Mobile banking and economic development: Linking adoption, impact, and use

m-banking

Mobile banking and economic development: Linking adoption, impact, and use
Jonathan Donner, Microsoft Research India
Camilo Andres Tellez, London School of Economics and Political Science

Reference Information:

This is a preprint of an article whose final and definitive form has been published in the Asian
Journal of Communication, © 2008 Asian Media and Communication Center; available online at
informaworldTM at http://www.informaworld.com/openurl?genre=article&issn=0129-
2986&volume=18&issue=4&spage=318


Temporary Citation:
Donner, Jonathan and Tellez, Camilo. (2008). “Mobile banking and economic development:
Linking adoption, impact, and use”, Asian Journal of Communication, 18(4), 318-322.

Abstract
Around the globe, various initiatives use the mobile phone to provide financial services to
those without access to traditional banks. Yet relatively little scholarly research explores the use
of these m-banking/m-payments systems. This paper calls attention to this gap in the research
literature, emphasizing the need for research focusing on the context(s) of m-banking/m-
payments use. Presenting illustrative data from exploratory work with small enterprises in urban
India, it argues that contextual research is a critical input to effective “adoption” or “impact”
research. Further, it suggests that the challenges of linking studies of use to those of adoption and
impact reflect established dynamics within the Information and Communication Technologies
and Development (ICTD) research community. The paper identifies three crosscutting themes
from the broader literature—amplification vs. change, simultaneous causality, and a multi-
dimensional definition of trust—each of which can offer increased theoretical clarity to future
research on m-banking/m-payments systems.
Notes on contributors
Jonathan Donner is a researcher in the Technology for Emerging Markets Group at Microsoft
Research, where he studies the social and economic impacts of mobile communication
technologies in developing countries.
Camilo Tellez is a doctoral student in the department of Information Systems at the London
School of Economics and Political Science, where he specializes in IT for development issues.
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 1

---

m-banking


The spread of mobile phones across the developing world is one of the most remarkable
technology stories of the past decade. Buoyed by prepay cards and inexpensive handsets,
hundreds of millions of first-time telephone owners have made voice calls and text messages part
of their daily lives. However, many of these same new mobile users live in informal and/or cash
economies, without access to financial services that others take for granted. Indeed, across the
developing world, there are probably more people with mobile handsets than with bank accounts
(Porteous, 2006). Various initiatives use mobile phones to provide financial services to “the
unbanked.” These services take a variety of forms—including long-distance remittances,
micropayments, and informal airtime bartering schemes—and go by various names, including
mobile banking, mobile transfers, and mobile payments. Taken together, they are no longer
merely pilots; in the Philippines, South Africa, Kenya, and elsewhere, these services are broadly
available and increasingly popular.
Scholarly research on the adoption and socioeconomic impacts of m-banking/m-
payments systems in the developing world is scarce (Maurer, 2008). Even less attention has been
paid to the social, economic, and cultural contexts surrounding the use of these systems. This
paper‟s goals are threefold: first, it calls attention to this gap in the research literature and
emphasizes the need for research focusing on the context(s) in which m-banking/m-payments
systems are used. Second, it argues that, to the extent it helps reveal the myriad social,
technological, and economic influences on use, this contextual research is not simply a
complement but rather a critical input to effective “adoption” or “impact” research. Finally, the
paper argues that the challenges of generating interdisciplinary dialogue about m-banking in the
developing world are illustrative of long-running dynamics within the community of scholars
and practitioners concerned with Information and Communication Technologies and
Development (ICTD). The m-banking case adds new wrinkles to broader discussions about
technology and development, and about mobiles in society.
M-Banking and M-Payments Systems in the Developing World
The terms m-banking, m-payments, m-transfers, m-payments, and m-finance refer
collectively to a set of applications that enable people to use their mobile telephones to
manipulate their bank accounts, store value in an account linked to their handsets, transfer funds,
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 2

