Barriers of Financial Inclusion in Sub-Saharan Africa

Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

Journal of Economics and Sustainable Development www.iiste.

org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Barriers of Financial Inclusion in Sub-Saharan Africa


Dunstone W Ulwodi* Peter W Muriu
School of Economics, University of Nairobi, Kenya

Abstract
This paper examines barriers to financial inclusion across sub-Sahara Africa (SSA) using the 2014 Global
Findex dataset from the demand-side perspective. A two-level model is estimated with individual households
nested in the country level variables. Lower levels of income are associated with lower levels of access to formal
account. Similarly, literacy rates have significant effect on the level of account ownership. Estimation results
also reveals that as individuals grow old, they tend to switch from one form of account to the other. Owning a
debit card is more likely to increase account ownership. Another important barrier to account ownership is
proximity to the nearest financial services. These barriers have the potential to cause market failure and therefore
the right policy interventions are required to stabilize the market.
Keywords: Financial Inclusion, Multilevel Analysis, sub-Saharan Africa

1. Introduction
Financial inclusion is considered a key enabler to economic growth and therefore placed high on the policy
agenda of most developing countries. However, this concept has attracted varied definitions. For instance, a
definition by Sarma (2008) emphasizes the ease with which financial services could be accessed, their
availability and affordability. In addition, the author also finds usage as an important aspect of financial
inclusion. On the other hand, Fungacova and Weill (2014) view the concept within the context of usage rather
than quality of services provided whereas Claessens (2006) views it from the perspective of variety and
affordability of financial services. Demirguc-Kunt and Levine (2008) define financial inclusion as removal of
obstacles in form of price and non-price constraints from financial services. However, the basic understanding
across these definitions places greater emphasis on access, usage and less on quality of financial services1. This
paper adopts access2 and usage as the working definition for financial inclusion. Majority of the adult population
are locked out of the formal financial system that would guarantee better standards of living.
The main objective of this paper is to investigate barriers of financial inclusion across the sub-Saharan
Africa. The global agenda has recently focused on financial inclusion imperatives. The realization that the
financial sector plays a critical role that can spur growth and reduce prevalence of poverty has led to renewed
enthusiasm in financial innovations aimed at bringing more people to the system of formal finance. The global
attention aims at ensuring equitable participation in economic growth through financial inclusivity. Available
empirical evidence provides proof that access and usage of financial services 3 in their variety is welfare
improving (Oya et al., 2011). Conversely, inadequacy of regular and methodical data has been a major hindrance
to detailed examination of financial inclusion.
Although remarkable progress in quantitative access has been made, it is yet to translate into economic
inclusivity in most of the developing countries. Financial systems that are inclusive allow for a wide variety of
financial services to be accessed by most people majority of whom are poor. Historically, financial exclusion has
removed certain segments of the population from engaging within the context of mainstream financial sector.
Financial exclusion limits participation of households in economic growth and development. The potential
inherent in these segments of the population is likely to provide a positive and significant input into the economy.
This is because to achieve inclusive growth, financial inclusion is considered a critical building block.
Uptake of financial services has been rising in most developing countries alongside informal financial
arrangements. As a result, the financial services portfolio does not provide the broad risk-pooling and low-cost
services (Honohan 2005) consequently, exposing individuals to unlimited risks and higher costs associated with
non-standardized services. The informal financial services have continued to co-exist with formal financial
services owing largely to the inability of the financial system to accommodate poor people with low incomes.
Until 2011, little was known about the extent of financial inclusion in most developing countries around the
globe. This was because of the inadequacy of systematic data that could support cross-country analysis. With the
Global Findex data wave of surveys of 2011 and 2014, different aspects of financial inclusion have been
subjected to empirical investigations. However, the number of studies are growing but the focus varies from one
region to the other. For instance, Demirguc-Kunt et al., (2015) focuses on financial inclusion from the global
1
We recommend that the focus of future research should be on affordability (price) and quality of financial services provided when
appropriate data becomes available.
2
The paper uses the terms access and usage interchangeably but in reality, access means the possibility of using a service while usage refers
to the actual use of the service.
3
Financial services refer to the following: credit services and products, savings services and products, payments or transactions services,
insurance and risk management services.

66
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

perspective by providing a cross-section of issues in general. Zins and Weill (2016) focus on some aspects of
financial inclusion in Africa by concentrating on what influences financial inclusion. Their definition of financial
inclusion relate to banking services. Beck and Cull (2015) show that most banks in Africa do not promote
inclusive finance like their counterparts in other parts of the world. On their part, Mlachila et al., (2013) view
distributional skewness as the main reason for slow penetration of financial inclusion. By understanding
financial inclusion differentials across the SSA region, the relevant policies can be formulated to target specific
areas of concern. This is supported by the view that SSA has a myriad of bottlenecks that vary from one country
to the other. This study seeks to address the key determinants and barriers of financial inclusion in sub-Saharan
Africa.

2. Previous Evidence
The relevance of financial sector development is not only being experienced in most advanced economies but in
emerging economies as well (Anderloni and Carluccio 2007). This is mainly because of its role in facilitating
access to credit through widening sources of borrowing, savings mobilisation and cushioning consumers against
risks of exposure. Legal and institutional strength play a critical role towards creating an enabling environment
where financial sector institutions are stable and trusted. Although most developing economies are characterized
by unstable and disjointed financial sector, it is imperative to create and empower institutions that will drive the
reform agenda. The neoclassical economics views institutional theory in the context of price mechanism and
production efficiency but ignores other pertinent parameters such as social norms, behaviour and rules (North
2000; Joskow 2008). Focusing on positive economic theory is not likely to yield the desired results because it
fails to incorporate the morality aspects. It is against this rational thinking that New Institutional Economics
(NIE) has set new benchmarks for determining public reform and policy agenda that will promote efficacy
sectors.
Financial inclusion perception falls within the broader concept of social inclusion whose aim is to ensure
social inclusivity and equal treatment. Social inclusion therefore promotes positive values of equal opportunities
for all while reducing the ever widening gap between the rich and the poor. Therefore, financial inclusion brings
about the total transformation of socioeconomic and financial realities surrounding individuals. Social inclusion
encompasses financial inclusion which affects the daily lives of the people in a positive way. For instance, access
to a wide variety of financial products promotes smoothing of consumption (Honohan and King 2012), savings
mobilisation and reduce exposure to risks.
Literature on development economics has shown that financial progress is key to economic development
(Levine 2005). Consequently, it leads to poverty reduction through provision of financial services (Banerjee and
Newman 1993). Affordable and trusted financial system is vital to facilitating efficient financial market
operations. It is more likely to reduce information costs as well as transaction costs, influence savings
mobilisation and promote better investment decisions (Beck et al., 2009). Further, the financial system
characterised by trust is more likely to promote financial long-run growth and innovations prospects while
facilitating expenditure smoothening as well as risk-pooling effect (Honohan 2005) that will cushion consumers
against unprecedented and unforeseen shocks. It is also a means for accessing other basic utilities such as quality
education and health as well as clean and safe water (Peachey and Roe 2004). Thus, a stable and trusted financial
scheme will control positive strategic decisions while creating a multiplier effect in other areas through its
communication mechanisms.
Financial inclusion imperative is known in academic and policy circles. It has been recognised that for the
growth benefits to trickle down to the lower spectrum, it is vital for financial inclusion to target the people who
are under-served. Financial services access is likely to ease poverty growth if it targets majority of consumers
(Galor and Zeira 1993). Further, constrained access to and utilization of financial services exposes individuals to
risks, decelerates savings rate and leads households to switch to informal network alternatives and access to
credit (Kimenyi and Ndung'u 2009). Love and Bruhn (2009) opine that more financial services access has the
potential to improve the macroeconomic environment and in turn financial inclusion.
Demirguc-Kunt and Klapper (2012) argue that investments have drawn individuals closer to services
provided by commercial banks while expanding individuals' financial literacy. At all income levels, it is
generally acceptable that not all households are sufficiently serviced given the available formal financial
institutions (Shankar 2013). Consequently, utilization of financial services is still limited primarily due to the
supply side constraint. Although access to credit is progressively informing key policy aspects of poverty
alleviation in developing countries, its theoretical underpinning is not apparent. Theoretical literature holds that
the association between access to credit and growth depends largely on the nature of credit in question. It has
also been shown that household investment that results in improved productivity may be driven by access to
credit according to the most popular hypothesized view. Access to savings and credit could potentially reduce
the effects of negative shocks arising from volatilities in income (Eswaran and Kotwal 1990). Access to credit
undermines vulnerabilities to negative shocks thereby enhancing household capability to smooth consumption

