Empirical Economics
https://doi.org/10.1007/s00181-024-02567-2

Foreign direct investment and inclusive finance: do
financial markets and quality of institutions matter?

Joshua Yindenaba Abor1 · Richard Adjei Dwumfour2 ·
Elikplimi Komla Agbloyor1,3 · Lei Pan2,4

Received: 14 March 2023 / Accepted: 10 February 2024
© The Author(s) 2024

Abstract
We examine the impact of foreign direct investment (FDI) on financial inclusion. To
identify the causal effect of FDI on financial inclusion, we use plausibly exogenous
source of variations in bilateral investment treaties (BITs) as a novel instrumental
variable (IV) for net FDI inflows. Using annual data for a broad panel of 90 countries
over the period 2004 to 2017, our results show that FDI improves financial inclusion
for both “access to finance" and “use of financial services". This impact is more
pronounced for relatively poor countries and developing countries compared to rich
and developed countries. We also find that higher financial market development and
quality institutions improve financial inclusion directly. Moreover, financial market
development and institutional quality can serve as potential channels and moderating
variables through which FDI affects financial inclusion. Our results are robust to

We are grateful to one associate editor and two anonymous referees for their valuable comments, which
helped to improve the quality and readability of the article. We also would like to thank seminar
participants at the Centre for Global Finance (CGF), University of London for helpful comments. All the
remaining errors are our own.

B Lei Pan
lei.pan@curtin.edu.au

Joshua Yindenaba Abor
joshabor@ug.edu.gh

Richard Adjei Dwumfour
r.dwumfour@postgrad.curtin.edu.au

Elikplimi Komla Agbloyor
ekagbloyor@ug.edu.gh

1 Department of Finance, University of Ghana, Ghana, Ghana

2 School of Accounting, Economics and Finance, Curtin University, Kent Street, Perth, WA 6102,
Australia

3 University of Stellenbosch Business School, Cape Town, South Africa

4 Centre for Development Economics and Sustainability (CDES), Monash University, Melbourne,
Australia

123

http://crossmark.crossref.org/dialog/?doi=10.1007/s00181-024-02567-2&domain=pdf
http://orcid.org/0000-0002-1054-981X


J. Y. Abor et al.

various estimations and sample splitting and have important implications for policy
on financial inclusion.

Keywords FDI · Financial inclusion · Financial market development · Institutional
quality · Bilateral investment treaties

JEL Classification E44 · F21 · G21

1 Introduction

Finance literature abounds on inequality and how the gap between population subsets
could be bridged. In this study,we concentrate on a type of inequality that has the poten-
tial of further worsening poverty and inhibiting national growth: financial inclusion
(exclusion). For many parts of the world, especially low-income countries, financial
inclusion remains a luxury. Even though global financial inclusion improved by about
18 per cent between 2011 and 2017 (Demirgüç-Kunt et al. 2018), more people espe-
cially the poor are still financially excluded. For instance, globally, about 31% of adult
population have no account with a financial institution or mobile money with about
200 million adults in rural China alone out of the formal financial services (Demirgüç-
Kunt et al. 2018). The literature on financial inclusion is not in dearth. Many of these
studies have, however, centred on the measurement and the impact of inclusion on
income inequality (Dabla-Norris et al. 2015), poverty reduction (Awaworyi Churchill
and Marisetty 2020; Neaime and Gaysset 2018), financial development (Anarfo et al.
2019) and economic growth (Allen et al. 2016; Dermirgüç-Kunt and Klapper 2012;
Sharma 2016).

Foreign direct investment (FDI hereafter) has, for a very long time, been a policy
goal of many economies, especially for developing countries. The argument for FDI
has, in most cases, been anchored on growth and income-related outcomes. However,
defending FDI with only growth, and income-related utility may be limiting and too
simplistic. If so, it seems to suggest that in countries where no positive relationship
exists between FDI and growth, governments should be discouraged from investing
in promoting FDI. We posit that the benefit of FDI may be indirect on growth even
if growth is argued to be the ultimate measure of welfare. FDI can drive several
economic outcomes aside economic growth (Agbloyor et al. 2014) to include wages
and employment (Zhao 1998), institutional quality (Long et al. 2015), and income
inequality (Liu et al. 2023). Chief among them yet somehow neglected by the literature
is financial inclusion.

FDI can help expand the financial sector and increase financial inclusion. Multi-
national firms are seen to bring technology and knowledge transfer that can spillover
to and improve the efficiency of domestic firms; for financial firms, this presents an
opportunity to improve their efficiency and extend their services to the financially
excluded. Therefore, FDI has the potential to increase financial inclusion. There are,
however, little empirical evidence that explore the potential impact of FDIs on financial
inclusion. Moreover, we know little about the transmission channels through which
FDI influences financial inclusion. Together, these are significant shortcomings in the

123



Foreign direct investment and inclusive finance...

existing literature on the relationship between FDI and financial inclusion when policy
makers are increasingly cared about the role of FDI.

In the present study, we investigate the role of FDI in driving financial inclusion
using data on a panel of 90 countries from 2004 to 2017.We contribute to the literature
in four ways. First, this is the first study, to the best of our knowledge, that uses cross-
country panel data to investigate the causal link between FDI and financial inclusion.

Second, we develop a composite index for financial inclusion that captures both
dimensions of access and use of financial services to examine howFDI affects financial
inclusion. This gives a broader scope of the study to capture how FDI affects both
dimensions of financial inclusion.

Our third contribution is that we use bilateral investment treaties (BITs) as a novel
instrumental variable (IV) for net FDI inflows. Studies such as Bhagwat et al. (2021)
found that BITs have a positive impact on cross-border mergers—a form of capital
flows—by protecting the property rights of foreign acquirers. The authors found a
doubling of the probability and dollar volume of mergers after the signing of a BIT.
The study found that BITs are effective in improving capital flows especially from
developed countries to developing countries. We therefore find BIT is a plausible and
novel instrument for FDIs.

Our fourth contribution is to examine the role of financial market development
and institutional quality as a potential mediator of the relationship between FDI and
financial inclusion. We argue that FDI through the financial market has the ability
to increase the supply of loanable funds within the financial market. Such increase
could lessen the stringent credit requirements in times of shortage of loanable funds;
thereby increasing the likelihood of credit being advanced to the financially excluded.
On the other hand, capital flows even to non-deposit taking institutions could increase
investment and savings, which would eventually flow to the financial sector. With
the flow of funds into the sector, financial institutions can extend their services to
areas that hitherto were not serviced. Moreover, capital is likely to flow into countries
with strong institutional framework that protects investor rights. We therefore explore
whether financial market development and the quality of institutions play a role in the
FDI-financial inclusion nexus.

Foreshadowing the main results, we find that FDI has a direct positive impact on
inclusive finance. Our results also suggest that domestic financial markets and insti-
tutional quality can serve as a channel through which FDI affects financial inclusion.
Moreover, countries with higher level of domestic financial market development and
stronger institutional quality amplify the positive effect of FDI on financial inclusion.

The rest of the paper is organized as follows. Sect. 2 provides a brief review of
related literature. Section 3 describes the dataset. Section 4 presents the empirical
methodology used in this study. Section 5 discusses empirical findings and conducts
mechanism analysis. Section 6 concludes with policy recommendations provided.

123



J. Y. Abor et al.

2 Definitions, measures of financial inclusion and review of related
studies

There is a large body of research on financial inclusion—mainly on its definition, deter-
minants and its impact on economic outcomes. In a broader sense, Naceur et al. (2015)
define financial inclusion as the share of the population who use financial services.
This definition is also adopted by theWorld Bank in the 2014 Global Financial Devel-
opment Report (World Bank 2014). Demirgüç-Kunt et al. (2017) more specifically in
the 2017 Global Financial Findex Database define financial inclusion to involve adults
having access to and use of financial services. This shows two main components of
financial inclusion—financial access and usage. These financial services like access to
credit helps people to be able to undertake critical financial transactions like financing
education, health or business which can translate into poverty reduction and growth.

The World Bank (2014) in the 2014 Global Financial Development Report show
severalmeasures of financial inclusion. These have largely been categorized into finan-
cial access and usage. On financial access, The World Bank (2014) provide various
measures which include having a bank account or a mobile money account, the num-
ber of bank branches and ATMs, and owning a debit or credit card. Financial services
usage, however, relate to the use of these financial services like sending or receiving
money using the bank account or mobile money wallet, making payment with debit
or credit card, saving money in the bank account, applying for credit as an account
holder among others. The literature has either used one or more of these measures. For
instance, Grohmann et al. (2018) used account ownership, and debit card ownership
as measures of financial access while using savings with a formal financial institution
and debt card usage as measures of financial usage. Anarfo et al. (2019), however,
criticized the frequent use of single financial inclusion variable in the literature; argu-
ing that such measurements are insufficient in capturing overall financial inclusion.
Other studies like those of Liu and Walheer (2022) and Sarma (2012) use a composite
measure of financial inclusion that captures different dimensions of financial inclu-
sion. We therefore use a composite financial inclusion measure borne out of principal
component analysis of multiple inclusion variables in our estimation.

Indeed, Demirgüç-Kunt et al. (2017) note the increased efforts by governments to
increase financial inclusion in their respective countries given the developmental role
of financial inclusion evidenced by the extant literature that find positive impact of
financial inclusion on development outcomes. However, literature on the determinants
of financial inclusion is limited.

The focus of our study is on the relationship between FDI and financial inclusion
which has received little attention in literature. To achieve this, we postulate that the
level of financial development and institutional quality in a country plays a significant
role in this relationship. Given limited studies on FDI and financial inclusion, we
review some of the existing literature on the effect of financial markets and institutions
on financial inclusions. We then discuss how our study differs from these existing
literature as we look at how these variables (financial markets and institutions) could
play a moderating in the relationship between FDI and financial inclusion.

123



Foreign direct investment and inclusive finance...

2.1 Effect of financial markets development on financial inclusion

There is a large stream of literature that has shown that well-developed financial mar-
ket has a strong positive link with easy access to finance leading to economic growth
in the long run. For instance, Beck et al. (2007) demonstrated that poor access to credit
in the financial markets can widen inequality between the developed and developing
world. This is due to the fact that access to financial tools enhances consumption,
savings and investment (Dupas and Robinson 2013). According to Beck et al. (2007),
countrieswith strong financialmarkets are able to access credit at low cost and improve
on the welfare of the individual and economy at large. Swamy (2012) also demon-
strated that robust financial market system increases access to finance by the poor
resulting in poverty reduction, inclusive growth and economic development. Anarfo
et al. (2019) recommended the pursuit of financial sector development in achieving
financial inclusion. From the literature, there seem to be a consensus that development
of the financial sector helps promote financial inclusion.