---

m-banking

or even access credit or insurance products. This paper uses the compound term m-banking/m-
payments systems to refer to the most common features.
The first targets for these applications were consumers in the developed world. By
complementing services offered by the banking system, such as checkbooks, ATMs,
voicemail/landline interfaces, smart cards, point-of-sale networks, and internet resources, the
mobile platform offers a convenient additional method for managing money without handling
cash (Karjaluoto, 2002). For users in the developing world, on the other hand, the appeal of these
m-banking/m-payments systems may be less about convenience and more about accessibility and
affordability (Cracknell, 2004; infoDEV, 2006). An exploration is underway—between banks,
mobile operators, hardware and software providers, regulatory agencies, donors, and users—to
determine the shape of m-banking/m-payments services in the developing world (infoDEV,
2006; Ivatury, 2004; Ivatury & Pickens, 2006; Porteous, 2006). Mobile phone operators have
identified m-banking/m-payments systems as a potential service to offer customers, increasing
loyalty while generating fees and messaging charges (infoDEV, 2006). Financial institutions,
which have had difficulty providing profitable services through traditional channels to poor
clients, see m-banking/m-payments as a form of “branchless banking” (Ivatury & Mas, 2008),
which lowers the costs of serving low-income customers. Government regulators see a similar
appeal but are working out the legal implications of the technologies, particularly concerning
security and taxation.
There is no universal form of m-banking; rather, purposes and structures vary from
country to country. The systems offer a variety of financial functions, including micropayments
to merchants, bill-payments to utilities, P2P transfers between individuals, and long-distance
remittances. Currently, different institutional and business models deliver these systems. Some
are offered entirely by banks, others entirely by telecommunications providers, and still others
involve a partnership between a bank and a telecommunications provider (Porteous, 2006).
Regulatory factors, which can vary dramatically from country to country, play a strong role in
determining which services can be delivered via which institutional arrangements (Mortimer-
Schutts, 2007).i
Most m-banking/m-payments systems in the developing world enable users to do three
things: (a) Store value (currency) in an account accessible via the handset. If the user already has
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 3

---

m-banking

a bank account, this is generally a question of linking to a bank account. If the user does not have
an account, then the process creates a bank account for her or creates a pseudo bank account,
held by a third party or the user‟s mobile operator. (b) Convert cash in and out of the stored value
account.ii If the account is linked to a bank account, then users can visit banks to cash-in and
cash-out. In many cases, users can also visit the GSM providers‟ retail stores. In the most
flexible services, a user can visit a corner kiosk or grocery store—perhaps the same one where he
or she purchases airtime—and transact with an independent retailer working as an agent for the
transaction system. (c) Transfer stored value between accounts. Users can generally transfer
funds between accounts linked to two mobile phones, by using a set of SMS messages (or menu
commands) and PIN numbers.
The new services offer a way to move money from place to place and present an
alternative to the payment systems offered by banks, remittance firms, pawn shops, etc.. The
uptake of m-banking/m-payments systems has been particularly strong in the Philippines, where
three million customers use systems offered by mobile operators Smart and Globe (infoDEV,
2006); in South Africa, where 450,000 people use Wizzit (“the bank in your pocket”) (Ivatury &
Pickens, 2006) or one of two other national systems (Porteous, 2007); and in Kenya, where
nearly two million users registered with Safaricom M-Pesa system within a year of its
nationwide rollout (Ivatury & Mas, 2008; Vaughan, 2007).
Current Research
The practitioner community may frame the discussion as being about “Transformational”
M-payments (Gamos, 2008); the popular press describes a “leap from the world of cash to
cellular banking” (The Economist, 2006); and researchers speak about the potential of m-
commerce to “close the digital divide” (Dholakia & Kshetri, 2004). There are a variety of
perspectives from which to view the technology, and as Maurer (2008), illustrates the
assumptions associated with an embrace of an “empowerment” or “market share story” (pp. 8-
9), for example, will impact the claims and research programs of those interested in the
technology.
The current research literature can be classified into three types of studies: (a) those that
explain or predict the adoption of m-banking/m-payments systems; (b) those that assess the
systems‟ impact on people and on economies; and (c) a relative few that try to understand the use
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 4

