The socio-cultural dynamics of mobile money adoption

Kusimba's anthropological analysis of MMT in family and social networks among the Bukusu of Western Kenya shows MMT's embeddedness within circuits of value among siblings, cousins and mothers, and its use for a wide range of purposes and for the expression of affective ties. This leads her to argue that MMT is an ‘amplifier of enduring practices around reciprocity, obligation, belonging and exclusion’ (Kusimba et al., Reference Kusimba, Yang and Chawla2015: 12; Reference Kusimba, Yang and Chawla2016). Both of her studies strongly resonate with Shipton's focus on cultural norms and values as key to understanding Luo practices of interpersonal resource exchange, demonstrating that a culturally defined relationality of finance is its critical feature (Shipton Reference Shipton2007).

Maurer, on the other hand, presents the emerging mobile money payments space as not just the product of these ‘on the ground’ practices but of ‘new socialities (that) are remaking money’ (Maurer Reference Maurer2012: 591). Using actor-network theory, Maurer draws on his engagement with the mobile money industry, to consider how mobile money is being created and formatted by a range of actors and the networked relations between them. He explains how airtime is at one moment a commodity, at another a technologically mediated relationship as talk or text, and at another an alternative currency means of exchange or store of value such that ‘it can continuously pass into and out of each of these moments – it is unstable and reversible’ (Maurer Reference Maurer2012: 591). He argues that the ‘payments space’ is being actively configured by technological and human actors in a way that is neither a solely ‘top down affair (emanating from corporate and donor agencies) or a bottom up affair (emanating from people's practices ‘on the ground’)’ but rather that this distinction – along with analytical separations of producer and consumer, and person and commodity – are confounded and ‘relationality itself plays a significant role in the creation of this new social space’ (Maurer Reference Maurer2012: 591). Maurer's analysis suggests a mode of relationality that overcomes the disjunctures identified by Guyer and which has the potential to usher African monetary and financial systems into a new phase of development, while apparently thriving on a fundamental level of open-endedness and indeterminacy of their operation.

This concern with relationality appears to be central to a deeper understanding of the socio-cultural dynamics of MMT adoption which have been identified as affecting variation in adoption rates across the continent. It offers a route to exploring the ways in which this latest wave of technological change may be provoking forms of market engagement which challenge the disjunctures of previous financial sector development and, in turn, fuel the aspirations and efforts of global policymakers for financial inclusion.

Exploring relationality through repertoires and cultural values

This paper takes as its starting point the financial practices of everyday resource exchange which MMT appears to amplify, and investigates their less conscious – conceptual – ‘fertile ground’ (Heyer & Mas Reference Heyer and Mas2011) as providing the enabling environment for MMT adoption. In doing so, it highlights the role of more enduring elements of financial practices, such as cultural values. This contrasts with Maurer's starting point of the mobile money sector itself and Kusimba's focus on the social networks involved, instead building on the contributions of Guyer and Shipton by examining the implicit socio-cultural concepts that people relate to when they engage in everyday practices of resource exchange.

We use the concept of repertoires developed by Guyer (Reference Guyer2004) which captures the way that patterns of transactional behaviour are embedded in various social domains and flexibly combined into transaction resolutions. However, while she suggests these repertoires can belong to less conscious levels of cognition so enabling a sense of normality and continuity to be maintained in the face of new circumstances, she says little about how these can be explored. We therefore complement the concept of repertoires with the concept of cultural values, which are characteristically enduring and normative (and therefore normal) among members of the same ethnic group. Drawn from applied ethnolinguistics, the concept enables the exploration of central cultural values through analysis of everyday language use. Combining these disciplinary approaches offers a sound theoretical perspective on both continuity and change in interpersonal financial practices.Footnote ⁵ Specifically, we explore predominant domains of resource exchange in three ethnic groups through a narrative analysis, and additionally for one through the lens of everyday language use. Together this reveals what meanings and values people consciously and unconsciously attribute to resource exchanges and how relationality is defined through them.

We argue that there are two divergent repertoires: one is relationship-focused; the other resource-focused. The identification of the resource-focused repertoire provides a counterfoil to the relationship-focused repertoire, enabling us to more fully grasp the set of valuations that people implicitly make when prioritising the relational repertoire. Indeed, we find that even the resource-focused transactions of the formal sector are understood within this relational repertoire when they sufficiently conform to its characteristics. We conclude that discontinuities between local and institutionally mediated forms of finance continue to be present. This also indicates potential boundaries to adoption of MMT across socio-cultural contexts. It suggests that where financial repertoires have stronger relational dynamics, they are likely to offer more ‘fertile grounds’ (Heyer & Mas Reference Heyer and Mas2011) for adoption, while also suggesting that attempts to configure the financial services space that are not strongly resonant with these repertoires and which seek to more specifically delineate the transactions involved are likely to continue to face more limited uptake. We recognise, however, that this approach opens up further dimensions of MMT adoption that may be explored through this socio-cognitive lens, such as the identity associated with the mobile network operator (Safaricom) through both its institutional history, branding and marketing; the acceptance of mobile phone technology as communicative devices and of ‘virtual’ money conversions.