2.2 Effect of institutions on financial inclusion

According to North (1990, p3) institutions are “the rules of the game in a society or,
more formally, are the humanly devised constraints that shape human interaction."
Based on North (1990), such institutions comprise the society’s incentive structure
and in particular the political and economic institutions form the basis of economic
growth. Extensive literature has established the stylized fact that improved institu-
tional qualities lead to higher economic growth and development (see e.g. Acemoglu
et al. 2001; Hall and Jones 1999; Knack and Keefer 1995; Mauro 1995). A plethora
of studies have also established the connection between institutional factors and cap-
ital flow (see e.g. Okada 2013; Rivera-Batiz and Oliva 2002) with focus mainly on
political and legal factors. However, the institutional quality (such as rule of law and
political and socio-economic stability) in facilitating financial inclusion and creating
the enabling environment for FDI has not yet been studied systematically. Meanwhile
these institutional qualities have been proven to be critical drivers of capital inflows
especially in environments characterized by declining regulation and weak financial
integration systems (Rivera-Batiz and Oliva 2002). Nkoa and Song (2020) in a more
recent study examined the impact of institutional quality (governance indicators) on
financial inclusion in 51 African countries. Using the generalized method of moments
(GMM) approach, their results show a positive impact of quality institutions on finan-
cial inclusion. Muriu (2020) examined the role of institutional quality in promoting
financial inclusion by using panel data from 125 countries over the period 2004 to
2015, and found that the institutional context non-negligibly drives financial inclu-
sion. The study suggested the sequencing of institutional reforms in ways that could
promote financial inclusion. In this study, we introduce institutions in our empirical
model and examine their relationship with financial inclusion.

From the literature, we see that financial development and institutional quality
help increase financial inclusion. What is lacking in the literature is the relationship
between FDI and financial inclusion. As we note earlier, we posit that FDI can have a

123



J. Y. Abor et al.

direct impact on financial inclusion as well as an indirect impact through its effect on
financial development and institutional quality. On the direct impact, we conjecture
that higher FDI inflows—especially those targeted at the financial sector—can help
expand financial services to the financially excluded. Thus, FDIs which are in the form
of Fintech or investment in banking or financial sector can directly increase access and
usage of financial services provided by Fintech firms and banks. On the other hand,
FDI can have an indirect effect on financial inclusion through its effect on financial
development. More generally, higher FDI in the country would lead to higher inflows
of funds which can increase the savings in banks and hence credit supply. As these
multinational companies (MNCs) operate in the country, they will access and use
financial services. This can be in the form of savings or financial transactions. Access
to and use of financial services by these MNCs would directly increases financial
inclusion while at the same time the savings of these multinational companies would
expand the credit supply of banks. As banks receive more savings, the available credit
supply increases leading to potential supply of credit to the financially excluded.
Moreover, banks revenue are likely to increase as they provide financial services to
the MNCs. Increased revenue would help banks finance their expansion to provide
financial services to the financial excluded.

In another breadth, a well functioning institutional environment can facilitate the
impact of FDI on financial inclusion. Indeed, host countries have an incentive to
improve their institutional environment in order to accommodate more foreign firms.
In order to keep improving the institutional framework, higher FDI inflows would
encourage host countries to improve their institutional environment to meet interna-
tional standards. Indeed, as argued earlier, multinational companies typically come
from developed countries where there are quality institutions that encourage innova-
tion and business success. Host countries are more likely to review their regulatory
environment as well as have a stable political environment in order to maintain foreign
investments in the country. This may involve meeting the growing needs of MNCs by
adjusting their business regulatory environment to be more business friendly. More
specifically, this could be in the form of reducing bureaucracies in government ser-
vices; fighting corruption; having a politically stable environment; and strengthening
the legal system by increasing investment in commercial courts that handle interna-
tional transaction among others.

The presence of these MNCs can also encourage strategic partnerships between
the host and origin countries in technical areas that can help improve the regulatory
environment of the host country. For instance, the growthof theFintech sector drivenby
MNCswould encourage the host country to have amore favourable Fintech regulatory
sector that encourages innovation learning from international standards where the
MNCs originate.

The increased attention to the institutional environment and their development
there-off can have positive spillovers on financial inclusion. As the institutional envi-
ronment improves and becomes business friendly, MNCs and other local firms in the
country can operate well and be profitable. If these companies are either banks or Fin-
tech firms, with higher revenue and profit, they can expand their services to include
the financially excluded. Moreover, as other firms grow, their demand for financial
services like credit will also increase leading to higher financial inclusion. As indi-

123



Foreign direct investment and inclusive finance...

cated above, more revenues would also mean more savings leading to higher credit
supply which would lead to higher financial inclusion.

Following from the literature and the theoretical background given, the current
study seeks to examine the direct impact of FDI on financial inclusion as well as the
indirect impact through financial development and institutional quality.

3 Data and sources

3.1 Data

We use annual unbalanced data from 2004 to 2017 for a total of 90 countries.1 For the
financial inclusion data used in this study, we include four measures that have been
widely used in the literature: (i) bank accounts per 1000 adults; (ii) bank branches per
100,000 adults; (iii) ATMs per 100,000 adults; and (iv) depositors with commercial
banks per 1000 adults. The key components of financial inclusion are access and use
of financial services. Financial access includes having a bank account with a financial
institution, bank branches and number of ATMs. These have been widely used in the
literature (Brune et al. 2016;Grohmann et al. 2018;Karlan andMorduch2010).Having
access to a bank account with a financial institution, for instance, affords users the
opportunity to easily and safely use financial services. Aside the measures of access,
we also include the number of depositors with commercial banks to indicate usage.
Making deposits at a financial institution affords the opportunity for depositors to take
advantage of other financial products and services that may be available at the bank.
The number of depositors shows that people do not only have a bank account which
may be dormant but actually makes use of it for financial transactions. Our financial
inclusion index therefore captures both dimensions of access and use.

Based on these indicators, we then build a composite index of financial inclusion
using principal component analysis (PCA). The PCA helps to reduce highly correlated
variables into one or set of smaller variables that are uncorrelated so called the principal
component(s). These components represent a considerable variation in the original
dataset. In this study, we use the Kaiser (1974) and Jolliffe (2002) criterion who
indicate that only common factors with an eigenvalue greater than one should be
retained. From Table 1, we retain component one as the financial inclusion index
as is has an eigenvalue of 2.951 which explains about 74% of the total variation or
information contained in the four financial inclusion variables. We scale our financial
inclusion index to fall between 0 and 1 with higher values indicating higher financial
inclusion.

Table 2 presents the correlation matrix between the composite financial inclusion
index and the four measures of financial inclusion. The composite index shows a
high correlation between the variables from 69 to 94%, suggesting that our index
appropriately represents the four indicators of financial inclusion.

1 The list of countries used in this study is provided in Appendix. Note that the total 0f 110 countries is
used in the cross-sectional analysis provided in Appendix.

123



J. Y. Abor et al.

Table 1 Principal component analysis for financial inclusion (FINIC)

Component Eigenvalue Difference Proportion Cumulative

Comp1(FINIC) 2.951 2.288 0.738 0.738

Comp2 0.664 0.283 0.166 0.904

Comp3 0.381 0.376 0.095 0.999

Comp4 0.005 – 0.001 1.000

FINIC denotes financial inclusion index

Table 2 Correlation between
financial inclusion index and
financial inclusion indicators

Variable FINC

FINIC 1.000

Depositors 0.939

Accounts 0.939

ATM 0.840

Branches 0.694

FINIC denotes financial inclusion index. Depositors refers to deposi-
tors with commercial banks per 1000 adults. Accounts represents bank
accounts per 1000 adults, ATM stands for ATMs per 100,000 adults,
and Branches is bank branches per 100,000 adults

Table 3 Summary statistics of
financial inclusion (FINC)

Stats FINC

25th Percentile 0.066

50th Percentile 0.206

75th Percentile 0.412

Again, when we observe the financial inclusion variable, we see that the median
index is around 0.206 which is quite low indicating that most countries within the
sample fall below 0.206. However, 75th percentile of our index records a value of
0.412 still below the 0.50 mid-point value. This shows that countries still need to
improve their efforts in increasing financial inclusion in their countries.

Also, from Fig. 1, we see that generally financial inclusion has been increasing
over the decades. This shows that countries have generally made progress towards
improving inclusive finance. Meanwhile, when we look at Fig. 2, we observe that on
average, Ukraine has the most inclusive financial system with Chad being the least
financially included country. We also observe that, most of the countries with weak
financial inclusion are countries in Sub-Saharan Africa (SSA).

FDI is measured by net inflows of FDI as a percentage of gross domestic product
(GDP). For domestic financialmarket development,weuse private sector credit toGDP
ratio as a proxy. We use four institutional quality measures of Kaufmann et al. (2010).
They are the rule of law, control of corruption, government effectiveness, and regula-
tory quality. These indicators ranges from approximately−2.5 (weak) to 2.4 (strong).
Thus, higher values indicate better or stronger institutional quality. Institutional quality

123



Foreign direct investment and inclusive finance...

Fig. 1 Trend of mean financial inclusion index (FINC)

can have a significant influence on financial inclusion. For instance, regulatory quality
captures perceptions of the ability of government to formulate and implement sound
policies and regulations that permit and promote private sector development. Hence,
the quality of regulations in the country is likely to promote the activities of banks to
help extend financial access to people.

Our study also uses a number of control variables that have been identified in the
existing studies that can affect financial inclusion. First, we control for economic
development which is captured by natural log of real GDP at purchasing power parity
(PPP) per capita. This is to compare progress in economic development across different
countries and how this affects financial inclusion. We also control for adult population
which is the ratio of population of ages between 15 and 64 to total population. As
indicated by Klapper et al. (2015), there is high possibility of people of working age
to gain financial knowledge and thus be financially included. Furthermore, we also
add mobile cellular subscriptions (per 100 people) to control for information and
communication technology (ICT) penetration. The use of mobile phones and other
technology tools being championed especially by financial technology (Fintech) firms
is revolutionizing the progress in achieving financial inclusion. People can access
financial services on their phones and thus ICT would be a good complement to
improve financial inclusion.

