Concerning distributive labour: Exploring the pragmatics of globalised interdependence

Abstract

The promotion of smallholder dairy farming in Rungwe District, Tanzania has been enormously successful, with the vast majority of households now in possession of productive dairy cows. This article compares supposedly traditional loans with self-help groups directly established by non-governmental organisations (NGOs) that both reveal ‘big’ recipients rather than distributors of resources, and that those carrying out development activities often benefit most from them. Rather than advocating a moral judgement, the article suggests the value of thinking with and against the concept of distributive labour to explain the pragmatics of interdependent relations that are key to doing development. Detachment between beneficiary and donor is essential, and mediates salutary claims that distributive labour is a means to advocate for shared values about the distribution of wealth across the globe.

James Ferguson’s (2015) conceptualisation of distributive labour has been held up as a new vision for more just global relationships with emancipatory potential for billions of people as well as theoretical significance within the academy (Hickel, 2016: 211; Nilsen, 2021). Distributive labour is the work that goes into generating and sustaining social relations that enable people to make successful claims on others with resources they want (Ferguson, 2015: 94–100). The concept blurs the dichotomy of ‘working for a living’ versus ‘begging for handouts’ and questions the intrinsic and exclusive link between production and labour. Ferguson has a hopeful and ambitious aim to ‘find ways to restore value in people (and not just their labour), and to build a new dispensation within which people could truly count, once again, as the most precious form of wealth’ (Ferguson, 2013: 238). The diversification of perspectives on ways of relating in a deeply unequal world is welcome, as are approaches that sustain living for all, rather than a small subset of people (Hart, 2015 [2013]). However, within this important move to carve out a position that is not merely ‘anti-’ (Ferguson, 2010), there is less attention to the pragmatics and limitations of what globalised interdependence offers and involves today. As an anthropologist working with smallholder farmers in sub-Saharan Africa, with international non-government organisations (INGOs) and private development funders who would change their lives (Eyre, 2021), I see a gap in attention to the messy reality of contemporary distributive labour. Who does it work for, and how?

Ferguson draws on and reinvigorates theories of wealth-in-people developed by anthropologists of Africa to explain patron–client relations and bridewealth, where unequal relationships are not merely a means to an end but a goal in themselves for both the dependant and the dependee (Kusimba, 2020: 173). Ferguson’s optimism seems a long way from these pioneering models of complexity, ambiguity, and enduring injustice associated with ‘big men’ and gerontocracy (Bledsoe, 1980; Guyer, 1995; Miers and Kopytoff, 1977). Recently illuminating the constrained agency afforded through interdependence in contemporary Kenya, Sibel Kusimba (2021: 51–2) defines wealth-in-people as the ability to do things with, for, and through other people through rights in their labour, loyalty, property, or reproductive capacity. Others actively seek to be wealth-in-people, that is, belonging to as well as in relations that offer them economic, social, political, and other resources (Miers and Kopytoff, 1977). Dependence offers personhood and agency through belonging. This contrasts with the liberal Euro-American association of agency with independence and autonomy (Hickel, 2014). Unlike Kusimba’s (2021) account of interdependence, in which people can play multiple roles over time and in different relationships, my focus is on Ferguson’s conceptualisation of ongoing efforts to secure resources dedicated to social assistance through strategies of dependence. I do not suggest that Ferguson is ‘analytically flawed’ because he does not conceptually engage with a political economy critique of global capitalism and neoliberalism, as Nilsen (2021) does. Neither does my argument focus on cash transfers in themselves, nor the politics of distribution within southern African states that generate tax revenues to pay for them, but on Ferguson’s assertion that distributive labour might be the basis of an argument that compels different parties to agree on a better ways to share the world’s wealth.

After briefly describing my methods, I describe key strands of Ferguson’s argument about distributive labour, a ‘rightful share’, and globalised interdependence, as well as literature that broadens the scope of enquiry beyond cash transfers. I then turn to an important early pioneer of wealth-in-people who demonstrated that the proliferation of cattle in Tanzania raised the problem of distribution (Schneider, 1981). This helps to situate my ethnography of the spread of dairy cows in Rungwe District, Tanzania. A decentralised loan system has been one important mechanism. But all dairy cattle originate with their deliberate introduction by outsiders as a development programme. I compare cattle loans with outside NGO-sponsored projects that have worked in Tanzania for over 40 years. Here distributive labour has also been important. But the process reveals several limitations in terms of the reproduction of existing hierarchies, and the transformative potential of the resources actually available. Finally, I explore claim-making by putative beneficiaries of these and other dairy development programmes, and its limited results in terms of support or solidarity between actors separated by lived experience, geography, and wealth.

