The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0306-8293.htm
Financial inclusion and Financialinclusion
poverty: evidence from and poverty
developing economies
Eunice Stella Nyarko
Department of Accounting and Finance, Faculty of Business Administration,
Methodist University College Ghana, Accra, Ghana and Received 14 January 2022
Revised 3 January 2023
Department of Finance, University of Ghana, Accra, Ghana Accepted 15 May 2023
Kofi Amoateng
School of Business, North Carolina Central University,
Durham, North Carolina, USA, and
Anthony Qabitoo Quame Aboagye
University of Ghana, Accra, Ghana
Abstract
Purpose – This paper examines the impact of financial inclusion on poverty through access to mobile money
in developing economies.
Design/methodology/approach – The authors employ the principal component analysis to construct an
index of financial inclusion using demand and supply indicators, including mobile accounts. The authors use
the two-step system GMM estimator for the analysis because of its efficiency and robustness in addressing
heteroscedasticity and autocorrelation.
Findings – The main finding is that financial inclusion generally increased and significantly reduces poverty
in the sample period. Furthermore, income inequality worsens poverty.
Research limitations/implications –This study has few limitations. First, the empirical analysis of the
study is restricted to macroeconomic factors only because of limited Household Finance Survey data set
and time availability. Second, the study is limited to developing countries and the results cannot be
generalized.
Practical implications – Financial inclusion is a significant policy tool for poverty reduction. There is the
need to enhance strategies that further improve financial inclusion by expanding and improving the use of
mobile money accounts.
Social implications – The paper sheds light on how developing countries can harness financial inclusion to
reduce poverty.
Originality/value –The paper differs from the previous studies in twoways. Firstly,mobilemoney account is
included in the computation of financial inclusion index over the sample period. It also determines the impact of
financial inclusion on poverty for short-run and long-run periods.
Peer review – The peer review history for this article is available at: https://publons.com/publon/10.1108/
IJSE-11-2021-0690
Keywords Financial inclusion, Financial access, Financial usage and poverty
Paper type Research paper
1. Introduction
Income poverty is a challenge confronting most developing countries. World Bank (2018)
reports that though extreme poverty has reduced globally, nearly half of the population of the
world still live in poverty. About half of the population in developing countries live on less
than US$1.90 a day. Those who live in extreme poverty usually suffer depravation and have
less access to financial resources, good health, adequate food, and education.
International Journal of Social
Economics
© Emerald Publishing Limited
0306-8293
JEL Classification — G18, O11, O16 DOI 10.1108/IJSE-11-2021-0690
IJSE One of the causes of poverty in developing countries is inadequate access to financial
resources, and related opportunities (Abebe and Quaicoe, 2014). The World Bank and some
researchers have pointed out that limited access to financial resources contribute to poverty
(Beegle et al., 2016). It becomes imperative to empirically investigate how financial inclusion
impacts poverty in developing countries.
Financial inclusion is defined as the delivery of financial services at an affordable cost
to those in need especially the marginalized, underprivileged and low-income groups.
Global Findex (2017) database shows that globally about 1.7 billion adults remain
unbanked and only 69% of adults have bank accounts. Most of these unbanked are in
developing countries.
Financial inclusion is a multidimensional concept and its measure needs to capture all its
dimensions. Some studies have measured it using a single indicator on the demand or supply
sides (World Bank: G20 Basic Set of Financial Inclusion Indicators). Others have measured
the concept using an index of indicators (Sarma, 2008; Wang and Guan, 2016; Park and
Mercado, 2018). How to weigh indicators when computing an index has been an issue as the
weighting approach affects the size of the index. Some researchers use equal weighting,
suggesting that all indicators have equal relevance to the index (Sarma, 2008). Others
compute weights of indicators (Wang and Guan, 2016). Still others use factor analysis and
principal component analysis to determine the weights of indicators based on their
eigenvalues. Using principal component analysis tends to assign weights based on the
relevance of an indicator to the index and does not assume that all indicators are of equal
importance to the index.
This study seeks to compute an index of financial inclusion using both demand and
supply indicators, including mobile money accounts to ascertain the extent of financial
inclusion an observed period. A measure of financial inclusion that captures mobile money
accounts is relevant in evaluating poverty because mobile money accounts make it possible
for the poor who do not have access to mainstream financial institutions to have access to
financial services which help them to meet their basic needs. This study focuses on the
dimensions of access and usage but not quality because there is currently no data on the
quality dimension.
Furthermore, it examines factors that affect financial inclusion in developing countries in
order to throw light on specific areas that need attention in the quest to achieve greater
financial inclusion. Finally, it seeks to analyze how financial inclusion affects poverty in
developing countries using a panel analysis. A panel study is likely to provide more insights
on how financial inclusion affects poverty across different economies over time. The study
uses income poverty indicators such as Poverty Head Count Ratio 2011 Purchasing Power
Parity (PPP) and Poverty Gap 2011 PPP. It uses income poverty indicators instead of the
multidimensional poverty index that captures education, health and general welfare because
income serves as the basis to access quality health care and education.
This study contributes to the literature by examining the level of financial inclusion over a
period in one hundred and forty-two developing countries. This helps to track how financial
inclusion has progressed over period especially with the use of mobile money transactions
unlike other measures in the literature that focus on single or few countries at a point in time.
The study also adds to the literature by demonstrating that financial inclusion has both short
run and long run effect on poverty reduction.
The rest of the paper is organized as follows: Section 2 provides the relevant literature
review. Section 3 provides the methodology for constructing the index of financial
inclusion; investigating the factors that affect financial inclusion; and for examining the
relationship between financial inclusion and poverty. Section 4 presents the results
and findings of the study. Section 5 concludes the study and highlights some policy
implications.
2. Literature review Financial
Different authors have defined financial inclusion to cover three dimensions of access/ inclusion
availability/outreach/penetration; usage and quality of financial services/products. Quality of and poverty
financial services refers to financial products serving the purpose for which they are
accessed. Generally, financial inclusion is the process of ensuring that appropriate financial
products/services such as savings, credit, payments, insurance, and remittances offered by
formal financial institutions are effectively accessible to individuals and businesses at
affordable costs; and the usage meets needs in a cost effective and sustainable manner.
