Content uploaded by Lira Peter Sekantsi
Author content
All content in this area was uploaded by Lira Peter Sekantsi on Aug 03, 2017
Content may be subject to copyright.
1
THE FINANCIAL INCLUSION CONUNDRUM IN LESOTHO: IS MOBILE MONEY
THE MISSING PIECE IN THE PUZZLE?
Lira P. Sekantsi
*
and Sephooko I. Motelle
†
Central Bank of Lesotho, Working Paper No.02/16
July, 2016
Abstract
The prolific use of mobile telephones in developing countries has given birth to financial
innovations such as mobile money. As a result, the use of mobile money has expanded the grid of
financial services to include previously unbanked populations in Africa. This development is a
harbinger for increased financial intermediation and positive spill-overs in terms of credit growth to
entrepreneurs and faster economic growth. Based on monthly data for the period 2013m7 –
2015m12, this study employs time series techniques to unpack the proliferation of mobile money
and its attendant impact on financial inclusion in Lesotho. The findings reveal existence of long-run
steady state relationship between financial inclusion and mobile money in Lesotho and that mobile
money Granger causes financial inclusion both in the short-run and long-run in Lesotho. Therefore,
financial inclusion policies should be directed towards leveling the playing ground for mobile money
to flourish to create a more financially inclusive society in Lesotho.
JEL classification: C01, C22, C87, G23, O10, O30
Key words: Financial inclusion, mobile money, credit growth
*
Lira P. Sekantsi, Analyst-National Payment System(NPS) Oversight Section, Department of Operations, Central Bank of Lesotho,
email: psekantsi@centralbank.org.ls, Tel: (+266) 2223 2180
†
Corresponding Author: Sephooko I. Motelle, Deputy Director of Supervision, Central Bank of Lesotho, email:
smotelle@centralbank.org.ls, Tel: (+266) 2223 2041
2
1. INTRODUCTION
Many developing countries especially in Africa are characterized by financial exclusion in the
form of low access to financial services. This is mainly a result of banking infrastructure gaps that
hinder an all -inclusive financial system (Andrianaivo & Kpodar, 2012). According to Beck &
Maimbo (2013) approximately 2.5 billion people in the world lack access to financial services and
have to rely on cash or informal financial services which are typically unsafe, inconvenient and
expensive. However, more than half of households in developing countries do not have an account
with a financial institution. In the case of Lesotho, approximately 38% of the adult population has a
bank account, which indicates that the majority of the adult population still lacks access to basic
financial services (Ketly & Kasi, 2015).The mainstream banking sector fails to deliver financial
services to millions of consumers especially those residing in rural areas. Banks are biased in favour
of affluent consumers due to high costs of physical infrastructure and operational costs as well as
low profits associated with serving the low income consumers (Dube, 2014). This lack of access to
financial services not only limits the ability of the poor to save, repay debts and manage risk
responsibly but also indirectly exposes them to poverty (see Donovan, 2012).
The development of mobile money has provided a glimpse of hope for the financially
excluded members of the population. Mobile money is perceived as a solution that can circumvent
poor banking infrastructure and geographical isolation and can offer low-cost distribution of
financial services through the mobile phone network. In addition, the surge and near-universal use
of mobile phones and the huge number of airtime distributors that can act as access points make
mobile money a cost-effective solution to financial access (Ketly & Kasi, 2015).Following the
successful adoption of mobile money in Kenya, numerous countries have scaled up their efforts to
implement it in an effort to increase access to financial services. Hence, it is crucial to study the
impact of mobile money on financial inclusion. This study pursues this objective in the context of
Lesotho using monthly data for the period 2013m7 – 2015m12.
The rest of the paper is organized as follows: Section 2 reviews the literature on mobile money
and financial inclusion while section 3 discusses mobile money developments in Lesotho. Section 4
describes the data and presents the analytical framework and section 5 discusses the empirical
results. Section 6 concludes the paper and offers a menu of recommendations.
3
2. REVIEW OF THE LITERATURE
2.1 The Definition of Financial Inclusion and Identification of its Constraints
Financial inclusion is defined as a process that ensures ease of access, availability and usage of
the formal financial system for all members of an economy. It entails access to financial services
such as payment services, remittance facilities, savings, loans as well as insurance services offered by
the formal financial system at costs that can be afforded by the poor and disadvantaged social
groups. An inclusive financial system has several benefits. First, it facilitates efficient allocation of
productive resources and can potentially lower the cost of capital. Second, access to appropriate
financial services can considerably improve the management of finances on daily basis. Third, it can
also help in reducing the growth of informal sources of credit such as money lenders that are often
found to be exploitative. An all-inclusive financial system enhances efficiency and welfare by
providing avenues for secure and safe saving practices and it also facilitates availability of a whole
range of efficient financial services (Sarma & Pais, 2011). According to Dube et al (2014) financial
inclusion does not only ensure access to basic financial services by all, but also promotes economic
growth, reduces poverty and inculcates a savings culture in rural areas.
The issue of inclusive financial system has attracted greater attention both in academic and
policy circles and has become a policy priority in many countries around the globe. Therefore,
financial regulators, governments and the banking industry over the world have beefed up efforts to
develop and implement various initiatives that deepen financial inclusion. For instance, the United
States (US) has developed legislative measures that require banks to offer credit throughout their
areas of operation without discriminating between the rich and the poor. France has also developed
measures that emphasize a person’s right to have a bank account. Furthermore, the banking industry
has introduced products such as “no-frills” accounts and “General Credit Cards” for low deposits
and launched low cost bank accounts to promote financial inclusion. In addition, micro-finance
institutions and “Self-Help Groups” have also been promoted in some countries such as India to
take care of the excluded groups (Sarma & Pais, 2011). These efforts are meant to ensure access and
affordable financial services to the poor to allow them to plan for routine expenses, cope with
external shocks and better cover unanticipated expenses. In addition, they contribute to increased
4
access to more stable and productive activities (Gwalani & Parkhi, 2014). This, not only enhances
economic growth and reduces poverty, but also promotes social inclusion
1
.
The degree of financial inclusion differs among countries depending on their stages of
economic and financial development. Developed countries such as the US and United Kingdom
(UK) have managed to provide financial services to the vast majority of their populations. However,
in developing countries particularly in Africa, the issue of financial inclusion still remains a challenge
as most countries are severely constrained by limited infrastructure and the other difficulties of
accessing financial institutions, which leave large proportions of the population, especially those
who reside in remote areas, with low access to affordable financial services or completely excluded
from financial services (see Kempson,2006 and Oji,2015).
The literature identifies both demand and supply side constraints to financial inclusion. For
example, people may choose not to use formal financial services because they do not need such
services due to religious and cultural reasons, and/or lack of trust in formal financial institutions.
Lack of trust may be a result of, among others, fear of bank failure or fraud. In addition, people who
wish to use formal financial services may face several barriers. First, inaccessibility due to difficulty
in reaching service points or absence of such services in the vicinity. Second, unaffordability as
formal financial services are often too costly for low income persons. Third, inappropriate product
design, which results in products that do not meet the needs of excluded customers. Fourth, inability
to meet eligibility criteria, for example not having sufficient assets to meet conditions for the
extension of a loan or being unable to provide documentation evidencing identity. In addition, other
demand side constraints include cumbersome documentation and procedures that customers have
to undergo when opening a bank account, limited literacy and numeracy skills, information
asymmetry due to lack of awareness, branch operating hours, which may be inflexible for some
sections of the population (see De Koker & Jentzsch, 2013; Gwalani & Parkhi,2014 and
Kempson,2006).
1
Social inclusion is defined as the degree to which people are and feel integrated in the different relationships, organizations, sub-
systems and structures that constitute everyday life. As a process, it refers both to integration into social, economic and civic life and
the pursuit of active citizenship as well as a means to counter poverty understood in the sense of capability deprivation (see Cardo,
2014).
5
On the supply side, some people who had initially utilized formal financial services may opt
to withdraw from such services due to high costs inherent in maintaining them, lack of trust or faith
in the banking system, bad credit records, difficulties associated with the management of their
spending and inappropriate product design as well as complex procedures for availing financial
services. Apart from that, regulatory requirements which require financial service providers to adopt
stringent disclosure requirements, which must be met by the customers before the service is
provided, deter customers from participating in the formal financial system (De Koker & Jentzsch,
2013). In addition, financial service providers may decide not to offer some services to customers if
they feel that the environment does not protect their interests (Central Bank of Lesotho, 2013).
Thus, in spite of intensified efforts to increase access and use of financial services, many developing
countries still have the vast majority of their populations unbanked
2
.
2.2 The Role of Telecommunication Technologies in improving Financial Inclusion
The conventional banking system has not been able to provide financial services to a large
number of low-income and poor people, especially in remote areas, due to high costs of physical
infrastructure, operational costs and unprofitability arising from serving low income consumers (see
Boston Consulting Group, 2011; Goss, Mas, Radcliffe & Stark, 2011). However, the diffusion of
information and communication technologies (ICT) and mobile telephony have the potential to
significantly reduce barriers to financial inclusion and therefore allow millions of people who were
otherwise excluded from the formal financial system to perform financial transactions relatively
cheaply, securely, and reliably through their mobile phones (Dube at al,2014). Mobile money,
defined as the provision of a range of financial service such as mobile banking, mobile payments and
mobile transfers to consumers through mobile devices, is one of many possibilities arising from
advancement in technology. It encompasses common functions such as balance checks, funds
transfer, depositing and withdrawing cash (cash-in and cash out), savings, access to credit, bill
payments, airtime purchase and long distance remittance of funds (Donner and Tellez, 2008;
Kasseeah & Tandrayen-Ragoobur, 2012; Jenkins, 2008).
2
According to Gross et al (2012) and Breitbach & Walstad (2014/2015) the unbanked are individuals or households without checking
or savings accounts and operate largely outside the banking system when making financial transactions. On the other hand, the
underbanked are individuals or households that have a bank account(checking, savings or money market account), but supplement
the account with alternatives to traditional banking services such as non-bank money orders, non-bank check-cashing services, payday
loans, rent-to-own agreements, payday loan, payroll card or pawnshops.