Methods

This article draws on 17 months of participant observation in Rungwe among smallholder farmers. From February 2018 to May 2019 and August to September 2019 I lived in a village I call Sukulu, 10 km from Rungwe’s largest town, Tukuyu, with a family who grew banana and kept cattle. During many unstructured discussions with them and their neighbours, and in 80 semi-structured interviews, I discussed what people thought of their economic prospects, and outside projects that aimed to transform them. I attended five weekly and bi-weekly meetings of dairy development groups. I traced the movement of cattle connected to four projects that operated over the previous twenty years. I took detailed life histories from five farmers and analysed patterns of cattle exchange in one area around my host’s home (totalling 61 households). I also conducted a village-wide household survey of 222 households (20% of the village), spread evenly across all eight sub-villages (vitongoji). ¹ I also undertook interviews and participant observation with NGO workers and philanthropy professionals, although this is not the basis of the article, which focuses on the lived experience of putative beneficiaries of dairy development.

Distributive labour

Ferguson (2015) argues that chronic un(der)employment in southern Africa has created a niche for a new understanding of proper economic relations best exemplified by cash transfers. He proposes they reveal new solidarity about ideas of deserving between different actors. Following Bayart (2000: 218), Ferguson insists that dependence is a ‘mode of action’ for those who have no access to employment opportunities. Distributive labour is this pressing of a claim to distribute resources (Ferguson, 2015: 101). Ferguson offers four examples: the unwanted window washer at traffic lights who is not providing a service but pressing a claim; the ‘panhandler’ who solicits cash donations on the street; the pickpocket who steals from people; and a rural mother visiting an employed urban son to show her love and needs, and request support from his wages (Ferguson, 2015: 100–1). This variety is important because, while Ferguson’s argument is steeped in debates about cash transfers, he explicitly relates the concept of distributive labour to welfare payments (and development) in general and the potency of new sources of claim-making within international transfers of resources (Ferguson, 2015: 26). He argues for political and ethical framing that attends to distributive labour as a mechanism for determining a share in resources to be enjoyed by all, irrespective of their productive capacity or nationality. Acknowledging that this politics of the ‘rightful share’ is only partially realised, he nonetheless argues that it is an increasingly compelling obligation (Ferguson, 2015: 184). Abstracted beyond the case of cash grants, he suggests that INGOs and private foundations re-framing of food and medical concerns as matters of ‘global health’ for humanity as a whole, rather than nation states suggests: ‘a new rationality and a new ground for certain forms both of care and of material distribution’ (Ferguson, 2015: 196). It is less clear how distributive labour connects those at extremes of wealth at a global scale. For example, his conclusion refers to the US-based international NGO GiveDirectly and the Bill and Melinda Gates Foundation as potential participants in the new global politics of distribution (Ferguson, 2015: 194–6) that works beyond the nation state. Is the idea that the perspectives of these actors should (or do?) converge with those of the poor people they wish to benefit? There is an ethnographic gap regarding how distributive labour constitutes a mode of action for poor people in such relations: how their ideas about the proper distribution of resources impact on others who possess them.

A growing number of ethnographers share Ferguson’s interest in shifting discussions about dependence towards dependants. They move beyond the focus on cash grants provided by southern African states generating tax revenues through extractive industries and a growing middle class. Beresford (2021) shows how distributive labour that enables South African entrepreneurs is constrained by class and racial hierarchies. Kusimba (2021) explores the interplay between technology and interdependence afforded by the explosion of mobile money in Kenya. Mario Schmidt (2020) engages with unexpected problems experienced by GiveDirectly when making supposedly ‘unconditional’ cash payments in Western Kenya that are rejected by almost 50% of intended beneficiaries. China Scherz (2014) attends to the distributive labour of would-be beneficiaries of different NGOs in contemporary Uganda. Scherz compares and contrasts the different moral logics of two organisations. One, Christian, charity accepts ongoing claims of need from dependants. The other, a sustainable development NGO, aims for those it supports to become independent (and on that basis denies repeated claims). Scherz evaluates both, alongside the Kiganda ethic of omutima omuyambi (Luganda: ‘heart for helping’). She proposes there is a ‘moral valence’ (connection) from Bugandans’ points of view between charity and Kiganda ethics of self-care through attaching oneself to patrons. On this basis, Scherz suggests that charity is not wounding but a morally superior intervention by outsiders (compared to sustainable development) because it better fits the expectations of its target beneficiaries. Questions remain about the significance of this valence in pragmatic terms: what does it offer, and to who?