67
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

during periods of low income by providing an opportunity to take part in less riskier investments. Financial
services use across most developing countries is subject to supply- and demand-side factors. This paper, however,
looks at demand-side factors and how they are impacted by the supply-side dynamics. According to Ayenew and
Zewdie (2010), availability and accessibility to financial services provides ingredients to nurturing economic
growth and ensuring a decline in income inequalities among households.
Multi-level theories can be used to address multi-level problems which are commonly found in our societies.
Although the multi-level techniques are a recent phenomenon (Huttner and van den Eeden, 1993), they have
nevertheless occupied an important space that was initially void. In a multi-level approach, specification of the
boundaries of variables is necessary so that a clear criterion of assigning measures to sub-groups suffice. Multi-
level theory in effect postulates situations where there are direct effects between individuals and cross-level
effects through relational variables. The advantage of this approach is that no group telepathy is assumed but
more focus is on the organisational structure and its attendant communication channels and processes inherent in
group formations. The multi-level approach is still evolving since the times of Durkheim's sociological concepts.
Overall, this approach is more suitable where the subject of investigation involves different levels in-build into
each other with individual entity influencing the group entity and in turn the group entity influencing the
individual entity. In most multi-level modelling, the outcome variable is to be found at the lowest level (level 1)
of the hierarchy. However, fewer studies applying this approach have investigated how individual variables may
influence group outcomes (DiPrete and Forristal 1994).
A growing body of knowledge has laid more emphasis on how financial inclusion should be measured.
Consequently, more definitions have emerged with each one taking a narrow perspective depending on the stage
of development of the country. With this view, financial inclusion becomes a supply side problem. However, this
perspective fails to recognise individual limitations of households and therefore ignores the role played by the
demand side. More recently, several initiatives have been made by most developing countries to collect
consistent and systematic data on financial inclusion indicators. For instance, The World Bank's Global Findex
data is a demand side data covering many economies around the world. The Global Findex is so far the most
recent data that is comparable across different economies. In the SSA region, a few data collection initiatives
have been spearheaded by Central Banks, Fin Mark Trust Fund and Financial Sector Deepening (FSD) Trust.
While these initiatives are key to understanding the extent of financial inclusion, many of the indicators remain
un-standardized and inconsistent.
The concept of financial inclusion is increasingly gaining worldwide recognition in many policy debates not
just because it brings people to using the financial services but also to choosing among the available options such
as payments, credit and insurance. Burgess and Pande (2005) investigated the relationships among financial
inclusion, poverty and income inequalities and found out that expansion in rural bank branches was more likely
to lead to reduction in poverty incidences. Similarly, empirical evidence indicates that availability and
affordability of financial services is likely to lead to greater utilization which in turn would lead to improving
poor people's well-being (Dupas and Robinson 2012; Caskey et al., 2006). However, the poor would need to
have a better understanding of the financial concepts to be able to enjoy the full benefits. Globally, the variation
in financial services access between male and female adults is explained by among other factors gender-based
discriminations that limit women’s participation in economic activities. In their study of financial inclusion and
discrimination against women in 98 countries, Demirguc-Kunt et al., (2013) identified major gender-based
differences in access to credit, savings, and ownership of bank accounts. The study attributed cross-country
differences in financial inclusion to gender norms and legal discriminations against women. In particular, female
adults were less likely to save, access credit, and own a bank account compared to their male counterparts in
countries where women faced legal restrictions in their attempt to work, decide on residence location, inherit
wealth, and head their households. In addition, an increase in violence against women and the incidence of early
marriage reduced access to financial services among women significantly.
According to Arun and Kamath (2015), financial inclusion is a process that occurs in stages, with payments
being the first stage. Therefore adopting a payment platform of financial product is the entry point to financial
inclusion. In this case, adopting the product is driven by a payment need such as receiving a salary. The
implication of this argument is that measures geared towards promoting financial inclusion should promote
economic activities that encourage the use of payment and other financial products. Adoption of a payment
product has promoted access to other financial services like insurance, savings, and credit. Use of electronic
payment systems enables financial institutions to collect vital transaction information about a customer to assess
credit worthiness. This promotes access to credit services. The product adopted after payment products, however,
varies from state to state depending on the needs of the customers. For instance, in Brazil customers often adopt
savings/ investment services, whereas in Italy they adopt credit services.
Creating awareness about availability and benefits of financial services is expected to improve inclusion,
especially in low income countries where the levels of education and financial literacy are still low. Indeed
Bayero (2015) in his study of cashless policy in Nigeria found that awareness of electronic channels of accessing