Research methodology and data analysis

The analysis is based on two distinct data sets collected in 2012. The first study used in-depth interviews to investigate the nature, purpose and relationships within which interpersonal resource transactions – whether via MMT or not – are carried out, along with the valuations attributed to these transactions. This opens a potentially vast terrain of transactions extending from dependence on family members for daily survival to occasional interactions with strangers; and from those that are more clearly symbolic to more commercial and instrumental ones. The focus was on how resource exchange – not just financial exchange – was important and meaningful in the context of a variety of relationships, while capturing a variety of types of giving/lendingFootnote ⁷ and receiving/borrowing. Interviews in Swahili asked respondents about the support that they had found ‘significant or important’ to them in the last year, although some responded with older examples. The term usaidizi was used, which means ‘help’ or ‘support’ and refers to assistance received from others in the broadest sense. This was then probed for the relationships involved and the history of interaction and support exchanged with that person. In this context, significant and important did not necessarily mean the scale of support, but could also refer to that which was most meaningful.

Research sites were three towns and their environs, representing ethnic diversity and terciles of Kenya's district poverty rankings (Kenya National Bureau of Statistics 2007). Mathira (top tercile) on the slopes of Mt Kenya, is ethnically Kikuyu; Nyamira (second tercile) is ethnically Gusii; and Kitui (bottom tercile) is ethnically Kamba. Town-based livelihoods involved business as well as employment, while rural on-farm livelihoods, including tea and coffee production in Mathira and Nyamira, were supported by a range of off-farm income generation activities as well as temporary and more permanent migration. A total of 62 interviews were undertaken with individuals in 40 households purposefully chosen from respondents to a random survey two years earlier and reflecting a variety of household poverty levels and financial service use. The earlier study (Johnson Reference Johnson2016) highlighted the extensive role of inter-personal resource exchange and, along with related research (Zollmann Reference Zollmann2014), describes the context and provides in-depth analysis of the variety of financial practices and service use, including banks and informal groups. The interviews were recorded, transcribed, then coded and analysed using an inductive and interpretive approach. This was combined with analysis to highlight the narrative elements that individual respondents bring in to describe their experiences and hence a layer of signification which includes the non-rational – e.g. the normative, affective and less conscious valuations.

The second study focuses specifically on language and ethnographic data. It explores the meanings of words and phrases as well as non-verbal signs in the everyday domain of informal interpersonal resource exchanges among the Kamba of Kitui, to draw out cultural values underpinning these practices (Wierzbicka Reference Wierzbicka1997; Astbury & Leeuw Reference Astbury and Leeuw2010; Peeters Reference Peeters, Peeters, Mullan and Béal2013). Due to the study being time consuming, it was necessary to restrict it to one geographic area but with people in all three sites speaking a Bantu language, lexical similarities seem likely. Kitui area was chosen out of the three sites of the first study for its social characteristics: poverty, food insecurity and (rurally) minimal access to services. These were expected to be conducive to a particularly variegated network of informal interpersonal exchanges and less affected by terminologies introduced by the presence of formal financial institutions.

First, a list of Kamba words in the everyday domain of ‘borrowing’/‘lending’ and ‘saving’Footnote ⁸ was produced using a Kamba dictionary (Mũtĩsya & Ndũnda Reference Mũtĩsya and Ndũnda2003) and complemented and adapted by informants. Sixty informants varying in age, gender and economic status contributed explanations of the words and indicated when, where and how the word is used as well as the valuations attributed to it (e.g. degree of importance, negative or positive associations; emotional echoes). Three assistants translated English questions into Kamba and informants’ responses back into English. Since some words appeared to be more relevant in the domains of ‘borrowing’/‘lending’ and ‘saving’ than others, a shortlist of the most salient words and their meanings – plus a description of their use and attributed valuations – resulted. Ethnographic data involved interviews eliciting informants’ personal histories in the Kamba domains of ‘borrowing’/‘lending’ and ‘saving’, complemented by observations of ‘resource exchange’, role-plays, informal discussions and observations of financial group meetings which were also transcribed and translated. Back translation into Kamba was used to elicit relevant socio-cultural information that was lost in translation. Video recordings were made of all fieldwork activities, thus enabling an analysis of non-verbal in addition to verbal communications. Finally, participation in the everyday lives of three different households enabled observation of local practices of ‘lending’/‘borrowing’ and ‘saving’ and the values and norms of behaviour embedded in them.