Finally, measures of education attainment are included as control variables. More
specifically, we use gross secondary enrolment and gross tertiary enrolment ratios
as the proxy. Including both allows us to capture how various levels or education
affects financial inclusion. We expect higher levels of education at the tertiary level to
improve financial inclusion. We conjecture that students at tertiary levels are normally

123



J. Y. Abor et al.

Fig. 2 Mean financial inclusion index (FINC) by country. Note: Country labels adjusted with 2 intervals

of adult age and are also more likely to access financial services unlike students at the
secondary level who are more likely to be younger. Hence, having higher proportion
the population as students at the secondary level would mean a negative impact on
financial inclusion considering that it is the adult population that are captured in the
financial inclusion measures. Data on financial inclusion measures were taken from
the Global Financial Development Database of the World Bank. All other controls
except the institutional quality measures (taken from World Governance Indicators
of the World Bank) were taken from the World Development Indicators of the World
Bank.

123



Foreign direct investment and inclusive finance...

3.2 Descriptive statistics

Table 4 presents the summary statistics. For financial inclusion measures, bank branch
penetration is seen to have a mean of 12.376 bank branches per 100,000 adults ranging
from aminimumof 0.392 to 59.123 per 100,000 adults. ATM, on the other hand, shows
higher financial access than bank branches with a mean of 28.824 ATMs per 100,000
of adults. Deposits and accounts show the widest variations with deposits having
a minimum of 1.17 and a maximum of 3383.357 deposits per 1000 adults, while
accounts ranged from a minimum of 1.166 to a maximum of 3379.810 number of
bank accounts per 1000 adults. The mean accounts recorded is 610.041 bank accounts
per 1000 adults. This shows that the number of ATMs provides the widest access to
financial services, while accounts shows the least. Our financial inclusion variable
ranges from 0 to 1 with a mean of 0.211 generally showing low levels of financial
inclusion.

In regard to FDI, it has a mean value of 6.8% of GDP. This indicates the important
role of FDI flows to the global economy. For the domestic financial market measure,
the private sector credit to GDP ratio has the mean of 36% of GDP. The institutional
quality indicators show that globally countries on average have poor institutions with
the lowest mean value in the rule of law (Law). This suggests that much effort is
needed to make progress in countries’ institutional qualities especially those related
to the rule of law. Real GDP per capita has a mean of 8.870 with a minimum of
6.337 and a maximum of 11.861. The mean share of adult population is 62.09 with a
minimum of 47.18 and amaximum of 86.40. ICT on the other hand show an average of
approximately 82 mobile phones per 100 people. The indicators on education shows
greater average enrolment at the secondary level (74%) than at the tertiary level (32%).
This shows the needed progress that needs to be made at educating more people at the
tertiary level to realize the benefits of higher education.

4 Empirical methodology

4.1 Model specification

Our basic econometric model that relates FDI and financial inclusion can be specified
as below:

FINICi,t = α0 + α1FDIi,t + α2FMi,t + α3INSTi,t + ϕXi,t + εi,t (1)

where the subscript i = 1, 2, ..., N represents countries; t = 1, 2, ..., T denotes the
time span in years; FINICi,t refers to the financial inclusion variable and is measured
by a composite financial inclusion index obtained by applying a PCA on four finan-
cial inclusion variables namely, bank accounts per 1000 adults, bank branches per
100,000 adults, ATMs per 100,000 adults, and depositors with commercial banks per
1000 adults. Our key variable of interest is FDI (FDIi,t ) measured by net inflows of
FDI (% of GDP); FMi,t represents domestic financial market development which is
captured by domestic credit to private sector ratio (% of GDP); INSTi,t stands for

123



J. Y. Abor et al.

Table 4 Descriptive statistics

Variable Obs Mean Std. dev Min Max

Panel A: Financial inclusion

Branches 977 12.376 11.537 0.392 59.123

ATM 977 28.824 30.264 0.000 140.421

Deposits 977 619.045 556.745 1.170 3383.357

Accounts 977 610.041 557.469 1.166 3379.810

FINIC 977 0.255 0.211 0 1

Panel B: Foreign direct investment and treaties

FDI 968 4.627 6.844 −8.574 105.667

BITs 936 19.983 20.652 0 120

Panel C: Institutional quality

Law 977 −0.382 0.746 −2.032 1.825

Government 977 −0.350 0.774 −2.484 2.437

Regulatory quality 977 −0.305 0.787 −2.344 2.261

Corruption 977 −0.358 0.759 −1.773 2.326

Panel D: Macroeconomic and infrastructure

Natural log of real GDP per capita 958 8.870 1.171 6.337 11.861

Adult pop 977 62.093 7.649 47.183 86.398

ICT 973 81.730 44.903 0.190 212.639

Private credit 874 36.426 29.511 0.513 156.230

Secondary education 827 73.708 27.425 9.689 128.930

Tertiary education 859 31.591 20.643 0.502 89.959

Braches refers to the number of bank branches per 100,000 adults; ATM represents the number of ATMs
per 100,000 adults; deposits stands for the number of depositors with commercial banks per 1000 adults;
accounts denotes the number of bank accounts per 1000 adults; FINIC is the financial inclusion variable
derived from the principal component analysis; FDI is the net inflows of foreign direct investment as a
percentage of GDP; BITs is the number of Bilateral Investment Treaties-note that we use the BITs per
100,000 adult population in the analysis; private credit represents private sector credit to GDP ratio (% of
GDP); law is the rule of law; government refers to government effectiveness; regulatory denotes themeasure
of regulatory quality; corruption is the control of corruption; adult pop stands for the share of population
ages 15 to 64 of total populations; ICT denotes mobile cellular subscriptions per 100 people; secondary
education is the secondary school enrolment (% gross); tertiary education refers to tertiary school enrolment
(% gross)

the institutional quality variables that contain rule of law, control of corruption, gov-
ernment effectiveness, and regulatory quality. These are included separately in our
estimations. Xi,t denotes a vector of control variables which includes the natural log
of real GDP per capita, population ages from 15 to 64 (% of total population), ICT
penetration measured by mobile cellular subscriptions (per 100 people), secondary
school enrolment and tertiary school enrolment.

As a benchmark exercise, we first use ordinary least squares (OLS) method to
estimate Eq. (1). However, net inflows of FDI are likely to be endogenous when
estimating the relationship between FDI and financial inclusion. Possible reasons
include reverse causality running from financial inclusion to FDI (Ćihák et al. 2016).

123



Foreign direct investment and inclusive finance...

Indeed, FDIs are likely to flow into countries where a larger share of the population is
financially included. Meanwhile, it is possible that the inflow of FDIs may particularly
be the reason to have the needed investment in the financial sector that promotes
financial inclusion. there is also possibility of omitted variables bias given that it is
unlikely to control for all variables that can affect financial inclusion. To overcome
the endogeneity issue, we rely on the two-stage-least squares (2SLS) instrumental
variables approach to identify any causal effect between FDI and financial inclusion.

4.2 Identification strategy

We instrument FDI using the total number of bilateral investment treaties (BITs)
ratified and come into force by a country with other countries.

The corresponding first-stage IV estimation regression is as below:

FDIi,t = δ0 + δ1BITsi,t + δ2FMi,t + δ3INSTi,t + ϕXi,t + vi,t (2)

where BITsi,t refers to the number of bilateral investment treaties signed and actually
come into force by a country at time t with other countries. Given that our financial
inclusion measures capture how the adult population is financially included, we scale
the number of treaties per 100,000 of adult population. This allows us to capture the
number of treaties a country signed to allow for FDI inflows in the country relative to
its adult population. BITs are treaties that are voluntary in nature signed between two
countries to encourage foreign investment inflows and to protect those investments.
They typically operate by having clauses that give protection to foreign investors
against political risk. This suggests that BITs is a plausible instrument for FDI given
that these treaties will encourage the inflows of FDI and hence their impact on financial
inclusion can only be through encouragingmore FDIs. Therefore, BITs is uncorrelated
with vi,t satisfies the exclusion restriction. Xi,t is a vector of covariates in the structural
regression and vi,t is the stochastic error term. Having the predicted values of ˆFDIi,t ,
we estimate second-stage regression that has the same form as Eq. (1).

5 Empirical results and discussions

Wefirst show some sample results to better illustrate the data. In Fig. 3, we show scatter
plots of financial inclusion composite index versus FDI. Clearly, we can see that net
inflows of FDI is positively correlated with financial inclusion. In Sects. 5.1, 5.2 and
5.3, we conduct more rigorous analysis using a variety of estimation techniques.

5.1 Baseline results

The baseline OLS results are presented in Table 5. From the table, most of the columns
show that FDIhas a positive impact onfinancial inclusionwith at least 10%significance
level. However, as we discussed earlier, FDI may be endogenous hence the need to

123



J. Y. Abor et al.

Fig. 3 Association between financial inclusion and FDI

control for this in an instrumental variables approach. The next section shows the
results of our IV estimations.

5.2 Main IV results

The results of our main IV estimations are presented in Table 6. From the table, we
see that FDI has a significant positive impact on financial inclusion in all estimations
at 5% level. In particular, a standard deviation increase in FDI leads to an increase
in the financial inclusion index ranging from 0.06 to 0.07. This shows the important
role of FDI in promoting inclusive finance. If efforts are made at driving FDI inflows
into the country, the financial sector seeks to gain as more capital flows into the
financial ecosystem. This can happen as more FDI inflows would mean more capital
investment into the economy where these multinational corporations (MNCs) can
demand financial services as well as employ more people who would seek financial
services. As the financial sector manages the transactions of MNCs, they can have
more available funds to expand credit to the otherwise financially excluded.Moreover,
FDI can have a rippling effect in the economy through higher direct and indirect
employment as other firms benefit in the value chain; these individuals and firms are
more likely to seek financial services as they benefit from these investments.

We also see that all the institutional quality variables have the expected sign at
1% significance level. We institutional quality has a positive correlation with financial
inclusion. This shows that quality institutional framework could lead to higher financial
inclusion. As we noted earlier, the existence of better institutional framework where
there is rule of law that protects property rights, less corruption, quality regulatory

123



Foreign direct investment and inclusive finance...