68
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

financial services had a positive effect on financial inclusion. Further, payment infrastructure was found to have
a positive effect on financial inclusion. An efficient and secure payment infrastructure reduces transaction costs,
which in turn motivates adoption of electronic payment products. Bayero (2015) also found that improving
customer value proposition enhanced financial access. This is consistent with Terpstra and Verbeeten (2014)
who concluded that customer satisfaction improved access to financial services. The results means that product
development initiatives should focus on satisfying the felt financial needs of the target population so as to
enhance inclusion. The business model adopted by financial institutions did not have any statistically significant
effect on financial inclusion.
In India, Gwalani and Parkhi (2014) found that expansion of bank branch network had little contribution to
financial inclusion. This was attributed to several social and economic barriers experienced by the rural poor
despite expansion of bank branch network. The barriers identified included high cost of financial services,
inefficiencies at the bank level that prevented delivery of relevant and affordable financial services, and
inadequate incentive structure to promote use of financial products such as savings bank accounts. Inadequate
staff and lack of appropriate technology to deliver appropriate services were also major barriers to financial
inclusion. Although Arun and Kamath (2015) argue that adopting a payment system should be the first step in
promoting financial inclusion, Gwalani and Parkhi (2014) opine that creating awareness about existing products
should be the first step. Specifically, the focus should be on financial literacy and eradicating poverty to increase
financial inclusion. Banks being the main beneficiaries of inclusion should invest upfront in financial literacy in
order to create a large market for their products in the long term. According to Gwalani and Parkhi (2014)
providing adequate income opportunities to the poor will improve inclusion by enhancing the use of products
such as electronic payment systems. This means that the use of technology to deliver financial services must also
improve to enhance inclusion.
In their study of the effect of access to financial services in India’s manufacturing sector, Beck and Hoseini
(2014) concluded that financial deepening and bank outreach reduce informality. The use of private firms as
proxy to financial deepening and bank branch penetration as proxy to bank outreach reduced the entry barriers to
the formal sector while increasing the productivity of formal manufacturing firms. Financial depth also increased
production efficiency, especially among firms that relied heavily on external funds to finance their operations.
Bank outreach, on the other hand, had no effect on production efficiency.
Ayyagari and Hoseini (2013) supported the premise that financial deepening alleviates poverty, especially
in rural areas by promoting inclusive growth in India. Household data analysis revealed that commercial bank
credit to state domestic product (SDP) used as proxy to financial depth had a negative effect on poverty. This
implies that an improvement in financial deepening reduced poverty in rural areas. By contrast, financial
deepening had no statistically significant effect on poverty in urban areas. Overall, financial depth had a greater
impact on rural poverty than financial outreach. Entrepreneurship was the main channel through which financial
deepening reduced poverty. This finding is attributed in part to the fact that credit supply in rural areas mainly
benefits the self-employed who use it to expand their businesses, thereby reducing poverty.
Firm characteristics are likely to determine the extent to which they respond to financial obstacles. Firms
with a large capital/ asset base are likely to have greater access to internal funds to finance their operations than
small firms. This suggests that small firms are more vulnerable to financial obstacles than medium growth firms.
This discourse is supported by Beck et al., (2007) who found that the degree to which legal and financial
underdevelopment and corruption hinder firm growth mainly depended on size of the firm. In addition, the
growth of small firms is consistently affected negatively by legal, financial and corruption related bottlenecks.
The effect of financial and legal obstacles was higher in countries where corruption was rampant. The financial
obstacles that adversely affect firm growth included bank paper work, bureaucracies, and the need to have
special relationship with a bank. This obstacles limit access to financial services, which in turn constrain firm
growth.
Djankov et al., (2005) investigated private credit in 129 countries and found out that creditor protection
rights and availability of information sharing institutions promoted inclusive financial growth. Private and public
credit bureaus were found to promote private credit development in poorer rather than rich and developed
countries. Credit bureaus provide vital information that lenders use to assess the credit worthiness of prospective
customers. An increase in credit reference bureaus, thus, is expected to improve access to relevant customer
information, thereby increasing access to credit, especially in poor countries where transaction information is
difficult to obtain through alternative means such as credit cards.
Financial intermediation plays an important role in promoting financial inclusion by channelling funds from
savers to borrowers. However, intermediation is associated with costs that if not controlled can impede rather
than enhance financial inclusion. Poghosyan (2013) provided evidence in support of a negative effect of interest
rate margin on credit supply used as proxy to intermediation costs. At the firm level, intermediation costs
increased with higher riskiness of credit portfolio, low bank capitalization, and small bank size. In addition, lack
of competition measured by concentration in the banking industry increased intermediation costs. This suggests

69
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

that measures geared towards promoting competition can reduce intermediation costs in terms of low interest
margins. As a result, financial inclusion is likely to increase in terms of access to loans and ownership of bank
accounts.
Adamo (2013) distinguishes two noteworthy classifications of individuals who are financially included:
Those who assume a dynamic part and those that are uninvolved. The dynamic part of financial inclusion is
played by strategy creators, sentiment shapers, specialists in financial sector while the inactive part is played by
individuals in low income and social orientation. Peach and Van de Werff (2013) contend that changes in the
approach are critical to macroeconomic dependability and mitigates the unfavourable outcomes of financial
exclusion. Policy fortifies the connections among growth, democracy and integrated society. This perspective is
supported by a number of studies that view advancement in financial sector as more likely to promote growth in
economy (Rajan and Zingales 2003; Aghion et al., 2003). Utilizing information from 109 countries both from
developing and developed nations, Calderon and Liu (2003), observed that the direction of causality is from
financial development to economic growth. Yet other studies have established economic development to be the
cause of financial development (Luintel and Khan 1999; Demetriades and Hussein 1996). The direction of
causality notwithstanding, economic development has greater benefits to the poor and the lower segments of the
society (Beck et al., 2007).
Available evidence shows that a large proportion of the rural population's interest is captured in the
informal financial market (Rutherford 2000). This is mainly because the activities in these markets within
economies of developing countries outweigh those in developed economies (Khandker and Faruqee 2003). The
recent evidence reveals that with availability of a variety of financial services, less rural populations are using
informal channels1.
A few studies have investigated the determinants of financial inclusion and found that robust explanatory
variables of access to formal financial services are varied by income level and its distribution across income
quintiles. According to Honohan (2007), distributional efficiency of income affects social inequality such that
the lower the income inequality the greater are the financial inclusion indicators. Using Honohan (2008) access
indicators, Park and Mercado (2015) concluded that a higher per capita income increases financial inclusion.
They constructed the financial inclusion indicators for 37 Asian countries using macro data from the World
Development Indicators. Their findings are corroborated by a similar study by Demirguc-Kunt and Klapper
(2013), which investigated the financial inclusion in multiple countries. They established that a higher-income
quintile was associated with deeper penetration of financial services. Therefore, empirical findings predict a
positive relationship between income and financial inclusion, although studies on the distributional aspects of
income show a completely different scenario.
Education indicator for households has remained a big challenge despite the existence of a clear relationship
between educational achievement and financial inclusivity (Atkinson and Messy 2013). Education
accomplishment has been shown to rise linearly in relation to consumption of financial services. This view is
supported by Camara and Tuesta (2014) in their analysis of a nationally representative sample in Peru using the
probit models and found out that education plays a significant role in financial inclusion. They used a
dichotomous depended variable for formal financial services. This analysis had one fundamental flaw in that the
data used is not amenable to cross-country comparisons. Numerous family units burdened by training restrictions
will probably endure absence of job and in this manner be prevented from securing access to financial services.
The education variable is predicted in many empirical findings to be positively related to financial inclusion
although a variation on the levels of education has shown quite different results. Numerous countries classified
as developing nations are affected by ethnicity bias and social exclusion although there is limited data to conduct
full analysis (Khan 2008). Empirical findings show that black and minority groups (Goodwin et al., 1999) are
more likely to be excluded from financial services not because of their ethnicity per se or gender but largely
because of their low incomes. Some studies however have paid attention to the location of the banks measured in
terms of residence. Accordingly, Burgess and Pande (2005) observed that expansion policy on bank branches to
diverse rural settings had a welfare improving effect on the households in India.
Honohan and King (2012) carried out a cross-country analysis of causes and effects of financial access.
This study surveyed on several African countries using a nationally representative sample that found out that
location measured by urban and rural residency was a key factor for financial access. The analysis employed the
probit and Ordinary Least Squares (OLS) regression approaches with location variable being highly significant.
Different studies have considered other variables, for instance; work status, financial knowledge, numeracy and
risk aversion metric being the essential factors affecting financial services access. In another study, cross country
gender analysis showed less likelihood of women utilizing formal financial services compared to their male
counterparts. The gender crevice is more conspicuous in developing countries where poverty is rife (Demirguc-
Kunt et al., 2012). This gender gap has tended to persist relative to different income sub-groups.