In Tanzania, development remains ‘anticipatory’ for many (Green, 2014: 132–50). Some urban and middle-class Tanzanians establish their own NGOs with written constitutions. They emulate ‘development speak’ in funding proposals for initiatives that can be framed as development projects, although they often only exist on paper. Thus an appeal for funding to extend a fast food business is written under the English title ‘Appeal for sponsorship of financial aid for empowering the socio-economic development status of the Star restaurant’ (Green, 2014: 137). For most, development never arrives because they do not have relationships with those who have access to resources. In particular, because wealth-in-people is most associated with rural Tanzanians who have limited access to outsiders, ethnographic questions remain about distributive labour at the conjuncture of externally funded development projects and their intended poor rural beneficiaries.

Cash cows

My approach also de-centres enquiry into distributive labour from cash transfers. It concerns dairy cows. Of course, cash and cows are different. But the pertinence of my case to distributive labour is two-fold. First, Ferguson’s manifesto for a ‘rightful share’ through distributive labour is illustrated with policy debates about cash transfers. But he argues its broader relevance to international development and global social policy (Ferguson, 2015: 196). Second, Ferguson’s conceptual foundation in wealth-in-people is itself grounded in the qualities of cattle in East Africa that make them money-like. A simple contrast between cash and cows neglects that both are often framed as catalytic capital (Ferguson, 2015: 16, 27–8). It also relies on reductive assumptions about money. Harold Schneider’s (1981) pioneering work on wealth-in-people argued that cattle was a form of money that was alive and expanding for the Turu of Singida Region in Tanzania. As herd size increased, successful cattle-keepers were faced with constraints on space for grazing and challenges of protecting cattle, as well as expenses brought on by wealth. This led people to loan cattle to others, or give them as bridewealth, to build debts and alliances through which they could exercise rights-in-people (Schneider, 1981: 218). This does not mean I take for granted a 40-year-old description of cattle as money, nor underestimate the important differences between different forms of livestock, of husbandry techniques, and money. But simply that cash and cows can and should be compared: a point Ferguson has long made!

The Nyakyusa of Rungwe (and Kyela) District are known in anthropology as cattle-keepers. Monica Wilson (1951: 60) says that cows were their most prized possession. People distinguish between Zebu cows, which they refer to as local cattle (ng’ombe wa kienyeji) and dairy cows, which they call modern cows (ng’ombe wa kisasa). Local cows typically produce just 1 or 2 litres per day, whereas modern cows can produce 30 litres (Makoni, 2014: 124–5). Although this total is rarely achieved outside optimum conditions, their higher productivity is valued. There are other differences. Modern cows are more vulnerable to disease and must be kept in cattle sheds, unlike hardy local cows which can be left to pasture. Rungwe has seen a major shift over forty years from local to modern cows. In Sukulu, there are now no local cows, whereas modern cows are almost ubiquitous: 81% of households have a modern cow, and 91% have owned one at some point. Furthermore, because all modern cows originated with development actors outside the country (particularly the INGO Heifer International, to which I return later), tracing their proliferation is ethnographically viable, material evidence of connection to outsiders, and illustrative of how distribution happened.

When asked what benefit they received from cows most people said manure (89%), milk (71%), and offspring (54%). Twenty-two per cent specified milk for drinking, 6% milk for sale, 15% specified for sale in the case of misfortune, 2% for school costs for children. Cattle are particularly valuable because of their circular relationship with the land. As well as milk, cows mean more banana (and other crops) and fertilise their own main foodstuff. Cows remain a source of food and form of money. This can be further interrogated because other important aspects have changed since Schneider’s account. Cattle are never butchered at home for funerals, which are common and lengthy (Marsland, 2015), or weddings, which are uncommon. The same was true of other life-cycle celebrations, although I witnessed more Christian events than ‘Nyakyusa traditions’ known in anthropological literature (Wilson, 1951). For large celebrations, people bought beef from butchers in Tukuyu, which had one abattoir that seemed to be the source of all carcasses. Due to the difficulty of transporting cattle on foot there was no live cattle market in Rungwe, nor any established place for transactions of cattle outside large-scale state-run and private farms, which were outside the district. Instead, people sold cows in two circumstances. If their cow was irrevocably ill (or died) they sold it to the abattoir in Tukuyu, who would collect it. ² People did sometimes sell live and healthy cattle when they had a problem of their own that required a larger amount of money than they could raise by selling chickens or pigs. Many referred to this common hierarchy of livestock, with cows always at the top. If they really had to sell a cow, people would let neighbours and friends know, often specifically approaching wealthy people or asking others to on their behalf, and could expect to receive offers within days or perhaps weeks. Prominent within any notice of sale would be the reason they needed money (such as a medical emergency or a desire to expand a house). Demonstration of need was required for prospective buyers, because people said either the cow or the person selling it had a problem. Decentralised sale made it hard to gather information in a systematic way about the prices of cattle sold, but these varied according to the need, the relationship between the buyer and seller and (in some cases) the wealth of the buyer. There was limited assessment by buyers of the productivity of the cow through its lineage or past output, and even less of price relative to supply and demand of cattle elsewhere in Rungwe, Tanzania, or internationally. Even where the distribution of cattle has become more ‘market-based’ than Schneider (1981) found, such negotiations for exchange focused on the respective needs and capacities of both parties to inform an acceptable price. Furthermore, in Sukulu only 34% of cows had been purchased. Combined with 28% born to farmers’ own herds, and a small number inherited, this left over one third of cows that were procured through alternative means.