Everyday resource exchange

Frequent exchanges among neighbours involved food, essential household items and small amounts of money – most often between women and including between in-laws as a result of patrilocal residence patterns. These everyday exchanges occurred in the more food secure areas as well as in the particularly food insecure area of Kitui. Due to physical proximity, this domain is one in which MMT is rarely used.

To express a request for this type of resource exchange – with the implicit agreement that the borrowed item or its equivalent is returned – a Kamba person from Kitui will say ‘Ngovethye’ (lend me) (from the verb kũkovethya (lend), which is derived from the verb kũkova (borrow)). These verbs generally refer to ‘borrowing’/‘lending’ without interest in relationships of neighbours, in-laws, (business) friends and shopkeepers. However, the lack of interest does not imply a non-committal arrangement. Either there is a tacit shared understanding based on earlier borrowing experiences, or an explicitly expressed agreement. Although a willingness to help starts, and may guide, the resource exchange interaction on the part of the lender, the agreement – which puts the resource at the centre of the interaction – defines the interaction as resource focused. Socio-emotive valuations are not entailed in Kamba understandings of a kũkova (borrow)/kũkovethya (lend) arrangement, although in practice it may give rise to different emotions in lender and/or borrower.

Resource exchange with expression of need

The second exchange domain involves expression of need. For example, food was requested in the hungry season to feed young children, and respondents reported that they gave some even where they only had a kilo of maize themselves. Where hunger led to requests for money from more distantly located relatives to be sent via MMT and with an offer to repay, there were instances where this repayment was strongly rejected. However, such support from relatives was not taken for granted and appreciation was frequently expressed when it was received, so confirming the relationship.

Kamba vocabulary data shows that this type of exchange is perceived to be a different conceptual category from that above. Here the relationship-focused verbs kũvoya (pray, ask for) and kũtetheesya (assist) are used. The person generally introduces the request by saying: ‘Ni thina’ (I have a problem) – used in the case of a not so intimate relationship – or ‘Nenge’ (imperative, give) – in the case of friends or relatives. Kamba people indicated this as their most common resource exchange practice. The verbs seem to indicate that there is no expectation of a return of the item or its equivalent. The reality is that kũvoya and kũtetheesya (‘ask for and assist’) create an ambiguous situation regarding the expectation of return. In any case, it is never a ‘pure’ gift, rather it is part of a wider system of exchanges, in which the ‘entrustment’ of resources produces an obligation on the part of the receiver.Footnote ⁹ With this indeterminacy, it seems likely that the accompanying repertoire includes values that are more widely distributed in Kamba society. One such value is ‘upliftment’ (m ũkĩlye). Indeed, uplifting and being uplifted materially or immaterially –for instance in case of illness, shortage of money or food, or emotional hardship – are an integral part of the understanding of being human among the Kamba.Footnote ¹⁰ Both the noun m ũkĩlye (the uplifting) and the verb kũkĩlya (take across, lift up) seem to resonate with traditional Christian and African images of support (cf. the biblical good Samaritan; the African stronger person taking a weaker person on his back). Indeed, informants explained the term mũkĩliye by lifting their arms as if they were raising something in the air. They argued that the one who is being lifted has the experience of lack of burden and hence of ‘lightness’ or ‘happiness’. Moreover, this value was found to be related to other Kamba social values such as interdependence, belonging and concerted action which add persuasiveness to ‘upliftment’. This comes out particularly clearly in the context of Kamba informal financial groups discussed below.

With the indeterminacy of the ‘ask for and assist’ arrangement, adherence to the above values enables two people to maintain their good (i.e. inspired by these values) relations. These will be strengthened or weakened depending on the evolving interaction. While the situation enables the expression of additional moral or affective qualities, such as closeness, generosity and obligation, it may equally put the relationship under strain if the norms of behaviour that go with these values are breached by one or the other's actions.

Resource exchange for education

A third domain is support for children's educational expenses which is frequently received from siblings or cousins (of the parent).Footnote ¹¹ This ranged from one-off gifts to ongoing support throughout the child's education, especially where children have lost parents. The reasons given for school fees and related assistance encompassed: need relative to parents’ ability to afford it, the academic potential of the child, and the potential for the child to support the giver or others in the future. Notably when a young person is earning and does not have the burden of school fees, rather than saving for their own future, the norm is to support the education of the children of a sibling – frequently through MMT when the older sibling has migrated. Considered from an economic perspective, the potential returns are likely to be much greater than saving in a bank account – especially if that child secures a good job and is then able to reciprocate support or connections in future.