1
FinAccess 2013, FinScope 2013

70
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Focusing on the Mexican economy, Ambrosius and Cuecuecha (2015) provided strong evidence in support
of the premise that foreign remittance has a positive effect on financial deepening in the domestic economy. The
analysis of Mexican household level data further revealed that remittances had a positive effect on access to
credit, savings, and ownership of savings bank account. However, remittance was found to promote access to
credit from the informal rather than the formal banking industry. The implication is that remittances create
demand for financial services such as lending and savings. However, the formal financial sector does not fully
cater for the demand, thereby forcing the recipients of remittances to rely on the informal financial sector.
Munyegera and Matsumoto (2016) investigated household welfare in Uganda and concluded that access to
mobile phone financial services had a positive effect on remittances. Specifically, households that had access to
mobile phone based money transfer services were more likely to receive remittances than those who did not. In
addition, the frequency and value of remittance increased with access to mobile phone based financial services.
The increase in remittance was attributed to the reduction in transaction costs associated with mobile phone
money transfer services, especially in the rural areas where financial institutions such as banks are not easily
accessible. An increase in remittances improved welfare in terms of increased real per capita consumption.
Percentage of individuals in a population with formal bank account and savings used as proxy for financial
inclusion has been found to promote economic growth and development by enhancing access to education and
promoting entrepreneurship. However, the level of financial inclusion varies from one country to the other
mainly due to socio-economic and demographic factors such as the level of income and age of an individual.
Indeed Fungacova and Weill (2014) in their financial inclusion study in China found out that financial inclusion
was deeper in China than in other BRICs. A key factor explaining the difference in the level of financial
inclusion was that voluntary exclusion was higher in China than in other BRICs where barriers such as high bank
charges and lack of trust in the banking system prevented access to financial services. Overall, education, income,
age, and gender had a positive effect on financial inclusion.
Financial services have to meet the expectations of the target population in terms of cost, quality, and
availability for their uptake to increase. Thus, financial institutions often focus on improving customer
satisfaction in a bid to attract more customers, thereby increasing financial inclusion. According to Terpstra and
Verbeeten (2014), improving customer satisfaction is associated with costs that can potentially prevent access to
financial services, especially among the low-income earners. They found a positive relationship between the cost
of service to customers and the satisfaction of customers. Additionally, they found out that the satisfaction of
customers creates higher returns in the most profitable customer segment. This means that financial institutions
are more likely to ignore the low income earners who often constitute the least profitable segment of the
customer base in their effort to improve profits. The resulting reduction in customer satisfaction is attributable to
barriers to financial inclusion. Moreover, financial services will remain inaccessible to the low income earners
due to their inability to pay a premium to access high quality services. This paper however, does not use the
customer satisfaction indicator mainly because it was not considered in the survey instrument.
According to Allen et al., (2012), account ownership, frequent account use, and using a bank account to
save provide an enabling environment for accessing financial services through the cost channel. Allen et al.,
(2012) concur with Fungacova and Weill (2014) that individuals who are poor, young and unemployed are likely
to be financially excluded. Additionally, financial inclusion increases with education and urbanization. This is
based on the fact that people who are highly educated are likely to have higher incomes, which enable them to
afford financial services. Urbanization has led to increased financial inclusion due to availability of financial
institutions that provide financial services in most urban areas.

3. Methodology
Most countries in Africa and especially the SSA region experience erratic growth characterized by high volatility
and low investment potential. This paper utilized the second wave of the Global Financial inclusion dataset
collected by World Bank in 148 economies around the world. The dataset is a global survey on the purpose for
use of certain financial services. The key areas of focus include: savings, credit, insurance and payments. The
paper also examines the barriers to financial inclusion across the SSA region. The paper focused on different
account ownership components namely: Bank account, account in financial institutions other than banks and
mobile phone account. The definition of financial inclusion adopted in this paper relates to financial services
consumed by households.
The paper adopts a modified conceptual framework based on World Bank (2014) model which assumes that
financial inclusion is affected by not only economic factors but non-economic factors as well. In the modified
framework, economic, financial as well as demographic factors that affect financial inclusion are utilized. The
empirical model therefore takes these components into account.

3.1 Empirical Model


The paper recognises that individual households are nested within countries and therefore the data is treated as

71
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

consisting of different layers. The natural grouping of households in countries tends to differentiate individuals,
implying that countries and their citizens influence each other. This explains the observed behavioural
differences of individuals in different countries. If this relationship is not taken into account, the country effects
may render the results invalid. This calls for the use of the multilevel approach.
Multilevel analysis has been widely used in sociological, demographic, education fields to depict an
approach that permits concurrent study of the effects of level one and level two variables on individual-level
results. Overtime, there has been greater need to use this approach to explain similar situations in other fields of
study. Additionally, the developments of statistical methods and softwares that incorporate multilevel structures
have led to greater demand for the approach. The fact that individuals and groups may interact to cause an
outcome has been the subject of intense debate in empirical literature (DiPrete and Forristal, 1994). The rationale
for multilevel models has also been prompted by unavailability of individual level variables that explain the
effect at that level. In this case, group level variables are only used as proxy variables. For a long time, studies
have continued to use the assumption of homogeneity which is not appropriate for addressing multi-level
problems. Single-level models are anchored on the assumption of independence of different population sub-
groups yet it is obvious that this assumption is more likely to be violated due to the existence of the nested data
structures (Goldstein 2003). The assumption of homogeneity is relaxed by the multi-level model essentially
rendering single-level model inadequate in handling hierarchical data structures. The key advantage of using
multi-level framework is in its ability to examine the extent of heterogeneity among different sub-groups
(Srholec 2010). Heterogeneity arises from differences between entities (firms; countries; schools) in terms of
size, population, technological advancement, firm-specific processes among others.
This paper considered variables at two levels, individual level and country level. Individual level variables
were separated from country level variables. Models that focus on individual households ignoring group
membership may be represented as follows:
yi = α + βxi + µi .............................................................................................................1
where yi is the dependent variable, i is the categories of the response variable, xi is a vector of explanatory
variables, α is the intercept, β is the slope and µi is the error term. This model cannot be used to describe
simultaneous relationships involving complex survey data. For instance, to describe the relationship between
individual households and group membership requires a two-level model. The model is thus augmented as
follows;
yij = α j + β j xij + γ ij
.....................................................................................................2
Where j is a subscript that refers to level 2 units (different countries) and i is a subscript that refers to level 1
units (individual households). yij is the response variable and xij is a vector of independent variables. For
consistency, this study adopts Goldstein (2003) notation standard. Hence, we let αj = β0j and βj= β1j. The paper
also notes that β0 and β1j are level 2 specifications. The group j intercept can be estimated by β0j= β0+ µ0j where
β0 is the population intercept estimate and µ0j is the variation around the individual groups. On the other hand,
β1j= β1+µ1j1 represents the slopes of groups which is assumed to be identical only when the subscript j is omitted.
µ1j represents random variables that is normally distributed.
Combining level 1 and level 2 units yields the two-level model of the form:
yij = β0 + β1xij + (µ0 j + µ1 j xij + γ 0ij )
............................................................................................3
Where var(γ0ij)=σγ02
Equation 3 expresses the response variable yij into two parts: fixed and random. In general, equation 3 can be
expressed as follows:
p
yij = β0 + β1 x1ij + ∑ β h xhij + (µ0 j +µ1 j x1ij + γ 0ij )
h=2 .........................................................................4
Based on equation 3, the empirical model is specified as:
finInc i j = β 00 + β 10 Edu ij + β 20 Gender ij + β 30 Age ij + β 01 Income ij + β 11 debitcard ij

+ β 21 saved ij + β 31 Borrowed ij + β 41 Govttransf er ij + β 51 Domremit tan ce ij + ... + ( µ ij + γ ij )


......5
Where finInc denotes categorical indicators of financial inclusion, i represents the index for individuals and j is
the index denoting the country. The predictor variables captures individuals’ characteristics found in the survey
dataset. Equation 5 assumes that the variables do not interact. It has both the fixed part and the random part.
Level 1 variables include age, education and gender while level 2 units are income, saved, borrowed,
government transfers, domestic remittances and debit card ownership as depicted by the notation in equation 5.
1
µ0j and µ1j are random variables. Their corresponding parameters are E(µ0j)= E(µ1j)=0; var(µ0j)=σ2, var(µ1j)=σ2, cov(µ0j, µ1j)= σ

72
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Although technological and cultural factors are not explicitly stated, some of the socioeconomic and
demographic factors are in-build within those factors. In addition, the dataset does not capture cultural and
technological aspects within the data. Equation 6 estimates barriers to financial inclusion. The estimation is
based on the probit regression.
finIncBarr ier i = β 0 + β 1 Edu i + β 2 Gender i + β 3 Age + β 4 debitcard i + β 5 Domremit tan ce i
+ β 6 Income i + β 7 Govettrans fer i + β 8 saved i + β 9 borrowed i +γi …
………........................................................................................................................................6
Where i represents the various barriers to financial inclusion identified in the Global Findex data. Equation 7
estimates the intra-group correlation. This model takes into account variations arising from group membership.
Intra-group correlation is thus measured as follows:
σ µ02
ρ=
σ µ02 + σ e2
...........................................................................................................................7
The intra-group correlation is indicative of the proportion of variation explained by the organisational structure
of the population.