Effective access involves convenient and responsible service delivery, at a cost affordable to
the customer and sustainable for the provider with the result that financially excluded
customers use formal financial services rather than existing informal options. There are two
main participants in the process of financial inclusion -those who demand and those who
supply financial services.
Similarly, there is no widely accepted definition of poverty, but the consensus is that any
definition of poverty needs to recognize specific social, cultural, income contexts, and well-
being. Poverty can be viewed from the absolute or relative perspectives. Absolute poverty
considers the number of people living below an income threshold while relative poverty
compares a group of poor people with others in each population. The World Bank defines
poverty in absolute terms as living on less than an income threshold per day. Living below
US$1.90 per day (PPP) to meet basic needs is considered extreme poverty and less than $3.10
a day asmoderate poverty. TheUnitedNations (2012) describes poverty in terms of a range of
interrelated and mutually reinforcing deprivations, and draws attention to the stigma,
discrimination, insecurity and social exclusion. It views poverty as a condition characterized
by severe deprivation of basic human needs, including food, safe drinking water, sanitation
facilities, health, shelter, education, and information. UNDP views poverty as
multidimensional which goes beyond income and captures the reality of living with less
than the essentials needed to lead a decent life which does not allow the poor to develop their
full potential. This paper adopts theWorld Bank’s definition of poverty and looks at poverty
from an income point of view because it is usually income that gives people access to the other
non-income dimension of poverty.
The relationship between financial inclusion and poverty is important because when the
poor have access to financial services, they are able to engage in productive economic activities
or expand their business activities which enable them to cater to their basic needs. Some studies
have examined how access to finance impacts poverty at the individual country level. Ayyagari
et al. (2013) suggest that financial inclusion contributes to poverty alleviation in the rural areas of
India by strengthening the productive base of the poor, enhancing the productive capacity and
promoting entrepreneurship. These studies examined only an aspect of financial inclusion, for
instance the access dimension and in a limited jurisdiction either at the community or country
level. In order to providemore insights into the impact of financial inclusion on poverty, it would
be helpful to consider access and usage dimensions of financial inclusion.
Lal (2017) reports that financial inclusion through cooperative banks reduces poverty among
the poor because access to finance enables the beneficiaries to raise their standard of living,
obtain quality food, purchase basic needs, increase the quality of flooring materials, and
consumption level. Shaban et al. (2019) report that financial inclusion is positively and
significantly associated with GDP per capita, employment, bank competition, human
development, government integrity, and Internet usage. This current study investigates how
financial inclusion affects poverty in 142 developing countries. This study differs from the cited
studies in the literature in two ways. First, it constructs a composite measure of financial
inclusion using both demand and supply side indicators including mobile money transactions
over a period. Second, it evaluates the short- and long-term impact of financial inclusion on
poverty.
IJSE On the measurement of financial inclusion, some studies use single indicators of access
and usage as a proxy for financial inclusion. Amid
zic et al. (2014) use saving/deposit and
loans/borrowings as a measure of financial inclusion. Nonetheless, financial inclusion is a
multidimensional concept; hence, using a single indicator to proxy for it does not capture the
other dimensions. Sarma (2008) computes a composite measure of financial inclusion using
indicators of access and usage dimensions. In Sarma’s approach, each of the indicators is
normalized to lie between 0 and 1 in order to remove the scale effect. Weights are then
assigned without a clear justification of the weightingmethod. Beck et al. (2009) use variables
that relate to physical access, affordability, and eligibility (deposits, loans, and payments).
Their study ranked the performance of countries by dimensions and a country had different
rankings in different dimensions which makes it challenging to determine the extent of
financial inclusion in each country and to compare across countries. Honohan (2008) uses
econometric model to measure the share of individuals that have access to financial products
to the total populace. In that study, the usage dimension of financial inclusion was not
analyzed, and econometric estimates arguably provide a one-time estimate of financial
inclusion which cannot be used to measure variations across time and countries (Sarma,
2012). Demirguۍ-Kunt andKlapper (2012) focus on demand side indicators that suggest usage
to measure financial inclusion. The core indicators address five areas of activity: borrowing,
saving, insurance, making payments, and operating an account. The limitation of this
measurement is that it does not account for the supply side factors such as use and access to
banking facilities and mobile money transactions, number of access points, and percentage
administrative units.
Further, Chakravarty and Pal (2013) compute a composite financial inclusion index using
indicators of access, usage, and availability of banking services. They use equal weighting of
variables in the dimensions in their analysis without providing a basis for the weighting
method. Amid
zic et al. (2014) criticize an index that assigns equal weights to all variables and
dimensions because applying the same weights suggest that all dimensions have the same
relevance on financial inclusion which may not reflect the reality in practice. They use factor
analysis to derive the weights for their analysis which makes the weighting approach
relatively objective compared to the equal weighting.
Finally, Park and Mercado (2018) use demand side indicators such as savings, borrowing
and financial literacy to construct an index of financial inclusion following the approach
Sarma (2008) but use principal component analysis to determine weights for the indicators.
Income poverty is measured using absolute, relative and multidimensional measures
(Alkire et al., 2015). An example of an absolute measure is poverty headcount ratio at $1.90 a
day (2011 PPP) (%). The relativemeasure like the poverty gap index considers the intensity of
poverty in a country. The multidimensional measure of poverty considers both the income
and non-income dimensions of poverty such as health, education and standard of living. This
paper uses income poverty measure because the other non-income dimensions of poverty
such as health, education, living standard and quality of well-being to a large extent are
contingent on income level.
Themeasure of financial inclusion in this paper is an improvement over existingmeasures
because it broadly captures financial inclusion to include mobile money transactions
compared to Honohan (2008) who uses only demand side indicators such as bank and MFI
account numbers from household cross-sectional data to construct a measure of financial
access in a limited number of countries.
The factors that affect financial inclusion emanate from demand and supply sides.