6
Mobile phones have a great potential for delivering financial services to a broader base of
customers due to their enormous uptake by large number of the unbanked and the poor in
developing countries. Moreover, mobile phone systems can be placed anywhere as long as there is
wireless phone connection and this overcomes the problem of distance and lack of bank branches in
remote areas. In this regards, it enables the possibility of ubiquitous access to financial services.
Furthermore, mobile money financial services are commonly set up with infrastructure provided by
a network of “cash merchants” (or “agents”), who may be located all over the country, as well as a
host of other supporting businesses such as banks, agent aggregators and liquidity management
firms (Donovan, 2012; Ramada-Sarasola, 2012). Therefore, it does not only enable new entrants to
the banking system but also offers such services at lower costs because it does not incur the costs of
physical roll-out and faces lower costs of handling low-value transactions( Flores-Roux & Mariscal,
2010).
The fact that mobile money uses existing mobile infrastructure to deliver all services online
brings cost efficiency to the provision of cash-in and cash-out services to the poor(Flores-Roux &
Mariscal, 2010). These it lowers transaction costs, which translate into savings for the poor.
Consequently, this assists the poor to reallocate their resources efficiently to smoothen their
consumption patterns (Donovan, 2012 and Dube, 2014). In addition, it reduces transportation costs
3
and improves information flows between transacting parties while allowing efficiency gains (see
Bhatia et al, 2008 & Sife et al, 2010).It can also be viewed as the most reliable, accessible and
convenient medium for the delivery of financial services by poor households due to its speed and
liquidity as well as its ability to act as a store of value since mobile money value does not decline with
time
4
.
Mobile money also increases the large scale financial connectedness among distant households
and individuals. This allows users/customers to benefit from remittances from either family
members or friends living in remote areas within the same country or abroad. Assuming other things
remain the same, this alone improves the economic well-being as this acts as source of income for
the poor (Hinson, 2011; Morawczynski & Pickens, 2009; Alleman & Rappoport, 2010), According
to Morawczynski (2010) using mobile money also increases money circulation, boosts local
consumption for the rural people as well as spurs economic activity by enabling “just-in-time”
3
This is relevant in cases where users or traders in rural areas would need to travel from to urban areas to send and receive money.
4
Unlike cash, mobile money does not attract charges, which ultimately reduces its value. However, its value remains the same until it
is used.
7
transfers that make capital available whenever and whenever it is needed. Apart from that, by acting
as a channel through which households and individuals receive remittances, mobile money often
enables households and individuals to absorb shocks arising from job losses and poor harvests, loss
of relatives, health problems and so on(Donovan, 2012).
Mobile money serves as a form of savings account for people without a formal bank account.
Thus, it enables them to engage in a safer and more efficient savings mechanism and improves
efficiency and regularity of savings (Nandhi, 2012). In connection with this, by acting as remittance
channel mobile money increases the income of the rural users, which leads to increase in savings.
Apart from that, these mobile money accounts have the potential of adding social value to low-
income people, who usually face constraints with respect to opening a formal bank account (Jack &
Suri, 2011). In addition, mobile money account has the potential of integrating the mobile money
users into the formal financial services grid by providing access to other accounts that cover a wide
range of other financial services’ needs (Alexandre & Eisenhart, 2013 and Flores-Roux & Mariscal,
2010). Furthermore, the fact that mobile money is less visible than other alternatives including cash
enables mobile money users to keep their money safe from dangers of theft and accessibility by
other family members (Jack & Suri, 2011). In this regard, it promotes privacy and individual
autonomy within the family while also makes it possible to facilitate or enable financial transactions
that either did not occur before or that were conducted at a higher risk and price (Donovan, 2012).
Moreover, when it has reached large scale and there is large customer base, provision mobile
money services can also prove to be commercially viable. It generates considerable revenue for both
service providers and cash agents, the success of which may lead to increased labor demand and
employment generation for the poor. In recent years, access and use of more sophisticated financial
services such as savings, credit, and insurance has proved to be far more beneficial to the poor. In
light of this development, financial institutions, banks, governments, and other institutions have
taken advantage of the payment services that are deployed by mobile money operators to actively
innovate and develop these financial services and offer them to customers. For instance, in Lesotho
some Alliance Insurance Company has partnered with
Econet Telecom Lesotho (ETL) to offer
funeral insurance covers for consumers whereas in Kenya, Equity Bank has partnered with
Safaricom to offer micro savings account, credit and insurance. In addition, some governments have
already adopted mobile money electronic payment services platform for cash transfers to reduce
leakage, transaction costs and overheads. For instance, in Tanzania mobile money has also been
8
adopted by the government to collect all levies, fees and taxes paid by the public. This helps to
enhance the government’s ability to monitor financial flows, collect tax revenue, and reduce illicit
corruption and fraud (Donovan, 2012).
While mobile money reduces the dependency on cash and contributes to the development of
an electronic ecosystem of financial services, it also generates data
5
in the form of financial
transaction records. These transaction records can be efficiently used to analyze creditworthiness,
enhance credit monitoring as well as facilitate access to micro-loans or other financial services (see
Andrianaivo & Kpodar, 2012 and Mutsune, 2015). Furthermore, the data generated by mobile
money can also act as tool that can be used to report suspicious financial transactions that mobile
money operators or banks can identify in an effort to combat money laundering and terrorism
financing (Alexandre & Eisenhart, 2013 and De Koker & Jentzsch, 2013). This is pertinent given the
rampant increase in money laundering and terrorism financing activities in recent years.
From the business perspective, the payment behavior data of the unbanked and poor
customers can shed some light on how poor customers transact, their payment behavior as well as
their financial service needs. Therefore, the providers of highly data-dependent areas such as credit,
insurance can use the payment data of customers to build business case to serve this new and
diverse segment of the market that has largely been ignored by the many financial service providers.
In addition, banks can use this information to develop opportunities to cross-sell additional
products such as credit, long-term savings accounts, which enhance the business case for low-value
bank accounts. On the other hand, the mobile money operators can generate revenue by selectively
selling the data to other parties that can utilise it to market products and services that assists in terms
of broadening the services that keep customers loyal to their existing mobile money schemes
(Alexandre & Eisenhart, 2013)
Mobile money has several unique attributes that make mobile based transactions attractive.
However, it also presents inherent risks, including money laundering, privacy and security, consumer
protection, fraud, and liquidity risks just like any retail payment system. In many mobile money
implementations, proportionally risk adjusted anti-money laundering (AML) procedures have been
applied to extend the service to the underserved populations. These adjusted AML requirements are
5
According to Alexandre and Eisenhart(2013), data is a strong asset for both financial inclusion and financial integrity.
9
usually counterbalanced by transaction volume and value restrictions placed on the account.
However, rogue actors circumvent these controls by dividing a large transfer of funds into small
ones, which fit within the definition of the restrictions applied using multiple mobile phones and
accounts and then transfer the funds. This is possible because unlike traditional banking, which
require the face-to-face interaction, mobile technology-enabled payments create a more opaque and
anonymous experience that may permit the opportunity for criminal activity. This is increasingly
plausible as mobile retail payments can occur rapidly and in cross-border environments. In addition,
there are numerous schemes for money laundering and terrorist financing that may migrate to the
mobile channel. For instance, the runners of the so-called “digital value smurfing” scheme bypass
banks and regulatory reporting requirements by exchanging ill-gotten funds for digital value through
mobile devices and thereby enable the proceeds of crime or terrorist financing to be transmitted
over airwaves to anywhere the runners intend to take funds(see Merritt ,2010 and Lake, 2013).
Mobile money may also compromise sufficient elements of the customers’ information and
privacy. This may not only allow another party to replicate the customer’s identity in the system and
use it to fraudulently conduct transactions but also exposes the customer to other risks such as lost
payments through faulty transmissions, or criminal activity on the part of the mobile operator, agent,
or other payment service providers. In connection with this, the recent surge in smart phone
applications may introduce vulnerabilities to malware attacks, which may increase payments risks as
bad actors gain access to personal information stored in the handset or accessed through a phone
application (Lake,(2013).In addition, lack of cash or electronic float at the agent outlet may
temporarily or permanently disable a client wishing to deposit or withdraw money to or from the
system. On one hand, poor network coverage and insufficient service points may make it difficult
for customers to undertake transactions, leading to withdrawal from the service. This can
consequently compromise revenue generation by the agents. On the other hand, system technical
errors and transaction delay by the network usually leave customers and agents in a difficult position
to know whether or not the transaction has been delivered and therefore unsure whether or not to
re-submit the transaction. Moreover, transactions within mobile payment network travel through
many communication systems to reach to the mobile money backend. Any breakage in this chain as
well as lack of literacy by the customer can lead to inability to transact by such a customer. The
length of the chains of message handling within the mobile money operation may also delay balance
10
updates for any given transaction. This exposes the customer to possibility of incorrect decline in
future transactions due to insufficient funds (see Lake, 2013).
According to Lake (2013) mobile money products are often delivered by consortia of mobile
network operators (MNOs), banks, agent network managers as well as agents. Therefore, any
significant relationship difficulty among these parties within this consortium could result in service
unavailability to the client. Consequently, this could not only cause unnecessary inconvenience on
the part of the customer but also lead to inability to transact. In addition, lack of clarity as to who
holds customer’s money may make it difficult for the customer to enforce rights whenever
necessary. Therefore, the mobile payments landscape
demands a collaborative effort among
different stakeholders to balance intervention for risk mitigation with market innovation. These
include mobile networks operators, banks, airtime sales agents, retailers as well as regulators.
2.3 Adoption of Mobile Money and Developments in Financial Inclusion: African
Country Experiences
Mobile money services have become popular in developing countries predominantly due to
large unbanked populations and low levels of financial inclusion. Among the countries that
implemented this service, Kenya is a global leader in mobile payments implementation and adoption
with its M-Pesa (mobile money in Swahili) service. Initially launched in 2007 by Safaricom, a
subsidiary of Vodafone, for person-to-person transfers, M-pesa has become probably the most
renowned and successful mobile money service to date. In May 2008, 14 months after its launch, M-
Pesa had 2.7 million users and almost 3,000 agents (GSMA, 2012). Within five years of its launch, it
had 15 million customers
6
and more than 18,000 agents and was processing $10 billion per year (Lal
& Sachdev, 2015). It has become so successful to the extent that almost all households use it (IOS
Press, 2012).