3.2 Definition and Measurement of Variables


Dependent Variable
The dependent variables used in this study include: Account Ownership (Bank), Account Ownership (Mobile
phone) and Account ownership (other financial institution other than bank). Previous studies that have used
similar variables include Demirguc-Kunt and Klapper (2012)

Independent Variables
The explanatory variables used have also been used in other studies that sought to understand the dynamics of
financial inclusion (Fungacova and Weill 2014; Beck et al., 2009). However, this paper considers Education,
Income, Gender, debit card possession, government transfer received through account, domestic remittances
received through mobile phone, saving and borrowing behaviour, and Age as relevant explanatory variables.
Table 1 shows variable description and variable measurements.

73
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Table 1: Definition, measurement and predicted sign for each explanatory variable
Variable Notation Measure Predicted Effect Source of data
LEVEL ONE
VARIABLES AGE Years lived (Continuous) - Global Findex
Age
Education EDU 0=completed primary or less + Global Findex
1=secondary
2=completed tertiary or more

Gender GENDER 0=female +/- Global Findex


1= male
LEVEL TWO
VARIABLES INCQUIN 0=poorest 20% Global Findex
Income quintiles 1=second 20% +
2=middle 20%
3=fourth 20%
4=richest 20%
Saved in the last Global Findex
year
0-No +
Borrowed in the 1-Yes
last year Global Findex
0-No +
1-Yes
Debit card
possession 0-No + Global Findex
1-Yes
Domestic
remittances
received through 0-No + Global Findex
phone 1-Yes

Government Global Findex


transfers received 0-No +
through account 1-Yes

3.3 Sources of data


The demand-side Global Findex dataset covers access and use indicators of more than 140 economies of the
world. This dataset consists of a nationally representative sample randomly selected among adults of 15 years or
above. The data contains a record of data collected in 2014 calendar year. An aggregate of 34 nations from sub-
Saharan Africa were analysed. On average about 1,000 individuals 15 years or above in each country were
considered. The data collection procedures were standardized across all the countries. The margin of error
considered in this survey was not significantly different from one country to the other. The database is the first
one that provides insights into households' use of financial services across different economies. Additionally, the
database contains information on financial exclusion. Although financial exclusion separates people from
benefiting from the usage of financial services, individual characteristics do hinder households from using such
services. Efforts towards gathering standard data on regular basis at all levels of usage have been made in most
emerging economies although the scope has remained limited (Claessens, 2006). Qualitative surveys on usage of
financial services have been undertaken within the context of SSA since 2002. Data collected from the surveys
remains inadequate as it does not place focus on other critical variables such as savings, credit and risk
management. The individual level variables included are; age, education, gender, while the country level
variables included are income quintiles, saving, borrowing, debit card ownership, received government transfers
through account, and received domestic remittances through mobile phone. These factors are important for
examining the extent of financial inclusion in SSA region.

4. Empirical Results
This section is based on a variance components model which addresses the question of whether or not there is
sufficient variance represented at the individual level that warrants adoption of the mixed approach. The rule of
the thumb states that more than 10 percent of the total variance needs to be represented at a given level.

74
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Applying the multilevel modelling with individual households at level 1 and country being level 2, the approach
was justified. Using the null model with nothing but constant term gives the variance estimates at level 2(country
level) and level 1 residuals. Approximately 17 percent of the total variance in mobile phone account ownership
is represented at the country level being greater than 10 percent which is the minimum expected proportion for
further multilevel modelling. Similarly, the total variances explaining differences in account ownership at
financial institution and account ownership in general are above the minimum threshold of 10 percent.

4.1 Descriptive Statistics


Table 2 presents the summary statistics for the variables used. The dependent variables are measured as 0 if the
individual has no access to an account and 1 if the individual has access to an account. Age is measured in single
years. The minimum age considered in this paper is 18 years and the highest is 70 years. The rest of the variables
are indicated with minimum and maximum categories. Table 2 also shows the mean values and standard
deviations. The total observations sum up to 33,930 respondents each country contributing on average 1000
respondents. Intuitively, the statistics show that the variables are fairly spread around the mean an indication that
there is no clustering.
Table 2: Summary Statistics
Variable Obs Mean Median Std. Dev. Min Max
Account Ownership 33,930 0.35 0.00 0.48 0 1
Account financial institution 33,930 0.29 0.00 0.46 0 1
Mobile account 31,937 0.14 0.00 0.34 0 1
Income quintile 33,930 2.24 2.00 1.43 0 4
Education 33,930 0.51 0.00 0.59 0 2
Age 30,074 35.30 32.00 13.13 18 70
Gender 33,930 0.51 1.00 0.50 0 1
Debit card ownership 33,489 0.18 0.00 0.38 0 1
Domestic remittances received through mobile
phone 12,313 0.29 0.00 0.45 0 1
Government transfers through account or card 2,384 0.46 0.00 0.50 0 1
Saved in the last one year 33,930 0.58 1.00 0.49 0 1
Borrowed in the last one year 33,930 0.53 1.00 0.50 0 1
Figure 1 shows the random effects illustration by country. Mauritius, South Africa and Kenya are the
leading countries in access to financial services according to World Development Indicators 2012 which gives
credence to our findings. These economies have embraced innovation in the financial sector by providing people
with a variety of financial services (Ikhide, 2015). In addition, the governments have not put much restriction in
terms of regulating the financial sector especially the mobile money transfer platform. The model used by the
three leading lights is the ideal although inherently risky. Where mobile money has succeeded, to a large extent
the mainstream banking services have also improved. Therefore, the three leading countries in SSA also have a
well-functioning financial sector capable of interfacing with the mobile money technology and indeed any other
financial innovations.
Figure 1: Random intercepts by Country

Angola
Benin
Botswana
Burkina Faso
Burundi
Cameroon
Chad
Congo, Dem. Rep.
Congo, Rep.
Côte d'Ivoire
Ethiopia
Gabon
Ghana
Guinea
Kenya
Madagascar
Malawi
Mali
Mauritania
Mauritius
Namibia
Niger
Nigeria
Rwanda
Senegal
Sierra Leone
Somalia
South Africa
Sudan
Tanzania
Togo
Uganda
Zambia
Zimbabwe