One of these was bridewealth. All married men calculated it in cattle, with an average of two cows. However, 45% had paid no cows (yet), which is a considerable increase on long-standing options for delayed or reduced payment (Schneider, 1968; Wilson, 1951). A further 17% had paid some bridewealth, while 38% had paid all. Men married before 1980 paid more cows, and were more likely to have handed some or all of those cows over. But even some men in their 70s told me that they had not finished paying bridewealth: one wizened and infirm old man told me ‘Bado!’ (Not yet!) with a mischievous laugh when I asked if he had paid all his cows following a wedding in the 1960s. Despite the fact that non-payment of cattle was common, rising to almost 100% for those married since 2000, all interlocutors insisted that bridewealth (always calculated in cattle) was essential because it showed how much women were valued, and it compensated families for loss of their daughters. Cattle remains the unit of account for bridewealth, even where no cattle actually change hands. But its importance as a mechanism for distribution is considerably less than Schneider (1981) noted. Only 1.8% of cows were obtained as bridewealth.

Instead, another mechanism that Schneider (1981: 216) noted – of cattle loans – has become enormously significant in Rungwe due to the growing supply of cows. In Sukulu this is boosted further by the husbandry techniques required for modern cattle, which are kept in stalls and fed on a diet of specially planted grass supplemented by grain and other bought inputs (in theory) and the offcuts of banana trees (in practice). A farmer with an average sized landholding of 2.18 acres struggles to keep more than three or perhaps four cows because they run out of freely available food that makes cattle farming viable.

Kufufya loans

In Nyakyusa this loan system is called ukufufya . In my household survey, I found over 35% of households were currently in receipt of cattle through kufufya , totalling 122 cows. Additionally, 28% had received their first cow through kufufya . A further 22% of the surveyed households were engaged in kufufya by providing 84 cows. This suggests that up to 77% of households had engaged in this practice, including over half of households currently involved in a kufufya loan.

The terms of ukufufya are not fixed. Normally, someone borrows a cow and they keep the first calf, then the owner of the adult cow receives the second calf. Upon the birth of the second calf, the adult cow is either returned, or a new loan is agreed. There are several possible exceptions. If the would-be borrower is particularly desperate to borrow a cow, he may have to accept receiving the second calf only. This is usually, but not always, negotiated before the loan. It is also possible to provide bulls on loan. Either the bull is reared and, as an adult, swapped for a female, or the borrower fattens up the bull, sells it for meat, and then receives a smaller share of the profits than the owner. The owner can recall his cow if they are somehow afflicted, for example, by the death of another cow, an emergency, or other family expense. Many interlocutors often said that owners would claim to have a difficulty or need, if they noticed that a cow out on loan produced a lot of milk. The borrower would have little power to stop this under any pretext.

The actual operation of kufufya is highly dependent on hierarchical relationships between borrower and lender. Exploring how they work in practice illuminates concrete examples of distributive labour. My friend Mwakambako provided one account of a lifetime of kufufya loans he received between 1999 and 2019. As a young man of 23, living alone in a house he built at one corner of his parent’s plot of land, Mwakambako helped a rich neighbour called Isaka to transport two cows by foot to neighbouring Kyela, well over a day’s walk away. Following this Isaka, who had 12 cows of his own, agreed to loan one to Mwakambako. After the return of this cow, Mwakambako secured a loan from a friend who attended the same church called Godian, in 2003. During the same year, Mwakambako received another loan from another wealthy neighbour, Mwaitege. As a boy, Mwakambako had worked for Mwaitege, helping him to operate a profitable maize flour-milling machine. He felt this personal history and prior trust was important to securing the loan. Altogether, by 2009 Mwakambako owned three cows (having paid three further calves to their owners, and returned the adult cows). One born from Isaka’s cow before 2003, one born from Godian’s cow in 2004, and one calf born to Mwaitege’s cow in 2009. In 2010 he sold two cows to buy farmland. Then in 2011 he sold one cow to build additional space to house his growing family. He then struggled to access cattle for several years until he was able to obtain a loan from the pastor of his church in 2015. This pastor, who was considered wealthy and with access to a greater range of resources beyond the village through the church, was a friend that Mwakambako cultivated through making himself useful as a helper, messenger, and church elder. He received a final loan in 2016 from a wealthy ‘brother’ of his wife. Although he refers to him as brother-in-law (shemeji), he was actually a childhood neighbour of his wife. This ‘brother’ is extremely wealthy, and Mwakambako thinks he may be allowed to keep all the offspring of the cow, who has given birth once in 2018. Mwakambako currently has four cows, including a pregnant one. He borrowed his first cow as a young and relatively poor man. He is now older (with a wife and children) and relatively wealthier; thanks, he says, in large part to the cows he has borrowed. His distributive labour has never involved broader expectations of ongoing political loyalty associated with patron–client relations in older ethnographies of Rungwe (Hekken and Velzen, 1972). However, he still works hard to maintain relations with benefactors, frequently taking me to see his fictive kinsman, and encouraging me to attend the Moravian Church, for example.