Demand side indicators relate to what enables demanders of financial services to access
products. These indicators are education, health, electricity, roads, transport, and marketing
opportunities. The supply side variables enable providers of financial services to meet the
needs of especially small borrowers at a cost-effective manner. The supply side indicators
include regulation, rule of law, products, accessibility, and cost of service delivery. Some Financial
studies have examined either demand or supply factors that affect financial inclusion at inclusion
individual country levels. Kaur (2017) examines factors that affect financial inclusion in and poverty
Pakistan and finds that literacy rate and gross domestic product significantly impacts
financial inclusion positively. Akudugu (2013) evaluates the determinants of financial
inclusion in Ghana and reports that age of individuals, literacy levels, wealth, lack of
documentation, and trust for formal financial institutions are significant determinants of
financial inclusion. Sahoo (2017) investigates the determinants of financial inclusion in tribal
districts of Odisha in India and finds that years of education by the household head, size of
private-owned land, income of the household, and participation in a guaranteed scheme are
significant determinants of financial inclusion.
3. Methodology
This study analyzes 142 developing countries for the sample period of 2004–2018. It computes
an index of financial inclusion using five indicators of access (three) and usage (two). It also uses
pooled ordinary least squares, least square dummy variable, and fixed effects estimation
techniques to examine factors that affect financial inclusion. The reason for using different
estimation techniques to analyze factors that affect financial inclusion is to ascertainwhether the
resultswill be comparable to other techniques. Finally, this study applies the systemgeneralized
method of moments (SGMM) regression proposed by Arellano and Bover (1995) and (Blundell
and Bond, 1998) to evaluate the relationship between financial inclusion and poverty.
This study computes each normalized indicator of financial inclusion using the following
formula:
Xij ¼ ðAij –mijÞ = ðMij –mijÞ (1)
where Xij is the transformed value of each indicator j in the dimension (Access, Usage); Aij is
the actual value;Mij and mij are the maximum and minimum of each indicator respectively.
The variables are transformed to normalize them to lie between 0 and 1. The purpose of the
normalization is to remove the scale effect and to allow for easy comparison of country
performance across indicators. The Index of Financial Inclusion in dimension i is computed
as follows: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼  w
2
i1 * ð1 2x 2i1Þ þ w12 * ð1 2xi2Þ þ . . .þ 2w2in * ð1 xinÞIFIi 1 þ (2)w2 2 2i1 wi2 þ . . .þ win
where Xij is the transformed value (0 ≤ Xij ≤1). wij denotes the weights (factor loadings)
derived from PCA of indicator j in dimension i.
Access is measured by account at a financial institution or with a mobile service provider
(% of population ages 15þ); ATM per 10,000 population and Mobile cellular subscriptions
(per 100 people). The usage dimension is measured by savings and borrowings.
The paper employs pooled ordinary least squares; least squares dummy variable; random
effects and fixed effects estimation techniques to examine factors that affect financial inclusion.
Fixed effects model is usually used when one is interested in analyzing the impact of variables
that vary over time. The least square dummy variable model offers a means to appreciate fixed
effects. Including a dummy for each country helps estimate the pure effect of the independent
variable. Each dummy absorbs the effects specific to each country. The random effects model
assumes that the variation across entities is random and uncorrelated with the independent
variables. Pooled OLS provides unbiased and consistent estimates of parameters even when
time invariant attributes are present.
IJSE The factors that affect financial inclusion are modelled as follows:
IFIit ¼ β1LRit þ β2 SECSCHCRit þ β3 INFRASTRUCit þ β4 AGEDRit þ β5 GNIPCit
þ β6 BANKRESIit þ β7 POPit þ β8RLAWit þ εit
and εit ¼ ϑt þ ɧi þ eit
where IFI is index of financial inclusion, LR is Loan rate (interest rate); SECSCHCR is
secondary school completion rate; INFRASTRUC is infrastructure (technological); AGEDR is
age dependency ratio; GNIPC is income; BANKRESI is bank resilience; POP is population;
and RLAW is rule of law. εit is the error term which decomposes into time effect), (country
specific effect) and eit is the stochastic error term assumed to be independently and identically
distributed.
Index of financial inclusion (IFI) is the compositemeasure of financial inclusion. Interest rate
on loans is a proxy for the cost of accessing credit. It is expected that higher interest rates on
loans will hinder financial inclusion. Cost to income ratio measure the efficiency with which
bank deliver financial services. Pasiouras and Kosmidou (2007) note that the more efficient and
cost effective a bank is, the more it yields a positive effect on financial inclusion. Thus efficient
and effective cost reduction improves financial inclusion. Secondary school completion rate
shows the average level of education of the population. Higher level of education can lead to
more access to financial services.Access tomobile phone is a proxy for infrastructure.Access to
mobile phonesmay enable people to access their account on their phones to facilitate payments,
borrowings, and getting information from the financial institutions. It is expected that financial
inclusionwill increase if a greater number of people havemobile phones. Age dependency ratio
shows the percentage of the population that do not earn income and must rely on the working
population for their survival. A lower dependency ratio is likely to improve access to financial
services and a higher ratio may hinder access to financial services as they do not earn income.
Income tends to increase financial inclusion in twoways. Firstly, people with higher income are
more likely to be included in the financial system as they can save. Secondly, economies with
higher income are more inclined to have developed financial sectors which help provide the
public with better access to financial services.
Bank resilience measures how sound a bank is in relation to performing its key functions.
The more resilient a bank is, the better it is positioned to serve a greater percentage of the
population and hence, increase financial inclusion. Population density indicates a larger
market size, and this is likely to enhance financial inclusion. Rule of law is a proxy for quality
of institutions and how effective and efficient the systems of the institutions work in each
economy. Better rule of law gives indication of enforcement of financial contracts which is
likely to improve financial inclusion. Table 1 describes the variables.
We present themodel to examine the relationship between financial inclusion and poverty
as follows:
POVi;t ¼ α0 þ α1POVi;t−1 þ α2IFIi;t þ α3GINIi;t þ αnXi;t þ εit
and εit ¼ ϑt þ ɧi þ eit
where POVi;t is the dependent variable(s) capturing poverty, POVi;t−1 is the lag of poverty.