-.4 -.2 0 .2
Random intercepts by Country

75
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Table 3 depicts a null and the fuller model for mobile phone account ownership as shown in equation 5.
Empirical analysis reveals that more people had mobile phone accounts compared to owning an account with a
financial institution. Table 3 also assumes that the first category is the reference category and therefore omitted
since it forms the basis for comparisons. The multilevel regression results in which the main dependent variables
are Account ownership (Bank Account), Account with a financial institution other than bank, and account with
mobile service provider (MSP) are shown in Table 3. The results show that owning a formal account is related to
the level of income of an individual. Based on the income quintiles variable, the results show that the lowest
income quintiles for the fuller model are all significant and negative. We observe that as you ascend from the
lowest quintile (poorest 20 percent) to the highest quintile (richest 20 percent), the magnitude of the coefficients
tend to reduce an indication that lower quintiles are associated with lower account ownership. The findings
corroborate Demirguc-Kunt and Klapper (2013) and Fungacova and Weill (2014) who established a similar
pattern between income and the financial inclusion.
This paper also examined education variable and found it to have negative coefficients that were significant.
Having primary education or less compared to having tertiary education reduces the likelihood of owning an
account. The coefficient for primary education or less is large and negative compared to the coefficient of
secondary education indicating that education variable is positively related to account ownership. As you move
from low levels of education to the highest level of education, the magnitude of the coefficient reduces, that is, it
tends to move towards the positive side (smaller negative number). These findings are consistent with Atkinson
and Messy (2013) who found out that financial literacy is key to promoting financial inclusion. This paper has
also established that having tertiary education or more influences individuals to own formal accounts whether
with the financial institutions or with mobile service providers. The findings may not be surprising given that in
sub-Saharan Africa, majority of the people have low levels of education.
Across all the account ownership categories, age is found to be negatively related to account ownership. As
individuals advance in age, they tend to be less likely to own an account at any given category. However,
observing the coefficients across all the account ownership categories, as individuals advance in age, they are
more likely to switch from holding an account in a formal financial institution to having a mobile account with
the mobile service providers. The age at which this switch occurs is necessary for policy intervention. Allen et al.,
(2012) investigated this relationship and established that ownership of formal account and age were non-linearly
related.
The paper also investigated the effect of owning a debit card on account ownership. Individuals with debit
cards were more likely to own an account with a financial institution compared to owning a mobile account with
a mobile service provider. The coefficient for debit card ownership is larger for account ownership with a formal
financial institution than for mobile account ownership. According to Damodaran (2012), individuals with debit
or credit cards, were also found to own formal accounts in India in the formal financial institutions.

76
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Table 3: Multilevel linear regression model with Account Ownership


Account Account in financial Mobile account
Ownership(Bank ) institution
(Other than Bank)
VARIABLES Null Model Fuller Null Model Fuller Model Null Model Fuller
Model Model
Fourth 20% -0.0240*** -0.0140** -0.0128**
(0.00665) (0.00591) (0.00582)
Middle 20% -0.0575*** -0.0391*** -0.0303***
(0.00655) (0.00583) (0.00573)
Second20% -0.0905*** -0.0666*** -0.0445***
(0.00641) (0.00570) (0.00560)
Poorest 20% -0.150*** -0.123*** -0.0879***
(0.00631) (0.00562) (0.00552)
Secondary -0.121*** -0.107*** -0.0519***
(0.00467) (0.00416) (0.00409)
Completed primary or -0.188*** -0.178*** -0.131***
less
(0.0102) (0.00905) (0.00929)
Age -0.0115*** -0.0101*** -0.00366***
(0.000586) (0.000521) (0.000512)
Female 0.0327 0.0195 0.0202
(0.00392) (0.00349) (0.00343)
Debit card ownership 0.518*** 0.591*** 0.108***
(0.00556) (0.00495) (0.00491)
Constant 1.649*** 0.405*** 1.706*** 0.376*** 1.862*** 1.780***
(0.0344) (0.0923) (0.0333) (0.0855) (0.0260) (0.0295)
lns1_1_1 -1.608*** -1.674*** -1.641*** -4.191 -1.921*** -0.868***
(0.122) (0.205) (0.122) (5.955) (0.126) (0.147)
lns1_1_2 -0.900*** -0.716*** -1.763***
(0.138) (0.132) (0.218)
lns1_1_3 -1.923*** -1.950*** -2.699***
(0.301) (0.302) (0.640)
lns1_1_4 -2.607*** -2.739*** -19.60***
(0.597) (0.708) (4.660)
lns1_1_5 -2.171*** -2.106*** -1.428***
(0.625) (0.446) (0.356)
lnsig_e -0.837*** -1.242*** -0.886*** -1.277*** -1.166*** -1.072***
(0.00384) (0.0226) (0.00384) (0.0227) (0.00396) (0.0215)
Observations 33,930 1,221 33,930 1,221 31,937 1,201
Number of groups 34 34 34 34 32 32
Standard errors in parenthesis
*** p<0.01, ** p<0.05, * p<0.1
Although the findings are largely corroborated by other studies across the world, it is imperative to point
out that contradicting findings have also been documented in the literature. For instance, this paper considers
gender to be insignificant contrary to other studies which view gender and financial inclusion to be strongly
associated. Such studies include (Fungáčová et al., 2014, Tuesta et al., 2013).

4.2 Barriers to Financial Inclusion


This paper also sought to investigate the barriers to financial inclusion in sub-Saharan Africa. Table 4 shows
estimation results based on equation 6. From the available data, the barriers to account ownership show that
income, education, age, government transfers and having saved affect the barriers to financial inclusion and are
significant at different levels. Being in the richest 20 percent category of income quintile implies that being too
far away from the financial service provider is not a major problem to this category. This finding is consistent
with other studies because richer individuals have greater mobility and therefore distance is not a major factor
(Brevoort and Wolken 2009). "Lack of money" and "too expensive" explain why individuals cannot own a
mobile phone account. These barriers are affected by income, age, government transfers, having saved, and

77
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

having borrowed and are consistent with the findings from previous studies (Fungacova and Weill 2014). The
findings further show that lack of documentation, lack of trust in financial institutions, religion, a family member
having an account and having no need for financial services were affected by most of the variables in the dataset.
These findings provide a view that is consistent with the current state of acquiring ownership of a mobile phone
account. However, it is surprising that there exist some individuals who perceive mobile phone account as “hard
to get” as depicted by “cannot get one” barrier. These findings however are consistent with Allen et al., (2012)
who established that more educated and wealthier individuals are more likely to own a mobile phone account
than the reverse.
Table 4: Determinants of barriers to financial inclusion
VARIABLES Too far Too Lack of Lack trust Religious Lack Family Cannot get No need for
away expensive documentation reasons money member one financial
already has service