Mwakipole, a farmer in his early 60s and a neighbour of Mwakambako’s, helps illustrate the motivation described by Schneider (1981) to loan out cattle through kufufya . Mwakipole had loaned out four of his six cows. This was not because he was extremely wealthy. He explained that he loaned them out because his own farm is barely big enough to feed two cows, and so he regularly cuts banana leaves from his neighbours’ trees. Any new additions to the herd will have to be loaned out. This is better than selling cows because loaned cattle remain his property. He can sell them in case of emergencies. They also meet the ‘illiquidity preference’, whereby people value assets that can be taken properly out of circulation and on which others cannot make permanent claims (Shipton, 1990: 16–19). Mwakipole knows the potential profits of successfully breeding cows, having bred one cow to a total of more than 10 cows owned (including those sold) over 20 years. For Mwakipole and the 22% of farmers in Sukulu who have cows out on loan, the prevalence of ukufufya today is driven by a combination of factors. These include the productive properties of cows and physiological requirements for proper care, but also an economic preference for illiquid assets.

The experiences and opinions of Mwakambako and Mwakipole are not universal. Two things are important about the pattern of ukufufya in the area in which they lived. Although several wealthy men did provide kufufya loans, the most common source of cows was not from ‘big men’ at all, but from widows. Three widows in an area of 61 houses around my host’s home were providers of 35% of kufufya loans. They inherited cows when their husbands died. This was surprising because widow’s inheritance rights through customary law, as elsewhere in Tanzania, are extremely limited. Many women are forced to leave their homes and abandon their property when their dead husband’s families claim family land (Ibrahim, 2017). In the cases I identified, however, the women were widowed at an old age, when their husband’s (male) siblings were already dead, and when their own children were fully grown. In two cases these children were influential and therefore able to protect their mothers’ interests. The three widows, aged 60-plus, had a total of 17 cows loaned to 14 different people. But they kept no cows themselves. Those they lent cows to conspicuously helped them with small farming or household tasks, but these widows could call on any young, fit, male neighbours for such support. Therefore the extent to which those who borrowed from them performed exceptional or additional service was questionable.

The most noticeable thing about borrowers of these and other cattle was the number of ‘big’ recipients. By this I mean older men, church elders, and others with village government or voluntary positions connected to outsiders. They were already wealthy or respected individuals: one owned a motorcycle and all had more than the average amount of land and/or livestock. All but one of these big men in the area received cattle through ukufufya . This sole exception did not enjoy keeping cattle. He owned one cow and loaned it out because he did not want the trouble of tending to it. At least two big men were both the recipients of cattle through ukufufya and themselves lenders, of the same or different cattle. Several others who borrowed cows already approached the maximum they could feed from their own farms in the way described earlier by Mwakipole. They were typically the most positive about the institution of ukufufya of all my interlocutors. Younger and poorer farmers thought cattle lenders were able to exploit them (kunonya, literally to suckle) because they could easily recall cow once it got pregnant or if it produced a lot of milk. ‘Big’ borrowers were less vulnerable to the whims of those who lent them cows because they could mobilise against any attempt to take productive cattle from them unfairly thanks to their age, wealth, and broad networks. Their social standing meant that it would more difficult to treat them unfairly. Multiple existing and overlapping hierarchies were often reinforced by ukufufya , and so its transformative potential was limited. This problematises the notion that shared frameworks for the transfer of resources are important to distributive labour, as Ferguson and Scherz suggest. Furthermore, borrowers often regarded distance between themselves and the lender, in physical, kinship, and wealth terms, as positive. Mwakatundu felt his ‘brother-in-law’ was less likely to ask for his cow back because he was so rich. Others suggested that a lender who lived far away from them could not observe their cattle and therefore they would not suffer if they kept the cow well and it produced a lot of milk.

