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 access and the finance– Financial

access and the
growth nexus: evidence from finance–
growth nexus
developing economies
Hazwan Haini 693
School of Business and Economics, Universiti Brunei Darussalam,
Bandar Seri Begawan, Brunei Darussalam Received 16 August 2020
Revised 31 December 2020
Accepted 10 February 2021
Abstract
Purpose – This study examines the impact of financial institutions access and financial institutions depth on
economic growth in 51 low- and lower–middle-income countries from 1996 to 2017.
Design/methodology/approach – The study employs an index of financial institutions depth and financial
institutions access that considers the multidimensional nature of finance. The study employs a generalised
least squares model as the baseline fixed effects model suffers from serial correlation. In addition, the study
examines the marginal impact of financial development on growth at varying levels of financial access.
Findings – The results show that both financial access and financial depth are positive to growth. However,
the marginal impact of financial depth is negative at low levels of financial access, while the finance–growth
relationship becomes positive at higher levels of financial access. Results suggest the importance of developing
inclusive financial systems that emphasise quality rather than quantity to promote economic growth.
Research limitations/implications – The major limitation lies in the measurement of financial access as it
focusses more on financial system penetration and overlooks the other aspects of financial inclusion such as
financial literacy and cultural differences.
Practical implications – Developing countries should continue to develop an inclusive financial system that
supports the Universal Financial Access 2020 initiative.
Originality/value – This study provides further empirical evidence on the finance–growth literature
focussing on the impact of financial inclusion which is scarce. Furthermore, the study employs an index of
finance that captures the multidimensional nature of finance.
Keywords Financial access, Access to finance, Economic growth, Finance–growth, Financial development
Paper type Research paper

1. Introduction
In 2017, an estimated 1.7 bn adults lacked access to a financial transaction account (Demirg€ uç
-Kunt et al., 2018). It is suggested that the majority of the population that are excluded from
the financial system cannot afford to obtain an account or find an account which fits into their
low-income lifestyle. As a result, in 2015 at the World Bank Group–International Monetary
Fund (IMF) Spring Meetings, the Universal Financial Access 2020 initiative was
implemented, which commits to increasing the number of adults with a transaction
account. The literature on financial inclusion has emphasised the importance of inclusive
financial systems as finance can help improve income distribution and poverty reduction
over time, which leads to economic growth (Beck et al., 2009; Hajilee et al., 2017; Kim et al.,
2018). The literature on financial access is part of a larger field that examines the role of
financial development on economic growth. Previous and recent studies find financial
development to be positive to growth (King and Levine, 1993). Additionally, financial
development can directly reduce poverty (Ho and Iyke, 2018; Sehrawat and Giri, 2016; Sirag

The author acknowledge the direction of the handling editor, Dr Julian Roche, and the suggestions International Journal of Social
provided by two anonymous referees. Economics
Vol. 48 No. 5, 2021
Funding: The author received no direct funding for this research. pp. 693-708
Conflict of Interest: No potential conflict of interest was reported by the author. The usual disclaimer © Emerald Publishing Limited
0306-8293
applies. DOI 10.1108/IJSE-08-2020-0549
IJSE et al., 2018). On the other hand, there are a number of studies that find evidence against the
48,5 finance–growth relationship, where it is suggested that financial development has a non-
linear relationship with economic growth (Breitenlecher et al., 2015). In fact, some studies find
financial development to be insignificant or even detrimental to growth (Bijlsma et al., 2018).
Consequently, re-examining the finance–growth nexus alongside the impact of financial
access in developing countries makes a compelling study.
This study examines the impact of financial institutions access and financial institutions
694 depth on economic growth in 51 low- and lower–middle-income countries from 1996 to 2017
using panel estimators. The study employs a fixed effects model with post-estimations to test
the validity of the model. The final model is estimated using a generalised least squares
estimator as the fixed effects model exhibits serial correlation. More specifically, the study
seeks to address the following research questions: first, how do financial institutions depth
and financial institutions access impact economic growth in developing economies? The
finance–growth literature has generally focussed on overall financial development; however,
the literature on financial access and economic growth is relatively scarce which is important
for low- and lower–middle-income countries (Sethi and Sethy, 2019). Secondly, what is the
moderating impact of financial access on the finance–growth nexus? Figure 1 highlights the
theoretical considerations of the study. There are many studies that find that financial
development affects economic growth; however, how does financial access affect this
relationship. The overall aim is to examine the marginal impact of financial depth on varying
levels of financial access as recent studies find evidence against the finance–growth nexus.
The study contributes to the literature on financial development and economic growth in
several ways. The study employs indices of financial development that better reflect the
multidimensional nature of the financial system as proposed by Svirydzenka (2016). Previous
studies in the finance–growth literature generally focus on financial development measures
that reflect financial depth in a one-dimensional manner such as credit volumes or M2 supply
(Bijlsma et al., 2018). This study addresses the issue by employing a set of financial indices.
Secondly, the study contributes to the literature providing new empirical evidence on the
impact of financial access and economic growth in developing economies. It is suggested that
the literature on financial inclusion and economic growth is scarce (Sethi and Sethy, 2019).
Finally, the study contributes further empirical evidence on the growing literature that
postulates the weakening effects of finance as it examines the marginal impact of financial
depth on growth at varying levels of financial access.
The estimated results show that both financial institutions depth and financial
institutions access are positive and significant to growth, supporting recent empirical
studies that examine the finance–growth nexus in developing economies (Haini, 2020; Inoue
and Hamori, 2016). On the other hand, the results indicate that the impact of financial
institutions access is larger in magnitude than financial depth. More interestingly, the
interaction between financial depth and financial access provides some evidence of a non-
linear relationship between finance and growth. The marginal effects show that the
relationship between finance and growth is negative at low levels of financial access, and this

Figure 1.
Does financial access
moderate the impact of
financial depth on
economic growth?
relationship becomes positive at higher levels of financial access. The estimated results Financial
suggest that financial access is an important requirement for financial depth to be effective in access and the
developing economies. Policymakers should continue to support the Universal Financial
Access 2020 initiative to improve the quality and accessibility of the system rather than
finance–
quantity as low levels of financial access lead to an unproductive use of finance. growth nexus
The rest of the study is organised as follows. Section 2 provides a brief literature review on
the importance of financial inclusion for economic development. Section 3 describes the data
sample, variables employed and the econometric strategy, while Section 4 presents the results 695
and provides a brief discussion on the reported results. Finally, Section 5 concludes the study
with policy implications and directions for future research.