At its launch as money transfers service, M-pesa adopted the slogan “send money home”. This
positioned it to serve as an “urban-rural” remittance corridor to take advantage of significant
domestic remittance market in Kenya. It allowed many urban migrants to remit to their relatives in
rural areas (Gugler, 2002 & Donovan, 2012). In addition, it has since grown to provide many other
financial services including bill payments, loan transactions, international remittances and public
6
This represents 37.5% of the country’s population.
11
transport payments. The overwhelming dominance of Safaricom in the Kenyan market and high
mobile telephony penetration rate as well as increased demand for additional services paved the way
for M-Pesa’s great success. In addition, an enabling regulatory environment and the relatively high
availability of decision-making data continued to support its development (IOS Press, 2012). Of
course, M-pesa still faces challenges. These include lack of universal mobile phone access (Jack &
Suri, 2011a) and difficulties with liquidity management by agents and raising start-up capital
(Eijkman, Kendall & Mas, 2009).Therefore, in order to develop further, Kenya needs to further
enhance its institutional and market environments and develop consumer protection provisions.
(Bilodeau et al, 2011).
Following the successful launch of M-pesa in Kenya, many mobile network operators (MNOs)
became eager to launch such products in their jurisdictions. Therefore, one year after the Kenyan
launch, Vodacom
7
launched M-pesa in April 2008 in Tanzania. Nonetheless, the user uptake of this
service in Tanzania has been much slower compared to Kenya. In June 2009, 14 months after the
launch, M-pesa had 280,000 users and 1,000 agents in Tanzania (Rasmussen, 2009). The slow M-
pesa uptake in Tanzania was due to the fact that Vodacom may not have carefully judged the unique
country context prior to implementing it in Tanzania. Therefore, it could not contextualize
advertising to suit the level of financial literacy in the country for customers to understand the
product. In addition, there were weaknesses in terms of contextualizing the nature of remittance
market, which is urban-rural, rural-urban, urban-urban and rural-rural.
Still following the Safaricom’s M-pesa in Kenya, Mobile Telephone Network (MTN) Uganda,
launched MTN money in March 2009. In June of the same year, Airtel launched Airtel money. In an
effort to increase their market share, other MNOs also launched their mobile money services. For
instance, Uganda’s Telecom launched M-sente in March 2010, Warid Telecom launched Warid Pesa
in December 2011 and Orange Telecom launched Orange money in the first half of 2012. Since its
launch, the subscriber base has increased steadily with over 9 million people using mobile money in
2012. Similarly, the number of mobile money transactions also reached 242 million during the same
year while the total value exchange recorded US$4.5 billion during the same time period. Of this
large subscriber base country wide, MTN money has the largest market share with over 15,000
agents and it remains one of the most successful mobile financial services deployments in East
7
A subsidiary of South African Vodacom Pty (Ltd)
12
Africa. The successful expansion of this service is partly attributed to both high mobile phone
network roll-out and mobile phone adoption rates (Munyegera and Matsumoto,2014; Orotin et
al,2013).
With 15 million adults and mobile penetration rate of 74% of the population, Ghana had five
mobile money services in 2010, the sixth of which was not operating in spite of acquiring a licence.
These include AfricXpress(txtNpay) launched in 2008,MTN mobile money launched in July 2009
with nine partner banks, Airtel money launched in April 2010 with six partner banks and Tigo Cash
established in October 2010 with three partner banks. Among these mobile money services, MTN
money is the most successful mobile money deployment in Ghana. In 2011, it had 1.8 million
registered customers, 4,000 trained agents in Accra and 2,000 in other parts of the country. Four
years after the launch, the service had registered 4.8 million customers, 19, 500 merchants and 18.5
million transactions. Its success is underpinned by heavy investment in above-the line marketing and
a primary marketing message focused on domestic remittance. This is due to the fact that Ghana has
a large number of households that depend on domestic remittances because of increase in
urbanization in city centres and constant migration (see CGAP, 2011 and Tobbin, 2010).
As the largest and most populous economy in Africa with a largest proportion of the
population remaining unbanked, Nigeria is a promising market for mobile financial services. In
March 2014, Nigeria had licensed 18 MNOs. These include Guaranty Trust Bank (GTBank), United
Bank of Africa (UBA/Afripay), Stanbic IBTC, Ecobank, Fortis MFB, Pagatech, Paycom, eTranzact,
Eartholeum, M-Kudi and Virtual Terminal Network (Phillips Consulting, 2013). Since
commencement of operations in 2012, mobile money had 9,989,297 subscribers, 67,494 enrolled
agents and conducted over 11 million transactions worth over N105 billion. However, mobile
payments market in Nigeria is still in its infancy. This is in part due to rapid deployment and rollout
of the system, which has been inhibited by a number of challenges. These include inadequate capital
outlay on the part of the MNOs, basic infrastructural challenges – power, telecommunications
network etc, lack of awareness/customer education which has slowed down the adoption rate and
lack of wide-spread agent network (see Ingba, 2014; Yakub et al, 2013 and Grameen Foundation,
2014)
13
In Zimbabwe, Ecocash was launched by Econet Services on the 30th September 2011. Initially,
Ecocash invested heavily on upfront capital to build mass of agents and active subscribers. In this
case, Ecocash provided incentives such as high commissions (80% of transaction revenues) and
performance based rewards to their agents. In addition, they ensured adequate float liquidity for
transactions and cash-on –hand to encourage investment in their business. They invested a lot of
money on consumer marketing utilizing above-the-line to raise awareness and below-the-line to
educate consumers on the service and drives registration. This service initially focused on the
underserved segments of the population in semi-urban and rural areas and offered person-to-person
transfers but over time the service expanded to other multiple products. The service achieved great
success to the extent that it reached an agent network of 4,000 agents, 2.3 million customer
registrations (which is equivalent to 31% of the country’s adult population) just within 18 months
after launch, with 1 million of them active as well as an annualized transaction volume valued at 22%
of the country’s GDP (Lal & and Sachdev, 2015).
In South Africa, Vodacom tried to replicate the model deployed in Kenya. However, service
was launched as a mobile alternative to existing financial and payments infrastructure and was rolled
out largely to better-off parts of the country, where there is robust banking environment. The launch
plan was not prepared based on the identified target market or an analysis of financial flows/
potential use cases of such market. Therefore, the service effectively did not serve the large
remittance corridors of the lower income and rural populations. In addition, the service experienced
challenges relating to customer registration and sustaining usage for registered customers. This was
due to poor marketing to consumers, which resulted in poor understanding and lack of trust in the
service Furthermore, the registration process was lengthy and slow, and this resulted in weak
adoption. The service used retail stores as agents, but it was only available at a limited number of
locations after launch. Inadequate float management led to suspension of cash-out transactions by
many outlets until a certain amount of sales had been registered. At the end, many retailers decided
to discontinue acting as M-pesa agents as the service was disrupting their retail business. As a result
of these problems the service only had 1.2 million registered users two years after launch, with 7%
annual growth, and of which only 1% appeared active. With this mediocre performance, Vodacom
opted to discontinue the service at the end of 2013. However, it has since been re-launched with a
new banking partner and new model, with a focus on serving the unbanked and low income
14
segments of the population (Lal and Sachdev, 2015). Consequently, the service was discontinued in
2016.
3. FINANCIAL EXCLUSION, MOBILE TELEPHONY AND MOBILE MONEY
DEVELOPMENTS IN LESOTHO
3.1 Financial Exclusion
Lesotho has four licensed banks; namely Standard Lesotho Bank (SLB), Nedbank Lesotho,
First National Bank Lesotho (FNBL) and Lesotho Postbank (LPB), which form the core of the
financial system. Among these banks, SLB is the largest in terms of assets. In addition, there are 6
credit-only Micro Finance Institutions (MFIs) namely Letshego Financial Services, Alibaba Financial
Services, Net Loans, Edu Loans, Lesana Lesotho Limited, Blessings Financial Services and Thusong
Financial Services with Letshego Financial Services only the largest. There also 70 formal money
lenders and approximately 250 savings and credit cooperatives (SACCOs) as well as one large
financial cooperative i.e. Boliba Savings and Credit Cooperative. Furthermore, there are 30 licensed
insurance brokers and 10 licensed insurance companies that generally specialize in both general
insurance and life insurance. However, the analysis focuses on the banking sector as it forms the
largest part of the financial sector in Lesotho and it has been the primary distributor of financial
services products in the country.
Access to banking is relatively low in Lesotho
8
due to limited banking infrastructure- bank
branches and devices infrastructure (ATMs and POSs). Table 1 shows that in 3 years, from 2013 to
2015, banking infrastructure has expanded by only 2 branches, 39 ATMs and 365 point of sales
(POSs). Therefore, the country has not seen much change in terms of financial inclusion. This
situation is exacerbated by the fact that large proportion of this banking infrastructure is also
mainly situated in urban areas though two-thirds of the country’s population resides in rural areas.
This mismatch implies that 55% of the population travels more than an hour to get to a bank
branch while only 24% of Basotho live within 30 minutes travel time from a bank branch.
According to Jefferis and Manje (2014), only 29.5% of the rural population in Lesotho is banked
compared to 57.9% in urban areas. The provision of financial services in Lesotho is made even
more difficult by the mountainous terrain of the country, which makes the banking infrastructure
difficult and expensive to distribute to most of Basotho. Despite the low banking access in Lesotho,
8
This is the lowest in the Southern African Customs Union (SACU).
15
it is unlikely that banks will significantly increase the number of bank branches and device
infrastructure (ATMs and POSs) on account of low population densities, financial viability, small
financial markets and mountainous terrain of the country (Ketly and Kasi, 2015).
Table 1: Banking Infrastructure
Branches
ATMs
POS
2013
2014
2015
2013
2014
2015
2013
2014
2015
SLB
17
17
17
73
82
86
307
448
529
FNBL
6
6
8
42
45
59
393
474
423
Nedbank
9
9
9
24
28
27
68
85
149
LPB
13
13
13
2
7
8
35
38
37
Total
45
45
47
141
158
180
803
1045
1168
Source: Central Bank of Lesotho
According to the FinScope surveys conducted among 15 African countries, the level of
financial inclusion in Lesotho stands as high as 80.9% of the adult population, using the traditional
measure of financial inclusion. However, only 45.8% of the adult population uses bank and non-
bank formal services, of which 38% has bank account with a financial institution. The use of
informal services is relatively high, with 62.4% of the adult population using these services
particularly funeral insurance, which is a limited product compared to the wide range of financial
services needs of the people in the economy (Lesotho FinScope, 2011 and Jefferis and Manje, 2014).