Distributing ‘modern’ cows

This leads to a consideration of the distant ultimate source of all modern cows in Sukulu, and exploration of distributive labour across development relations. Because people in Rungwe were barred from owning dairy cattle by the German and then British authorities before independence, it is possible to trace their origins with clarity and describe mechanisms for their dispersal in detail. The importation of dairy cows for the benefit of Tanzanian farmers began in 1973 as part of a World Bank and USAID project to stimulate commercial dairying in the country, with over 1000 cows imported by a US INGO Heifer International. These were farmed collectively at Kitulo Farm in present-day Njombe region to provide milk for Tanzania’s growing urban population and used for distribution to ujamaa villages, at the heart of President Nyerere’s vision of ‘African socialism’. Collective cattle farming failed dismally (Murnyak and Kinsey, 2006: 94). In 1978, Heifer piloted an alternative method for promoting dairy farming, distributing heifers directly to low-income farmers that soon spread across the country. It is known internationally as heifer-in-trust (HIT).

By the early 1990s, HIT was active across Rungwe District. Heifer worked with the Moravian Church and established several farmer cooperatives to help distribute cattle and organise the sale of milk. Heifer paid for the secondment of a Project Supervisor, Marambo, an officer from the District Livestock Office, to oversee HIT in the area. As well as working at the Moravian Church headquarters, he forged links with the local Pentecostal hierarchy. Churches, as well as the government, were essential to the way that HIT worked across Rungwe.

Three HIT groups were started in Sukulu between the 1990s and my fieldwork. The first was started in 2005 through the Pentecostal Church. The second was started in 2008 through the district government. It was based on HIT, and used cattle originating in Kitulo. It was linked materially to Heifer through these cows. The third group was started in 2013 by a smaller NGO that used the HIT model to help families caring for orphans, largely those whose parents died of HIV/AIDS. They had no formal connection to Heifer, although it is likely that many of their cows did have a connection to the Heifer’s own cows because they were bought in-country rather than importing them. They were also clearly inspired by Heifer’s approach. Marambo was responsible for oversight and training of the first and second group, but uninvolved in the third group. The groups worked in slightly different ways, as I explore below. The first and second began with 10 female cows plus one bull. The third group began with 16 female cows, and one male.

The first group began when the retired Pentecostal pastor in Sukulu, Mzee, approached one of his colleagues in a neighbouring village who was on the Pentecostal Church’s HIT committee in charge of distributing cows. Mzee had heard the cows brought big benefits (faida kubwa) and he was able to use his strong relationship with Pastor Halisi to persuade him to petition the Pentecostal HIT committee to start a group in Sukulu. Several subsequent members of the group referred to Mzee and Halisi by name as the source of cows in the village. However, Halisi told me: ‘I simply carried out my responsibilities (majukumu yangu)’ by putting Mzee’s request to the committee. He said he would have done exactly the same for a Pentecostal representative from any village that had not already received cows. He further suggested that, although he knew Mzee, they did not have a special relationship beyond being fellow pastors of the same church. Mzee’s representation of his role as distributive labour was accepted by those who received cows through him, but not by the man from whom he secured cows. The second group was instigated by Marambo and by the village’s livestock extension officer, rather than anyone who lived in the village and wanted to receive a cow. However, the founding chair of the group had worked in a voluntary capacity helping the livestock officer to carry out his duties in Sukulu for some time. The importance of this relationship is suggested by the fact that the chair resigned his position after a short time when members of the group planned to change the way the group worked in defiance of the livestock extension officer and Marambo. The third group was also instigated by outsiders, the Tanzanian staff of the NGO, who approached the village government to select beneficiaries of cows as I discuss below.

The three groups worked in similar ways. In each case, members were required to attend a small number of seminars about cattle husbandry. They committed to build cattle sheds to a similar design, and not let cows graze openly. They also committed to plant grass to feed their cow via cut-and-carry methods. Despite the fact all group members signed up to these terms I only found two cattle sheds built to the approved designs. No one planted a sufficient amount of grass to feed a single cow. Many explained to me that this was because the requirements were not feasible. Some built stipulated cattle sheds in order to obtain cows but let them fall into disrepair before replacing them with smaller and cheaper sheds. Several never built a good cattle shed (banda bora), despite the requirement, because of the cost. In recognition of this, Group Three provided members with one bag of cement, whereas neither of the other groups provided any resources beyond the cow. In fact, members of the first two groups had to pay. Group One members paid TSH 5000 to join the group, and then TSH 20,000 for the transport of their cow. Group Two members paid TSH 25,000. All group members also promised to ‘pass on’ the gift of the cow. This worked in different ways. Within the first and second groups, recipients committed to return the first calf to the group to pass on to a new member, and the second calf to the organisation behind the group (the Pentecostal Church, and the government livestock office). I discuss the operation of this below, but note first that the second group quickly decided after receiving their cows that they would pass on only one calf, to a new group member, and not return a cow to the government. The third group differed by design. Members had to return the first and third calves to the NGO responsible for the project, but could keep the second and subsequent cows themselves. Several noted that being allowed to keep the second calf rather than ‘pass on’ the first two calves was a significant improvement on earlier HIT groups.