2. Financial inclusion and economic development

Financial inclusion can be defined as a process that serves to remove the barriers of specific
societal groups and individuals. These not only include the poor but also those with
disadvantaged access to use low-cost, fair and safe formal financial services. It should ensure
ease of access, availability and usage of the range of formal financial instruments for all
members of an economy. Thus, measuring inclusion is a multidimensional variable that
includes banking penetration, availability of banking services and usage of banking services
(Claessens, 2006). As a result, financial inclusion is emphasised as an important policy option
aimed at alleviating poverty, minimising social exclusion and enhancing economic growth. In
addition, financial inclusion offers incremental and complementary solutions to tackle
poverty and to promote inclusive development (Chibba, 2009).
Although overall financial development is important to economic growth, it is crucial to
ensure that financial inclusion is taken into account. Beck et al. (2009) suggest that lack of
access to finance is often the critical mechanism for generating persistent income inequality,
as well as slower economic growth. Globalisation of finance can also play a part in improving
access as it allows the increase of funds and efficiency of capital allocation. However, in-depth
financial systems still offer limited outreach that offers the core functions of a system such as
savings, payments and risk management. Furthermore, academics suggest that access to
affordable finance may enable the poor to undertake economic activities and to take
advantage of growth opportunities necessary for financial development (Claessens, 2006).
Thus, expanding access to finance remains an important challenge for many government and
organisations.
In addition, financial development’s effect on fostering economic growth can help reduce
poverty indirectly, augment self-employment and enable households to smooth consumption
expenditure over time. Evidence suggests that the effects of financial sector deepening on
poverty are greater through the indirect effects of growth and labour markets than via direct
effects of employment through credit provision (Johnson and Arnold, 2012). This effect on
poverty may be important in countries where financial inclusion is naturally high. For
example, in Africa, Beck and Cull (2014) showed that enterprises in the African region are less
likely to have a loan than other regions. As small and medium-sized enterprises (SMEs)
constitute an important component of the private sector in the developing world, having
financial exclusion means having a higher obstacle to operation and growth.
Governments and other organisations need to be proactive in promoting financial
inclusion. Microfinancing institutions (MFIs) can play a significant role in bridging the gap
between the formal financial institutions and the rural poor. MFIs are generally overseas non-
profit organisations that make microcredit loans to villagers, micro entrepreneurs and poor
families with the main role of improving access to finance. These are institutions that serve
the poor in order to improve access to finance. Economists are increasingly considering the
potential of MFIs as it has managed to reach millions of clients at an impressive repayment
rate, despite serving the poorest individual, in addition to being self-sufficient. In addition to
IJSE tackling the poverty issue, MFIs support economic growth as an institution that promotes
48,5 savings, improves access to credit and supports gender equality. Khandker (2005) discusses
the role of microfinancing in Bangladesh and how it allows the poor and women to engage in
informal activities.
Yet, the question remains as to how exclusion towards the poor occurs in the first
place. There are various reasons that suggest why the poor do not have access to formal
finance. Financial exclusion can exist due to physical distance from the formal financial
696 sector (Beck et al., 2009). Rural areas are always underserved by the formal banking
sector and thus simply cannot access the financial services they need. Many of the
underserved people are in countries that are less developed, may lack education and thus
may have difficulties in understanding services during the application process (Beck
et al., 2009).
In addition, less developed countries may have a lack of steady jobs and collateral, making
it more expensive for financial institutions to price risk and thus making it costly. This is
supported by Sarma and Pais (2011), who identified that financial exclusion is a manifestation
of social exclusion and that employment can affect participation in the financial system. The
lack of steady jobs in less developed countries may be explained by the fact that the informal
economy accounts for a large share of employment, where salaries and wages are not paid
through the formal banking system. Furthermore, similar observations are identified where
the level of human development, such as education, is strongly positively correlated with
financial inclusion (Sarma and Pais, 2011). Thus, social and personal deprivation can be
viewed as key barriers to financial inclusion (Cnaan et al., 2012). Therefore, encouraging
inclusion must deal with the several barriers that affect different groups of the underserved
population.
Broad institutional framework plays an important role in expanding financial outreach
and inclusion. Adverse impacts of financial inclusion are not only lost in terms of opportunity
cost but also higher transaction costs. Therefore, financial inclusion is always a desired
outcome as it can help poor people access financial services at a lower cost and reduce the
consequences of poverty (Cnaan et al., 2012). However, more and better data are needed on
financial outreach and inclusion as well as analysis to better understand the channels
through which financial deepening and inclusion help reduce income inequality and poverty
(Beck and Demirg€ uç-Kunt, 2008). Similarly, due to data availability, this study focusses on the
accessibility of financial inclusion, and as a result, this needs to be interpreted as such.
However, as there are various dimensions to access and consequently various dimensions in
which access may be deficient (Claessens, 2006); thus, the study contributes new evidence as
it employs an index of financial institutions access.