Within the Southern African Customs Union (SACU) and Southern African Development
Community (SADC) regions, Lesotho has a high level of financial inclusion in the region; it is
exceeded by only South Africa and Namibia (see Ketly and Kasi, 2015).
Gough
3.2 Mobile Telephone Industry
The launch of the first commercial mobile telephone services three decades ago saw a
phenomenal growth in mobile communications around the world. According GSMA (2012b) the
total mobile penetration has more than doubled in all regions of world since 2005. This can be
attributed to a number of factors including a fall in handset and usage costs and an improvement in
service quality and network. On the other hand the use of fixed lines has decreased as they remain
undeveloped and unavailable to the majority of the population in developing countries. Low-income
countries are experiencing faster growth rates– more than twice as fast as in high income countries
in the 21st century (GSMA, 2012b).As of mid-2015; Sub-Saharan Africa (SSA) region had 367
16
million unique subscribers and 680 million connections (GSMA, 2015). This significant mobile
phone penetration has increased the availability of basic mobile services such voice, texts and basic
text-related services to millions of people across all income groups in SSA.
Figure 1: Sector Teledensity Trends 2004-2014
Source: Lesotho Communications Authority Annual Report 2013-2014
In the context of Lesotho the mobile phone industry is dominated by two mobile network
operators (MNOs), namely Econet Telecom Lesotho (ETL) and Vodacom Lesotho (VCL).The
former came into being following the merger between Telecom Lesotho and Econet Ezi ~ Cel
Lesotho in April 2008 while the latter is a subsidiary of South Africa-based Vodacom and began
operating in Lesotho in 1996. There is effective competition in the mobile sector between these two
MNOs and they provide network services to their subscribers. Figure 1 shows the mobile sector
teledensity trends over a period of ten years from 2004 to 2014. In 2014, the mobile subscribers
reached a total of 1,753,323 from 1,580,713 reported in the previous year. This translates into a
teledensity of approximately 93% of the population (98% of which are the prepaid subscribers while
only 2% are post-paid subscribers).However, the fixed line subscriber base remained constant at 3%
during the same year. The mobile subscribers accounted for 97% of telecommunication market
share compared to fixed subscribers at 3%.Overall, based on the 2006 population census figure of
1,880,661 for Lesotho, the teledensity for both fixed and mobile subscribers increased from 87% to
96%. In addition, the geographic coverage area with access to communications service (most of
which is driven by mobile services) has also increased. This is reflected in the coverage maps of the
two major network operators, depicted in appendices, 1 and 2 (LCA Annual Report, 2013/2014).
17
3.3 Mobile Money
The success of mobile money in East Africa especially M-pesa in Kenya saw many countries
around the globe adopting the same model to launch similar products in their jurisdictions. This is
because M-pesa has allowed millions of people who were otherwise excluded from the formal
financial system to perform financial transactions relatively cheaply, securely, and reliably. In
the same manner, ETL and VCL in Lesotho launched mobile money services in an effort to narrow
financial exclusion and drive economic development in Lesotho. ETL launched its mobile money
service, Eco-cash, in October 2012 while VCL launched M-pesa in July 2013. Since its launch until
December 2015, M-pesa signed up to 745,242 customers with 1999 agents. On the other hand, Eco-
cash has accumulated 318,786 customers and 1480 agents countrywide during the same period
Figure 2: Number of Registered Mobile Money Customers and Agents
Source: Central Bank of Lesotho, 2015
The number of registered mobile money customers in Lesotho kept increasing since mobile
money inception as indicated by Figure 2. Based on 2006 population census figure of 1,880,661
inhabitants, the number of registered mobile money customers increased from 10% in June 2013 to
approximately 57% of the population in December 2015.On the other hand, the number of agents
increased exponentially from 337 in June 2013 to 3654 in December 2015. Figure 2 indicates the
number of registered mobile money customers and agents in Lesotho since June 2013. Based on
2006 population census figure of 1,880,661 inhabitants, the number of registered mobile money
customers increased from 10% in June 2013 to approximately 57% of the population in December
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0
200
400
600
800
1,000
1,200
Jun
Aug
Oct
Dec
Feb
Apr
Jun
Aug
Oct
Dec
Feb
Apr
Jun
Aug
Oct
Dec
2014 2015
Number of Registered Customers (000s, LHS) Number of Agents (000s, RHS)
18
2015. As a proportion of MNO subscribers, the two MNOs achieved approximately 48% market
penetration in December 2015. This is reflected in Figure 3 below.
Figure 3: Number of Subscribers and Mobile Money Users during April 2014-
2015(in thousands)
Source: Central Bank of Lesotho, 2015
Consistent with the increasing market penetration, mobile money transaction volumes,
especially customer’s withdrawals, bill payments, domestic money transfers and airtime purchases,
keep an upward trend since the inception this service. Figure 4 shows that as of December 2015,
mobile money has processed a total of 751,743 airtime purchases, 243,169 customer cash
withdrawals, 321,768 bill payments and 221,257 domestic money transfers. However, its use case in
the processing of salaries has remained significantly limited despite its potential to reduce the hurdles
of salaries processing of contract and casual workers by different organizations in the country
including the Government of Lesotho (GoL). The high uptake and usage of mobile money is also
consistent with the growth of trust account balances of the two MNOs (see Figure 5). The growth in
adoption and usage of mobile money in Lesotho is attributed to the gradual appreciation of the
product especially by the urban based users and heightened efforts by MNOs in advertising and
educating customers about this product offering.
0
500
1,000
1,500
2,000
2,500
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
2014 2015
Total no. of MNO subscribers No. of Mobile Money Customers
19
Figure 4: Volume of Mobile Money Transactions (in thousands)
Source: Central Bank of Lesotho, 2015
Figure 5: Trust Account Balances (in thousand Maloti)
Source: Central Bank of Lesotho, 2015
However, as indicated in Figure 6 below, mobile money is used mainly in urban districts as
opposed to rural districts
9
, where there is high financial inclusion gap due to limited banking
infrastructure. This is attributed to failure by MNOs to reach remote areas as a result of lack of
9
This is because the physical presence of agents in the vicinity not only actually drives knowledge about the product but
could also increase its usage by customers. Product appreciation by customers increases its usage.
0
100
200
300
400
500
600
700
800
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
2013 2014 2015
Airtime Purchases Domestic Money Transfers
Bill Payments Customer Withdrawals
Salary Payments
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
2013 2014 2015
Total Trust Account Balances
20
customer education and poor agent network in rural areas and inaccessibility of some rural areas due
to mountainous terrain. Therefore, there is need for more efforts on the part of MNOs scale up
customer education on mobile money and increase agent recruitment in rural areas. In addition,
MNOs should promote the use of mobile money in its electronic form in conducting transactions.
This will solve agent liquidity problem as customers would not need to convert mobile money into
cash prior to undertaking transactions.
Figure 6: Agent Network per District (Number of Agents per district)
Source: Central Bank of Lesotho, 2015
As indicated before, approximately 48% of adult populations who own mobile phones use
mobile money services. However, a large number of mobile phone users still do not use mobile
money services. Therefore, there is potential for other mobile phone users who have not yet signed
up for mobile money services to do so as time progresses. This is possible because the banking
infrastructure is limited and primarily situated in urban areas, although the majority of Lesotho’s
population resides in rural areas. In addition, it is unlikely that banks will significantly expand their
bank branches and devices (ATMs and POS) because establishing such infrastructure in rural areas
would not be financially viable. Therefore, the difficulty of accessing financial infrastructure by the
majority of rural population creates an opportunity for mobile money to provide a solution. This is
because it provides lower cost and convenient alternative to traditional banking products, which the
poor in rural areas can afford.
4. METHODOLOGY
0
200
400
600
800
1000
1200
1400
1600
1800
2000
No. of Agents in 2013 No. of Agents in 2014 No.of Agents in 2015
21
4.1 Data Type and Sources
Financial inclusion and mobile money cannot be directly measured; therefore, they need to be
measured by means of proxies. Table 2 below gives the description and measurement of variables.
These data are available at the Central Bank of Lesotho (CBL) and covers the period 2013m7
through 2015m12. This sample period was chosen as comprehensive data on mobile money started
to be collected in 2013.
Table 2: Description of Variables
Variable
Measurement
Description
Source
Total deposits per 1000 adults
Aggregate deposits held by commercial banks
divided by 1000 adult population.
CBL
Total credit per 1000 adults
Total credit extended by commercial banks
divided by 1000 adult population.
CBL
Number of automatic teller
machines(ATMs) per 1000
adults
Number of ATMs divided by 1000 adults
CBL
Number of Point of
sales(POSs) per 1000 adults
Number of POSs divided by 1000 adults
CBL
Mobile money transaction
(values) divided by 1000 adults
These include values of the following
transactions; airtime purchases, domestic
money transfers, customer withdrawals, and
customers deposits and float10 (each divided
by 1000 adults)
CBL
Trust account balance per 1000
adults
This is the current cumulative balance in
the mobile money trust accounts held by
commercial banks in the name of mobile
money agents and customers divided by 1000
adults
CBL
The number of registered
mobile money customers
The total number of customers(users) who
have registered for both M-pesa and Eco-
cash
CBL
Broad money
Broad money (M2) in Lesotho consists of M1
and quasi money.
CBL
4.2 Model Specification
This study adopts the approach followed Andrianaivo and Kpodar (2011, 2012) and Lundqvist
and Erlandsson (2014) who examine the relationship between financial inclusion and mobile money
by using the following bivariate econometric model:
10
The sum of customer withdrawals and deposits.
22
where denotes natural logarithms of financial inclusion, represents natural
logarithms of indicators of mobile money and denotes the random error term.
4.3 Estimation Strategy: Autoregressive Distributed Lag (ARDL) Bounds Testing
Procedure and Granger Causality
4.3.1 Primary Model( Bivariate Model)
In examining the relationship between financial inclusion and mobile money, the study
estimates equations (1). This is done in three steps. First, the paper determines the order of
integration of the variables using the Augmented Dickey Fuller and Phillips and Perron (1988) tests.