Poverty is measured by two indicators – Poverty Head Count Ratio 2011 PPP (P0) and
Poverty Gap 2011 PPP (P1). IFI is a measure of financial inclusion, GINI account for income
inequality and Xit is a vector of all other control variables that have the potential to affect the
dependent variable. These are banking sector credit to private sector (CRETOPSB) as a
measure of financial depth; GDP per capita (GDPPC) as a measure of income, inflation (INFL)
as a measure of macroeconomic stability; rule of law (RLAW) a measure of institutional
Financial
Variable Notation Description Data
inclusion
Financial access ACCM Account at a financial institution or with a mobile service WDI and poverty
provider
Financial access ATM ATM per 10,000 populations WDI
Financial access MOBILE Mobile cellular subscriptions (per 100 people) WDI
Financial usage SAV Savings WDI
Financial usage BORR Borrowings WDI
Income INCOME GNI per capita WDI
Education SECSCHCR Secondary school completion rate WDI
Age dependency AGEDR Age Dependency ratio WDI
ratio
Infrastructure IFRASTRUC Access to mobile phones WDI
Cost of loan/credit INTL Interest rate on loan/credit WDI
Bank resilience BANKRESI Capital adequacy ratio WDI
Institutional quality RLAW Rule of law Index WDI Table 1.
Population POP Population WDI Description of
Source(s): Compiled by Authors, 2021 variables
quality; INFRASTRUC is technological infrastructure proxied by access to mobile phone;
population (POP) as a measure of population density; and secondary school completion rate
(SECSCHCR) as a measure of education. The paper also controls for time using a dummy
variable for the countries. εit is already explained above. Table 2 describes the variables and
expected signs.
The SGMM overcomes a potential weakness in the Arellano and Bond (1991) estimator
where lagged levels are shown to be weak instruments for first differenced variables. SGMM
estimator comprises the lagged levels and lagged differences. It has been argued that SGMM
ismore useful in generating efficient resultswhen the panel units (countries) are large, and the
time periods are relatively small. This estimation technique adds additional moment
Expected
Variable Notation Description Data sign
Dependent P0 Poverty headcount ratio at $1.90 a day PovcalNet
variable (2011 PPP) (% of population)
P1 Poverty gap (2011 PPP) (% of PovcalNet
population)
Independent IFI Index of financial inclusion Computed by Negative ()
variables authors
GINI Income inequality, GINI coefficient WDI Positive (þ)
Control variables CRETOPSB Financial depth, Banking sector credit to WDI Negative ()
private sector to GDP
GDPPC Income, GDP per capita
INFL Inflation, Consumer prices (annual %) WDI Positive (þ)
RLAW Rule of law index World Justice Positive (þ)
Project
INFRASTRUC Technological infrastructure, access to WDI Negative ()
mobile phones
Table 2.
POP Population, Population density WDI Positive Table of variables and
SECSCHCR Education, Secondary school completion WDI Negative () expectations for the
rate relationship between
Dummy Year 5 1; otherwise 5 0 WDI poverty and financial
Source(s): Created by Authors, 2021 inclusion
IJSE restrictions and allows lagged first differences to be employed as instruments in the level
equations and this corrects for any bias in the DGMM by Arellano and Bond (1991). In
addition, SGMM is more efficient in the presence of weak instruments as it uses more
conditions compared to DGMM. Like the DGMM, SGMM further requires that the
unobserved effects and error terms do not correlate over cross-section units.
The SGMM has gained much attention in the growth literature. Levine et al. (2002)
examined the impact of financial expansion on growth using a linear dynamic panel model.
They used the SGMMmodel to capture the unobserved, country specific effects in the face of
the lagged dependent variable. The main assumptions about the data generating process of
these estimators are some regressors may be endogenous; the process is usually dynamic
with current dependent variable being influenced by the previous ones.
4. Results
The financial inclusion index for the countries is grouped into three main categories – low,
medium, and high – based on the Index of Financial Inclusion (IFI) values. These values are
categorized as follows: high financial inclusion, 0.5 < IFI ≤1; medium financial inclusion,
0.3 < IFI≤0.5; and low financial inclusion, 0 < IFI≤0.3. The result shows that generally, most
countries (113) exhibited IFI values between medium (0.3 < IFI ≤0.5) and high (0.5 < IFI ≤1);
and these are increasing over the period. This suggests that financial inclusion has been
increasing over the period in developing countries. Twenty-nine (29) countries showed IFI
values that range from low, medium and high but did not show consistent pattern in the
increase of financial inclusion over the period. It is worthy of note that these countries were
predominantly in Africa. The graph in Figure 1 shows that generally financial inclusion
increased over the period 2005 to 2018 except for 2011 and 2017 where there was a decline
relative to the previous years. The cause of this decline is yet to be ascertained to ensure that
financial inclusion improves steadily in subsequent years. Refer to Appendix for the index of
financial inclusion for the countries over the period.
The results of the estimations of the factors that affect financial inclusion in the sample is
presented in Table 3. It uses the computed financial inclusion indicator for the period as the
dependent variable partially following Honohan (2008). The analysis of the factors that affect
financial inclusion is done using pooled ordinary least squares (OLS), least squares dummy
variables (LSDV), random effects and fixed effects estimation techniques. The results of the
estimations using the four approaches look similar. The dependent variable is financial
Plot of Index of Financial Inclusion (IFI) (2005-2018)
Figure 1.