Fourth20% -0.214** -0.274*** -0.195* -0.123 -0.222** -0.181* -0.166 -0.216** -0.200*
(0.106) (0.103) (0.107) (0.106) (0.107) (0.107) (0.107) (0.106) (0.107)
Middle 20% -0.335*** -0.354*** -0.428*** -0.343*** -0.386*** -0.375*** -0.411*** -0.372*** -0.386***
(0.104) (0.102) (0.104) (0.103) (0.105) (0.105) (0.104) (0.104) (0.104)
Second20% -0.439*** -0.409*** -0.469*** -0.371*** -0.470*** -0.453*** -0.495*** -0.460*** -0.454***
(0.103) (0.101) (0.103) (0.102) (0.103) (0.103) (0.103) (0.103) (0.103)
Poorest 20% -0.569*** -0.481*** -0.565*** -0.502*** -0.567*** -0.555*** -0.586*** -0.569*** -0.577***
(0.0994) (0.0975) (0.0997) (0.0986) (0.0996) (0.0996) (0.0997) (0.0995) (0.0996)
Secondary -0.655*** -0.654*** -0.699*** -0.676*** -0.670*** -0.705*** -0.674*** -0.697*** -0.687***
(0.0736) (0.0721) (0.0740) (0.0736) (0.0738) (0.0740) (0.0740) (0.0738) (0.0738)
Completed primary or less -1.231*** -1.125*** -1.197*** -1.176*** -1.187*** -1.209*** -1.182*** -1.217*** -1.228***
(0.146) (0.141) (0.143) (0.143) (0.143) (0.143) (0.143) (0.144) (0.144)
Age -0.0407*** -0.0396*** -0.0438*** -0.0414*** -0.0406*** -0.0412*** -0.0447*** -0.0430*** -0.0410***
(0.00904) (0.00886) (0.00909) (0.00901) (0.00907) (0.00907) (0.00909) (0.00906) (0.00906)
Female -0.0901 -0.0800 -0.111* -0.0823 -0.112* -0.0879 -0.0969 -0.0728 -0.106*
(0.0628) (0.0616) (0.0630) (0.0626) (0.0630) (0.0630) (0.0630) (0.0628) (0.0630)
Has debit card -0.392 -0.400 0.147 0.161 -0.369 -0.150 0.0166 0.314 -0.361
(0.344) (0.316) (0.292) (0.286) (0.333) (0.312) (0.297) (0.289) (0.341)
Domestic remittances 0.00607 -0.0279 0.0112 -0.0122 0.0291 -0.000337 -0.00423 -0.00395 0.0126
(0.0642) (0.0629) (0.0642) (0.0639) (0.0644) (0.0643) (0.0643) (0.0641) (0.0643)
Government transfers 1.319*** 1.235*** 1.324*** 1.324*** 1.332*** 1.334*** 1.333*** 1.317*** 1.322***
(0.0644) (0.0634) (0.0644) (0.0641) (0.0644) (0.0643) (0.0645) (0.0643) (0.0643)
Saved 0.698*** 0.637*** 0.692*** 0.661*** 0.715*** 0.665*** 0.680*** 0.650*** 0.667***
(0.0718) (0.0699) (0.0720) (0.0714) (0.0721) (0.0719) (0.0720) (0.0716) (0.0718)
Borrowed -0.124* -0.131* -0.105 -0.126* -0.114* -0.127* -0.124* -0.0740 -0.0809
(0.0689) (0.0675) (0.0690) (0.0686) (0.0691) (0.0691) (0.0691) (0.0688) (0.0690)
Constant -0.518 -0.408 -0.951** -1.041*** -0.576 -0.678 -0.786* -1.130*** -0.517
(0.443) (0.420) (0.405) (0.402) (0.432) (0.419) (0.410) (0.403) (0.438)
Observations 2,384 2,384 2,384 2,384 2,384 2,384 2,384 2,384 2,384
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

5. Conclusion and Policy Implications


The use of a multilevel approach was motivated by the nature of the data which formed a hierarchy at two levels:
individual level and country level. Estimation results showed that country differences explain more than 10
percent of the observed variations in account ownership. Income was found to influence account ownership.
Lower levels of income would negatively impact account ownership. Majority of mobile phone owners are likely
to be less endowed with resources and therefore influence their choice to use available financial services. The
need to package financial products at different ages is critical. This allows transition from one product to another
to happen in a seamless manner. This paper has also shown that there exist differences in financial inclusion
across countries. The key differences are largely explained by individual differences although country
differences also play a critical role. Many barriers to financial inclusion were investigated with the most critical
ones being lack of money resources, too expensive to use the services and the distance to the nearest financial
service provider. Having scarcity of money income and distance to the nearest financial service provider play the
most critical role in influencing the barriers to financial inclusion in Africa. Specifically, the main barriers to
financial inclusion were associated with lower income levels and lower educational attainment. Individuals will
less income are more likely to be served by mobile phone account than any other form of account. Distance
expressed in terms of proximity to the nearest service provider also affects account ownership pattern.
Overall, the findings raise important questions. There is a threshold on age at which households switch from
one aspect of financial inclusion to another. Different income quintiles play different roles in influencing
decisions regarding account ownership and therefore participation in financial inclusion matters. The question
would be to establish a clear pattern of events across the income quintiles that would lead to certain specific
outcomes. In general, the most important barriers that policy interventions should focus on include lack of
money income, too far away (distance) and too expensive (cost of accessing financial services). These barriers
are critical to unlocking the problem of financial exclusion especially in rural areas where services are greatly on

78
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

demand. Longer distances to the service provider could prove a major constraint if the level of income of
individuals is low. Similarly, lack of money income and the cost of financial services work to reduce dependence
on formal financial services. This perhaps explains why individuals tend to hold onto a portfolio of both formal
and informal financial services.

References
Adamo, R., (2013), “Financial Inclusion”, in Pizzo, G. and Tagliavini, G., (eds), Dizionario di microfinanza,
Carocci Editore, Rome, pp. 153-158.
Aghion, P., Howitt, P., and Mayer-Foulkes, D., (2003), “The Effect of Financial Development on Convergence:
Theory and Evidence,” The Quarterly Journal of Economics (1), pp.172-222.
Allen, F., Demirguc-Kunt, A., Klapper, L., and Pería, M., (2016), The foundations of financial inclusion:
Understanding ownership and use of formal accounts. Journal of Financial Intermediation.
Ambrosius C., and Cuecuecha A., (2013), Are Remittances a Substitute for Credit? Carrying the Financial
Burden of Health Shocks in National and Transnational Households. World Development, (46) American
Statistical Association, (78), pp. 605-610.
Anderloni, L., Carluccio, M., (2007), Access to Bank Accounts and Payment Services, in: Anderloni, L., Braga,
M.D. and E.M. Carluccio (eds.), New Frontiers in Banking Services, pp. 5-105.
Arun, T., and Kamath, R., (2015), Financial inclusion: Policies and practices. IIMB Management Review, 27(4),
pp.267-287.
Atkinson, A., and Messy F., (2013), "Promoting Financial Inclusion through Financial Education: OECD/INFE
Evidence, Policies and Practice", OECD Working Papers on Finance, Insurance and Private Pensions, (34),
OECD Publishing, Paris.
Ayenew, Z., and Zewdie S., (2010), Do urban poor benefit from microfinance
institutions? National Monthly Preferred Journal of Research in Commerce and Management, pp. 57- 63.
Ayyagari M., Beck T., and Hoseini M., (2013), “Finance and Poverty: Evidence from India”. Discussion Paper
(9497).
Banerjee, V., and Newman, F., (1993), Occupational choice and the process of development. The Journal of
Political Economy, (4), pp. 274-298.
Bayero, A., (2015), Effects of Cashless Economy Policy on financial inclusion in Nigeria: An exploratory study.
Procedia-Social and Behavioral Sciences, 172, pp.49-56.
Beck, T., Cull R., (2015), Banking in Africa. In: Berger, A.N., Molyneux, P., Wilson, J.OS (Eds), The Oxford
University press, pp. 913-937.
Beck, T., Demirgűç-Kunt A., and Honohan P., (2009), “Access to Financial Services: Measurement, Impact and
Policies.” World Bank Research Observer. (1), pp. 119-145.
Beck, T., and Hoseini, M., (2014), Informality and Access to Finance: Evidence from India.
Beck, T., Levine, R. and Loayza, N., (2000), Finance and Sources of Growth, Journal of Financial Economics,
(7), pp. 261-300.
Beck, T., Asli D., and Levine, R., (2007),“Finance, Inequality, and the Poor.” Journal of Economic Growth,
(11), pp. 27-49.
Brevoort P., and Wolken D., (2009), “Does Distance Matter in Banking?” The Changing Geography of Banking
and Finance, (34), pp. 1540-1565.
Burgess, R., and Rohini P., (2005), "Do Rural Banks Matter? Evidence from the Indian Social Banking
Experiment.” The American Economic Review, pp. 780-795.
Calderon, C., Liu, L., (2003), The direction of causality between financial development and economic growth.
Journal of Development Economics, pp. 321–34.
Camara, N., Peña, X., and Tuesta, D., (2014), Factors that Matter for Financial Inclusion: Evidence from Peru
(1409).
Caskey, J., Duran, R., and Solo, M., (2006), “The Urban Unbanked in Mexico and the United States”, Policy
Research Working Paper (3835), World Bank, Washington, DC.
Claessens, S., (2006), Access to financial services: a review of the issues and public policy objectives. World
Bank Research Observer, pp. 207–240.
Damodaran, A., (2012), Financial inclusion: Issues and challenges. AKGEC International Journal of Technology,
4(2), pp.54-59.
Demetriades, O., and Hussein Khaled A., (1996), Does Financial Development Cause Economic Growth? Time-
series Evidence from 16 Countries. Journal of Development Economics, pp. 51.
Demirgüç-Kunt, A., and Klapper, L., (2012), “Measuring Financial Inclusion: The Findex Database”, World
Bank Policy Research Working Paper (6025).
Demirgüç-Kunt, A., Klapper, F., and Singer, D., (2013), Financial inclusion and legal discrimination against
women: evidence from developing countries. World Bank Policy Research Working Paper, (6416).