---

m-banking

of such systems in social, economic, and cultural contexts. Variants of this trichotomy, which
distinguishes adoption studies from impact studies and from “use” studies, have been
documented before (Fischer, 1992; Markus & Robey, 1988; Orlikowski & Iacono, 2001; Sein &
Harindranath, 2004) and are a reflection of the disciplines that take an interest in communication
technologies. Donner (2008) applied the trichotomy in a review of the research literature on
mobile telephony in the developing world.
M-banking/m-payments systems have all the markers of an “innovation” waiting to be
“diffused” to or adopted by a subset of mobile users in the developing world (e.g., Rogers,
1983). Brown, Cajee, Davies, and Stroebel (2003) used a statistical model combining elements
of the theories of diffusion of innovation (Rogers, 1983) and of planned behavior (Taylor &
Todd, 1995) to predict mobile banking take-up in South Africa, finding high levels of perceived
risk to be a major barrier to further adoption. To date, it is one of the few evaluations of an m-
banking/m-payments system in the developing world explicitly applying a theoretical lens. Two
studies from the economic development/practitioner literature (Ivatury & Pickens, 2006;
Porteous, 2007) suggested that mobile banking users in South Africa are wealthier and better
educated than the average South African with a bank account, let alone the average unbanked
South African. Porteous suggests that the profile of the typical m-banking user in South Africa
still resembles that of the “early adopter” (see also, Ivatury & Mas, 2008). Drawing on
representative survey data, Porteous cites mistrust and unawareness among the primary reasons
South Africans might choose not to adopt m-banking.
Studies of the impact of m-banking/m-payments systems in the developing world are also
scarce because the systems are so new. The best impact assessment to date is (Porteous, 2007), in
which impact is operationalized using an “access frontier,” which divides those who have the
wherewithal—a monthly income from a formal source—to open the most basic of conventional
bank accounts. Those below the frontier who use m-banking/m-payments systems do so as an
alternative or addition to other choices. Those from above the frontier have done so by necessity.
Porteous concludes that the “transformational” impact of m-banking/m-payments services in
South Africa has been small (so far) because virtually all of the users are from below the frontier.
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 5

---

m-banking

Studies on Use (would be useful)
Additional adoption and impact studies are sure to follow, but the research community
should also pursue studies of the context and use of m-banking/m-payments systems in the
developing world. This section presents three important examples of non-technical (social and
economic) contextual factors: comfort with electronic money, the availability of alternatives, and
the social context of transactions. Each influences the dynamics of m-banking/m-payments‟
adoption and impact, currently unfolding around the world.
Conceptualizing Electronic Money
Even the simplest handsets have features buried deep in menu structures. If navigating an
m-banking/m-payments interface is difficult for experienced mobile users with bank accounts,
even greater is the difficulty for first-time users in the developing world, many of whom will
have only been using a mobile for a year or two (Cracknell, 2004; Peevers, Douglas, & Jack,
2008). However, the challenges may run deeper than interface design. People coming to banking
for the first time via the mobile handset require a command of abstract concepts about
invisible/virtual money. Consider the lack of ways to wrap or “gift” a digital money transfer
(Singh, 2007). Beliefs, misunderstandings, habits, and concerns must be addressed if people who
are used to storing money in cash are asked to store it “in” a handset; the analogy remains
strained—the mobile is not yet a wallet (Chipchase, Persson, Piippo, Aarras, & Yamamoto,
2005).
Existing Payment Mechanisms
The role of existing mediated transfers and other financial services also deserves scrutiny.
A large proportion of the volume of m-transactions may reflect existing transactional
relationships, shifted over to the new channels. This is not to say that a shift is not itself
valuable—there are significant benefits of cost, reliability, safety, flexibility, and immediacy
associated with m-banking/m-payments systems. However, it is important for industry,
researchers, and policymakers to understand the transactional networks and behaviors that
already exist. An antecedent to this argument comes from the microfinance sector. Arguing that
“it is no longer acceptable for prospective providers not to inform themselves of what their future
clients are already doing and what services they appear to need,” Ruthven (2002, p. 269)
identified a broad array of “money mosaics” operating in a Delhi slum. These “financial relations
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 6