The latter test is also used in addition to the former as it caters for serial correlation, endogeneity of
regressors and allows for the possibility of heteroskedastic disturbance terms (Hamilton, 1994).
While acknowledging the fact that the autoregressive distributed lag (ARDL) bounds testing allows
for the presence of I(0), I(1) or mixed integrated variables in the estimation, pre-testing of the order
of integration of the variables to ensure the absence of I(2), whose presence would nullify the
procedure.
Second, after establishing the integration properties of variables, the study employs ARDL
bounds testing approach to cointegration developed by Pesaran and Shin (1999) and advanced by
Pesaran et al (2001) to study the existence long-run relationship between mobile money and financial
inclusion. This procedure is preferred to other cointegration techniques due to the several
advantages. For example, ARDL bounds testing is applicable irrespective of whether the underlying
regressors are I(0), I(1) or mutually cointegrated. In addition, this procedure still remains robust for
cointegration analysis in empirical macroeconomic studies where small samples size is a common
phenomenon. Furthermore, it also has finite-sample critical values as opposed to other cointegration
approaches for which the distribution of the test statistics may be unknown in finite-samples. In
particular, Narayan (2005) develops a set of sample-specific critical value bounds for the sample
sizes ranging from 30 to 80 using the same approach and GAUSS code used by Pesaran et al (2001)
in generating the asymptotic values. Furthermore, this technique generally provides unbiased
estimates of the long-run model and valid -statistics even in the presence of endogenous regressors.
.
The paper transforms the financial inclusion model (equation 1) into an ARDL framework as
follows:
23
where all variables are as previously defined, Δ is the first difference operator, is the lag
length, ’s are parameters to be estimated, and is a white-noise error term. Similarly, the other
variable in equation (2) as a dependent variable, the other equation can also be estimated.
According to the ARDL bounds testing procedure, cointegration test between the variables is
conducted using the Wald test (F-statistic). The test imposes restrictions on the estimated long-run
coefficients of one period lagged level of the variables to be equal to zero. The two sets of critical F-
values (lower and upper bound values) for a given level of significance are reported by Pesaran et al
(2001) for large sample sizes and Narayan (2005) for small sample data. The lower bound values
assume that all variables in the ARDL model are integrated of order zero, or I(0), while the upper
bound values assume that the variables are integrated of order one, or I(1). Therefore, if the
computed F-statistic is below the lower bound value, I(0), the null hypothesis of no cointegration
cannot be rejected. Conversely, if the computed F-statistic exceeds the upper bound value, I(1), the
null hypothesis is rejected and it is concluded that the variables are cointegrated. Nevertheless, the
result becomes inconclusive if the F-statistic falls between the two bounds.Once cointegration has
been established between the variables using ARDL bounds testing procedure, then the next step is
to estimate the long-run and short-run error correction models from the established cointegration
regression. The long-run model and the associated error correction model are given by:
Where all variables are as previously defined, ’s and ’s are the parameters to be estimated,
and are the lag lengths and is the coefficient of the error correction term, which measures
the speed of adjustment to the long-run equilibrium following a shock to the system.
According to Granger (1969 and 1988) cointegration among the variables may imply the
existence of causality between the variables at least in one direction. However, it does not indicate
24
the direction of causality between the variables. Therefore, once cointegration has been established
between mobile money and financial inclusion using ARDL bounds testing procedure, then the
third step is to employ a dynamic Granger causality test to determine the short-run and long-run
causal effects between mobile money and financial inclusion. For this purpose, the error correction
model (equation 4) is used to examine Granger causality from mobile money to financial inclusion.
In this test, the short-run causality is implied by the significance of the statistic (or Wald statistic)
on the first differences of lagged independent variables. On the other hand, the long-run causality is
captured by the significance of the -statistic on the coefficient of the lagged error correction term.
Nevertheless, if there is no cointegration between the variables only short-run causality can be
determined.
4.3.2 Robustness Checks(Trivariate Model)
In addition to estimating the bivariate relationship between financial inclusion and mobile
money, the study also estimates this relationship in the context of a trivariate model where broad
money (M2) is used as a control variable. This is done to avoid omission of variable bias inherent in
a bivariate model, which may lead to unreliable results (Lütkepohl, 1982). The use of broad money
(M2) is motivated by its high correlation with gross domestic product (GDP)
11
because data for the
latter is only available annually in Lesotho. For this purpose, cointegration between financial
inclusion, mobile money and M2 is established using the following ARDL model:
Where denotes natural logarithms of broad money (M2), Δ is the first difference
operator, is the lag length, ’s are parameters to be estimated, and is a white-noise error term.
This test is conducted using the steps discussed earlier. Once this is done, the following trivariate
long-run and short-run models are estimated.
11
The correlation between M2 and GDP is approximately 99%.
25
Where all variables are as previously defined, ’s and ’s are parameters to be estimated, and
is the speed of adjustment to the long-run equilibrium following a shock to the system. The
estimated trivariate error correction model (equation 7) is also used to examine the short-run and
long-run Granger causality from mobile money to financial inclusion.
5. ANALYSIS OF EMPIRICAL RESULTS
5.1 Unit Root Test Results
As a standard practice in time series analysis, the unit root properties of the each series are
studied. The test results are presented in Table 3. The results show that all the variables used in the
study are integrated of order one, that is except the log of the number of mobile money
customers, log of domestic money transfers per 1000 adults and log of airtime purchases per 1000
adults, which are Therefore, the case of a mixed order of integration of the variables, and
, has been established.
Table 3: ADF and PP Unit Root Test Results
Variable
Variable in levels
Variable at first differences
Conclusion on
order of
integration
ADF statistic
PP statistic
ADF statistic
PP statistic
-1.1503
(0.6817)
-1.1247
(0.6922)
-6.6015*
(0.0000)
-6.8463*
(0.0000)
I(1)
-1.7384
(0.4022)
-1.4626
(0.5379)
-6.7076*
(0.0000)
-12.0432*
(0.0000)
I(1)
-0.0362
(0.9474)
0.1257
(0.9623)
-8.0843*
(0.0000)
-7.8044*
(0.0000)
I(1)
-1.0243
(0.7309)
-0.7022
(0.8309)
-4.9448*
(0.0005)
-9.3211*
(0.0000)
I(1)
-2.2781
(0.1853)
-3.2690**
(0.0260)
I(0)
-4.4050*
(0.0017)
-4.3937*
(0.0017)
I(0)
-1.3231
(0.6045)
-2.1980
(0.2112)
-5.6164*
(0.0001)
-5.7529*
(0.0001)
I(1)
-3.7548*
(0.0083)
-4.1114*
(0.0035)
I(0)
-1.9792
(0.2936)
-1.9792
(0.2936)
-7.6065*
(0.0000)
-8.9371*
(0.0000)
I(1)
-2.4652
-2.4652
-6.0511*
-6.5952*
I(1)
26
(0.1340)
(0.1340)
(0.0000)
(0.0000)
-1.6307
(0.4537)
-1.6307
(0.4537)
-6.3914*
(0.0000)
-7.0360*
(0.0000)
I(1)
-2.0451
(0.2689)
-7.4735
(0.0000)
-1.9559
(0.3035)
-11.3808
(0.0000)
I(1)
Note: Values in parentheses are probability values. * and ** denote the level of statistical significance at 1 and 5% ,
respectively. The variables = log of total credit per 1000 adults, log of total deposit per 1000 adults, log
of the number of ATMs per 1000 adults, log of the number of POSs per 1000 adults, log of trust account
balances per 1000 adults log of mobile money customers, log of domestic money transfers per 1000 adults,
log of airtime purchases per 1000 adults, log of customer deposits per 1000 adults, log of customers
withdrawals per 1000 adults log of amount of float per 1000 adults. In addition, denotes the log of M2.
5.2 The Relationship between Financial Inclusion and Mobile Money
5.2.1 The Long-run Relationship – Cointegration Results
Appendix 3 presents the results of ARDL bounds testing between financial inclusion and
mobile money tested in the context of a bivariate model. The results indicate that the calculated F-
statistic is greater than the upper bound critical value at either 1% or 5% levels of significance when
financial inclusion is a dependent variable in each model. Hence, the null hypothesis of no
cointegration is rejected in all models. Similarly, the existence of long-run relationship is also
obtained even in the case of trivariate models; four models where log of total credit is a dependent
variable and two models where the log of the number of POSs is a dependent variable (see appendix
5)
12
. Therefore, there is a strong evidence of long-run steady state relationship between financial
inclusion and mobile money both in a bivariate and trivariate setting in Lesotho.
The results of the long-run estimates of the bivariate models are presented in Tables 4 and 5
(also see Appendix 4) and those from a trivariate model are presented in Appendix 6. The results
show that the long-run coefficients are not only positive but also statistically significant at 1% level
of significance in all models, and therefore consistent with a priori expectations. Thus, this finding
suggests that all explanatory variables representing mobile money determine financial inclusion in
the long-run in Lesotho in the bivariate model. For instance, an increase in trust account balances
avails more funds to the banking industry, which can be used for credit extension. In addition, the
higher the proportion of mobile money users in the country, the higher the number people with
access to some form of financial services. The long-run estimates of a trivariate model support the
established long-run relationship with positive and statistically significant coefficients.
12
However, the rest of the trivariate models where the logarithms of total credit and number of POSs are dependent variables, which
were reported in the bivariate case, did not produce robust results and therefore are not reported in the paper. In the same manner,
the trivariate models where logarithms of total deposit and number of ATMs are dependent variables are not reported by the paper
because did not produce robust results.