3. Empirical methodology
This section discusses the econometric strategy employed in the study, followed by a brief
discussion on the data and variables employed. Since the study employs panel estimators,
prior to modelling, the study employs panel unit root tests to assess the stationarity of the
variables employed. Due to space constraints, the results are available as requested. The
panel unit root tests are non-stationary at levels and the panel unit root tests at first difference
reject the joint null hypothesis at the 1% level.
3.1 1Panel estimators
There are various methods to examine the impact of financial access and financial
development on economic growth. As this study examines 51 low- and lower–middle-income
countries over a sample time period, panel analysis is suitable in this case. Previous empirical
studies that examine the finance–growth nexus generally employ a Barro (1991)-styled
regression or a Cobb–Douglas production function approach. Following previous empirical
studies, this study employs a Barro (1991) neoclassical-styled growth regression that allows Financial
the model to be parsimonious and augmented with financial variables and other variables of access and the
interest.
yit ¼ α þ βXit þ εit finance–
εit ¼ μi þ νit (1) growth nexus

Equation (1) presents a standard growth regression in panel form, where the observations are
measured across ith country and at time t. Here, yit represents the dependent variable gross 697
domestic product (GDP) per capita, while the independent variables of interest are denoted by
Xit. The error term εit consists of the country-specific time-invariant error μi and the
idiosyncratic error νit , which is assumed to be independent and identically distributed.
There are several methods of estimating Equation (1). On the one hand, Equation (1) can be
estimated using the ordinary least squares (OLS) estimator. However, the OLS estimator can
be biased as the country-specific time-invariant error can be correlated with the independent
variable as the OLS estimator ignores the panel structure of the model. To address these
issues, a common approach is to employ a panel regression with fixed effects. The fixed
effects estimator can control for endogeneity and omitted variables biasedness as it assumes
that country-specific effects are correlated with the independent variables (Wooldridge,
2002). However, if this assumption does not hold, then the random effects estimator is more
efficient. The Hausman test is reported to discriminate between the two estimators and the
estimated results below show that the fixed effects estimator is the preferred model.
In contrast, linear panel data models are susceptible to serial correlation. This is an issue
as serial correlation in panel models leads to biases in the standard errors and leads to less
efficient estimates (Drukker, 2003). As a result, the fixed effects model must be tested for
serial correlation. This study employs the Woolridge test for serial correlation in panel
models, which has good size and power properties and is based on fewer assumptions leading
to a more robust test (Woolridge, 2002).
Furthermore, panel data models are also sensitive to cross-sectional dependence in the
error terms. Cross-sectional dependence can arise due to the presence of common shocks that
are unobserved but uncorrelated with the independent variables leading to less efficient
estimates for the fixed effects model (De Hoyos and Sarafidis, 2006). This is important to
examine, especially in the context of financial development, as many economies have
increased economic and financial integration across borders, leading to interdependencies.
This study employs Pesaran’s (2004) and Friedman’s (1937) test of cross-sectional
dependence for the fixed effects model. This is crucial as the impact of cross-sectional
dependence is severe where the efficiency of fixed effects do not differ much from the OLS
approach (De Hoyos and Sarafidis, 2006). This study does not employ the Lagrange multiplier
test to examine cross-sectional dependence as the sample data as a larger set of countries
relative to time. The study also conducts several tests for normality and examines the
skewness and kurtosis of the error term proposed by Alejo et al. (2015).
The results below indicate that the fixed effects estimates display serial correlation. On a
positive note, the results do not suffer from cross-sectional dependence. Thus, the study
estimates the final model by employing the generalised least squares estimator to address
serial correlation issues. Furthermore, as this study employs a group of developing
economies, the data sample is aggregated, potentially leading to clustering and policy
autocorrelation problem. Meanwhile, the generalised least squares estimators address these
issues and asymptotically provide a more efficient estimate and avoid misleading inferences
(Hansen, 2007). The generalised least squares estimator considers serial autocorrelation
within the panels, which addresses the issues arising from the fixed effects estimator.
Finally, the study examines the non-linear impact of financial development and economic
growth. Many previous studies examine the non-linearity effect of finance by including the
IJSE squared term of the variable or by using threshold effects (Breitenlecher et al., 2015; Haini,
48,5 2020). This study examines the non-linear impact of financial development at varying levels
of financial access. The hypothesis is as follows: at low levels of financial access, financial
depth is hypothesised to have an insignificant or negative impact on growth. However, as
financial access increases, the impact of financial depth should turn positive. This approach
provides new evidence on the non-linear impact of finance by focussing on the marginal
impact of financial access. The following subsection justifies and describes the variables
698 employed.
3.2 Data and variables
The study employs annual-level data from a balanced panel data set of 51 low- and lower–
middle-income countries. The low- and lower–middle-income countries are defined using the
World Bank Atlas method with a gross national income of $1,035 or less for low-income
countries and a gross national income of $1,036 and $4,045 for lower–middle-income
countries (World Bank). The sample countries were chosen as the study focusses on a
homogeneous set of countries that face issues with financial inclusion and access, which are
generally lower in developing economies. Meanwhile, the sample time period was chosen
based on data availability. The data are compiled from various sources including the Penn
World Table 9.1 (Feenstra et al., 2015), the World Governance Indicators (World Bank), the
World Development Indicators (World Bank), the KOF Globalisation database (Gygli et al.,
2019) and the IMF Financial Development Index database (Svirydzenka, 2016).
The variables employed are summarised in Table 1, while Equation (2) represents the full
specification of the estimated model. All variables are log-transformed, except for the indices
employed. It is suggested that the logarithm specification provides more robust estimates
even after correcting for publication bias (Bijlsma et al., 2018). Furthermore, all variables are
normalised to have a mean value of 0 with a standard deviation (SD) of 1 prior to the panel
estimations. This allows for the marginal impact between financial access and financial depth
to be interpreted correctly.
yit ¼ α þ β1 fdit þ β2 fasit þ β3 ðfd 3 fasÞit þ β4 fgit þ
(2)
β5 govit þ β6 opnit þ β7 popit þ β8 hcit þ β9 lawit þ εit

The dependent variable, economic growth, is measured by real GDP per capita. This follows
previous empirical studies that examine the relationship between economic growth and
financial development (Bijlsma et al., 2018; King and Levine, 1993). Real GDP per capita,