Different practices may have reflected the selection criteria for inclusion within the three groups. The first group was connected to the Pentecostal Church and had no formal selection criteria, although it was connected to Heifer International who aimed to benefit poorer members of the community and, in particular, women as both financially marginalised and responsible for household expenses and welfare (Sumberg and Lankoandé, 2011). Marambo simply came to the church and told people they could join and would receive cattle. In practice, the first group’s members were senior figures within that church. They included the former pastor, the wife of the current pastor, and at least two church elders, as well as three members of a prominent family within the church. Mzee was seen as the instigator of the group and two of his neighbours also received cows through the group. The second group was formed by the government and had a chairman who had served for a substantial amount of time as a voluntary helper for the village livestock extension officer. Several members of his family, and neighbours, joined initially and later. Members noted to me that the existence of the group was not widely publicised and it was through friendships with existing members that new members found out about it. In addition, the selection criterion they emphasised was expertise (utaaram) in cattle husbandry rather than poverty or need, which meant that many older, relatively wealthy, men benefited. In contrast to the other groups, the third had strict and clear eligibility criteria. People could not request to receive a cow or join the group. Instead, recipients were selected by a committee who were appointed through an election organised by the village government. Each sub-village had one member, and between them they selected the 17 most deserving people in the village. This was largely according to the rationale of the NGO supplying funds for this initiative. Their mission was to support orphans (watoto yatima) and the families of orphans, who were often grandparents. Because there were many more families looking after orphans than cows available, they had to use their own judgement as to whose need was greatest. There was some evidence that those with the most need were not considered because the committee felt they would not be able to build a cattle shed or to look after cows properly. But most people in the village felt that the committee did a good job. The chair of the committee herself told me: ‘We received more cows after the first 12 because we did a good job of spreading the cows across the whole village.’

Outsiders’ categories of need were important to all groups, but mediated by local understanding. Women were meant to be priority target beneficiaries for all three groups. I found that 40% of the named beneficiaries of cattle from the groups were women. However, it was not that simple. All three groups had named female beneficiaries of calves who had no ownership rights, decision-making power, or control over income, whether from milk or selling a cow. In contrast to smaller livestock, neither did women care for cattle except in exceptional circumstances. I asked the Pentecostal pastor and his wife if she had ever milked the cow she received from the first cattle group in 2005 (along with his cow). They both laughed, and he replied simply, ‘Mimi ni baba’ (I am the man [literally: father]). She had never milked the cow, and did not know how. Only one women in a male-headed household told me that she did sometimes milk those cows, and even then I never saw it.

Representations of need, such as women’s economic empowerment, are important to the distribution of cattle through HIT. But this does not mean they benefit most women who are named beneficiaries, at least directly. The ability to broker between those who have resources and others who want them by managing different ethical rationalities and micro-political realities is part of distributive labour. Intersecting with this is the cultivation of key relationships. This includes membership of church, volunteering with government officials, and the expected duties of a junior family member to a senior. These relations cannot be explained as a means to an end because the material advantage of membership of the Pentecostal Church, for example, cannot be disentangled from its adherents’ beliefs, nor is the care provided by a grandparent for orphaned grandchildren undertaken simply to receive cattle. But such roles bring material and immaterial rewards.

Many people noted that leaders of groups benefited in various ways, including from the sale of cattle. An almost complete lack of record-keeping meant that no audit of any group was possible. But people expected that leaders who brought benefits should be able to profit from their activities themselves. The truly immoral act was ‘eating the money’ by not distributing resources. The one outright criticism of leaders I heard was in a neighbouring village, where one person suggested that representations of success had been too great so that those providing HIT projects had decided that they no longer needed to distribute cows although many people still lacked them.

In Sukulu I did not see this happen, and HIT continues to this day. I found strong evidence of 95 people who directly benefited from cows through this approach in the last 15 years, and anecdotal evidence of up to 125 people receiving cows. In a village of around 1100 households this is a significant proportion, but far less than the number I have shown received cows through kufufya .

Comparison allows me to reflect on the pragmatics of distributive labour in Sukulu, before turning in my concluding section to a discussion of Ferguson’s suggestion that distributive labour at a global scale could build consensus on how to share wealth. The proliferation of modern cows was widely seen as a good thing in Sukulu but had important limits. HIT was always seen as preferable to ukufufya. This was because when cattle were obtained through HIT this was considered a permanent transfer, not a temporary loan. Despite this, kufufya was much more significant in terms of the number of cows distributed, and it was much easier to secure a cow through kufufya than through HIT. The problem was that kufufya did not easily transform lives. Even HIT often did not help the most vulnerable. Distributive labour on its own was not a viable strategy for survival and only made sense (like cattle-keeping) within a multiplicity of livelihood strategies. It clearly did not achieve the full redistributive aims of all who undertook it, but many felt it was worthwhile anyway.