Plot of index of 2005 2010 2015 2020Year 
financial inclusion (IFI) Plotted Coefficients of the year dummies
for 2005–2018
Source(s): Created by authors (2021)
IFI
0 0.05 0.1 0.15 0.2
Financial
Variables OLS LSDV FE RE
inclusion
Dependent variable: Financial inclusion and poverty
Loan Rate, LR 0.0001*** 0.0006*** 0.0006*** 0.0006 *
(0.24) (2.08) (2.08) (1.99)
Sec. Sch. Compl. Rate, SECSCHCR 0.0004 0.0004 0.0004 0.0004
(0.98) (1.27) (1.27) (1.28)
Infrastructure, IFRASTRUC 0.0007** 0.0025 0.0025** 0.0024***
(1.96) (7.94) (7.94) (7.91)
Age Dependency Ratio, AGEDR 0.0002 0.0015 0.0015 0.0015
(0.31) (0.81) (0.81) (1.08)
Gross National Product Per Capita, GNIPC 0.0000*** 0.0000*** 0.0000** 0.0000***
(14.64) (9.11) (9.11) (10.06)
Bank Resilience, BANKRESI 0.0108 0.0014 0.0014 0.0007
(4.31) (0.61) (0.61) (0.33)
Population, POP 0.0003 0.0006 0.0006 0.0005*
(4.10) (1.83) (1.83) (2.18)
Rule of Law, RLAW 0.1155* 0.2322* 0.2322* 0.2076***
(4.89) 5.43) (5.43) (5.65)
Constant 0.4849 0.6838 0.4236 0.3910
(7.82) (3.03) (3.00) (3.33)
Year Dummies
2005 0.0163
2006 0.0137
2007 0.0012
2008 0.0253
2009 0.0349
2010 0.0337
2011 0.0103
2012 0.0857
2013 0.1081
2014 0.1310
2015 0.1501
2016 0.1664
2017 0.1553
2018 0.2198*
Constant 0.3910** 0.2996 0.4236 0.3910***
Number of Observations 2130 2130 2130 2130
R-squared 0.1936 0.1917 0.1889 0.1936
Wald χ2/F-Statistic 42.52 66.02 5.17 341.41
Note(s): Legend: *p < 0.1, **p < 0.05, and ***p < 0.001
OLS – Ordinary Least Squares; LSDV – Least Squares Dummy Variable; FE – Fixed Effects; RE – Random Table 3.
Effects Factors that affect
Source(s): Created by Authors, 2021 financial inclusion
inclusion which is proxied by the index. The coefficients of the independent variables loan
rate, level of infrastructure, level of income, and rule of law have the predicted signs.
Loan rate shows a statistically significant positive impact on financial inclusion at 1%
level. The finding on loan rate is consistent with that of Olaniyi and Babatunde (2016), and
Sarma and Pais (2011) who find that the level of interest rates has a significant positive
impact on financial inclusion.
As expected, income shows a statistically positive effect on financial inclusion at 1%
significance level. This suggests that as income level increases, financial inclusion increases
all things being equal. These findings are consistent with that of Park and Mercado (2018),
Zins and Weill (2016), and Sarma and Pais (2011).
IJSE Technological infrastructure also impacts financial inclusion positively and statistically
significant at the 5% level. This indicates that as technological infrastructure in an economy
improves, financial inclusion increases ceteris paribus. Lyons et al. (2017) find similar results.
Finally, rule of law shows a positive impact on financial inclusion at a statistically
significant level of 10%. Enhanced rule of law increases financial inclusion as administration
of financial transactions improves. This finding is consistent with Park and Mercado (2018).
The results of the impact of financial inclusion on poverty are presented in Tables 4–7.
The dependent variable poverty is proxied by two indicators – poverty head count ratio 2011
PPP (P0) and poverty gap 2011 PPP values (P1). The results of the estimations using poverty
head count ratio 2011 PPP values as the dependent variable is presented in Table 4.
Table 4 reports the two-step variant of Blundell and Bover estimator which gives more
robust andmore efficient estimates. The results show that the coefficient of the first lag of the
dependent variable, poverty head count ratio 2011PPP has statistically significant
association with the dependent variable at the 1% level which suggests that current level
of poverty is significantly influenced by previous level.
Financial inclusion is inversely and statistically significant with poverty at the 1% level.
This indicates that a unit increase in financial inclusion is associated 1.2572 reduction in
income poverty in the short run, on average holding all other factors constant. This finding is
consistent with the works of Park and Mercado (2018) which suggests that access and usage
of financial services enable people to come out of poverty.
In addition, income inequality is directly related with poverty at 1% level of significance.
Hence, a unit increase in income inequality leads to 0.2411 increase in poverty in the short run,
other factors remaining constant. This implies that as income inequality widens more people
become poor. This finding lends support to that of Fosu (2010) which shows that a more
equitable distribution of income is effective in reducing poverty.
Credit to private sector by banks, a measure of financial sector development is inversely
related with poverty. As financial sector develops by a unit, poverty reduces by 0.0138
Two-step sys GMM Corrected std.
Variables (coefficients) Errors P-values
P0_1 0.0557*** (3.23) 0.0172 0.002
Financial Inclusion, IFI 1.2572*** (3.04) 0.4134 0.003
Income Inequality, GINI 0.2411*** (6.49) 0.0372 0.000
Credit to Priv. Sector by Banks, 0.0138 (1.01) 0.0136 0.314
CRETOPSB
GDP Per Capita, GDPPC 0.0524 (0.91) 0.0573 0.363
Inflation, INFL 0.0063 (0.45) 0.0140 0.651
Rule of Law, RLAW 0.3038 (0.42) 0.7226 0.675
Infrastructure, IFRASTRUC 0.0428*** (3.03) 0.0141 0.003
Population, POP 0.0011 (0.77) 0.0014 0.442
Sec. Sch. Compl. Rate, SECSCHCR 0.0122 (1.53) 0.0080 0.128
Year Dummies YES
Number of Observations 1861
F-Statistic 4.56
Groups 133
Instruments 37
AR(2) 0.968
Table 4.
System GMM results Hansen Statistic 0.015
(dependent variable: Note(s): *, **, *** are statistical significance at the 10%, 5 and 1% levels respectively; t-Statistics are based on
P0 - poverty head count white heteroscedasticity-consistent standard errors, p-values are reported for AR (2) and Hansen statistic
ratio, 2011PPP) Source(s): Created by Authors, 2021
Financial
Two-step sys GMM Corrected std.
Variables (coefficients) Errors P-values inclusion
and poverty
P1_1 0.0491*** (3.05) 0.0161 0.003
Financial Inclusion, IFI 0.4278** (2.56) 0.1670 0.012
Income Inequality, GINI 0.0877*** (5.66) 0.0155 0.000
Credit to Priv. Sector by Banks, 0.0060 (1.12) 0.0054 0.265
CRETOPSB
GDP Per Capita, GDPPC 0.0202 (1.27) 0.0159 0.208
Inflation, INFL 0.0001 (0.01) 0.0054 0.993
Rule of Law, RLAW 0.0731 (0.27) 0.2740 0.790
Infrastructure, IFRASTRUC 0.0160** (3.11) 0.0052 0.002
Population, POP 0.0003 (0.45) 0.0006 0.654
Sec. Sch. Compl. Rate, SECSCHCR 0.0046 (1.55) 0.0030 0.123
Year Dummies YES
Number of Observations 1862
F-Statistic 3.88
Groups 133
Instruments 37
AR(2) 0.551
Table 5.