However, reciprocation was not necessarily expected to be to the same person. Support of nephews, nieces or other children with school fees was explained as ‘they will help us or somebody else’ in future (Z. 2012, Int.). This response again underlines the characteristic lack of specificity of the reciprocity involved. Whether the support will be reciprocated depends on how the child succeeds in education and in developing their livelihood. Nevertheless, respondents who had helped relatives with school fees in the past, reported that they were now able to appeal to those they had helped for support for their own children's education. A woman in Nyamira had supported her nephews who now had good jobs in Nairobi – one a doctor. Needing support for her own son both for school fees and transport, she received it from them via MMT, arguing ‘now is the time to … get the benefit’ (C. 2012, Int.). Indeed, such a reciprocal dimension of assisting was even used to justify support given to parents by grown-up children, with them referring to the past educational support they had received.

Supporting a child with school fees would be qualified as an act of kũtetheesya (assist) in the Kamba language. While entrustment and obligation are implicit associations, the verb does not evoke a sense of reciprocity. If a Kamba person wants to highlight this aspect of a future benefit, she uses the verb kũsuvia, which means ‘care for’, ‘keep safe’ and in a resource context expresses ‘saving’. Apart from ‘keeping safe’ being understood as an essential element of ‘saving’, the verb kũsuvia typically includes a sense of future benefit. However, it has implicit reference to value maintenance rather than value increase. One informant said that she was very much aware that she had to kũsuvia her son. This meant that she ensured she fed him well so that he would stay healthy and strong, and that she raised school fees for him to complete his secondary education. This would benefit her in the end, she argued, because he would be able to get a job that would pay him enough to also care for her.

Resource exchange in case of emergency

In a fourth domain involving cases of emergency – particularly health – patterns of support were less clear: there were cases where immediate family did not respond to needs for medical expenses and a more distant friend did, with an opportunity to respond referred to as an opportunity to confirm the strength of a relationship. The rationale for assisting can recognise an understanding of idiosyncratic shocks: ‘you know problems are for everyone, today it's you, tomorrow it's me’ (R. 2012, Int.). On the other hand, support was sometimes lent rather than given, even among close family members. In such cases the pressing circumstances can lead to a lack of clarity over the nature of the support as the terms on which it is given may not be explicitly discussed. But if it is agreed as borrowing, then the return date is recognised as needing to respond to the circumstances of the borrower.

This type of situation Kitui Kamba would refer to as one of kũvoya/kũtetheesya (‘ask for and assist’). In practice, they manifest themselves as kũkova/kũkovethya (‘borrow and lend’) situations, which Kamba informants attributed to the growing lack of solidarity between well-off and not so well-off relatives. Yet, even if the arrangement is understood in terms of ‘ask for and assist’, rural informants were hesitant to consider these types of need as falling within the interpersonal domain. In cases of prolonged illness and death where the financial commitment generally exceeds the household potential, they consider one or more collectives – such as a community, clan or financial group – as the general response mechanisms in the form of organised fundraising events (harambee), door-to-door collections and/or allocation of funds by clan or financial groups. The appropriate Kamba term in this context would be kũmbanya (put together; ‘save’), which generally refers to members of a group putting (e.g. money, food, ideas) together for a specific purpose.

Resource exchange as business capital

The final domain is where business people – mostly men – appeared to develop lending relationships with their siblings, cousins and friends who were also in business. For some, these transactions are so frequent that they do not even arrange when funds will be returned, trusting that the person will pay when she or he gets paid, usually in the space of days or weeks and using MMT when people are mobile. Even though funds are usually given interest-free, it is more likely that ‘something on top’ will be given because of the value increase obtained from the business purpose to which they are put. This extra was explained as ‘giving back thanks’, meaning expressing appreciation (in the emotive senseFootnote ¹²). The ‘something on top’ may depend on the scale of the gain and the gratitude of the borrower and is not necessarily small relative to the amount lent. One woman returned a kanga and leso Footnote ¹³ worth about twice the value of borrowed pigeon pea seedlings which had produced a very good harvest.

While the Kitui Kamba would qualify these interpersonal exchanges for business as a kũkova/kũkovethya (borrow/lend) arrangement, entailing a distinct expectation of return of the equivalent, they can refer to ‘something on top’ in two different ways. The two words are ũsyao (produce, interest) and vaita (profit, interest); the first word more common in rural and informal contexts and the second in urban and formal ones.

Ũsyao is embedded in general human experiences of production and derived from the verb kũsyaa, of which the first meaning is ‘to give birth’ (Mũtĩsya & Ndũnda Reference Mũtĩsya and Ndũnda2003). This ‘giving birth’ applies to an array of products, such as the children of a nuclear family, the chicks of a chicken, the fruits of a mango tree. The verb kũsyaa can equally be used in a financial context, referring to any increase of money on top of what is there. This implies that ũsyao does not need to refer to a fixed interest rate, but rather an addition of money on top of what has been borrowed. By contrast, the term vaita (Kiswahili faida (profit)) highlights economic value and is restricted to use in the financial domain.