Table 5 Financial inclusion and FDI, OLS estimation

Variable FINIC
(1) (2) (3) (4)

FDI 0.001 0.001** 0.001** 0.001*

(0.001) (0.001) (0.001) (0.001)

Law 0.063***

(0.009)

Government 0.058***

(0.010)

Regulatory quality 0.040***

(0.009)

Corruption 0.054***

(0.008)

Private credit 0.002*** 0.002*** 0.002*** 0.002***

(0.0002) (0.0002) (0.00002) (0.00002)

Natural log real GDP per capita 0.039*** 0.043*** 0.050*** 0.043***

(0.008) (0.008) (0.008) (0.008)

Adult population 0.003*** 0.002** 0.002** 0.002**

(0.001) (0.001) (0.001) (0.001)

ICT 0.001*** 0.001*** 0.001*** 0.001***

(0.0001) (0.0001) (0.000) (0.0001)

Secondary education −0.0001*** −0.000*** −0.000*** −0.000***

(0.00002) (0.00002) (0.00002) (0.00002)

Tertiary education 0.002*** 0.002*** 0.002*** 0.002***

(0.0003) (0.0003) (0.000) (0.0003)

R2 0.72 0.72 0.71 0.72

No. of countries 91 91 91 91

Obs 723 723 723 723

*, ** and *** denote statistical significance at the 10, 5 and 1% levels, respectively

framework including those related to financial institutions, and an effective governance
framework can ensure a favourable financial sector environment to encourage inclusive
finance.

The results also show that private sector credit toGDP ratio has a significant positive
correlation with financial inclusion in all the estimations at 1% significance level.
Specifically, when there is strong institutions in relation to the rule of law that protects
property rights, we see that a standard deviation increase in private sector credit toGDP
ratio raise financial inclusion by 0.03. This shows the important role of financialmarket
development on financial inclusion especially in countries where there is strong rule
of law that ensures the protection of private property. As the financial sector develops,
access is likely to be expanded to the financially excluded for them to benefit from the
financial services. Importantly, in these countries where there are strong institutions,
especially the rule of law, people are likely to be confident in the legal system and

123



J. Y. Abor et al.

hence would feel safe to access and use financial services with the confidence that
their funds are protected.

We also find that higher economic growth and adult population have positive cor-
relation with financial inclusion. ICT infrastructure also has a significant positive
correlation with financial inclusion at 1% level. Concerning education, we find that
while secondary education does not improve financial inclusion, higher education at
the tertiary level tend to have a significant correlation with financial inclusion at 1%
level. We also test whether we could treat FDI as exogenous. Both the Durbin and
Wu–Hausman test statistic are highly significant rejecting the null hypothesis that FDI
is exogenous justifying our use of IV approach.

5.3 Robustness checks

In this section, we conduct five types of robustness tests. We first re-estimate the
results using the Lewbel (2012) heteroscedasticity-based identification approach. We
then estimate the model using the two-step system generalized method of moments
approach.We thenproceed to examinewhether the impact of FDI onfinancial inclusion
differs from one regional bloc to the other and whether it differs based on the income
levels of the countries and the development status of the countries. Finally, we conduct
further sensitivity test using alternative measure of financial inclusion. These are done
to ensure that our results are robust to difference estimation techniques and sub-
samples.

5.3.1 Lewbel (2012) heteroscedasticity-based identification

We perform sensitivity analysis with alternative IV estimation by augmenting our
external instruments with heteroskedasticity-based instruments developed by Lewbel
(2012). The Lewbel (2012)’s method is appealing for the following two reasons: (i) It
provides IV estimates when there are lack of external IVs; (ii) In instances in which
it is doubtful whether the available IVs satisfy the exclusion restrictions, the Lewbel
(2012) method can be combined with potentially weak IVs to provide more robust
estimates. We construct instruments based on heteroskedasticity in error terms by
following Lewbel (2012). According to Lewbel (2012), the constructed instruments
can be used as a valid IV when there are lack of external IVs or can be used to augment
external IVs to test over identifying restrictions and improve efficiency.

The results based on Lewbel (2012)’s IV approach are presented in Table 7. Even
though the Hansen p-value is significant, We find that FDI still has a significant
positive impact on financial inclusion in all estimations except when rule of law is
used as institutional quality variable. These results are significant with at least 10%.

5.3.2 Using alternative approach: GMM estimate

Given that our data exhibit relatively large cross-sectional components compared to
time-series components, we use system generalized method of moments (GMM) esti-
mator specifically developed for this type of dataset. System GMM approach also

123



Foreign direct investment and inclusive finance...

Ta
bl
e
6

Fi
na
nc
ia
li
nc
lu
si
on

an
d
FD

I,
IV

es
tim

at
io
n

V
ar
ia
bl
e

FI
N
IC

(1
)

(2
)

(3
)

(4
)

FD
I

0.
00

9*
*

0.
01

0*
*

0.
01

0*
*

0.
00

9*
*

(0
.0
04

)
(0
.0
04

)
(0
.0
04

)
(0
.0
04

)

L
aw

0.
04

0*
**

(0
.0
11

)

G
ov
er
nm

en
t

0.
03

8*
**

(0
.0
12

)

R
eg
ul
at
or
y
qu

al
ity

0.
02

3*
*

(0
.0
10

)

C
or
ru
pt
io
n

0.
03

5*
**

(0
.0
12

)

Pr
iv
at
e
cr
ed
it

0.
00

1*
**

0.
00

1*
**

0.
00

1*
**

0.
00

2*
**

(0
.0
00

3)
(0
.0
00

3)
(0
.0
00

3)
(0
.0
00

3)

N
at
ur
al
lo
g
re
al
G
D
P
pe
r
ca
pi
ta

0.
04

4*
**

0.
04

6*
**

0.
05

1*
**

0.
04

6*
**

(0
.0
10

)
(0
.0
09

)
(0
.0
09

)
(0
.0
10

)

A
du

lt
po

pu
la
tio

n
0.
00

3*
*

0.
00

3*
*

0.
00

3*
*

0.
00

3*
*

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

IC
T

0.
00

1*
**

0.
00

1*
**

0.
00

1*
**

0.
00

1*
**

(0
.0
00

2)
(0
.0
00

2)
(0
.0
00

2)
(0
.0
00

2)

Se
co
nd

ar
y
ed
uc
at
io
n

−0
.0
00

1*
**

−0
.0
00

1*
**

−0
.0
00

1*
**

−0
.0
00

1*
**

(0
.0
00

02
)

(0
.0
00

02
)

(0
.0
00

02
)

(0
.0
00

02
)

Te
rt
ia
ry

ed
uc
at
io
n

0.
00

2*
**

0.
00

2*
**

0.
00

2*
**

0.
00

2*
**

123



J. Y. Abor et al.

Ta
bl
e
6

co
nt
in
ue
d

V
ar
ia
bl
e

FI
N
IC

(1
)

(2
)

(3
)

(4
)

(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

4)

F
ir
st
-s
ta
ge

re
gr
es
si
on

B
IT
s

3.
73

2*
**

3.
74

8*
**

3.
75

0*
**

3.
70

3*
**

(1
.0
38

)
(1
.0
01

)
(1
.0
02

)
(1
.0
35

)

A
dj
us
te
d
R
2

0.
28

0.
28

0.
28

0.
28

D
ur
bi
n

χ
2
-t
es
t

48
.0
7*

**
58

.6
8*

**
59

.9
6*

**
55

.1
9*

**

W
u–
H
au
sm

an
F
-t
es
t

50
.8
7*

**
63

.1
3*

**
64

.6
4*

**
59

.0
6*

**

1s
t-
st
ag
e
F

12
.9
3*

**
14

.0
1*

**
14

.0
1*

**
12

.7
9*

**

R
2

0.
66

0.
64

0.
64

0.
65

N
o.

of
co
un

tr
ie
s

88
88

88
88

O
bs

70
1

70
1

70
1

70
1

R
ob
us
ts
ta
nd
ar
d
er
ro
rs
ar
e
in

pa
re
nt
he
se
s,
**

an
d
**
*
de
no
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
5
an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



Foreign direct investment and inclusive finance...

Table 7 Financial inclusion and FDI, Lewbel (2012) IV estimates

Variable FINIC
(1) (2) (3) (4)

FDI 0.002 0.004*** 0.002* 0.004***

(0.001) (0.001) (0.001) (0.001)

Law 0.049***

(0.008)

Government 0.055***

(0.008)

Regulatory quality 0.043***

(0.008)

Corruption 0.038***

(0.008)

Private credit 0.002*** 0.002*** 0.002*** 0.002***

(0.0003) (0.0003) (0.0003) (0.0003)

Natural log real GDP per capita 0.057*** 0.060*** 0.061*** 0.059***

(0.009) (0.008) (0.008) (0.009)

Adult population 0.0007 0.0002 0.001 0.0004

(0.001) (0.001) (0.001) (0.001)

ICT 0.0002 0.0002 0.0001 0.0003**

(0.0002) (0.0001) (0.000) (0.0001)

Secondary education −0.0001*** −0.0001*** −0.0001*** −0.0001***

(0.00002) (0.00001) (0.00001) (0.00001)

Tertiary education 0.002*** 0.002*** 0.002*** 0.002***

(0.0003) (0.0003) (0.0003) (0.0003)

R2 0.71 0.70 0.71 0.70

No. of countries 88 88 88 88

Obs 701 701 701 701

Hansen J stat 71.60 65 87.2 62.90

Hansen J p-value 0.00 0.00 0.00 0.00

Robust standard errors in the parenthesis. *, ** and *** denote statistical significance at the 10, 5 and 1%
levels, respectively

combines in a system the regression in differences with the regression in levels and
takes into account the endogeneity issue. Moreover, the introduction of lagged finan-
cial inclusion, for instance, is necessary because previous year’s progress in financial
inclusion is likely to influence the following period’s financial inclusion levels. Fol-
lowing Roodman (2009), the lags of the independent variables are used as instruments.
Since this reduces the number of observations and power of regressions, we employ
the collapsing method of Holtz-Eakin et al. (1988) and Arellano and Bover (1995)
forward orthogonalization procedure to limit the number of instruments.