Hansen Statistic 0.082 System GMM results
Note(s): *, **, *** are statistical significance at the 10%, 5 and 1% levels respectively; t-Statistics are based on (dependent variable:
white heteroscedasticity-consistent standard errors, p-values are reported for AR (2) and Hansen statistic P1 -poverty gap,
Source(s): Created by Authors, 2021 2011PPP)
Variables Two-step sys GMM (coefficients) Std. Errors P-values Table 6.
Long-run GMM
Financial Inclusion, IFI 1.1909*** (3.05) 0.3909 0.002
coefficients for
Income Inequality, GINI 0.2284*** (6.45) 0.0354 0.000 significant coefficients
Infrastructure, IFRASTRUC 0.04053*** (3.05) 0.0133 0.002 (dependent variable:
Note(s): Legend: *p < 0.1, **p < 0.05, and ***p < 0.001 (P0 -poverty
Source(s): Processed Data by Authors, 2021 head count)
Variables Two-step sys GMM (coefficients) Std. Errors P-values Table 7.
Long-run GMM
Financial Inclusion, IFI 0.4019*** (2.58) 0.1559 0.010
coefficients for
Income Inequality, GINI 0.0825*** (5.76) 0.0143 0.000 significant coefficients
Infrastructure, IFRASTRUC 0.0149*** (3.12) 0.0048 0.002 (dependent variable:
Note(s): Legend: *p < 0.1, **p < 0.05, and ***p < 0.001 (P1 -poverty
Source(s): Created by Authors, 2021 gap 19PPP)
though this is not statistically significant. This finding corroborates that of Rashid and
Intartaglia (2017), Jeanneney and Kpodar (2011), and Akhter et al. (2010) who find that
financial development reduces poverty.
Technology infrastructure shows an inverse relationship with poverty as expected at a
significance level of 1%.A unit increase in technological infrastructure is associatedwith 0.0428
reduction in poverty in the short run, ceteris paribus. This finding is in linewithProdi (2015)who
argues that infrastructure, particularly information and communications technology (ICT) can
reduce poverty. This is possible because information and communications technology (ICT) can
IJSE improve people’s access to health, education, agricultural productivity and financial services. In
addition, technology enables people to mobilize each other more efficiently, enhance social
inclusion and interdependence which improve well-being.
Another variable that impacts poverty is human capital proxied by secondary school
completion rate. The result shows an inverse association between human capital and poverty
though not statistically significant. The coefficient shows that a unit increase in human
capital leads to a 0.0122 reduction in poverty, holding other things constant. Olopadea et al.
(2019) confirm that human capital development has a significant effect on poverty reduction
in selected OPEC member countries as improved human capital generates opportunities for
growth through its effects on total factor productivity and this tends to reduce poverty.
Table 5 presents the output of the relationship between financial inclusion and poverty
using the poverty gap 2011 PPP as the dependent variable. This result is very similar to the
results of using poverty headcount ratio as the dependent variable. It shows that financial
inclusion and infrastructure have statistically significant positive impact on poverty. Hence,
financial inclusion and infrastructure reduce poverty. Similarly, as indicated in the previous
results income inequality has a statistically significant negative impact on poverty which
indicates that income inequality worsens poverty. This clearly shows that each of the
measure of poverty can be used in poverty estimations to get comparable results.
The long run estimates verify whether the variables that impact poverty in the short run have
similar impact in the long run when all other factors change. The results of the long run
estimations are presented inTables 6 and 7.Table 6 shows the results for poverty head count ratio
as thedependentvariable. The coefficient of financial inclusion1.1909which suggests that aunit
increase in financial inclusion leads to 1.1909 decline in poverty in the long run, at the 1% level of
significance on average ceteris paribus. Financial inclusion has a relatively smaller positive impact
onpoverty in the long run (1.1909) compared to the short run coefficient of1.2572.Nonetheless,
financial inclusion has both short run and long run influence in reducing poverty.
Further, the coefficient of the long run estimate of income inequality on poverty is 0.2284.
This gives indication that a unit increase in income inequality leads to 0.2284 increase in
poverty at 1% significance level on average ceteris paribus. Income inequality has a slighter
smaller effect on poverty in long run (0.2284) than in the short run (0.2411); though income
inequality aggravates poverty in both the short and long run.
Technological infrastructure has a positive impact on poverty in the long run. The
coefficient of0.04053 suggests that a unit increase in technological infrastructure leads to a
0.04035-unit reduction in poverty. This is significant at 1% level on average. Comparing this
to the short run estimate, infrastructure has a slightly greater positive impact on poverty in
the short run (0.0428) than in the long run (0.04053).
To generate the long-run estimate, estimate the Stata command for the two-step system
GMM for the model that is estimated to generate the short run coefficients and identify the
significant explanatory variables. From the estimation, the variables with significant short-
term coefficients are financial inclusion (IFI), income inequality (GINI), and infrastructure.
Generate the long-run GMM estimates using the following commands: nlcom(_b[IFI])/(1-_b
[L1. P0]); nlcom(_b[GINI])/(1-_b[L1.P0]) and nlcom(_b[IFRASTRUC])/(1-_b[L1.P0]), where IFI,
GINI, and IFRASTRUC are the short-term significant explanatory variables and P0 is
Poverty Head Count Ratio 19 PPP, the dependent variable.
Lastly, the long-run coefficients for the explanatory variables that were significant in the
short run are estimated using poverty gap as the dependent variable. The result is presented
in Table 7. The significant variables that significantly impact poverty in the short run also
significantly influences poverty in the long run. It can be observed from the table that the
financial inclusion has a relatively smaller positive impact on poverty or ability to reduce
poverty in the long run (0.4019) compared to the short run (0.4278). However, financial
inclusion has both short run and long run impact of reducing poverty.
Similarly, the long-run effect of income inequality on poverty is slighter lower in the long Financial
run (0.0825) than in the short run (0.0877), though income inequality worsens poverty in both inclusion
the short and long run. Finally, the long-run impact of technological infrastructure on poverty and poverty
has a slightly greater positive impact in the short run (0.0160) than in the long
run (0.0149).