The distributive labour of dairy development

Many donors overestimate the effects of donations, especially when they concern the provision of training rather than the transfer of assets, such as cows (Swidler and Watkins, 2017). INGOs like Heifer International have sophisticated fundraising operations that create representations to persuade people to donate funds. Meanwhile people in Sukulu debate those to whom they might connect. For example, at a time of competition between two competing dairy development projects during my fieldwork, I witnessed arguments in drinking clubs about the relative wealth of benefactors (wafadhili), invoking several of them by name:

A: Bill Gates is the richest person in the world. He cannot fail to pay.

B: But he is an American. He supports Mradi [a dairy project]. They do not have so much money. Booths [connected to a wealthy Tanzanian individual] he has the money. He has wealth [uwezo: literally, capacity)].

Discussions such as these were common. Connections to benefactors, including relations with multiple mediators to the source of funds, were valued and clearly analysed by potential beneficiaries. Many people were highly critical of dairy development projects subsequent to HIT that did not give them cattle but offered only loans (alongside training). ‘Loans are not capital (mtaji), cows are capital,’ one young female household-head told me in frustration when she learned that a project she had joined, and walked more than 10 miles to visit, would not offer cattle through an HIT scheme. She did not reject out of hand such schemes that thwarted her preferences, but remained cautious and sceptical of claims they made. She showed no loyalty when asked to commit to the project with no clear route to material reward.

Even local elites who support ‘neoliberal’ programmes of assistance based on loans and training to increase productivity, accept and perpetuate rationalities for distribution that are at odds with such programmes. One ward councillor told me his own version of Ferguson’s favoured development cliché: ‘The Chinese say: “Give a man a fish and you feed him for a day, Give him a fishing rod and you feed him for many days.”’

The adaption from the better known version cited by Ferguson (that offers training in fishing rather than a fish) points to the need for distribution of material resources for legitimacy in Rungwe. Such ideas about the proper allocation of resources are important to those who instigate development projects there. This may seem like a recapitulation of Ferguson’s argument, but there is a vital difference. People in Sukulu debate the capacity of Bill Gates or other rich people to pay, as well as the willingness of mediators between these rich people and themselves to pass on transfers. They want material resources, proverbial fishing rods and real dairy cows, not loans. But such preferences have no legitimacy as the basis of claims in the minds of Bill Gates, his presumed competitor at Booths Dairies, who bemoaned to me that cattle-keepers in Rungwe always wanted handouts, or international NGOs (Murnyak and Kinsey, 2006). In the case of Heifer International, development relations are real and manifest in the material form of the dairy cow spread across the village that originated with the NGO. However, these relationships rely upon almost complete lack of knowledge of one another between different parties. This is detachment in the sense proposed by Yarrow et al. (2015): not an absence of a relationship but a relationship made possible through disengagement, distance, and disconnect in rationalities. Representations made by Heifer of their beneficiaries lives and effects of their interventions necessarily obviated dependence as a mode of action for the latter because it would be unacceptable to potential donors (Murnyak and Kinsey, 2006). Meanwhile, beneficiaries were unaware of their benefactors’ preferences (and limitations). For this reason they often found it hard to evaluate the relationship between cause and effect of their efforts to secure resources. They were neither particularly hopeful nor despondent about the limited opportunities afforded by distributive labour, but patient, determined, and pragmatic about trying to secure material resources.

Crucially, at an international scale, the distributive labour of people in Sukulu becomes somewhat unstuck from rights-in-persons. Potential beneficiaries of dairy development cannot successfully make claims on those who ultimately fund it and do not ‘belong’ in relationships with them (Miers and Kopytoff, 1977: 17). In contrast to Ferguson’s conceptualisation of cash transfers, dairy development does not work because it achieves a measure of solidarity between worlds about transfers between them. It works through the relational and representational mediation of actors operating within existing power structures at multiple scales, such as Heifer International, Pentecostal and other churches, Tanzanian villages, financial elites within the country and out of it, and the changing ‘development apparatus’. Their ideas about legitimate distribution of resources do not converge and are in fact largely invisible to one another. The process keeps the worlds of beneficiaries and donors apart in order to facilitate transfers between them. This vitiates a sense that distributive labour helps different and distant people value one another or establishes a moral order encompassing all. Distributive labour is important because the claim-making rationalities and activities of ‘beneficiaries’ impact on the fate of development projects and social welfare programmes. However, its limited rewards leave little room for optimism about a transformative new politics of distribution that persuades powerful global actors to accept others’ claims on their resources.