Variable Definition Source Mean SD Min Max

y Real GDP per capita (US$ 2011 Constant) PWT 1245.73 889.04 187.52 4343.44
fd Financial institutions depth index (0–100) IMF 9.65 7.99 0.35 42.72
fas Financial institutions access index (0–100) IMF 8.57 9.83 0.36 66.52
fg KOF financial globalisation index (0–100) KOF 47.18 11.34 20.56 75.56
gov Share of government expenditure to GDP (%) PWT 15.28 6.99 0.81 53.71
opn Ratio of imports and exports to GDP (%) PWT 33.24 23.92 3.59 191.99
pop Population growth (%) WDI 2.18 1.08 %1.69 8.12
hc Human capital index PWT 1.81 0.49 1.05 3.45
law Rule of law index WGI %0.74 0.49 %2.13 0.35
Note(s): N 5 1,122 observations from 51 low- and lower–middle-income countries (World Bank) 1996 to
2017.The countries included are shown in Appendix. The statistics presented are in levels and prior to log-
transformation and normalisation. PWT refers to Penn World Table 9.1, IMF refers to the International
Table 1. Monetary Fund, KOF refers to the KOF globalisation index, WGI refers to World Governance Indicators and
Summary statistics WDI refers to World Development Indicators
denoted by yit , captures economic growth and adjusts for total population and inflation Financial
(Haini, 2020). Real GDP per capita is generally used to compare standards of living between access and the
countries and over time.
Focussing on the financial development and financial access variables, the study employs
finance–
two finance variables suggested by Svriydzena (2016). The IMF Financial Development Index growth nexus
database provides indices for financial development that accounts for both financial
institutions and financial markets. As this study focusses on developing economies, the
financial institutions indices are employed. It is suggested that financial intermediaries 699
generally outperform a market-based system, particularly for developing countries as
financial intermediaries can lead to a more equitable distribution of income (Chakraborty and
Ray, 2006). Furthermore, financial intermediaries are suggested to promote physical and
human capital investment more effectively for poorer countries (Cull and Xu, 2013). As a
result, the financial markets indices are not employed in this study.
The study employs the financial institution depth index, denoted by fdit, and the financial
institutions access index, denoted by fasit. Recent studies have incorporated the use of indices
to measure financial development and financial inclusion as the variables are
multidimensional in nature, and as a result, the use of indices provides a more complete
picture of the impact of financial access and financial development on economic growth (Kim
et al., 2018; Serawat and Giri, 2016; Sethi and Sethy, 2019). The financial institutions depth
index includes the private sector credit to GDP, pension fund assets to GDP, mutual fund
assets to GDP and insurance premiums, while the financial institutions access index is
captured by bank branches per 100,000 adults and automated teller machines (ATMs) per
100,000 adults. Finally, the interaction term between financial institutions depth and financial
access, denoted as ðfd 3 fasÞit , is included to examine whether financial access and financial
depth are complementary and provide positive externalities.
In addition, the study employs an additional measure of financial development. The study
employs the KOF financial globalisation index, denoted as fgit, proposed by Gygli et al. (2019).
The financial globalisation index is an important addition to the finance–growth nexus
literature as many previous studies focus on overall financial depth. Meanwhile, financial
globalisation focusses on the extent of external financing and financial openness of an
economy. The KOF financial globalisation index includes foreign direct investment, portfolio
investment, international debt, international reserves and international income payments, as
well as the policies involved in financial openness such as investment restrictions, capital
account openness and international investment agreements. The inclusion of financial
globalisation should provide interesting findings for the finance–growth nexus in the
developing economies context.
The study employs several macroeconomic control variables that are used in previous
empirical studies examining finance and growth (Habib et al., 2019). The study employs a
human capital development index, denoted by hcit , that considers the multidimensional
nature of human capital as the index considers the average years of schooling and the rates of
return to education (Feenstra et al., 2015). The role of human capital in promoting financial
development is well established in previous and recent empirical literature; however, there
are measurement issues as different proxies of human capital lead to varying results
(Ibrahim, 2018). Thus, the study employs an index of human capital that encompasses the
varying nature of human capital.
In addition, the control variables also include openness to trade, denoted by opnit, which
measures the ratio of total imports and exports to GDP. The role of trade openness in
economic growth is well established in the empirical and theoretical literature and as such is
included as a control variable (Darku and Yeboah, 2018). The study also includes government
expenditure, denoted by govit . This is measured by the ratio of government consumption to
GDP. The impact of government expenditure on GDP is inconclusive as on the one hand,
IJSE some previous studies find a positive relationship between government expenditure on
48,5 growth (Easterly and Rebelo, 1993). Yet, other studies find large governments can retard
growth and be detrimental to the allocation of resources (F€olster and Henrekson, 2001).
Moreover, the study also includes population growth as a control variable, denoted by
popit. Population growth is measured by annual population growth in percentages and this
has been employed in previous empirical studies that examine the role of population in
economic growth (Darku and Yeboah, 2018). Finally, the rule of law, denoted by lawit , is
700 included as a control variable. The quality of institutional development in a country can be a
significant determinant of economic growth as it aligns the incentives of private and social
returns together that can indirectly affect financial development (La Porta et al., 1998). The
impact of institutional quality, such as the rule of law, has been well researched and
established in previous and recent empirical studies (Law et al., 2018).
The summary statistics of the variables employed (in levels) are presented in Table 1. The
annual GDP per capita shows some variation with a mean of $1,245.73. The low level of GDP
per capita is expected as the sample size consists of developing economies and the variance
suggests changes across the developing countries over time. This may be due to having a
sample set of countries that are industrialised such as Indonesia and India, with countries
that are in extreme poverty. Despite the differences, as the mean value is low, the
homogeneous nature of the sample is preserved. On the other hand, financial access and
financial depth are both low in their mean values and exhibit high levels of variation across
the sample and time. This is expected as many developing economies have underdeveloped
financial systems and have implemented policies to increase financial access in recent years.
On the other hand, human capital is low with low levels of variance as expected, while the
mean level of population growth stands at 2.18%. Meanwhile, the summary statistics show
that openness to trade has higher variance and again, this can be explained by the sample set
of countries which includes low- and lower–middle-income countries. Government
expenditure has lower levels of variance as expected as governments play an important
role in many developing countries. Finally, the rule of law is negative as expected as
institutional development is generally lower in developing countries.

