbolic) are inflicted upon them. Third, the mobility of these workers is severely restricted, with their movements being closely monitored and controlled (both in terms of mobility of employment and simple physical mobility). Fourth, there is an utter disregard for providing decent working conditions (including subsistence wages) or living conditions for the workers. Fifth, there is a close intertwining of this institution with another historical institution that has survived into the present through various mutations in history, namely, caste. Kara documents carefully the close correlation between bonded labourers and their origins in the oppressed castes, in the different contexts in which he has conducted his field research. Through this correlation, he shows that there is a particular systemic basis for the institution of bonded labour that cannot be easily dismissed as random, episodic, or as merely arising out of economic coercion. The unequal relations that define the caste system lie at the very foundation of the slave-like institution of bonded labour. Sixth, there are infringements of basic human rights such as sovereignty over the body; frequently, these involve physical or sexual abuse, and also child labour from a very early age.

Is this characterization of a “modern” institution, namely bonded labour, as slavery justified? In other words, is it possible to clearly identify the “unfreedom” inherent in these relations of servitude and liken it to slavery? From a broadly Marxist standpoint, if we attempt a classification of production relations of various kinds, we may arrive at five basic types, with a whole lot of hybrids among the five types. These five types and their hybrids may coexist at different levels of significance in different social formations. Three of these five types of production relations are fundamentally exploitative, in the sense that surplus labour is appropriated from the producers by other classes. These are the relations between master and slave, feudal lord and serf, and capitalist and worker. The first two involve what is usually defined as “unfree” labour, and the exploitation process (extraction of unpaid labour) is apparent in them. The logic of exploitation in these two cases is primarily governed by extra-economic coercion. The third, the relation of capitalist–worker, involves “free” labour that is also exploited, although this process is not apparent and extra-economic coercion is not usually present. The fourth and fifth types of production relations can be found in petty commodity production and communist production, which are not exploitative in their pure forms. However, petty commodity producers could get locked into various coercive structures within a contract through interlinked markets (such as product–credit), and lose their autonomy over the product or the production process or both – in which case they may be sliding into what Marx calls the formal subsumption of labour to capital. The production relations inherent in the institution of bonded labour in South Asia seem to come closest to slavery among these five types, since the producers frequently have no inalienable rights over any means of production or even over their own labour, or sometimes even over their own bodies, during the period their labour is pledged. While exploitation in bonded labour seems to originate mainly from economic coercion, there are also extra-economic factors that play a key role in the functioning of the institution (caste, for example). There may be many other relations that are similar in nature, even when the producers are much richer than the kinds of producers that the author is describing.⁵ How do we understand this specific hybrid relationship, where the entry into the relationship of bondage may be voluntary and primarily governed by economic coercion, and where exit is always a possibility as long as debts are repaid, while the perpetuation of the institution has unequal economic and non-economic coercive forces at work that do not appear to readily function in the domain of freedom?

The question of whether the institution of bonded labour that Kara describes is like slavery is not an easy question to settle. In the 1990s, there was a vigorous debate between Brass (1990) and Jodhka (1994), which is also discussed in detail in Rao (1999), that raised a similar question in the context of attached or tied labourers (as opposed to daily casual labourers) in what could be termed advanced agriculture, in Haryana and other parts of North India. After making a distinction between permanent attached labourers and temporary ones, and admitting for the possibility of heritability of debt, the question that was posed was: are attached labourers in these contexts unfree? Rao argued that unfreedom pertains to the politico-juridical conditions in which entry and exit are defined vis-à-vis the employment contract (this is the domain of market exchange), while positive freedom arises out of conditions in and of employment. Rao concluded that in the case of temporary attached labour without heritability of debt there is no strong basis to call the relationship unfree, although the inequality of property and the power that ensues from this inequality are real within the domain of employment.⁶ Such inequality in property may also occur in employer–worker relations in the classic or canonical forms of capitalism, making it difficult to distinguish between the attached worker and the regular worker on the formal axes of freedom or equality. At the same time, there may be several substantive differences, such as the nature of localized monopolies and power structures in rural India (as opposed to canonical capitalist relations), the absence of pledging one’s labour power for a loan in a regular capitalist system, and so forth.

Is this entire debate relevant for the book that is being discussed here? The author describes some relations, such as the selling of child labourers by parents or selling of women for sex work, and cases where there is heritability of debt, that are unambiguously slavery-like. There are other relations, however, where voluntary contracts are entered into for credit purposes and where there is always an option to exit once the debt is repaid, that are much more ambiguous and hybrid, and therefore could be construed broadly as relations of servitude. By conflating these somewhat different relations, the book probably loses some analytical clarity.