Following a common diagnostic test procedure in the literature, we report the
Hansen test of overidentifying restrictions and the second-order autoregressive, AR(2)

123



J. Y. Abor et al.

Table 8 Financial inclusion and FDI, system GMM estimates

Variable FINIC
(1) (2) (3) (4)

L.FINIC 0.964*** 0.960*** 0.964*** 0.961***

(0.007) (0.007) (0.007) (0.007)

FDI 0.0003*** 0.0003*** 0.0003*** 0.0003***

(0.0001) (0.0003) (0.0001) (0.0001)

Law 0.002**

(0.001)

Government 0.004***

(0.001)

Regulatory quality 0.001

(0.001)

Corruption 0.004***

(0.001)

Private credit 0.00003 0.00004 0.00005 0.00005

(0.00004) (0.00004) (0.00004) (0.00004)

Log GDP p.c. (PPP) 0.003*** 0.003*** 0.004*** 0.004***

(0.001) (0.001) (0.001) (0.001)

Adult pop. (15 to 64 years) 0.0003* 0.0003* 0.0003* 0.0003*

(0.0002) (0.0002) (0.0002) (0.0002)

ICT 0.00004** 0.00002 0.00003* 0.00003

(0.0002) (0.00002) (0.00002) (0.00002)

Secondary education −0.000002 −0.000002 −0.000001 −0.000001

(0.000002) (0.000002) (0.000002) (0.000002)

Tertiary education −0.00001 −0.00002 −0.00002 −0.00001

(0.0001) (0.0001) (0.0001) (0.0001)

AR (2) 0.405 0.405 0.407 0.411

Hansen test (overid) 0.145 0.191 0.209 0.233

No. of countries 90 90 90 90

No. of instruments 40 40 40 40

Obs 648 648 648 648

Robust standard errors in the parenthesis. *, ** and *** denote statistical significance at the 10, 5 and 1%
levels, respectively

tests. TheHansen test provides evidence of the validity of the instruments by evaluating
the entire set of moment conditions in satisfying the exclusion restriction.

Table 8 reports the systemGMMresults. In all the estimations, theAR(2) test and the
Hansen test statistic cannot reject the null hypothesis of no second-order serial correla-
tion and the validity of the overidentifying restrictions, respectively. More specifically,
the p-value for the second-order serial correlation in the system GMM estimation is
greater than 10% in all specifications, hence the null hypothesis of no second-order
serial correlation at the conventional significance levels (1, 5 and 10%) cannot be

123



Foreign direct investment and inclusive finance...

rejected. Furthermore, from the Hansen test, the null hypothesis of the validity of the
overidentifying restrictions at the conventional significance levels cannot be rejected.
Overall, the AR(2) tests for second-order serial correlation and the Hansen test of
overidentifying restrictions are both satisfied, indicating that our GMM estimates are
consistent and efficient. The estimates also indicate that the lag of the dependent vari-
able is significant in all the regressions. This suggests that financial inclusion persists.
The system GMM results are consistent with our earlier estimations with FDI having
a significant positive impact on financial inclusion at 1% level. Three out of four insti-
tutional quality variables also show a significant positive impact on financial inclusion
with at least 5% significance level.

5.3.3 Examining regional differences

Here, we examine the regional differences in the impact of FDI on financial inclusion.
We classify the countries based on the World Banks regional grouping, namely South
Asia (SA), Europe &Central Asia (ECA), Middle East &North Africa (MENA), Sub-
Saharan Africa (SSA), East Asia & Pacific (EAP), Latin America &Caribbean (LAC),
and North America (NA). We create dummy variable for the regions and interact with
FDI.

The results are presented inTable 9. From the table,wefind that FDI has a significant
positive impact on financial inclusion in the ECA,MENA, SSA, EAP and LAC regions
in all the regressions alternating the institutional quality variables. We observe that
the LAC region has the largest impact of FDI on financial inclusion followed by the
MENA, EAP, SSA and ECA regions, respectively. FDI has no significant impact on
financial inclusion in SA and NA regions. These results generally confirm our earlier
findings that FDI has a direct positive impact on financial inclusion.

5.3.4 Examining differences in income and development status of countries

Here, we examine the impact of FDI on financial inclusion by looking at the develop-
ment status and income level of the countries. To do this, we include dummies based
on the income level of the country according to the World Bank classification per
year. These are: (i) lower income countries (LIC), (ii) lower middle income countries
(LMIC), (iii) upper middle income countries (UMIC), and (iv) high income countries
(HIC).We also group the countries whether they are developed or developing based on
the United Nations (UN) Classification. The results are presented in Tables 10 and 11.
From Table 10, the results show that FDI has a significant positive impact on financial
inclusion across all income levels. We, however, observe that the impact is more pro-
nounced in LMIC followed by LIC, UMIC and HIC, respectively. It seems to suggest
that countries at the bottom income distribution benefit more from FDIs than those at
the top. This affirms the important development role of FDIs for relatively poor and
developing countries (see e.g. Reiter and Steensma 2010).We find similar results when
we group the countries into developing or developed countries. From Table 11, we can
see that developing countries seem to benefit more from net FDI inflows for financial
inclusion than their developed counterparts. This further confirms the important role
of FDI for financial inclusion in developing countries.

123



J. Y. Abor et al.

Table 9 Financial inclusion and FDI (regional dummies), IV estimation

Variable Law Government Regulation quality Corruption
(1) (2) (3) (4)

FDI × SA 0.171 0.052 0.072 0.116

(1.142) (0.039) (0.053) (0.096)

FDI × ECA 0.014* 0.009** 0.074** 0.011**

(0.008) (0.003) (0.004) (0.006)

FDI × MENA 0.095** 0.056*** 0.060*** 0.078***

(0.046) (0.016) (0.019) (0.030)

FDI × SSA 0.055* 0.033*** 0.034*** 0.046**

(0.028) (0.011) (0.011) (0.019)

FDI × EAP 0.068*** 0.045*** 0.049*** 0.059***

(0.026) (0.009) (0.011) (0.019)

FDI × LAC 0.132** 0.092*** 0.098*** 0.115**

(0.061) (0.028) (0.032) (0.047)

FDI × NA 0.769 −0.070*** −0.597 0.326

(0.765) (0.017) (0.779) (0.593)

Controls Yes Yes Yes Yes

Obs 701 701 701 701

No. of countries 88 88 88 88

First-stage regression

Durbin χ2-test 106.69*** 107.91*** 119.91*** 106.24***

Wu–Hausman F-test 17.26*** 17.49*** 19.84*** 17.17***

1st-stage F

FDI × SA 4.33*** 4.64*** 4.20*** 4.36***

FDI × ECA 17.57*** 24.21*** 28.77*** 21.17***

FDI × MENA 8.58*** 10.21*** 7.93*** 7.27***

FDI × SSA 13.79*** 14.26*** 16.41*** 14.21***

FDI × EAP 12.79*** 13.03*** 15.24*** 12.26***

FDI × LAC 8.79*** 8.90*** 9.90*** 8.79***

FDI × NA 0.94 234.68*** 0.47 0.55

Robust standard errors in the parenthesis. *, ** and *** denote statistical significance at the 10, 5 and 1%
levels, respectively. SA is South Asia, ECA is Europe & Central Asia, MENA is Middle East & North
Africa, SSA is Sub-Saharan Africa, EAP is East Asia & Pacific, LAC is Latin America & Caribbean, and
NA is North America

5.4 Mechanisms

In this section, we examine whether development of domestic financial market and
institutional quality can serve as potential channels throughwhich FDI affects financial
inclusion. Aswe argued earlier, FDI has the potential to increase the supply of loanable
funds as financial markets receive funds in the form of investment or savings through
FDIs. Financial institutions can therefore extend their reach to the financially excluded.

123



Foreign direct investment and inclusive finance...

Ta
bl
e
10

Fi
na
nc
ia
li
nc
lu
si
on

an
d
FD

I
(i
nc
om

e
le
ve
l)
,I
V
es
tim

at
io
n

V
ar
ia
bl
e

L
aw

G
ov
er
nm

en
t

R
eg
ul
at
io
n
qu

al
ity

C
or
ru
pt
io
n

(1
)

(2
)

(3
)

(4
)

FD
I
×

L
ow

er
in
co
m
e

0.
02

7*
0.
02

8*
0.
02

6
0.
02

7*

(0
.0
15

)
(0
.0
15

)
(0
.0
16

)
(0
.0
15

)

FD
I
×

L
ow

er
m
id
dl
e
in
co
m
e

0.
03

6*
**

0.
03

8*
**

0.
03

7*
**

0.
03

7*
**

(0
.0
06

)
(0
.0
06

)
(0
.0
06

)
(0
.0
06

)

FD
I
×

U
pp

er
m
id
dl
e
in
co
m
e

0.
01

9*
**

0.
01

9*
**

0.
01

9*
**

0.
01

9*
**

(0
.0
04

)
(0
.0
04

)
(0
.0
04

)
(0
.0
05

)

FD
I
×

H
ig
h
in
co
m
e

0.
00

7*
*

0.
00

7*
*

0.
00

8*
*

0.
00

8*
*

(0
.0
03

)
(0
.0
03

)
(0
.0
03

)
(0
.0
03

)

C
on

tr
ol
s

Y
es

Y
es

Y
es

Y
es

O
bs

70
1

70
1

70
1

70
1

N
o.

of
co
un

tr
ie
s

88
88

88
88

F
ir
st
-s
ta
ge

re
gr
es
si
on

D
ur
bi
n

χ
2
-t
es
t

97
.1
4*

**
11

1.
44

**
*

11
2.
77

**
*

10
5.
16

**
*

W
u–
H
au
sm

an
F
-t
es
t

27
.5
5*

**
32

.3
7*

**
32

.8
3*

**
30

.2
2*

**

1s
t-
st
ag
e
F

FD
I
×

L
ow

er
in
co
m
e

24
.0
1*

**
24

.7
5*

**
24

.0
0*

**
22

.8
8*

**

FD
I
×

L
ow

er
m
id
dl
e
in
co
m
e

55
.7
2*

**
54

.1
7*

**
52

.7
7*

**
58

.9
5*

**

FD
I
×

U
pp

er
m
id
dl
e
in
co
m
e

30
.5
2*

**
27

.5
3*

**
29

.0
9*

**
29

.1
7*

**
*

FD
I
×

H
ig
h
in
co
m
e

9.
10

**
*

9.
00

**
*

9.
20

**
*

9.
12

**
*

R
ob
us
ts
ta
nd
ar
d
er
ro
rs
in

th
e
pa
re
nt
he
si
s.
*,

**
an
d
**
*
de
no
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
10
,5

an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



J. Y. Abor et al.

Table 11 Financial inclusion and FDI (development status), IV estimation

Variable Law Government Regulation quality Corruption
(1) (2) (3) (4)

FDI × Developing 0.022*** 0.024*** 0.025*** 0.021***

(0.005) (0.005) (0.005) (0.005)

FDI × Developed 0.007** 0.008** 0.008** 0.008**

(0.003) (0.003) (0.003) (0.003)

Controls Yes Yes Yes Yes

Obs 645 645 645 645

No. of countries 78 78 78 78

First-stage regression

Durbin χ2-test 74.74*** 97.71*** 106.77*** 70.10***

Wu–Hausman F-test 41.48*** 56.51*** 62.78*** 38.59***

1st-stage F

FDI × Developing 38.95*** 36.38*** 38.40*** 38.57***

FDI × Developed 8.75*** 8.36*** 7.94*** 9.65***

Robust standard errors in the parenthesis. ** and *** denote statistical significance at the 5 and 1% levels,
respectively

Also, as FDIs flow to countries, the domestic authorities are inclined to improve the
quality of their institutions as a way to give more confidence to foreign firms. This
can also send a signal to other MNCs to follow the herd as they see their counterparts
operate well in these countries.