79
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Demirgüç-Kunt, A., Klapper, F., Singer, D., and Van Oudheusden, P., (2015), The Global Findex Database 2014:
measuring financial inclusion around the world. World Bank Policy Research Working Paper, (7255).
Demirgüç-Kunt, A., and Levine, R., (2008), Finance, financial sector policies, and long-run growth. World Bank
Policy Research Working Paper, (4469).
DiPrete A., Forristal D., (1994), Multilevel models: Methods and Substance. Annual Review Sociology. pp. 331-
57.
Djankov S., Miguel E., Qian Y., Roland G., and Zhuravskaya E., (2005), Who are Russia’s entrepreneurs?
Journal of the European Economic Association, (3), pp. 587-597.
Dupas, P., and Robinson, J., (2012), Savings Constraints and Microenterprise Development: Evidence from a
Field Experiment in Kenya. NBER Working Paper (14693).
Eswaran and Kotwal (1990), “Implications of Credit Constraints for Risk Behaviour in Less Developed
Economies”, Oxford Economic Papers, pp. 473-482.
Fungáčová, Z., and Weill, L., (2014), Understanding financial inclusion in China. China Economic Review.
Galor, O. and Zeira, J. (1993), Income Distribution and Macroeconomics. Review of Economic Studies, (60), pp.
35-52.
Goldstein D., and Johnson E., (2003), “Do Defaults Save Lives?” Science, pp. 1138-39.
Goodwin, D., Adelman, L., Middleton, S., and Ashworth, K., (1999), Debt, money management, and access to
financial services: Evidence from the 1999 PSE Survey.
Gwalani H., and Parkhi S., (2014), “Financial Inclusion- Building A Success Model in the Indian Context”,
ICTMS, Procedia, (133), pp. 372-378.
Honohan, P., and King, M., (2012), Cause and effect of financial access: cross-country evidence from the
Finscope surveys. Banking the world: empirical foundations of financial inclusion, pp.45-84.
Honohan, P., (2005), "Measuring Microfinance Access: Building on Existing Cross-Country Data." World Bank
Policy Research Working Paper No. (3606).
Honohan, P., (2007), Cross-country variation in household access to financial services. The World Bank, Trinity
College Dublin and CEPR. Prepared for the Conference.
Honohan, P., (2008), Cross-country variation in household access to financial services. Journal of Banking and
Finance, 32(11), pp.2493-2500.
Hüttner M., and van den Eeden P., (1993), The multilevel design. A guide with an annotated bibliography,
1980–1993. London: Greenwood Press.
Ikhide, I., (2015), The Finance and Growth Debate in Africa: What Role For Financial Inclusion?.
Joskow, L., (2008), Introduction to New Institutional Economics: A Report Card. Cambridge, Cambridge
University Press.
Khan, O. (2008) Financial Inclusion and Ethnicity: An Agenda for Research and Policy Action. Runnymede
report.
Khandker R., and Farqee R., (2003), The impact of farm credit in Pakistan. Agricultural Economics (3), pp.
197-213
Kimenyi, S., and Ndung’u, S., (2009), Expanding the Financial Services Frontier: Lessons from Mobile Phone
Banking in Kenya. Washington DC: Brookings.
Levine R., (2005), Finance and Growth: Theory and Evidence. NBER Working Paper (10766).
Love, I., and Bruhn, M., (2009), The Economic Impact of Banking the Unbanked: Evidence from Mexico.
Policy Research Working Paper (4981). The World Bank Development Research Group.
Luintel, B., and Khan M., (1999), A Quantitative Reassessment of the Finance-Growth Nexus: Evidence from a
Multivariate VAR. Journal of Development Economics, (60).
Mlachila M., Park G., and Yabara M., (2013), “Banking in sub-Saharan Africa: The macroeconomic context”
International Monetary Fund.
Munyegera, K., and Matsumoto, T., (2016), Mobile money, remittances, and household welfare: panel evidence
from rural Uganda. World Development, 79, pp.127-137.
North, D., (2000), ‘Understanding Institutions’. In: Ménard (2000) pp. 7–10.
Oya, A., Maximilien H., and Nataliya M., (2011), "Access to Financial Services and the Financial inclusion
Agenda around the world. A Cross-Country Analysis with a New dataset." WB Policy Research Working
Paper No. (5537).
Park, Y., and Mercado, V., (2015), Financial Inclusion, Poverty, and Income Inequality in Developing Asia.
Peach, D., and Van der Werff, D., (2013), A Cross-Country Analysis of Financial Inclusion within the OECD.
Consumer Interests Annual, (59).
Peachey, S., and Roe, A., (2004), Access to Finance: Analysis of the World Savings Banks Institute. Oxford
Policy Management, Oxford.
Poghosyan, T., (2013), Financial intermediation costs in low income countries: The role of regulatory,
institutional, and macroeconomic factors. Economic Systems, 37(1), pp.92-110.

80
Journal of Economics and Sustainable Development www.iiste.org
ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online)
Vol.8, No.14, 2017

Rajan, G., and Zingales, L., (2003), Saving capitalism from capitalists, Crown Business, New York.
Rutherford, S., (2000), The poor and their money. New Delhi: Oxford University Press and Department for
International Development.
Shankar, S., (2013), Financial Inclusion in India. Do Microfinance Institutions Address Access Barriers? ACRN
Journal of Entrepreneurship Perspectives. Vol 2. Issue 1, pp 60-74, ISSN, pp. 2224-9729.
Srholec M., (2010), A Multilevel Approach to Geography of Innovation. Regional Studies, (44) 1.
Terpstra M., Verbeeten M., (2014), Customer satisfaction: Cost driver or value driver? Empirical evidence from
the financial industry. European Management Journal, (32), pp. 499-508.
World Bank, (2014), Financial Inclusion. Global Financial Development report (2014). International Bank for
Reconstruction and Development.
Zins, A. and Weill, L., (2016), The determinants of financial inclusion in Africa. Review of Development
Finance, (1) 6, pp.46-57.

81

You might also like