In addition to the above issues, apart from the moral anger that the author may induce in the reader through the use of the word slavery, is there any analytical value in the use of terms like slavery or slave-like relations (whether or not the use of these terms is wholly appropriate) in the book? The term bonded labour might be adequate to bring out all the nuances in the book. However, there could be analytical uses of the hybrid characterization of the bonded labourer as possessing certain aspects of the institution of slavery that are only hinted at in the book but not teased out. A key feature that Marx discusses in the context of another kind of slavery (the chattel form) is the fact that apart from the daily subsistence that is given to the slave for daily reproduction, the modern slave (in the larger capitalist context) should be analytically included in constant capital (which includes physical capital in the Marxian framework). The capitalist deals with constant capital and variable capital in very different ways. He owns his constant capital and can use it in any manner and for however long he desires to use it. The case of variable capital (value of labour power) is different. While the capitalist may believe that by virtue of having bought the commodity of labour power, he owns the entire labour as well as the entire time of the worker, this belief is not necessarily reciprocated by the worker. The worker would resist, and this conflict is indeed one of the major manifestations of class struggle in the capitalist era, i. e., over the length of the working day. In the institution of bonded labour within a larger capitalist context, the employer who may sell his produce in a regular product market can be structurally more exploitative than his capitalist counterpart.  The treatment of labour as a part of constant capital (during the period of the labour pledge) allows the employer to extract very long working days from the worker under poor working conditions. The subjectivity of certain classes of bonded labourers towards their own labour during the period of their labour pledge may also be typically consistent with that of a modern slave; and this subjectivity, along with economic and other kinds of coercion, are the composite elements that allow the institution to be perpetuated. To reiterate this point in a different way, the subjectivity of the bonded labourer is one in which he or she is selling not labour power as a commodity but instead his or her labour itself. This is especially visible in the case of contracts that are defined over a fixed period of time. When the contract period is not fixed but the relation could extend until the debt is cleared, there may be a strong tendency on the part of the employer to deny bonded labourers their autonomy over the commodity of labour power.⁷ Therefore, the characterization of bonded labour as deriving certain aspects from the institution of slavery is analytically useful in understanding the very different labour process as well as working conditions, as compared to typically capitalist spaces.

However, slaves or slave-like workers revolt too (and there are several historical examples of this), when working conditions become unendurable or when the larger political context is conducive to developing a different kind of consciousness; and this is something we occasionally see among bonded labourers too. A part of this political struggle in a capitalist context involves the bonded labourer changing his or her own consciousness and that of his or her employer in making a distinction between labour power and labour. What is being pledged is not labour itself but labour power, which then allows the bonded labourer to fight for better working conditions, compensation, and better control over his or her own time and body. Political battles, or even the above-cited Supreme Court judgments, frequently gesture towards this kind of a consciousness. Kara’s account in the book under review is largely oblivious of (or deliberately ignores) the larger political struggles, and it appears as if he relies on and advocates, almost entirely, other means: state legislation and action, the acts and decisions of progressive jurists (such as the former Chief Justice of the Indian Supreme Court, P. N. Bhagwati), activism by non-government organisations, or the activism of well-meaning liberal actors.

There are certain other areas that have not adequately been addressed in the book. A key puzzle it generates is why such a huge proportion of slave-like relations persists in South Asia, and not in other parts of the global South, such as other parts of Asia, Africa, or Central and South America. There are hints at different places in the narrative. First, the author engages sketchily with historical processes in India such as the caste system and its continuity from the Vedic period onwards, notions of servitude that were prevalent during Muslim rule, and the heightening of certain kinds of servitude (such as indentured labourers in various British colonies) during the colonial period. Second, there is reference to the fact that no revolutionary transformations have occurred in this region of South Asia. Third, there seems to be an almost crude Malthusian story – to the effect that population pressures create a low land–labour ratio, and that this process in itself may drive people towards servitude in agriculture and other sectors of the economy. While the first two factors are indeed important, the third is somewhat questionable, and we can elucidate this by revisiting Domar’s famous hypothesis about servitude: “of the three elements of an agricultural structure relevant here – free land, free peasantry, and non-working landowners – any two elements but never all three can exist simultaneously” (see Domar 1970, p. 21).⁸ Domar’s view is that (in an equilibrium situation) if there exists abundant land relative to workers, there would be slavery or serfdom. This is because there would be an upward pressure on wages when there is a high land–labour ratio, leading to elimination of rents for a class of non-working landowners. This in turn would create an impetus for the non-working landowners to create relations of servitude in order to garner those rents. On the other hand, a low land–labour ratio would allow rents to exist and therefore there would be no logical necessity for the existence of relations of servitude. From the Domar perspective, then, the existence of servitude with an abundance of labour in contemporary South Asia would constitute a deep puzzle. While this requires deeper analysis and empirical investigation, a hypothetical explanation may be posed here. Assuming that the South Asian elites are not any greedier than the elites of the rest of the world (although the excess greed propensity of South Asian elites cannot be ruled out as a hypothesis, it is too behavioural or culturalist to be credible), this would have to be a careful explanation derived from deeper materiality. In an economy that has not generated viable alternatives to the current avenues for employment, which have not been welfare-enhancing, especially for the poor, why is there a propensity among a section of the elites to look for a secure or captive labour force that will ensure the generation of normal or super-normal profits? Perhaps the absence of viable alternatives creates a desperate pressure among the poor to migrate, and this is one of the many forces to which elites are reacting when they repeatedly constitute the institution of bonded labour. This is one of the areas where the book falls short in its exploration of the institution of bonded labour in South Asia.