27
Table 4: Total Credit Models (bivariate case)
Relation
Horizon
Explanatory
Variable
Dependent Variable, Log of total Credit
Short-run
-0.3387*
(0.0010)
-0.2881**
(0.0105)
-0.2933**
(0.0165)
-0.2993*
(0.0018)
-0.2819*
(0.0004)
-0.5872*
(0.0021)
-0.3659*
(0.0011)
0.0243***
(0.0913)
0.1151*
(0.0026)
0.0257***
(0.0797)
0.0172**
(0.0200)
0.0404
(0.7738)
0.2564**
(0.0208)
0.2676**
(0.0240)
0.0131**
(0.0215)
0.0463*
(0.0029)
0.0129*
(0.0042)
-0.0309*
(0.0000)
-0.0293*
(0.0000)
-0.0267*
(0.0000)
-0.0312*
(0.0000)
-0.0299*
(0.0000)
-0.0274*
(0.0000)
-0.0283*
(0.0000)
-0.0232*
(0.0000)
-0.0212*
(0.0000)
-0.0208*
(0.0000)
-0.0249*
(0.0000)
-0.0239*
(0.0000)
-0.0290*
(0.0000)
-0.0229*
(0.0000)
1.1881**
(0.0163)
Long-run
0.0764*
(0.0000)
0.1976*
(0.0000)
0.0545*
(0.0000)
0.0788*
(0.0000)
0.0594*
(0.0000)
0.0654*
(0.0000)
0.0316*
(0.0000)
-0.0743*
(0.0000)
-0.0653*
(0.0034)
-0.0522*
(0.0010)
-0.0862*
(0.0001)
-0.0751*
(0.0021)
-
0.0329*
(0.0043)
-0.0579*
(0.0005)
-0.0564*
(0.0000)
-0.0489*
(0.0062)
-0.0423*
(0.0001)
-0.0728*
(0.0000)
-0.0611*
(0.0002)
-
0.0383*
(0.0003)
-0.0488*
(0.0000)
28
6.3061*
(0.0000)
5.4774*
(0.0000)
6.3609*
(0.0000)
6.3739*
(0.0000)
6.3485*
(0.0000)
6.3591*
(0.0000)
Diagnostics
JB
0.3807
(0.8267)
0.4087
(0.8152)
1.2081
(0.5465)
0.3588
(0.8358)
0.5760
(0.7498)
0.6145
(0.7355)
0.8277
(0.6611)
BG-LM test
1.7161
(0.4240)
7.7595
(0.1008)
10.5822
(0.1022)
0.8073
(0.6679)
1.2754
(0.5285)
2.4491
(0.2939)
3.2395
(0.1979)
RESET
test(F-statistic
0.2733
(0.6061)
0.3278
(0.5725)
1.8590
(0.1859)
0.8073
(0.6679)
0.2162
(0.6465)
1.0866
(0.3103)
0.0596
(0.8093)
Note: *, ** and ***denote the level of statistical significance at 1%, 5% and 10%, respectively. The
variables,,,,, and are defined as previously. D2014M2 and D2014M5 are dummy
variables representing decline in total credit owing to stringent requirements by some banks. The values in parentheses
are the probability values.
Table 5: ATMs & POS Models (bivariate case)
Relation
Horizon
Explanator
y Variable
Dependent Variable, Log of the
number of ATMs
Dependent Variable, Log of the
number POSs per 1000 adults)
Short-run
-0.3964*
(0.0007)
-0.1635*
(0.0000)
-0.2925*
(0.0025)
-0.4349**
(0.000)
-0.4034*
(0.0000)
-0.3374*
(0.0001)
0.0248**
(0.0361)
0.0345*
(0.0071)
0.0131**
(0.0168)
-0.0174*
(0.0038)
0.0292**
(0.0336)
0.0356**
(0.0198)
0.0469
(0.7306)
0.1441
(0.3264)
0.0651
(0.6811)
0.0141**
(0.0112)
-0.0356*
(0.0000)
-0.0344*
(0.0000)
-0.0329*
(0.0094)
Long-run
0.0698*
(0.0000)
0.1184*
(0.0000)
0.0858*
(0.0001)
0.0797*
(0.0000)
0.1347*
(0.0000)
0.0697*
(0.0000)
-0.1077*
(0.0005)
-0.0999*
(0.0042)
-0.1109*
(0.0094)
-1.0876*
(0.0000)
-1.1598*
(0.0000)
-0.1378*
(0.0000)
-0.4780*
(0.0000)
-0.5678*
(0.0000)
-0.5948*
(0.0000)
Diagnostics
JB
3.3114
(0.1910)
1.2398
(0.5380)
4.2663
(0.1185)
1.8559
(0.3954)
1.2775
(0.5280)
2.5985
(0.2727)
BG-LM
Test
5.7391
(0.1250)
2.2872
(0.3187)
6.6705
(0.1544)
2.5594
(0.1096)
1.7548
(0.4159)
2.1869
(0.3351)
RESET
1.2528
1.6367
1.3367
1.6804
0.0066
0.1451
29
Test
(0.1300)
(0.1250)
(0.1150)
(0.2083)
(0.9359)
(0.7071)
Note: *, ** and ***denote the level of statistical significance at 1%,5% and 10%, respectively. The variables log of
the number of ATMs, the log of the number of point of sales (POS) devices and other variables are as
previously defined. D2015M9 is the dummy variable representing the decline in the number of POS as a result of
cancellation of contract between some merchants and one commercial bank, which owned the POSs. The values in
parentheses are the probability values.
5.2.2 Diagnostic Tests Results
Following the establishment of long-run estimates in each model, the next step involves the
estimation of the error correction model (ECM). The results of short-term elasticities estimated
within the ARDL framework together with their associated diagnostic tests also are presented in
Tables 4 and 5 (also see Appendices 4 and 6). Diagnostic tests were applied to the estimated ECMs
to ensure the reliability of the estimated parameters. The results show that all estimated ECMs pass
all specification tests. For example, the findings show absence of serial correlation, normality of
residuals and no heterokesdasticity (as the models were estimated using White’s heteroskedasticity
standard errors). In addition, Ramsey’s RESET test for the stability of the models together with
CUSUM and CUSUMQ tests (though not presented here) suggest that the models are stable over
the sample period.
5.2.3 The Short-run Relationship
Consistent with the long-run dynamics, the results of estimated short-run elasticities show that
mobile money influences financial inclusion in Lesotho. This is supported by the positive and
statistically significant coefficients of the explanatory variables in all bivariate error correction
models and triavariate error correction models that passed the robustness checks. This finding
provides evidence that in addition to influencing the dynamics of financial inclusion in the long-run,
mobile money also has significant impact on the dynamics of financial inclusion in the short-run.
The findings also show that the coefficient of the lagged error correction term, which indicates the
speed of adjustment to long-run equilibrium in the event of a shock to the system, is negative and
statistically significant at either 1% or 5% level of significance. This suggests that in the bivariate
setting, on overage, 16% to 85% (depending on proxies that are used) of the disequilibrium of
financial inclusion is corrected in the current month following a shock in the previous month on the
one hand. On the other hand, the speed of adjustment in the trivariate error correction models
implies that, on average, 14% to 50% of the disequilibrium from the previous month is corrected in
the current month. In addition, the fact that the coefficient of the lagged error correction term is
statistically significant and bears a correct sign (i.e. negative) in all models implies that the series are
30
non-explosive and that long-run equilibrium is attainable. Therefore, this is consistent with the
cointegration relationship between the variables in each model.
5.2.4 Granger Causality between Financial Inclusion and Mobile Money
The existence of a cointegrating relationship between the variables may suggest that there must
be Granger causality in at least one direction, but does not show the direction of temporal causality
between the variables (see Granger, 1969 & 1988). Therefore, the paper employs the estimated error
correction models to also examine both short-run and long-run Granger causality between financial
inclusion and mobile money. The short-run causality can be determined by the significance of the
Wald F-test (or t-statistic) on the first differences of the explanatory variables on one hand. On the
other hand, the long-run causality can be examined by the significance of the t-statistics on the
coefficient of the lagged error correction term. Granger causality can be unidirectional in either
directions or bidirectional. However, this paper specifically focuses on establishing unidirectional
Granger causality from mobile money to financial inclusion, which answers the research question in
this study.
Based on the estimated error correction models presented in Tables 4 and 5 (also see
Appendices 4 and 6), the coefficients of all the first differences of explanatory variables in each
model appear with expected positive signs and are also statistically significant at either 5% or 1%
levels of significance. This result provides evidence of short-run Granger causality from mobile
money to financial inclusion. Similarly, the negative and statistically significant coefficient of the
lagged error correction term in the same models supports long-run Granger causality from mobile
money to financial inclusion. Thus, in general the findings imply that indeed mobile money Granger
causes financial inclusion both in the short-run and long-run in Lesotho.
6. CONCLUSION AND POLICY RECOMMENDATIONS
The acquisition and use of mobile telephone in Sub-Saharan Africa has grown significantly in
recent years and now covers a large proportion of the region’s population. This has led to
emergence of financial innovations such as mobile money, which has expanded the grid of financial
services to include the previously unbanked and underbanked sections of population, who could not
access formal financial services on account of limited banking infrastructure. Empirical evidence has
shown that this new development has increased financial intermediation with positive spill-overs in
terms of credit growth to entrepreneurs and consequently leads to faster economic growth and
31
perhaps broader economic development. This study employs ARDL bounds testing approach to
cointegration and Granger causality test based on ECM to examine the impact of mobile money on
financial inclusion and the direction of causality between these variables in Lesotho using monthly
data from July 2013 to December 2015.
The findings suggest a strong evidence of long-run steady state relationship between financial
inclusion and mobile money in Lesotho with positive and statistically significant long-run
coefficients, which are consistent with apriori expectations. In addition, the estimated ECM models
provide evidence that mobile money also has significant impact on the dynamics of financial
inclusion in the short-run in Lesotho. For instance, the results suggest that, on overage, 16% to 85%
of the disequilibrium of financial inclusion is corrected in the current month following a shock in
the previous month. Furthermore, the findings show that mobile money Granger causes financial
inclusion both in the short-run and long-run in Lesotho.
The findings of this paper underscore the importance of mobile phone diffusion and hence
mobile money in extending financial services in Lesotho. This is because it has resolved the hurdles
of limited banking infrastructure by allowing the previously unbanked and under banked sections of
the population to access financial services. This could also serve as a breakthrough for these people
to build accounts history that would consequently help them to open formal bank accounts with the
banking industry in Lesotho. Therefore, policy makers in Lesotho should promote and facilitate
interaction and investments in mobile phone technology deployment and its related financial
services. In addition, financial inclusion policies should be directed to leveling the playing ground for
mobile money to flourish to create a more financially inclusive society in Lesotho. In this regard, the
legal and regulatory framework should be friendly and accommodative to enable more innovation in
mobile money and other digital financial services. This would contribute drastically to financial
development and consequently faster economic growth.