4. Results and discussion

This section begins with a brief discussion on the correlation matrix, followed by a discussion
on the estimated regression coefficients and the marginal impacts of financial depth and
financial access.
The estimated correlation matrix is presented in Table 2. The correlation matrix measures
the linear relationship between two variables. The variables are perfectly positively
correlated if they obtain a value of þ1 and are perfectly negatively correlated if they have a

Variable y fd fas fg gov opn pop hc law

y 1.000
fd 0.469 1.000
fas 0.698 0.638 1.000
fg 0.495 0.333 0.356 1.000
gov 0.042 0.024 0.078 0.013 1.000
opn 0.468 0.318 0.343 0.525 0.185 1.000
pop %0.384 %0.339 %0.527 %0.071 %0.057 %0.137 1.000
hc 0.566 0.443 0.608 0.374 0.128 0.275 %0.546 1.000
law 0.161 0.358 0.293 0.048 0.183 %0.053 %0.208 0.108 1.000
Table 2. Note(s): Definition of variables is in Table 1. Variables are log-transformed and normalised prior to
Correlation matrix estimations
value of %1. The estimates from the correlation matrix provide some insight prior to the Financial
estimation of the full model. The financial variables present interesting correlations. access and the
Financial institutions depth has a positive and moderate correlation to growth, while
financial institutions access has a positive and strong correlation to growth. This provides
finance–
the expectation that financial access is more effective in spurring growth than depth for the growth nexus
developing countries. Similarly, financial globalisation also has a positive and moderate
correlation to growth. As the final model includes all three financial measures, the
correlations between the three variables should be examined. The reported correlation 701
between financial access and financial depth is positive and moderate, which is acceptable.
Finally, the control variables show expected correlations; human capital and openness to
trade are positive and moderately correlated to growth; government expenditure and the rule
of law are positive and weakly correlated to growth, while population growth is negative and
moderately correlated to growth.
The estimated model coefficients for the OLS, fixed and random effects and the
generalised least squares estimators are reported in Table 3. As previously discussed in
Section 3, the OLS estimator is subject to biasness, while the estimated fixed effects model is
susceptible to serial correlation and cross-sectional dependence, as the study employs a panel
data set of a relatively homogeneous group of developing economies. Table 3 reports the
various estimators for comparative purposes and to justify the use of the generalised least
squares estimator. Since the OLS estimator is prone to biasness, the estimated OLS
coefficients can be overlooked.
Thus, focussing between the fixed and random effects model, it can be observed that the
fixed effects model is the preferred model as the Hausman test rejects the null hypothesis of
having the random effects as the preferred model. The details of the Hausman test coefficients
are reported in Appendix Table A1. Moreover, the serial correlation test shows that the
estimated fixed effects model is not robust. In detail, the null hypothesis of no serial
correlation is strongly rejected. On a positive note, Persaran’s and Friedman’s cross-sectional
dependence test fails to reject the null hypothesis of no cross-sectional dependence in all sets
of regressions. However, as the estimated fixed effects estimators are not robust, the results
are not discussed.
Consequently, the discussion focusses on the estimated generalised least squares
estimator. Focussing on the financial variables, financial institutions depth is positive at the
5% level. The results suggest that financial depth can promote economic growth in
developing economies and broadly support recent studies that find evidence of the positive
effect of financial development on economic growth in developing economies (Ho and Iyke,
2018; Sehrawat and Giri, 2016; Sirag et al., 2018). In contrast, the results indicate evidence
against the literature that postulates the weakening effects of financial development on
growth (Breitenlechner et al., 2015).
Similarly, the estimates show that financial access is positive and significant at the 1%
level to economic growth. Although both financial access and financial depth are positive, the
impact of financial access is greater in magnitude. This highlights the importance of
distinguishing the multidimensional nature of finance as financial depth and financial access
can have varying impacts on economic growth. Furthermore, financial access has a
complementary effect on overall financial development as the interaction term between
financial depth and financial access is positive and significant at the 1% level. This
interaction term will be decomposed to estimate the marginal impact of financial depth at
varying levels of financial development. Thus, the estimates support the literature that
highlights the importance of financial access for spurring economic growth in developing
economies as financial access allows finance to be utilised by a larger set of participants that
can contribute to growth (Kim et al., 2018; Rastogi and Ragabiruntha, 2018; Sethi and Sethy,
IJSE Variables OLS Fixed effects Random effects Generalised least squares
48,5
fd %0.041 0.077*** 0.074*** 0.028**
(0.026) (0.018) (0.018) (0.015)
fas 0.502*** 0.190*** 0.196*** 0.218***
(0.030) (0.020) (0.019) (0.021)
(fd 3 fas) 0.106*** 0.055*** 0.055*** 0.030***
702 (0.019) (0.009) (0.009) (0.011)
fg 0.178*** 0.023* 0.026** 0.019**
(0.024) (0.013) (0.013) (0.009)
gov %0.074*** %0.031*** %0.031** %0.019**
(0.020) (0.011) (0.011) (0.009)
opn 0.189*** 0.094*** 0.096*** 0.037***
(0.024) (0.015) (0.015) (0.008)
pop 0.024 0.032** 0.031** 0.004
(0.025) (0.014) (0.014) (0.012)
hc 0.170*** 0.360*** 0.353*** 0.393***
(0.027) (0.033) (0.032) (0.030)
law 0.041* 0.034** 0.032** 0.042***
(0.022) (0.015) (0.015) (0.012)
Constant %0.068*** %0.035*** %0.035 %0.021
(0.022) (0.008) (0.089) (0.031)
F-statistic 190.68 174.12
Wald χ 2, p 0.000 0.000
Hausman test, p 0.000 0.000
R2 within 0.596 0.597
R2 between 0.522 0.528
R2 overall 0.607 0.525 0.530
Woolridge test for serial-correlation in panels
prob > F 0.000
Cross-sectional dependence tests
Pesaran, p 1.732 1.710
Friedman, p 1.000 1.000
Normality tests
Skewness εit %0.003 Skewness μi 0.045
(0.005) (0.051)
Kurtosis εit 0.007 Kurtosis μi %0.032
(0.006) (0.044)
Prob > joint test εit 0.384 Prob > joint test μi 0.509
Table 3.
Regression coefficients Note(s): N 5 1,122 observations. Definition of variables is in Table 1. Variables are log-transformed and
and associated normalised. *, **, *** denote statistical significance at the 10%, 5% and 1% levels, respectively. Standard
parameters errors are in parenthesis