Another key issue that gets inadequate treatment through the whole book is the interaction between capitalist and non-capitalist sectors (the latter include the relations of servitude that the author tries to explore in his work). The only interactions the author highlights are: (i) that of global (relatively affluent) consumers and their complicity in the perpetuation of the institution of bonded labour; and (ii) the global crisis of 2008 and its aftermath in terms of increased poverty, creating push factors for an expansion in relations of servitude. However, there are other questions. The numbers of these bonded labourers that the author has computed – have they increased over the last thirty years or so in South Asia? There have been massive structural transformations in the South Asian economies in recent decades, with a clear turn from a state-oriented economic structure to a market-oriented one. There are two key processes that are fall-outs of these transformations that should have been woven into the narrative of the book. First, there has been an uneven outbreak of the phenomenon of agrarian distress across different regions in India that has devastated livelihoods, pushed small and marginal farmers into various kinds of indebtedness, and resulted in tragic farmer suicides. This has been caused, in large part, by the withdrawal of state support to agriculture after the advent of neoliberal economic reforms in the 1990s. How does this process intertwine with the perpetuation of bonded labour, especially the debt–bondage relations on which the author focuses? Second, while the high-growth, formal/organized sectors in India over the last two or three decades have not created too many jobs, there has been an increased impetus among the rural poor to migrate out (nowhere close to the Chinese levels but higher than earlier within India) into the urban informal sector. This migration has been both of a circulating kind and a more permanent one. How has the migration process in India (or, more broadly, South Asia) affected the institution of bonded labour? Without making these connections and looking for certain kinds of discontinuities in the way the institution is mediated by different kinds of capitalism (for example, state-oriented versus market-oriented), the overall narrative (especially in the post-colonial period) of bonded labour in the book becomes a bit too linear and seems like a long, continuing process, although there might be interesting transformations in the nature and the evolution of this institution over time.

I would like to add two brief comments in conclusion: first, about the calculation of the number of slaves; and second, on the general tone Kara adopts through the book. The author never makes it clear how he has produced his estimates of bonded labourers. Since there are no secondary data sources available on bonded labourers in South Asia or indeed the world, he merely mentions in passing that the estimates are derived from his field research. By using a statistical language that involves terms like confidence intervals, there is an attempt to impart some credibility to these numbers. However, without laying out the methodology of the calculation in detail (Appendix A of the book, where the author claims to have presented the methodology in detail, leaves the reader groping in the dark), the numbers are at best intelligent guesses made by a committed researcher. There was probably no need to come out with such estimates, which have no rigorous methodological basis, to convey the central message of the book.

Regarding the tone of the book, I will make an unconventional comment for a review essay. There is an anxiety that the author carries right through the narrative. One may guess from his tone that he perceives a freedom of choice: whether to go a step further and do something more concrete for bonded labourers than what he is already doing, or to continue in the mode of a researcher’s engagement with his subjects, interspersed with some activism, while retaining his own position of relative economic privilege. Perhaps this freedom, then, produces a deep anxiety in him. All the rich documentation and field descriptions of bonded labour in the book seem to be strongly intertwined with this existential anxiety of the author. I am adding this comment not because I have a moral judgment of how the author has structured his own life or because I feel that this anxiety should somehow be overcome, but because this tone distracts somewhat from the focus of the intended narrative.

Notwithstanding the inadequate basis for the author’s estimates of the number of slaves, the tone of his narrative, and the lack of sufficient critical engagement with notions of freedom and unfreedom, the book adds significantly to the literature on relations of servitude in the South Asian region. It may also motivate other researchers to further explore the intricacies of the nature, evolution, and spatial diversity of the institution of bonded labour, as well as motivate political and other activists to work harder in their efforts to root out this institution.