Similarly, the long-run coefficients for the significant explanatory variables in the short
run are generated using Stata. The results show that financial inclusion and infrastructure
reduces poverty in the long run while income inequality worsens poverty in the long run.
5. Conclusion
This paper empirically examines the impact of financial inclusion on poverty in developing
countries. It employs principal component analysis to construct an index of financial
inclusion with demand and supply indicators, including mobile accounts. For the first in the
literature, financial inclusion index is modelled using mobile money accounts for the poor in
the society.
It further examines factors that affect financial inclusion using pooled ordinary least
squares, least square dummy variable and fixed effects estimation. The system GMM
estimator is used to account for the dynamic nature of the data. In addition, this paper uses
the two-step system GMM estimator because it is more efficient and more robust to
heteroscedasticity and autocorrelation.
The main finding of this study is that most developing countries have financial inclusion
index between medium and high with increasing rates over the observed period. This
indicates that financial inclusion has generally increased over the period.
In addition, it finds that loan rate, technology infrastructure, level of income, and rule of
law have significant positive impact on financial inclusion. This suggests that improving rule
of law, income, infrastructure, and reducing loan rate would be effective strategies to enhance
financial inclusion significantly.
Another evidence is that financial inclusion significantly reduces income poverty in both
the short run and long run irrespective of the proxies for poverty. However, income inequality
exacerbates poverty. The results suggest that financial inclusion is an effective tool to
alleviate poverty in developing countries. Also, equitable distribution of income is effective in
reducing poverty.
The study has several policy implications for developing economies. Firstly, there is the
need to further expand financial inclusion in developing economies because it significantly
reduces poverty. To implement this, policymakers need to create a conducive environment to
expand technological infrastructure and make it affordable to all and sundry. Improving the
use of mobile accounts has the tendency to reach the poor in areas where there are no
traditional financial services providers. The enables the poor to have access to financial
products that helps them out of poverty as mitigation of poverty is key to sustainable growth
and development.
Secondly, it is imperative for policy bearers to enhance the quality of rule of law to instil
confidence in the populace and motivate those who are involuntarily excluded from the
financial system to participate in the system in order to increase financial inclusion. There
is also the need for policymakers to improve the level of income and livelihood of the poor
and vulnerable through economic empowerment initiatives. These initiatives are
important because rule of law and level of income are found to significantly reduce
poverty in the study.
Thirdly, financial inclusion is a policy tool for poverty reduction. Finally, government
organizationsmust double their efforts to narrow income inequality with the goal of reducing
poverty in developing countries.
IJSE References
Abebe, T. and Quaicoe, N. (2014), “Causes of poverty in Sub-Saharan Africa: a layered theory
approach to understanding significant factors”, Journal of Economics and International Finance,
No. 6, pp. 112-124.
Akhter, S., Liu, Y. and Daly, K. (2010), “Cross country evidence on the linkages between financial
development and poverty”, International Journal of Business and Management, Vol. 5
No. 1, pp. 3-19.
Akudugu, M.A. (2013), “The determinants of financial inclusion in Western Africa: insights from
Ghana”, Research Journal of Finance and Accounting, Vol. 4 No. 8, pp. 1-9.
Alkire, S., Foster, J.E., Seth, S., Santos, M.E., Roche, J.M. and Ballon, P. (2015), Multidimensional
Poverty Measurement and Analysis, Oxford University Press, Oxford.
Amid
zic, G., Alexander, M. and Andre, M. (2014), Assessing Countries’ Financial Inclusion-A New
Composite Index, IMF Working Paper, WP/14/36, International Monetary Fund,
Washington, DC.
Arellano, M. and Bond, S. (1991), “Some tests of specification for panel data: monte Carlo evidence and
an application to employment equations”, The Review of Economic Studies, Vol. 58, pp. 277-297.
Arellano, M. and Bover, O. (1995), “Another look at the instrumental variable estimation of error-
components models”, Journal of Econometrics, No. 68, pp. 29-51.
Ayyagari, M., Beck, T. and Hoseini, M. (2013), “Finance and poverty: evidence from India”, CEPR
Discussion Paper No. DP9497, pp. 1-70.
Beck, T., Demirgu€ç-Kunt, A. and Honohan, P. (2009), “Access to Financial Services: Measurement,
Impact, and Policies”, The World Bank Research Observer, No. 24, pp. 119-145.
Beegle, K., Christiaensen, L., Dabalen, A. and Gaddis, I. (2016), Poverty in a Rising Africa, The World
Bank, Washington.
Blundell, R. and Bond, S. (1998), “Initial conditions and moment restrictions in dynamic panel data
models”, Journal of Econometrics, Vol. 87, pp. 115-143.
Chakravarty, R.S. and Pal, R. (2013), “Financial Inclusion in India: an axiomatic approach”, The
Journal of Policy Modeling, No. 35, pp. 813-837.
Demirguۍ-Kunt, A. and Klapper, L. (2012),Measuring Financial Inclusion: The Global Findex Database,
World Bank Policy Research Working Paper, (6025), The World Bank, Washington.
Fosu, A.K. (2010), “Does inequality constrain poverty reduction programs? Evidence from Africa”,
Journal of Policy Modeling, No. 32, pp. 818-827.
Global Findex (2017), available at: http://www.worldbank.org/globalfindex
Honohan, P. (2008), “Cross-country variation in household access to financial services”, Journal of
Banking and Finance, Vol. 32 No. 11, pp. 2493-2500.
Jeanneney, S.G. and Kpodar, K. (2011), “Financial development and poverty reduction: can there be a
benefit without a cost?”, Journal of Development Studies, Vol. 47 No. 1, pp. 143-163.
Kaur, J. (2017), Rural-Urban Disparities in Financial Inclusion in India: A District Level Study, Post-
Doctoral Fellowship Report, ICSSR, New Delhi.
Lal, T. (2017), “Impact of financial inclusion on poverty alleviation through cooperative banks”,
International Journal of Social Economics, Vol. 45 No. 5, pp. 808-828, doi: 10.1108/IJSE-05-
2017-0194.