MNOs should work hard to scale up the use of mobile money in remote areas of the country,
where the majority of people still do not have access to financial services. This could be achieved
through more customer education, improving network coverage in rural areas of the country and
growing agent network in rural areas by negotiating with Chinese businesses, which have more reach
in rural communities, to become agents and hence act as cash-in and cash-out points. More
32
importantly, the MNOs should endeavor to promote the use of mobile money in its electronic form
in carrying out transactions. These would help resolve many of the hurdles related to liquidity
management by the MNOs. Lastly, MNOs should work towards forming many partnerships with all
commercial banks and other financial institutions in Lesotho to ensure interoperability between
MNOs and commercial banks
13
. This would lead to more access to banking services and allow
innovation of more services.
REFERENCES
African Development Bank. (2013). Financial Inclusion and Integration through Mobile Money and
Transfer. A Paper Sponsored by India-Africa Economic Cooperation Funds.
Alleman, J. & Rappoport, P. (2010). Mobile Money: Implications for Emerging Markets,
Communications and Strategies, Issue 79, pp. 15-28.
Alexandre,C. & Eisenhart,L.C.(2013). Mobile Money as an Engine of Financial Inclusion and
Lynchpin of Financial Integrity, Washington Journal of Law,Technology and Arts. Vol.8, Issue
3,pp.285-302.
Andrianaivo,M. & Kpodar,K.(2012). Mobile Phones, Financial Inclusion, and Growth, Review of
Economics and Institutions,Vol.3,No.2, pp.1-29.
Andrianaivo,M. & Kpodar,K.(2011). ICT, Financial Inclusion, and Growth: Evidence from African
Countries. IMF Working Paper, WP/11/73.
Beck,T. & Maimbo,S.M.(2013). Financial Sector Development in Africa: Opportunities and
Challenges. The World Bank,pp. 41-76.
Bhatia, D. Bhvanani, A., Chiu, R. W., Janakiram, S., & Silarszky, P.(2008). The role of Mobile
Phones in sustainable Rural Poverty Reduction, ICT Policy Division, Global Information and
Communication Departmment (CGICT).
Boston Consulting Group. (2012). The Socio-economic Impacts of Mobile Financial Services:
Analysis of Pakistan, Bangladesh, India, Serbia and Malaysia,pp.1-16. Available online at
http://www.telenor.com/wp-content/uploads/2012/03/The-Socio-Economic-Impact-of-
Mobile-Financial-Services-BCG-Telenor-Group-2011.pdf
Breitbach,E. & Walstad, W.B. (2014/2015). Financial Literacy and Bank Affiliation: Results for the
Unbanked, Underbanked, and Fully Banked, Perspectives on Economic Education Research, Vol. 9,
No.1, pp. 21-35.
Bilodeau,J. Hoffman,W. & Nikkelen, S.(2011). The Seven Pillars of Mobile Financial Services
Development, The Mobile Financial Services Development Report 2011. Available online at
http://www3.weforum.org/docs/WEF_MFSD_Report_2011.pdf
Cardo, M. (2014). Social Inclusion and Policymaking in South Africa: A Conceptual Overview, The
Journal of the Helen Suzman Foundation, Issue 73, pp. 9-15.
Central Bank of Lesotho. (2013). Annual Report. Central Bank of Lesotho, Maseru.
CGAP. (2011). Technology Program Country Note,pp.1-14,Available online at
http://www.cgap.org/sites/default/files/CGAP-Technology-Program-Country-Note-Ghana-
Jun-2011.pdf.
13
Of course, there are already interesting developments on the ground with Ecocash customers having access to their
bank accounts with SLB.
33
De Koker, L. & Jentzsch, N. (2013). Financial Inclusion and Financial Integrity: Aligned Incentives?,
World Development, Vol. 44, pp. 267–280.
Donovan,K.(2012).Mobile Money for Financial Inclusion, Information and Communication Development,
pp. 1-14.
Donner, J., & Tellez, A. (2008). Mobile banking and economic development: Linking Adoption,
Impact, and Use, Asian Journal of Communication, Vol. 18, Issue 4, pp. 318-332
Dube,T., Chtakunye, P. & Chummun, B.Z.(2014). Mobile Money as a Strategy for Financial
Inclusion in Rural Communities, Mediterranean Journal of Social Sciences, Vol.5, No.25, pp. 216-
224.
Eijkman, F., Kendall, J., & Mas, I. (2010). Bridges to Cash: The Retail End of M-pesa, Savings and
Development, Vol.34, No.2, pp. 219–252.
Finmark Trust (2011). FinScope Consumer Survey Lesotho 2011 Report. A Survey Commissioned
by the Central Bank of Lesotho. Available online at http://www.finmark.org.za/wp-
content/uploads/pubs/RepFSLESOTHO_FNL_2012.pdf
Flores-Roux, E.M. & Mariscal, J. (2010). The Enigma of Mobile Money Systems, Communications and
Strategies, No.79, pp.41-62.
Goss, S., Mas, I., Radcliffe, D., & Stark, E. (2011). The Next Challenge: Challenging Savings
Through Mobile Money Schemes, WorldEconomic Forum Report, April 2011.
Grameen Foundation. (2014). Mobile Money in Nigeria: The User Experience, pp. 1-17. Available
online at
http://www.grameenfoundation.org/sites/grameenfoundation.org/files/resources/Nigeria%2
0Landscape%20Report%20FINAL%20Dec%2013%202014.pdf.
Granger, C.W.J. (1988). Causality, Cointegration, and Control, Journal of Economic Dynamics and
Control, Vol. 12,pp. 551-559.
Granger, C.W.J. (1969) . Investigating Causal Relations by Econometric Models and Cross-spectral
Methods, Econometrica, Vol. 37,No.3,pp. 424-438.
Gross, M., et. al. (2012). Use of Financial Services by the Unbanked and Underbanked and the
Potential for Mobile Financial Services Adoption, Federal Reserve Bulletin,Vol. 98, No.4, pp. 1-
20.
GSMA.(2012a). What Makes a Successful Mobile Money Implementation? Learning from M-pesa in
Kenya and Tanzania, Mobile Money for the Unbanked.pp.1-10. Avalaible online at Mobil
http://www.gsma.com/mobilefordevelopment/wp-content/uploads/2012/03/What-makes-
a-successful-mobile-money-implementation.pdf.
GSMA.(2012b). What is the Impact of Mobile Telephony on Economic Growth, A Report for the
GSM Association. Available on line at http://www.gsma.com/publicpolicy/wp-
content/uploads/2012/11/gsma-deloitte-impact-mobile-telephony-economic-growth.pdf
GSMA.(2015).The Mobile Economy: Sub-Saharan Africa 2015. Available online at
https://gsmaintelligence.com/research/?file=721eb3d4b80a36451202d0473b3c4a63&downlo
ad.
Gwalani,H. & Parkhi,S.(2014). Financial Inclusion – Building a Success Model in the Indian
Context, Procedia -Social and Behavioral Sciences, Vol. 133, pp.372-378.
Gugler, J. (2002). The Son of the Hawk does not Remain Abroad: The Urban-Rural Connection in
Africa. African Studies Review, Vol.45, No. 1, pp. 21-41.
Hamilton, D. J. (1994). Time Series Analysis. New Jersey: Princeton University Press.
Hinson, E. R. (2011). Banking the Poor: The Roles of Mobiles, Journal of Financial Services Marketing,
Vol.15, pp. 320-333.
34
Ingba, S.I. (2014). Mobile Money Operation in Nigeria- The Journey so far. A paper
UNCTAD/ECOWAS Seminar on the Harmonisation of Cyber Legislation in ECOWAS ,
ACCRA (Ghana), 18-21 MARCH, 2014. Available online at
https://tftcal.unctad.org/mod/resource/view.php?id=6284.
IOS Press. (2012). E-Government Developments: Ensuring Financial Inclusion via Mobile Money,
Journal of E-Governance, Vol.35, pp.71-76.
Jefferis, K. and Manje, L. (2014). Making Access Possible: Lesotho Country Diagnostic Report. A
Report Commissioned by Finmark Trust, Centre for Financial Regulation and United National
Capital Development Fund. Available online at http://www.finmark.org.za/wp-
content/uploads/pubs/MAP-Lesotho_Diagnostic-report_final_2014.pdf.
Jenkins, B. (2008). Developing Mobile Money Ecosystems, Washington DC: International Finance
(IFC) and the Havard Kennedy School.
Jack, W. & Suri, T. (2011). Mobile Money: The Economics of M-PESA, National Bureau of Economic
Research (NBER) Working Paper, No.16721.
Kasseeah, H., & Tandrayen-Ragoobur, V. (2012). Mobile Money in an Emerging Small Island
Economy, ARPN Journal of Science and Technology, Vol. 2, No.5, pp.454-458.
Kempson,E.(2006). Policy Level Response to Financial exclusion in Developed Economies: Lessons
for Developing Countries. Report for Personal Finance Research Centre, University of Bristol.
Available online at http://www.bristol.ac.uk/
Ketly,R. & Kasi,S.(2015). Action Plan for Electronic Money (e-money) in Lesotho. Paper
Commissioned by Finmark Trust.
Lake,A.J.(2013). Risk Management in Mobile Money: Observed Risks and Proposed Mitigants for
Mobile Money Operators, Internation Financial Corporation Paper.Available online at
http://www.ifc.org/wps/wcm/connect/37a086804236698d8220ae0dc33b630b/Tool+7.1.+R
isk+Management.pdf?MOD=AJPERES
Lal,R. & Sachdev,I.(2015). Mobile Money Services – Design and Development for Financial
Inclusion. Working Paper 15-083. Harvard Business School. Avaiable online at
http://www.hbs.edu/faculty/Publication%20Files/15-083_e7db671b-12b2-47e7-9692-
31808ee92bf1.pdf
Lesotho Communications Authority(2013/2014). Annual Report 2013-2014. Maseru,Lesotho.
Available online at
http://www.lca.org.ls/images/documents/LCA%20Annual%20Report%202013_14.pdf.
Lundqvist,M. & Erlandson, F.(2014). The Diffusion of Mobile Phones and its Impact on Financial
Inclusion and Economic Growth in Africa. Master’s Thesis. Lund University, School of
Economic Management.
Lütkepohl, H. (1982). Non-causality due to Omitted Variables, Journal of Econometrics,
Vol. 19, pp.367-378.