To examine whether domestic financial market and quality of institutions qualify
as potential channels through which FDI impacts financial inclusion, we follow the
approach in the literature such as Alesina and Zhuravskaya (2011) and Ackermann
et al. (2021). There are two conditions that need to be satisfied for domestic financial
market and institutional quality can serve as potential channels. First, domestic finan-
cial market development and institutional quality need to be correlatedwith FDI. Table
12 presents results for the influence of FDI on the two potential channels. The results
indicate that FDI is positively associated with domestic financial market development
and institutional quality. Specifically, 1 percentage point increase in net FDI inflows
leads to 0.40 unit rise in private sector credit to GDP ratio and 1 unit increase in net FDI
inflows leads to an increase in institutional quality ranging from 0.011 to 0.016 unit.
This shows that FDI has a significant impact on both financial market development
and institutional quality suggesting that these variables are channels through which
FDI affects financial inclusion.

The second condition is incorporating private sector credit toGDP ratio or quality of
institution as an additional control variable in the regression relates FDI and financial
inclusion should reduce the scale of the coefficient on FDI or render it insignificant.
Tables 13 and 14 report the results. Column (2) in Table 13, and columns (2), (4), (6)
and (8) in Table 14 show that when private sector credit to GDP ratio or institutional
quality indicator is added as an additional control, the coefficient on FDI reduces
in magnitude or becomes insignificant. Our findings suggest that domestic financial

123



Foreign direct investment and inclusive finance...

Table 12 Effect of FDI on the potential channel

Variable (1) (2) (3) (4) (5)
Private credit Law Government Regulatory quality Corruption

FDI 0.402*** 0.016*** 0.011*** 0.012*** 0.014***

(0.118) (0.003) (0.002) (0.003) (0.003)

Controls Yes Yes Yes Yes Yes

R2 0.41 0.44 0.55 0.46 0.40

Obs 723 809 809 809 809

*** indicates statistical significance at the 1% level

Table 13 Effect of FDI and domestic financial market as a channel on financial inclusion

Variable (1) (2) (3)

FDI 0.0025*** 0.0018*** 0.0044***

(0.0006) (0.0006) (0.0012)

Private credit 0.0021*** 0.0023***

(0.0002) (0.0002)

FDI × Private credit −0.00003**

(0.00001)

Controls Yes Yes Yes

R2 0.66 0.70 0.71

Obs 809 723 723

** and *** indicate statistical significance at the 5 and 1% levels, respectively

market development and quality of institution can serve as potential channels through
which FDI affects financial inclusion.

As a further robustness check on our mediation (potential channel) results, we
also examine whether domestic financial market development and institutional quality
moderate the relationship between FDI and financial inclusion. To do so, we augment
Eq. (1) to include the interaction term between FDI and private sector credit to GDP
ratio (or institutional quality indicators). Column (3) in Table 13 shows that the inter-
action term between FDI and domestic financial market is negative and significant at
5% level. This implies that FDI and domestic financial market may act as substitutes.
Columns (3), (5), (7) and (9) in Table 14 show that the interaction terms between
FDI and institutional quality are also mostly negative and significant at 10% level or
higher. The marginal effect plots in Fig. 4 confirm the substitution effect between FDI
and financial markets, and FDI and our institutional quality variables. In summary,
our results suggest that domestic financial market and quality of institution also mod-
erate the relationship between FDI and financial inclusion. Specially, the relationship
between FDI and financial inclusion is lower in countries with higher level of financial
market development or stronger institutional quality.

123



J. Y. Abor et al.

Ta
bl
e
14

E
ff
ec
to

f
FD

I
an
d
in
st
itu

tio
n
qu
al
ity

as
a
ch
an
ne
lo

n
fin

an
ci
al
in
cl
us
io
n

V
ar
ia
bl
e

(1
)

(2
)

(3
)

(4
)

(5
)

(6
)

(7
)

(8
)

(9
)

FD
I

0.
00

18
**

*
0.
00

10
0.
00

18
**

0.
00

13
**

0.
00

22
**

*
0.
00

14
**

0.
00

28
**

*
0.
00

11
*

0.
00

12
*

(0
.0
00

6)
(0
.0
00

6)
(0
.0
00

8)
(0
.0
00

6)
(0
.0
00

8)
(0
.0
00

6)
(0
.0
00

8)
(0
.0
00

6)
(0
.0
00

7)

L
aw

0.
06

32
**

*
0.
06

70
**

*

(0
.0
08

7)
(0
.0
08

9)

FD
I
×

L
aw

−0
.0
01

3*

(0
.0
00

7)

G
ov
er
nm

en
t

0.
05

83
**

*
0.
06

36
**

*

(0
.0
10

0)
(0
.0
10

3)

FD
I
×

G
ov
er
nm

en
t

−0
.0
01

6*
*

(0
.0
00

7)

R
eg
ul
at
or
y
qu

al
ity

0.
03

97
**

*
0.
04

66
**

*

(0
.0
08

5)
(0
.0
08

7)

FD
I
×

R
eg
ul
at
or
y
qu

al
ity

−0
.0
02

1*
**

(0
.0
00

7)

C
or
ru
pt
io
n

0.
05

44
**

*
0.
45

2*
**

(0
.0
07

9)
(0
.0
70

)

FD
I
×

C
or
ru
pt
io
n

−0
.0
00

3

(0
.0
00

9)

C
on
tr
ol
s

Y
es

Y
es

Y
es

Y
es

Y
es

Y
es

Y
es

Y
es

Y
es

R
2

0.
70

0.
72

0.
72

0.
72

0.
72

0.
71

0.
72

0.
72

0.
72

O
bs

72
3

72
3

72
3

72
3

72
3

72
3

72
3

72
3

72
3

*,
**

an
d
**
*
in
di
ca
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
10
,5

an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



Foreign direct investment and inclusive finance...

Fig. 4 Marginal effects of FDI on financial inclusion (with 95% confidence interval)

6 Conclusion and policy recommendations

We examine the impact of FDI on financial inclusion and show the role of domes-
tic financial markets and institutions. The results show that higher net FDI inflows
improve financial inclusion. Higher level of financialmarket development and stronger
institutional quality also improve financial inclusion directly. Moreover, we find that
domestic financial market and quality of institution can be potential channels through
which FDI influences financial inclusion. We also find that financial market develop-
ment and institutional quality play moderating role in the relationship between FDI
and financial inclusion. In particular, financial market development and institutional
quality can be substitute to net FDI inflows in countries with more developed capital
markets and stronger quality of institutions.

The policy implication is that countries should put in appropriatemeasures to attract
more FDI while being mindful of their country specific characteristics. For instance,
as the results show a possible substitution between net FDI inflows and financial
markets (or institutions), countries with low development of financial markets need
more FDI flows than those with advanced financial markets if they want to get more
of their citizens in the financial ecosystem. Similarly, countries with poor institutional
quality need FDI to make progress in financial inclusion. As these countries attract

123



J. Y. Abor et al.

more FDI and as they develop their financial markets and build stronger institutions,
their dependence on FDI flows to improve financial inclusion would dwindle. It is
not surprising that developing countries in particularly are the most countries that are
financially excluded and needing the most FDI flows owing to the underdeveloped
financial markets and poor institutions. More effort is needed to improve financial
inclusion and FDI flows can be a key channel to achieve this.

Our results also show per capita income improves financial inclusion. As countries
improve in their income levels, more investment can be made in their financial system
to broaden the net to get more people included. As a policy direction, countries should
pursue the attainment of higher income levels as this can give them room to invest in
the financial system to improve financial inclusion.

Our results also reveal that higher share of adult population improves financial
inclusion. This suggests that as more people grow and gets to the working age, they
are more likely to appreciate and engage in financial transactions through the formal
sector. This could be a good policy area where the government invests more in the
financial sector by providing the necessary framework of favourable regulations for
banks and financial technology (Fintech) companies to bring out innovate financial
products and services targeted at the adults.When these adults get financially included,
they are more likely to encourage the younger ones and even their children in the
financial services. They can take up insurance for their children as well as invest in
financial products for their kids; for instance, investing in financial products that would
cater for the tertiary education of their kids. The adult population therefore presents a
good opportunity for the government, banks and Fintech companies to get the younger
ones included.

Furthermore, the results show that higher ICT penetration improves financial inclu-
sion. There is no doubt the importance of ICT to modern financial services and
products. The traditional means of reaching the unbanked through bank branches
and ATMs may not be enough to reach the financially excluded especially in deprived
communitieswhere it is unprofitable to set up a branch or to put anATMbecause of low
security. Mobile and internet banking is providing banking services on the go. People
can now perform their banking transactions on their phones anywhere in the world
either through the internet or without internet through mobile money or unstructured
supplementary service data (USSD) mobile banking. Governments especially should
invest in information technology (IT) infrastructure that other small private Fintech
companies can tap onto to provide modern financial services. Central banks in partic-
ular should also provide the favourable regulatory environment that would encourage
more firms to use digital financial services. For instance, deregulation could allow
telecommunication companies to partner with banks and other financial institutions
to provide mobile banking services like the mobile money services.