2019). Furthermore, the complementary nature of financial access suggests that financial
inclusion is said to play a major role in overall financial development (Hajilee et al., 2017).
Finally, the control variables show expected results. The reported estimates show that
human capital is positive and significant at the 1% level to economic growth. The role of
human capital in economic growth is well established, and the findings support recent
empirical studies that examine the interactive effects of human capital in a finance–growth
model (Ibrahim, 2018). Additionally, openness to trade is positive and significant to economic
growth at the 1% level, supporting previous studies that highlight the role of trade openness
in economic growth (Darku and Yeboah, 2018). However, the coefficients for openness to Financial
trade are very small in their magnitude when compared to the other control variables, and access and the
this may be due to the fact that developing economies may be weaker in terms of its export
performance when compared to industrialised countries.
finance–
On the other hand, the estimated results for government expenditure are negative and growth nexus
significant to growth at the 5% level. The role of governments in economic growth has been
well debated, and the negative impact of government expenditure has been found in earlier
studies (F€olster and Henrekson, 2001). This suggests that government spending in 703
developing countries may be inefficient. In the meantime, the rule of law is reported to be
positive and significant at the 1% level to economic growth as expected. It is suggested that
institutions have emerged as a fundamental determinant of economic growth and that
institutions in low-income countries are crucial in order for financial development to succeed
(La Porta et al., 1998; Law et al., 2018). Thus, the findings broadly support the role of
institutional development in promoting economic growth. Finally, the estimated results find
that population growth is insignificant to economic growth. The role of population growth
has been debated; on the one hand, excessive population growth can crowd out the positive
effects of other factors through overconsumption, while a drop in fertility can negatively
impact the labour force (Haini, 2020).
Overall, the estimated results broadly support many previous empirical studies that
examine growth determinants in developing economies. Meanwhile, the study is interested in
examining the marginal impact of financial depth at varying levels of financial access. Thus,
focussing on the interaction term, (fd 3 fas), Table 4 presents the marginal effects of financial
depth at varying levels of financial access as the interaction term cannot provide inferences.
Table 4 decomposes the interaction term and evaluates the marginal effects of the interaction
term at the minimum, mean and maximum level of financial access.
Although Table 4 reports the estimated marginal effects of the OLS, fixed and random
effects, the study focusses on the marginal effects of the generalised least squares estimator.
The estimated marginal impact of the interacted term shows that at the minimum level of
financial access of %2.459 (growth rate, normalised), an increase of 10% in financial
institutions depth decreases GDP per capita by 4.6%. In contrast, at the mean and maximum
level of financial access of 0 and 2.442 (growth rate, normalised), an increase of 10% in
financial institutions depth increases GDP per capita by 3.0 and 10.2%, respectively. The
results suggest that the impact of financial development becomes negative at low levels of
financial access, while the impact becomes positive as financial access increases. The results
provide new empirical evidence of a non-linear impact between financial development and
economic growth in a new perspective that focusses on financial access. The results broadly
support the view that financial development has a non-linear relationship with economic
growth (Bijlsma et al., 2018).

5. Conclusion
The empirical literature supporting the positive relationship between financial development
and economic growth is vast. However, in recent years, there is a growing body of literature
that postulates a non-linear relationship between the two due to excessive financial
liberalisation and financial crises. Deepening our understanding of financial development
and economic growth is important in the context of low- and lower–middle-income countries
as they have less developed financial systems. Consequently, policy prescriptions that
encourage excessive financial depth while compromising financial access can be detrimental.
Furthermore, many low- and lower–middle-income countries have a sizeable under banked
population that can benefit more. The World Bank Group and many partnering countries
have committed towards the Universal Financial Access 2020 initiative that aims to develop
an inclusive financial system that provides the population with an account for transactions.
 48,5
IJSE

704

Table 4.

financial access
varying levels of
financial depth at
Marginal effects of
OLS Fixed effects Random effects Generalised least squares
Minimum Mean Maximum Minimum Maximum Minimum Maximum Minimum Maximum
(%2.459) (0) (2.442) (%2.459) Mean (0) (2.442) (%2.459) Mean (0) (2.442) (%2.459) Mean (0) (2.442)