Levine, R., Loayza, N. and Beck, T. (2002), “Financial intermediation and growth: causality and
causes”, Central Banking, Analysis, and Economic Policies Series, No. 3, pp. 031-084.
Lyons, A.C., Grable, J.E. and Zeng, T. (2017), Infrastructure, Urbanization, and Demand for Bank and
Non-bank Loans of Households in the People’s Republic of China, ADBI Working Paper 767,
Asian Development Bank Institute, Tokyo, available at: https://www.adb.org/publications/
infrastructure-urbanization-and-demand-bank-and-nonbank-loans-households-prc
Olaniyi, E. and Babatunde, A. (2016), “Determinants of financial inclusion in Africa: a dynamic panel Financial
data approach”, University of Mauritius Research Journal, No. 22, pp. 310-336. inclusion
Olopadea, B.C., Okoduaa, H., Oladosuna, M. and Asaleyeba, A.J. (2019), “Human capital and poverty and poverty
reduction in OPEC member-countries”, Heliyon, Vol. 5 No. 2, pp. 02-27.
Park, C.Y. and Mercado, R. Jr. (2018), “Financial inclusion, poverty, and income inequality”, The
Singapore Economic Review, No. 63, pp. 185-206.
Pasiouras, F. and Kosmidou, K. (2007), “Factors influencing the profitability of domestic and foreign
commercial banks in the European Union”, Research in International Business and Finance,
Vol. 21 No. 2, pp. 222-237.
Prodi, R. (2015), Poverty Alleviation: A Role for Technology and Infrastructure?, Concept note,
Foundation for World Wide Cooperation, Bologna, pp. 1-21.
Rashid, A. and Intartaglia, M. (2017), “Financial development–does it lessen poverty?”, Journal of
Economic Studies, Vol. 44 No. 1, pp. 69-86, doi: 10.1108/JES-06-2015-0111.
Sahoo, A.K. (2017), “Determinants of financial inclusion in tribal districts of Odisha: an empirical
investigation”, Social Change, Vol. 47 No. 1, pp. 1-20.
Sarma, M. (2008), Index of Financial Inclusion, Working Paper 215, Indian Council for Research on
International Economic Relations (ICRIER), New Delhi.
Sarma, M. (2012), “Index of Financial Inclusion–a measure of financial sector inclusiveness”,
Berlin Working Papers on, Money, Finance, Trade and Development, Vol. 25 No. 1207, pp. 1-37.
Sarma, M. and Pais, J. (2011), “Financial inclusion and development”, Journal of International
Development, Vol. 23 No. 5, pp. 613-628.
Shaban, M., Girardone, C. and Sarkisya, A. (2019), “Financial inclusion: trends and determinants”,
in Frontier Topics in Banking, Palgrave Macmillan.
Wang, X. and Guan, J. (2016), “Financial inclusion: measurement, spatial effects and influencing
factors”, Applied Economics, Vol. 49 No. 18, pp. 1751-1762.
World Bank (2018), “The World Bank annual report 2018”, available at: www.worldbank.org
Zins, A. and Weill, L. (2016), “The determinants of financial inclusion in Africa”, Review of
Development Finance, Vol. 6 No. 1, pp. 46-57.
Further reading
Abor, J.Y., Amidu, M. and Issahaku, H. (2018), “Mobile telephony, financial inclusion and inclusive
growth”, Journal of African Business, Vol. 19 No. 3, pp. 430-453.
Aghion, P., Howitt, P., Howitt, P.W., Brant-Collett, M. and Garcıa-Pen~alosa, C. (1997), Endogenous
Growth Theory, MIT Press, Cambridge, MA.
Ahlburg, D.A. (1996), “Population growth and poverty”, in Ahlburg, D.A., Kelley, A.C. and Mason,
K.O. (Eds), The Impact of Population Growth on Well-Being in Developing Countries Population
Economics, Springer, Berlin, Heidelberg.
Fosu, A.K. (2017), “Growth, inequality, and poverty reduction in developing countries: recent global
evidence”, Research in Economics, Vol. 71 No. 2, pp. 306-336.
Inoue, T. and Hamori, S. (2016), “Financial access and economic growth: evidence from sub-Saharan
Africa”, Emerging Markets Finance and Trade, Vol. 52, pp. 743-753.
Kosmidou, K., Pasiouras, F. and Tsaklanganos, A. (2007), “Domestic and multinational determinants
of foreign bank profits: the case of Greek banks operating abroad”, Journal of Multinational
Financial Management, No. 17, pp. 1-15.
World Bank (2008), Finance for All? Policies and Pitfalls in Expanding Access, World Bank,
Washington, DC.
IJSE Appendix
Country 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Afghanistan 0.52 0.52 0.52 0.52 0.49 0.49 0.49 0.49 0.49 0.49 0.50 0.50 0.50 0.49 0.49
Albania 0.40 0.25 0.02 0.11 0.37 0.36 0.53 0.63 0.73 0.69 0.75 0.70 0.79 0.63 0.62
Algeria 0.31 0.33 0.33 0.29 0.28 0.23 0.23 0.20 0.17 0.15 0.17 0.15 0.20 0.18 0.15
American 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52
Samoa
Angola 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.40 0.39 0.38 0.37 0.33 0.35
Antigua and 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52
Barbuda
Argentina 0.54 0.56 0.59 0.61 0.65 0.71 0.73 0.76 0.78 0.83 0.83 0.84 0.86 0.88 0.90
Armenia 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52
Azerbaijan 0.52 0.35 0.23 0.02 0.13 0.05 0.16 0.40 0.75 0.45 0.89 0.52 0.52 0.52 0.52
Bangladesh 0.21 0.23 0.24 0.20 0.20 0.20 0.21 0.22 0.21 0.16 0.12 0.18 0.16 0.17 0.17
Belarus 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.54 0.58 0.60 0.64 0.65
Table A1.
Index of Financial Belize 0.54 0.55 0.58 0.60 0.64 0.73 0.73 0.76 0.78 0.81 0.82 0.81 0.85 0.73 0.75
Inclusion for some of Benin 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52 0.52
the 142 Developing Note(s): The full Table for all the 142 countries is available on request
Countries Source(s): Created by Authors, 2021
Corresponding author
Kofi Amoateng can be contacted at: amoateng@nccu.edu
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com