Our study also show that secondary school enrolment reduces financial inclusion.
As indicated earlier, most of the students enrolled at the secondary level are young and
thus are more likely to be financially excluded. These students are normally limited in
accessing key financial services because of age restrictions. They are mostly required
to come with an adult guarantor who may be their parent or sibling or a relative. As
indicated above, a good way to include these students would be through the adult
population. As more of their parents and adult siblings get financially included, they

123



Foreign direct investment and inclusive finance...

are more likely to be available to introduce and support the student to engage in
financial services.

In contrast, higher tertiary enrolment leads to improvedfinancial inclusion. Students
at the tertiary levels are most likely adults who are of working age. These students
are more likely to be exposed to financial services at this level. It is not surprising
that most banks take advantage of freshmen orientations to introduce their financial
products and services to students at the tertiary levels. At this level the students are
more likely to work to either finance their education or the students may already be
in the working class. As people work, the need for financial services goes up like the
need for a bank account to save and possible acquire loans in the future. As a policy
direction, governments can focus on these tertiary students as a conduit for scaling up
financial inclusion in their countries. These students are also mostly of age and can
thus be exposed to more financial education that can also increase their likelihood of
accessing diverse financial services.

Funding Open Access funding enabled and organized by CAUL and its Member Institutions

Declarations

Conflict of interest No potential competing interest was reported by the authors.

OpenAccess This article is licensedunder aCreativeCommonsAttribution 4.0 InternationalLicense,which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence,
and indicate if changes were made. The images or other third party material in this article are included
in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If
material is not included in the article’s Creative Commons licence and your intended use is not permitted
by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Appendix

Table 15 provides the list of countries used in the study (including cross-sectional
analysis).

A.1 Correlation analysis

Here, we present the correlation matrix of the independent variables in Table 16.
From the table, we see that the institutional quality variables are highly correlated and
hence we use them individually in separate regressions. All the other correlations have
absolute values of less than or equal to 0.70 indicating that their inclusion will not
present any problem of multicollinearity (Kennedy 2008).

123

http://creativecommons.org/licenses/by/4.0/


J. Y. Abor et al.

Ta
bl
e
15

L
is
to

f
co
un
tr
ie
s

A
fg
ha
ni
st
an

C
os
ta
R
ic
a

In
do

ne
si
a

M
on

go
lia

Si
er
ra

L
eo
ne

A
lb
an
ia

C
ot
e
d’
Iv
oi
re

Ir
el
an
d

M
on
te
ne
gr
o

Si
ng
ap
or
e

A
lg
er
ia

C
ro
at
ia

Is
ra
el

M
or
oc
co

Sl
ov
ak

R
ep
ub
lic

A
rg
en
tin

a
C
yp

ru
s

It
al
y

M
oz
am

bi
qu

e
Sl
ov
en
ia

A
rm

en
ia

C
ze
ch

R
ep
ub
lic

Jo
rd
an

M
ya
nm

ar
So

ut
h
A
fr
ic
a

A
us
tr
al
ia

D
en
m
ar
k

K
az
ak
hs
ta
n

N
am

ib
ia

Sp
ai
n

A
us
tr
ia

D
om

in
ic
an

R
ep
ub

l
K
or
ea
,R

ep
N
ep
al

Sr
iL

an
ka

A
ze
rb
ai
ja
n

E
cu
ad
or

K
uw

ai
t

N
et
he
rl
an
ds

Sw
ed
en

B
an
gl
ad
es
h

E
gy
pt
,A

ra
b
R
ep

K
yr
gy
z
R
ep
ub
lic

N
ew

Z
ea
la
nd

Ta
jik

is
ta
n

B
el
ar
us

E
lS

al
va
do
r

L
at
vi
a

N
ig
er

Ta
nz
an
ia

B
el
gi
um

E
st
on
ia

L
es
ot
ho

N
or
w
ay

T
ha
ila
nd

B
en
in

Fi
nl
an
d

L
ib
er
ia

Pa
ki
st
an

To
go

B
ol
iv
ia

Fr
an
ce

L
ith

ua
ni
a

Pa
na
m
a

T
un

is
ia

B
ul
ga
ri
a

G
eo
rg
ia

L
ux
em

bo
ur
g

Pe
ru

T
ur
ke
y

B
ur
ki
na

Fa
so

G
er
m
an
y

M
ad
ag
as
ca
r

Ph
ili
pp
in
es

U
ga
nd
a

C
am

bo
di
a

G
ha
na

M
al
ay
si
a

Po
la
nd

U
ni
te
d
A
ra
b
E
m
ir

C
am

er
oo
n

G
re
ec
e

M
al
i

Po
rt
ug
al

U
ni
te
d
K
in
gd
om

C
en
tr
al
A
fr
ic
an

G
ua
te
m
al
a

M
al
ta

R
om

an
ia

U
ni
te
d
St
at
es

C
ha
d

H
on

du
ra
s

M
au
ri
ta
ni
a

R
us
si
an

Fe
de
ra
tio

n
U
ru
gu

ay

C
hi
le

H
on

g
K
on

g
SA

R
,C

M
au
ri
tiu

s
R
w
an
da

U
zb
ek
is
ta
n

C
hi
na

H
un
ga
ry

M
ex
ic
o

Sa
ud
iA

ra
bi
a

V
ie
tn
am

C
ol
om

bi
a

In
di
a

M
ol
do
va

Se
rb
ia

Z
am

bi
a

123



Foreign direct investment and inclusive finance...

Ta
bl
e
16

C
or
re
la
tio

n
m
at
ri
x

V
ar
ia
bl
e

FD
I

L
aw

G
ov
er
nm

en
t
R
eg
ul
at
or
y
qu
al
ity

C
or
ru
pt
io
n
Pr
iv
at
e
cr
ed
it
ln
(r
ea
lG

D
P
pe
r
ca
pi
ta
)
A
du
lt
po
pu
la
tio

n
IC
T

Se
co
nd
ar
y
ed
uc
at
io
n
Te
rt
ia
ry

ed
uc
at
io
n

FD
I

1

L
aw

0.
24
5*
**

1

G
ov
er
nm

en
t

0.
23
6*
**

0.
91
2*
**

1

R
eg
ul
at
or
y
qu
al
ity

0.
24
3*
**

0.
89
0*
**

0.
90
3*
**

1

C
or
ru
pt
io
n

0.
24
1*
**

0.
92
3*
**

0.
88
1*
**

0.
83
1*
**

1

Pr
iv
at
e
cr
ed
it

0.
20
5*
**

0.
60
9*
**

0.
66
3*
**

0.
62
7*
**

0.
51
0*
**

1

ln
(r
ea
lG

D
P
pe
r
ca
pi
ta
)
0.
13
7*
**

0.
64
7*
**

0.
71
8*
**

0.
65
6*
**

0.
60
4*
**

0.
56
8*
**

1

A
du
lt
po
pu
la
tio

n
0.
13
9*
**

0.
49
0*
**

0.
59
3*
**

0.
51
4*
**

0.
47
**
*

0.
60
5*
**

0.
62
2*
**

1

IC
T

0.
08
2*
*

0.
49
8*
**

0.
56
4*
**

0.
54
5*
**

0.
46
5*
**

0.
49
5*
**

0.
70
4*
**

0.
60
1*
**

1

Se
co
nd
ar
y

ed
uc
at
io
n

0.
01
4

0.
03

−0
.0
16

−0
.0
04

−0
.0
08

0.
02
3

0.
06
4*

0.
02
5

0.
07
4*
*

1

Te
rt
ia
ry

ed
uc
at
io
n

0.
02
2

0.
36
0*
**

0.
43
2*
**

0.
42
3*
**

0.
29
0*
**

0.
44
1*
**

0.
53
4*
**

0.
44
5*
**

0.
49
3*
**

−0
.1
19
**
*

1

**
an
d
**
*
de
no
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
5
an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



J. Y. Abor et al.

A.2 Further robustness

A.2.1 Using lags of independent variables

Here, we use the lags of the independent variables and re-estimate our baseline model.
This is to provide some robustness checks for our initial results. Perhaps, FDI may
have some kind of gestation lag before they exert some impact on financial inclusion.
Again, the use of lags of the independent variables helps to reduce the likelihood of any
contemporaneous correlations and thus reduces any potential endogeneity concerns.
The results as shown in Table 17 are consistent with our main estimates that FDI has
a positive impact on financial inclusion with at least 5% significance level.

A.2.2 Cross-sectional analysis: alternative measure of financial inclusion

As further robustness check, we use the 2017 round of the global FINDEX survey
by the World Bank. Even though this represents a cross-sectional data, the data has
several measures of financial inclusion. As indicated by Suri and Jack (2016), financial
inclusion involves access to and use of financial services. We create two financial
inclusion indices: (i) financial access, which measures access to financial services; (ii)
use of financial services, which measures the actual usage of financial services. Thus,
respectively, these represent the proportion of adult population that have access and
used any form of financial services over the past year. We use PCA to extract these
measures and scale between 0 and 1. For financial access, we use four variables: (i) If
an individual owns an account with a financial institution or a mobile money account;
(ii) If an individual owns an account with a financial institution; (iii) If an individual
owns a debit card, and (iv) If an individual owns a credit card. Also, for financial usage,
we use four variables: (i) If the individual saved with a financial institution within the
study period; (ii) If the individual borrowed within that period; (iii) If the individual
used a debit or credit card, and (iv) If the individual received digital payments within
that period.

The descriptive statistics and description of the variables are indicated in Table
18. From the table, we see that the highest financial inclusion measure is account
ownership with about an average of 61% of people owning an account with the least
being owning a credit card. This shows thatmore people in SSA tend to own an account
but very few tend to use a credit card even though about 43% use a debit card. We
also generally see that more people use financial services compared to financial usage
considering the all the sub-components. We observe the country rankings of financial
access and usage this time with 142 countries for financial access and 123 countries
for financial usage, we see that Norway tops the list with the highest access to financial
services while Denmark uses financial services the most.