%0.303*** %0.041 0.216*** %0.058** 0.077*** 0.209*** %0.062** 0.074*** 0.207*** %0.046* 0.030*** 0.102***
(0.052) (0.026) (0.053) (0.027) (0.018) (0.030) (0.027) (0.018) (0.030) (0.024) (0.015) (0.035)
Note(s): Variables are log-transformed and normalised (mean 0, standard deviation 1). *, **, *** denote statistical significance at the 10%, 5% and 1% levels,
respectively. Standard errors are in parenthesis
With this in mind, understanding the role of financial access in the finance–growth nexus is Financial
the basis of the study. access and the
This study examines the impact of financial institutions access and financial institutions
depth on economic growth in 51 low- and lower–middle-income countries from 1996 to 2017
finance–
using panel estimators. This study differs from previous studies as it employs the financial growth nexus
development indices proposed by Svirdydzenka (2016), which measure the multidimensional
nature of financial institutions accessibility and depth. While previous studies only focus on
single variables such as financial depth, they do not capture the importance of financial 705
access which is more important for developing economies. As a result, the study provides new
empirical evidence that examines the role of financial access in the finance–growth nexus,
which is relatively scarce in the literature.
The estimated results show that both financial institutions access and financial
institutions depth lead to more growth. Meanwhile, the interaction between financial
access and financial depth presents interesting results as it suggests a non-linear relationship
between financial development and growth. More specifically, the marginal results suggest
that at low levels of financial access, the impact of financial depth on economic growth is
negative. However, the relationship becomes positive as financial access increases. The
findings support the literature that highlights the importance of financial inclusion and
financial access on economic growth (Hajilee et al., 2017; Kim et al., 2018; Rastogi and
Ragabiruntha, 2018). Thus, this study provides new empirical evidence on the role of
financial access and financial depth in economic growth.
There are a number of policy implications for the study. Policymakers in low- and lower–
middle-income countries should continue to develop an inclusive financial system that
supports the Universal Financial Access 2020 initiative if they are not part of the programme.
This includes developing access points to reach those who are geographically disadvantaged
and creating a regulatory environment which supports marginalised communities. The
importance of inclusive finance is well established in the study, where it is found that
excessive financial deepening can have detrimental consequences on economic growth than
an underdeveloped financial system. As a result, focussing on inclusive financial systems can
provide a more sustainable growth path.
There are a number of directions for future work. This study focusses on an index of
access point as a measure of financial access. There are various dimensions to financial
access and more importantly various dimensions to financial inclusion that are equally
important when considering the impact of financial inclusion on growth. As a result, future
work can examine other elements of financial inclusion in both developing and developed
economies such as the role of online banking, financial literacy and cultural issues in financial
system adoption and its impact on economic growth. Financial literacy and cultural issues are
important to consider as these marginalised groups are excluded from the formal financial
system due to lack of education or due to gender. These issues are important to examine in
developing countries and can deepen our understanding on the role of financial inclusion on
economic growth.

References
Alejo, J., Montes-Rojas, G., Galvao, A. and Sosa-Escudero, W. (2015), “Tests for normality in linear
panel-data models”, STATA Journal, Vol. 15 No. 3, pp. 822-832.
Barro, R.J. (1991), “Economic growth in a cross-section of countries”, Quarterly Journal of Economics,
Vol. 106 No. 2, pp. 407-443.
Beck, T. and Cull, R. (2014), “Small-and medium-sized enterprise finance in Africa”, Africa Growth
Initiative, Working Paper 16, Brookings Institution.
IJSE Beck, T. and Demirg€
uç-Kunt, A. (2008), “Access to finance: an Unfinished Agenda”, The World Bank
Economic Review, Vol. 22 No. 3, pp. 383-396.
48,5
Beck, T., Demirg€
uç-Kunt, A. and Honohan, P. (2009), “Access to financial services: measurement,
impact, and policies”, The World Bank Research Observer, Vol. 24 No. 1, pp. 119-145.
Bijlsma, M., Kool, C. and Non, M. (2018), “The effect of financial development on economic growth: a
meta-analysis”, Applied Economics, Vol. 50 No. 57, pp. 6128-6148.
706 Breitenlechner, M., G€achter, M. and Sindermann, F. (2015), “The finance-growth nexus in crisis”,
Economics Letters, Vol. 132, pp. 31-33.
Chakraborty, S. and Ray, T. (2006), “Bank-based versus market-based financial systems: a growth-
theoretic analysis”, Journal of Monetary Economics, Vol. 53 No. 2, pp. 329-350, doi: 10.1016/j.
jmoneco.2005.01.003.
Chibba, M. (2009), “Financial inclusion, poverty reduction and the Millennium development goals”,
European Journal of Development Research, Vol. 21, pp. 213-230.
Claessens, S. (2006), “Access to financial services: a review of the issues and public policy objectives”,
The World Bank Research Observer, Vol. 21 No. 2, pp. 207-240.
Cnaan, R.A., Moodithaya, M.S. and Handy, F. (2012), “Financial inclusion: lessons from rural South
India”, Journal of Social Policy, Vol. 41 No. 1, pp. 183-205.
Cull, R. and Xu, L.C. (2013), “Job growth and finance: are some financial institutions better suited to
the early stages of development than others?”, World Bank Economic Review, Vol. 27 No. 3,
pp. 542-572, doi: 10.1093/wber/lhs050.
Darku, A.B. and Yeboah, R. (2018), “Economic openness and income growth in developing countries: a
regional comparative analysis”, Applied Economics, Vol. 50 No. 8, pp. 855-869.
De Hoyos, R.E. and Sarafidis, V. (2006), “Testing for cross-sectional dependence in panel-data models”,
STATA Journal, Vol. 6 No. 4, pp. 482-496.
Demirg€
uç-Kunt, A., Klapper, L., Singer, D., Ansar, S. and Hess, J. (2018), The Global Findex Database
2017: Measuring Financial Inclusion and the Fintech Revolution, International Bank for
Reconstruction and Development/The World Bank, Washington.
Drukker, D.M. (2003), “Testing for serial correlation in linear panel-data models”, STATA Journal,
Vol. 3 No. 2, pp. 168-177.
Easterly, W. and Rebelo, S. (1993), “Fiscal policy and economic growth”, Journal of Monetary
Economics, Vol. 32, pp. 417-458.
Feenstra, R.C., Robert, I. and Timmer, M.P. (2015), “The next generation of the Penn world table”,
American Economic Review, Vol. 105 No. 10, pp. 3150-3182.
F€olster, S. and Henrekson, M. (2001), “Growth effects of government expenditure and taxation in rich
countries”, European Economic Review, Vol. 45 No. 8, pp. 1501-1520.
Friedman, M. (1937), “The use of ranks to avoid assumption of normality implicit in the analysis of
variance”, Journal of the American Statistical Association, Vol. 32, pp. 675-701.
Gygli, S., Haelg, F., Potrafke, N. and Sturm, J.E. (2019), “The KOF globalisation index – revisited”,
Review of International Organizations, Vol. 14 No. 3, pp. 543-574.
Habib, M., Abbas, J. and Noman, R. (2019), “Are human capital, intellectual property rights, and
research and development expenditures really important for total factor productivity? An
empirical analysis”, International Journal of Social Economics, Vol. 46 No. 6, pp. 756-774.
Haini, H. (2020), “Examining the relationship between finance, institutions and economic growth:
evidence from the ASEAN economies”, Economic Change and Restructuring, Vol. 53 No. 4,
pp. 519-542.
Hajilee, M., Stringer, D.Y. and Metghalchi, M. (2017), “Financial market inclusion, shadow economy
and economic growth: new evidence from emerging economies”, The Quarterly Review of
Economics and Finance, Vol. 66, pp. 149-158.
Hansen, C.B. (2007), “Generalized least squares inference in panel and multilevel models with serial Financial
correlation and fixed effects”, Journal of Econometrics, Vol. 140 No. 2, pp. 670-694.
access and the
Ho, S.Y. and Iyke, B.N. (2018), “Finance-growth-poverty nexus: a re-assessment of the trickle-down
hypothesis in China”, Economic Change and Restructuring, Vol. 51 No. 3, pp. 221-247.
finance–
Ibrahim, M. (2018), “Interactive effects of human capital in finance–economic growth nexus in Sub-
growth nexus
Saharan Africa”, Journal of Economic Studies, Vol. 45 No. 6, pp. 1192-1210.
Inoue, T. and Hamori, S. (2016), “Financial access and economic growth: evidence from Sub-Saharan 707
Africa”, Emerging Markets Finance and Trade, Vol. 52 No. 3, pp. 743-753.
Johnson, S. and Arnold, S. (2012), “Inclusive financial markets: is transformation under Way in
Kenya”, Development Policy Review, Vol. 30 No. 6, pp. 719-748.
Khandker, S. (2005), “Microfinance and poverty: evidence using panel data from Bangladesh”, The
World Bank Economic Review, Vol. 19 No. 2, pp. 263-286.
Kim, D.W., Yu, J.S. and Hassan, M.K. (2018), “Financial inclusion and economic growth in OIC
countries”, Research in International Business and Finance, Vol. 43, July 2017, pp. 1-14.
King, R. and Levine, R. (1993), “Finance and growth: Schumpeter Might Be Right”, Quarterly Journal
of Economics, Vol. 108 No. 3, pp. 717-737.
La Porta, R., Lopez de Silanes, F., Shleifer, A. and Vishny, R.W. (1998), “Law and finance”, Journal of
Political Economy, Vol. 106 No. 6, pp. 1113-1155.
Law, S.H., Kutan, A.M. and Naseem, N.A.M. (2018), “The role of institutions in finance curse: evidence
from international data”, Journal of Comparative Economics, Vol. 46 No. 1, pp. 174-191.
Pesaran, M.H. (2004), “General diagnostic tests for cross section dependence in panels”, Cambridge
Working Papers in Economics No. 0435, University of Cambridge, Faculty of Economics.
Rastogi, S. and Ragabiruntha, E. (2018), “Financial inclusion and socioeconomic development: gaps
and solution”, International Journal of Social Economics, Vol. 45 No. 7, pp. 1122-1140.
Sarma, M. and Pais, J. (2011), “Financial inclusion and development”, Journal of International
Development, Vol. 23 No. 5, pp. 613-628.
Sehrawat, M. and Giri, A.K. (2016), “Financial development and poverty reduction: panel data analysis
of South Asian countries”, International Journal of Social Economics, Vol. 43 No. 4, pp. 400-416.
Sethi, D. and Sethy, S.K. (2019), “Financial inclusion matters for economic growth in India: some
evidence from cointegration analysis”, International Journal of Social Economics, Vol. 46 No. 1,
pp. 132-151.
Sirag, A., SidAhmed, S. and Ali, H.S. (2018), “Financial development, FDI and economic growth:
evidence from Sudan”, International Journal of Social Economics, Vol. 45 No. 8, pp. 1236-1249.
Svirydzenka, K. (2016), “Introducing a new broad-based index of financial development”, IMF
working papers WP/16/5, International Monetary Fund, WA, DC.
Wooldridge, J.M. (2002), Econometric Analysis of Cross Section and Panel Data, 2nd ed., MIT Press,
Cambridge.