Merritt, C. (2010). The Next Phase in the Evolution in Person-to-Person Payments, Retail Payments
Forum White Paper. Federal Reserve Bank of Atlanta.
Motsune,T.(2015). No Kenyan Left Behind: The Model of Financial Inclusion Through Mobile
Banking, Review of Business and Finance Studies.Vol.6,No.1,pp.35-42.
Morawczynski, O., & Pickens, M. (2009). Poor People Using Mobile Financial Services: Observation
on Customer Usage and Impact from M-Pesa, CGAP.
35
Morawczynski,O. (2010). Examining the Adoption, Usage and Outcomes of Mobile Money
Services: The Case of M-Pesa in Kenya, PhD Thesis, Science and Technology Studies-
University of Edinburgh.
Munyegera, G.K. & Matsumoto, T. (2014). Mobile Money, Remittances and Household Welfare:
Panel Evidence from Uganda, GRIPS Discussion Paper 14-22.Available online at
http://www.grips.ac.jp/r-center/wp-content/uploads/14-22.pdf.
Nandhi, M., A. (2012). Effects of Mobile Banking on the Savings Practices of the Low-income
Users: The Indian Experience, Institute for Money, Technology & Financial Inclusion
(IMTFI), Working Paper No.7, Jesus and Mary College, University of Delhi, India.
Narayan, P.K. (2005). The Saving and Investment Nexus for China: Evidence from Cointegration
Tests, Applied Economics, Vol. 37, and pp. 1979-1990.
Oji, C.K. (2015). Promoting Financial Inclusion for Inclusive Growth in Africa, South African Institute
of International Affair Occasional Paper, No.210, pp. 1-18.
.
Oritin, P., Quisenbery, W. & Sun, T. (2013). A Study on Factors Facilitating Access to Mobile
Phone Money in Uganda, Greener Journal of Business and Management Studies, Vol.3, No. 3, pp.
279-291.
Pesaran, M.H. and Shin, Y. (1999). An Autoregressive Distributed-led Modelling Approach to Cointegration
Analysis, Cambridge University Press, Cambridge.
Pesaran, M.H., Shin, Y. & Smith, R.J. (2001). Bounds Testing Approaches to the Analysis of Level
Relationships, Journal of Applied Econometrics, Vol. 16, pp. 289-326.
Phillips Consulting. (2013). Current Trends in Mobile Money in Nigeria, Mobile Money Report.
Available online at http://www.phillipsconsulting.net/files/mobile-money-report2013.pdf.
Phillips, P. and Perron, P. (1988). Testing for Unit Root in Time Series Regression, Biometrica,
Vol.75, pp. 335-346.
Ramada-Sarasola, M.(2012). Can Mobile Money Systems have a Measurable Impact on Local
Development? Innovation & Research Multiplier and Social Trade Organisation for the
International Development Research Centre. Available online at
http://ict4cct.org/sites/dms/files/4%20%20%20MMS%20and%20Local%20development.pd
f.
Rasmussen, S. (2009). Delivering Successful Mobile Money Solutions to the Unbanked - Customer
Acquisition. GSMA Mobile Money Summit, Barcelona.
Sarma,M. & Pais,J.(2011). Financial Inclusion and Development, Journal of International
Development,Vol. 23, pp.613-628.
Sife, S., A., Kiondo, E., & Lyimo-Macha, G. J.(2010). Contributions of Mobile Phones To Rural
Livelihoods and Poverty Reduction in Mororgoro Region, Tanzania, Electronic Journal of
Information Systems in Developing Countries (EJISDC), Vol. 42(3), pp. 1-15.
Tobbin, P. E. (2010). Modeling Adoption of Mobile Money Transfer: A Consumer Behaviour
Analysis, Paper presented at the 2nd International Conference on Mobile Communication
Technology for Development, Kampala, Uganda.
36
Yakub, J.O. Bello, H.T. & Adenuga, I.A. (2013). Mobile Money Services in Nigeria: An Inquiry of
Existing Models, International Journal of Economics and Management Sciences, Vol. 2, No.9, pp. 94-
105.
37
LIST OF APPENDICES
Appendix 1: Econet Telecom Lesotho (ETL) Coverage Map as at 31st March 2014
Source: Lesotho Communications Authority Annual Report, 2013-2014
Appendix 2: Vodacom Lesotho (VCL) Coverage Map as at the 31st March 2014
Source: Lesotho Communications Authority Annual Report, 2013-2014
Appendix 3: ARDL Bounds Testing to Cointegration Results (bivariate case)
Total Credit Models
Model
-
statistic
Critical value bounds of the F-statistic
Evidence of
Cointegration?
b)
17.57*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
14.67*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
38
t)
9.72*
8.170
9.285
5.395
6.350
4.290
5.080
Yes
)
16.20*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
12.26*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
14.04*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
14.78*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
Total Deposit Models
)
6.24**
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
6.31**
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
5.16**
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
6.24**
6.027
6.760
4.090
4.663
3.303
3.797
Yes
ATMs and POS Models
)
9.47*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
9.37*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
7.76*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
17.56*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
15.19*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
)
13.34*
6.027
6.760
4.090
4.663
3.303
3.797
Yes
Note: 1) is the number of regressors and 2) * and ** denote the level of statistical significance at 1% and 5%,
respectively.
Appendix 4: Total Deposit Models (bivariate case)
Relation
Horizon
Explanatory
Variable
Dependent Variable, Log of Total Deposits per 1000 adults
Short-run
-0.8519*
(0.0002)
-0.8063*
(0.0002)
-0.7928*
(0.0005)
-0.8003*
(0.0004)
-0.0443
(0.1188)
0.0178
(0.4890)
0.0128
(0.2759)
0.1685***
(0.0552)
Long-run
0.0518*
(0.0000)
0.0446*
(0.0000)
0.0258*
(0.0000)
0.1700*
(0.0000)
3.7045*
(0.0000)
3.7429*
(0.0000)
3.7008*
(0.0000)
2.9298*
(0.0000)
Diagnostics
JB
0.3983
(0.8194)
0.2679
(0.8746)
0.3557
(0.8371)
0.3557
(0.8371)
BG-LM Test
0.8984
(0.6381
1.7444
(0.4180)
3.2587
(0.1961)
0.9878
(0.6102)
RESET test
0.0143
0.0323
0.1107
0.0177
39
(0.9057)
(0.8588)
(0.7421)
(0.8952)
Note: *, ** and *** denotes the level of statistical significance at 1%, 5% and 10%, respectively. The variables are as
previously defined. The values in parentheses are the probability values.
Appendix 5: ARDL Bounds Testing to Cointegration Results (trivariate Case)
Total Credit Models
Model
-
statistic
Critical value bounds of the F-statistic
Evidence of
Cointegration?
,)
9.65*
5.155
6.265
3.538
4.428
2.915
3.695
Yes
)
11.46*
5.155
6.265
3.538
4.428
2.915
3.695
Yes
)
10.73*
5.155
6.265
3.538
4.428
2.915
3.695
Yes
)
11.07*
5.155
6.265
3.538
4.428
2.915
3.695
Yes
POS Models
)
9.43*
5.155
6.265
3.538
4.428
2.915
3.695
Yes
16.28*
5.155
6.265
3.538
4.428
2.915
3.695
Yes
Note: 1) is the number of regressors and 2) * and ** denote the level of statistical significance at 1% and 5%,
respectively.
Appendix 6: Total Credit and POS Models (trivariate Case)
Relation
Horizon
Explanat
ory
Variable
Dependent Variable, Log of total credit
Dependent Variable,
Log of the number
POSs per 1000
adults)
Short-run
-0.3831*
(0.0004)
-0.4959*
(0.0001)
-0.3489*
(0.0048)
-0.4406*
(0.0009)
-0.2744*
(0.0001)
-0.1355*
(0.0000)
0.0118**
(0.0316)
0.0171*
(0.0086)
0.0158**
(0.0281)
0.0083***
(0.0649)
0.1122*
(0.0061)
0.0115*
(0.0088)
0.1722**
(0.0162)
0.1566**
(0.0248)
0.0412**
(0.0685)
0.1429**
(0.0346)
0.1493**
(0.0294)
0.1224***
(0.0668)
-0.1478***
(0.0652)
0.0004
(0.9979)
-0.0262*
(0.0000)
-0.0288*
(0.0000)
-0.0289*
(0.0000)
-0.0247*
(0.0000)
-0.0243*
(0.0000)
-0.0285*
(0.0000)
-0.0221*
(0.0000)
-0.0235*
(0.0000)
-0.0361*
(0.0000)
-0.0367*
(0.0000)
40
Long-run
0.0386*
(0.0015)
0.1104*
(0.0001)
0.0364**
(0.0387)
0.1055**
(0.0250)
0.1542*
(0.0005)
0.0244*
(0.0001)
0.4584**
(0.0153)
0.6598*
(0.0084)
0.2718***
(0.0894)
0.2949**
(0.0102)
0.6433**
(0.0294)
1.1955***
(0.0630)
-0.0583**
(0.0101)
-0.0598*
(0.0014)
-0.0573*
(0.0001)
-0.0477*
(0.0000)
-0.0521**
(0.0051)
-0.0578*
(0.0003)
-0.0461*
(0.0000)
-0.0436*
(0.0000)
-0.0361*
(0.0000)
-0.2930**
(0.0172)
3.2710**
(0.0108)
1.9126
(0.2223)
3.8384*
(0.0037)
4.3659*
(0.0013)
-4.9514**
(0.0128)
-8.6480**
(0.0474)
Diagnostics
JB
0.0250
(0.9876)
0.3364
(0.8452)
0.6649
(0.7171)
0.3387
(0.8442)
0.6063
(0.7385)
1.4491
(0.4846)
BG-LM
Test
0.0889
(0.9565)
0.7029
(0.7037)
3.1935
(0.2026)
1.6178
(0.4453)
2.1697
(0.3380)
1.9632
(0.3747)
RESET
Test
0.0924
(0.7640)
1.82E-05
(0.9966)
0.4036
(0.5318)
0.2120
(0.6497)
0.3896
(0.5396)
0.4011
(0.5328)
Note: * , ** and *** denote the level of statistical significance at 1%, 5%, and 10%, respectively. The values in
parentheses are probability values.