---

m-banking

are frequently embedded in other social relations which reflect the diversity of social, security,
and economic needs which people have. It highlights the relatively small role of commercial
transactions in people's financial lives, and the importance, extent and diversity of personal
networks” (2002, p. 267).
In the case of m-banking/m-payments channels, pawn shops, bus companies, the post
office, hand carrying by friends and family, underground money transfer mechanisms—such as
China‟s fei ch’ien („flying money,‟ a network of affiliates allowing users to put money into the
network in one city and have it available in another without the actual banknotes making the trip)
(Maurer, 2008)—and formal transfer services like Western Union all have their adherents, and
the list is longer when one includes alternative savings and credit mechanisms like chit funds and
moneylenders. There are communication issues, as well: transfers are exchanges at a distance,
and as Ruthven points out, there is an implicit or explicit network of communication and
information exchange embedded into almost every transaction. Remittances,iii in particular, are a
context in which it is difficult to separate financial transactions from symbolic meaning and
social bonding (Hart, 2000; Singh, 2007).iv
The Social Embeddedness of Economic Transactions
There is a litany of social/contextual influences on m-banking/m-payments use. Both
macro-level cultural factors and micro-level, locally-negotiated norms in families and among
peers—particularly about money—are at play (Zelizer, 1994). For example, respondents in focus
groups we conducted in Manila (Donner, 2007b) explained that, while they would certainly
transfer money to a family member (a gift), they would not do so to an acquaintance (a loan).
Technically, the actions are the same. Socially, they are miles apart.
Practitioners and policymakers are already concerned about validating m-transactions
under conditions of sharing behavior (infoDEV, 2006), in which two people use the same
handset. On the other hand, others suggest that m-banking/m-payments systems may alter
patterns of money sharing within families by giving women greater autonomy and control over
household savings (von Reijswoud, 2007).
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 7

---

m-banking

Cross-Cutting Themes in Studies of M-Banking/M-Payments Use
The use questions described above (skills and mental models, alternatives, and social
norms) each represent fruitful paths for future research. When such studies of m-banking/m-
transactions use in the developing world appear, it is likely that they will touch, implicitly or
explicitly, on crosscutting themes shared by studies of other mediated communication
technologies. There is little need for a new “theory of m-banking.” Rather, our existing toolkit of
theories of technology use—and particularly technology use in the service of economic and
social development (ICTD)—is sufficiently robust to handle the introduction of this new
technology (e.g., Sandvig & Sawhney, 2006). Rather, the task at hand for communication
researchers is to find ways to have the existing theory inform and strengthen the assessments of
impact and diffusion, of design and policy, which are occurring at this time as diverse
stakeholders are establishing the mobile payments landscape. While some periods of alternating
exuberance and introspection are to be expected in the realm of new technology development
and policy, the opportunity exists at this time for theory to temper some of the more dramatic
swings, at least as far as m-banking/m-payments as an “ICT4D” is concerned.
In this section, we introduce three crosscutting themes, each drawing on the existing body
of research on technology use, which illustrate social structures underlying m-banking/m-
payments and can guide decision-makers and practitioners as they build and evaluate these
systems over time. These themes are the bi-directionality of influence, amplification versus
change, and the multi-dimensionality of trust. As possible themes and theoretical perspectives,
these three are neither mutually exclusive nor collectively exhaustive. However, they do help
contextualize issues encountered in these early days of the technology. These three themes are
sufficient to illustrate this paper‟s primary argument: the importance of “use studies” as input to
the evaluation of adoption and impact of m-banking/ m-payments systems.
A willingness to examine the bi-directionality of influence between communication
technologies and the social structures in which they exist—a focus on the “dynamic interactions
between people and technology” (Orlikowski & Iacono, 2001)—is a hallmark of many studies
applying a “use heuristic” (Latour, 1987) or “ensemble” view (Orlikowski & Iacono, 2001).
Within the communication research tradition, two prominent examples of these approaches are
adaptive structuration theory (Poole & DeSanctis, 1990), usually applied to organizational
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 8