Further reading
World Bank (Various Years a), World Development Indicators.
World Bank (Various Years b), World Governance Indicators.
IJSE Appendix
48,5
Low- and lower–middle-income countries
Low-income countries ($1,025 or less): Benin, Burkina Faso, Burundi, the Central African Republic, the
DR of the Congo, Ethiopia, Gambia, Haiti, Liberia, Madagascar, Malawi, Mali, Mozambique, Nepal, the
Niger, Rwanda, Sierra Leone, Tajikistan, Tanzania, Togo, Uganda and Yemen Rep.
708 Lower–middle-income countries ($1,026 to $3,995): Angola, Bangladesh, Bolivia, Cambodia,
Cameroon, CongoRep., Egypt, El Salvador, Ghana, Honduras, India, Indonesia, Kenya, Lao PDR,
Lesotho, Mauritania, Moldova, Morocco, Myanmar, Nicaragua, Nigeria, Pakistan, the Philippines,
Senegal, Sudan, Tunisia, Ukraine, Vietnam and Zambia.

Variables Fixed effects (b) Random effects (B) Difference (b-B) Squared S.E.

fd 0.077 0.074 0.003 0.002

fas 0.190 0.196 %0.007 0.003
(fd 3 fas) 0.055 0.055 0.000 0.000
fg 0.023 0.026 %0.004 0.001
gov %0.031 %0.031 0.000 0.001
opn 0.094 0.096 %0.002 0.002
pop 0.032 0.031 0.001 0.001
hc 0.360 0.353 0.008 0.009
Table A1. law 0.034 0.032 0.001 0.002
Hausman test Note(s): b is consistent under the null hypothesis and alternative, while B is inconsistent under the alternative
coefficients hypothesis. The null hypothesis tests that the difference between the coefficients is not systematic

Corresponding author
Hazwan Haini can be contacted at: hazwan.haini@ubd.edu.bn

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