---

m-banking

settings, and domestication theory (Silverstone & Hirsch, 1992), usually applied unsurprisingly,
to domestic settings. Mobile telephones have also been the subjects of a wide range of studies
complicating the directionality of influence, from technological affordances to user choices to
social structures and back again (Donner, 2008; Haddon, 2003).
Studies that take “ensemble” approaches to assessing the spread of m-banking/m-
payments systems are sorely needed. Like text messaging (Ling, 2004), multimedia messaging
(Ling & Julsrud, 2005), and simple missed calls (Donner, 2007c), the set of social norms and
expected behaviors surrounding mobile-enabled financial services will evolve over time and
probably will differ from place to place. For example, respondents in the Manila focus groups
reported a norm of “gifts not loans,” based partially on their family structures and partially on
their experience with the m-banking/m-payments system. Wholly different norms might emerge
within another system, where perhaps early adopters were not dispersed families seeking a
channel for cost-effective remittances but rather traders looking for a way to protect money from
theft on deserted roads (as in, perhaps, Northern Kenya). In that case, transfers to near-strangers
might be perfectly acceptable, even expected.
A second crosscutting theme involves the capacity of a communication technology to
amplify existing relational and social structures as well as to alter them. When we speak of the
“impact” of a technology on a social structure—a particularly common frame within the ICTD
perspective—, we often seek insight into how technologies change social structures, but the
reverse can also be true (Harper, 2003). If we look with this lens, we may find that m-banking/m-
payments systems may increase the volume or frequency of existing transactional patterns rather
than alter the target of those transactions. For example, in the Manila focus group, of the 10
people using the system, only one had used it to trade with somebody that they had not
previously traded with using another channel—a women who had used the service to start her
own informal money lending business (Donner, 2007b). Additional findings of this kind may
qualify some of the “transformational” language used concerning mobile banking to this point.
A final crosscutting issue involves the introduction of “trust” as a factor in the analysis of
m-banking/m-payments use. Early evidence and intuition alike suggests that “trust” plays a role
in use (Ivatury, 2004; Porteous, 2007). For example. users feel more comfortable with at least
some face-to-face contact and assistance while using an m-banking/m-payments system like
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 9

---

m-banking

Wizzit (Ivatury & Pickens, 2006). Luarna and Lin (2005) proposed a modified technology
acceptance model that included a trust variable—perceived credibility—to predict m-banking
adoption in Taiwan. Yet their modification also included another variable, self-efficacy, a form
of trusting one‟s self.
Indeed, trust itself is a multifaceted concept, which must be handled carefully in any
analysis of m-banking/m-payments use (Benamati & Serva, 2007). Trust is a crosscutting
concept in that people can trust (or mistrust) their own skills. They can trust the interface, the
network across which their funds travel, the representatives of the institutions (channels) who
control their money, and/or trust the institutions themselves (Maurer, 2008). And, of course, they
can differentially trust various people in their networks: some might be eligible as exchange
partners using m-banking/m-payments systems while others might not. These forms of trust may
change over time with use of the system. People might become more or less trusting along any of
these dimensions as their experience of the system changes, relative to friends, family, and others
in the community.
The role of trust is a crosscutting issue because multiple research traditions examine
economic transactions in their social context—not as discrete acts but as markers and
reinforcements of a set of interrelated responsibilities, roles, and transactional networks in which
trust plays a central role (Burt, 1992; Geertz, 1978; Granovetter, 1985). Often these transactions
are seen as either being structured by or creating a form of “social capital” (Coleman, 1988).
These transactions need not be face-to-face; researchers have used social-capital/social-networks
lenses to explore how the information technologies generate and reinforce social/economic
relationships in ways that provide “returns” to actors (Huysman & Wulf, 2004; Sawyer,
Crowston, Wigand, & Allbritton, 2003). For example, Horst and Miller (2006) described the
practice of “link up” in Jamaica, where mobiles are used quite strategically to build and maintain
networks of resources for future assistance or loans.
Initial reports from an ongoing ethnographic project in Kenya elaborate these dimensions,
distinguishing the complexities of trust in the local m-banking middleman from trust in the telco
that runs it and from the government that (presumably for many users) controls the whole
operation (Morawczynski & Miscione, 2008). However, there is room for more work that
Pre-publication draft. Edits possible. © 2008 Asian Media and Information Center 10

---