The regression estimates are reported in Tables 19 and 20. Overall, the IV results
confirm our earlier findings that FDI has a positive impact on financial inclusion for
both access to and usage of financial services. The impact of FDI on access to financial
services seems to be larger than that of usage of financial services. This suggests that
higher FDI inflows has higher potential to extend financial services to the excluded

123



Foreign direct investment and inclusive finance...

Ta
bl
e
17

Fi
na
nc
ia
li
nc
lu
si
on

an
d
FD

I,
la
gs

of
in
de
pe
nd
en
tv

ar
ia
bl
es

V
ar
ia
bl
e

FI
N
C

(1
)

(2
)

(3
)

(4
)

L
.F
D
I

0.
00

78
**

0.
00

86
**

0.
00

89
**

0.
00

81
**

(0
.0
04

)
(0
.0
04

)
(0
.0
04

)
(0
.0
04

)

L
.L
aw

0.
04

29
**

*

(0
.0
11

)

L
.G
ov
er
nm

en
t

0.
04

20
**

*

(0
.0
12

)

L
.R
eg
ul
at
or
y
qu

al
ity

0.
02

49
**

(0
.0
10

)

L
.C
or
ru
pt
io
n

0.
04

21
**

*

(0
.0
11

)

L
.P
ri
va
te
cr
ed
it

0.
00

15
**

*
0.
00

15
**

*
0.
00

16
**

*
0.
00

16
**

*

(0
.0
00

3)
(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

3)

L
.N
at
ur
al
lo
g
re
al
G
D
P
pe
r
ca
pi
ta

0.
04

36
**

*
0.
04

51
**

*
0.
05

06
**

*
0.
04

35
**

*

(0
.0
10

)
(0
.0
09

)
(0
.0
10

)
(0
.0
10

)

A
du

lt
po

pu
la
tio

n
0.
00

30
**

0.
00

27
**

0.
00

30
**

0.
00

28
**

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

IC
T

0.
00

05
**

*
0.
00

05
**

*
0.
00

05
**

*
0.
00

05
**

*

(0
.0
00

2)
(0
.0
00

2)
(0
.0
00

2)
(0
.0
00

2)

123



J. Y. Abor et al.

Ta
bl
e
17

co
nt
in
ue
d

V
ar
ia
bl
e

FI
N
C

(1
)

(2
)

(3
)

(4
)

Se
co
nd

ar
y
ed
uc
at
io
n

−0
.0
00

1*
**

−0
.0
00

1*
**

−0
.0
00

1*
**

−0
.0
00

1*
**

(0
.0
00

02
)

(0
.0
00

02
)

(0
.0
00

02
)

(0
.0
00

02
)

Te
rt
ia
ry

ed
uc
at
io
n

0.
00

21
**

*
0.
00

20
**

*
0.
00

20
**

*
0.
00

22
**

*

(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

4)

F
ir
st
st
ag
e
re
gr
es
si
on

B
IT
s

4.
45

3*
**

4.
42

1*
**

4.
42

4*
**

4.
41

0*
**

(1
.1
73

)
(1
.1
28

)
(1
.1
26

)
(1
.1
72

)

A
dj
us
te
d
R
2

0.
31

0.
31

0.
31

0.
31

D
ur
bi
n

χ
2
-t
es
t

38
.7
2*

**
47

.5
2*

**
49

.6
7*

**
42

.7
0*

**

W
u–
H
au
sm

an
F
-t
es
t

40
.6
25

**
*

50
.6
1*

**
53

.1
1*

**
45

.1
0*

**

1s
t-
st
ag
e
F

14
.4
1*

**
15

.3
6*

**
15

.4
4*

**
14

.1
5*

**

R
2

0.
67

4
0.
66

0.
65

2
0.
67

2

N
o.

of
co
un

tr
ie
s

87
87

87
87

O
bs

62
3

62
3

62
3

62
3

R
ob
us
ts
ta
nd
ar
d
er
ro
rs
ar
e
in

pa
re
nt
he
se
s,
**

an
d
**
*
de
no
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
5
an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



Foreign direct investment and inclusive finance...

Table 18 Descriptive statistics for cross-sectional data

Variable Obs Mean Std. dev Min Max

Panel A: Financial access

Account ownership 142 61.05 26.69 8.57 99.92

Financial account ownership 142 57.75 29.31 8.57 99.92

Own a debit card 142 43.78 31.43 1.73 98.81

Own a credit card 142 18.86 20.86 0 82.59

Access 142 0.42 0.29 0 1

Panel B: Financial usage

Saved 142 23.35 19.44 1.63 79.33

Borrowed 142 23.58 18.62 2.36 82.83

Used a debit or credit card 123 38.83 31.58 1.98 95.69

Made or received digital payments 142 53.51 28.39 7.35 99.39

Usage 123 0.37 0.29 0 1

Panel C: Other variables

FDI 140 4.94 9.63 −10.62 81.10

Private credit 132 58.99 46.12 3.59 223.39

Law 142 −0.04 1.00 −2.26 2.03

Government 142 0.03 0.99 −2.48 2.22

Regulatory quality 142 0.07 1.00 −2.21 2.16

Corruption 142 −0.08 1.00 −1.71 2.24

Natural log of real GDP per capita 140 9.41 1.17 6.82 11.63

Adult population 141 63.53 6.33 47.32 84.46

ICT 141 111.35 35.02 25.56 251.77

Secondary education 118 90.83 28.15 17.14 158.54

Tertiary education 122 46.95 28.90 3.73 136.60

Account ownership is the percentage of respondentswho report having an account (by themselves or together
with someone else) at a bank or other financial institution or owning a mobile money account (% age 15+);
financial account ownership is the percentage of respondents who report having an account (by themselves
or together with someone else) at a bank or other financial institution (% age 15+); own a debit card is the
percentage of respondents who report having a debit card (% age 15+); own a credit card is the percentage
of respondents who report having a credit card (% age 15+); saved is the percentage of respondents who
report saving or setting aside any money at a bank or another type of financial institution in the past 12
months (% age 15+); borrowed is the percentage of respondents who report borrowing any money from
a bank or another type of financial institution, or using a credit card, in the past 12 months (% age 15+);
used a debit or credit card is the percentage of respondents who report using a debit or credit card to make
a purchase in the past 12 months; made or received digital payments is the percentage of respondents who
report using mobile money, a debit or credit card, or a mobile phone to make a payment from an account,
or report using the internet to pay bills or to buy something online, in the past 12 months. It also includes
respondents who report paying bills, sending or receiving remittances, receiving payments for agricultural
products, receiving government transfers, receiving wages, or receiving a public sector pension directly
from or into a financial institution account or through a mobile money account in the past 12 months (%
age 15+). All other variables are as defined earlier

123



J. Y. Abor et al.

Ta
bl
e
19

Fi
na
nc
ia
li
nc
lu
si
on

(a
cc
es
s)
an
d
FD

I,
IV

es
tim

at
io
n

V
ar
ia
bl
e

Fi
na
nc
ia
li
nc
lu
si
on

in
de
x
(a
cc
es
s)

(1
)

(2
)

(3
)

(4
)

FD
I

0.
00

45
0.
00

56
*

0.
00

56
0.
00

67
*

(0
.0
03

)
(0
.0
03

)
(0
.0
04

)
(0
.0
04

)

L
aw

0.
11

81
**

*

(0
.0
24

)

G
ov
er
nm

en
t

0.
14

85
**

*

(0
.0
39

)

R
eg
ul
at
or
y
qu

al
ity

0.
07

98
**

(0
.0
31

)

C
or
ru
pt
io
n

0.
09

58
**

*

(0
.0
23

)

Pr
iv
at
e
cr
ed
it

0.
00

02
0.
00

00
5

0.
00

05
0.
00

02

(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

4)

N
at
ur
al
lo
g
re
al
G
D
P
pe
r
ca
pi
ta

0.
11

52
**

*
0.
10

12
**

*
0.
14

60
**

*
0.
13

47
**

*

(0
.0
28

)
(0
.0
35

)
(0
.0
32

)
(0
.0
26

)

A
du

lt
po

pu
la
tio

n
−0

.0
03

4
−0

.0
04

8
−0

.0
06

1*
−0

.0
03

8

(0
.0
03

)
(0
.0
03

)
(0
.0
03

)
(0
.0
03

)

IC
T

−0
.0
00

6
−0

.0
00

6
−0

.0
00

8*
−0

.0
00

7

(0
.0
00

4)
(0
.0
00

4)
(0
.0
00

5)
(0
.0
00

5)

123



Foreign direct investment and inclusive finance...

Ta
bl
e
19

co
nt
in
ue
d

V
ar
ia
bl
e

Fi
na
nc
ia
li
nc
lu
si
on

in
de
x
(a
cc
es
s)

(1
)

(2
)

(3
)

(4
)

Se
co
nd

ar
y
ed
uc
at
io
n

0.
00

08
0.
00

09
0.
00

12
*

0.
00

10
*

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

Te
rt
ia
ry

ed
uc
at
io
n

0.
00

10
0.
00

09
0.
00

07
0.
00

09

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

F
ir
st
-s
ta
ge

re
gr
es
si
on

B
IT
s

3.
55

8*
**

(0
.8
77

)
3.
49

1*
**

(0
.8
66

)
3.
36

5*
**

(0
.8
78

)
3.
47

0*
**

(0
.8
59

)

A
dj
us
te
d
R
2

0.
21

0.
21

0.
21

0.
21

D
ur
bi
n

χ
2
-t
es
t

4.
96

**
7.
08

**
*

6.
20

**
9.
33

**
*

W
u–
H
au
sm

an
F
-t
es
t

4.
72

**
6.
88

**
*

5.
97

**
9.
27

**
*

1s
t-
st
ag
e
F

5.
83

**
5.
91

**
5.
09

**
6.
64

**

R
2

0.
86

0.
85

0.
83

0.
83

N
o.

of
co
un

tr
ie
s

11
0

11
0

11
0

11
0

R
ob
us
ts
ta
nd
ar
d
er
ro
rs
in

th
e
pa
re
nt
he
si
s.
*,

**
an
d
**
*
de
no
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
10
,5

an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



J. Y. Abor et al.

Ta
bl
e
20

Fi
na
nc
ia
li
nc
lu
si
on

(u
sa
ge
)
an
d
FD

I,
IV

es
tim

at
io
n

V
ar
ia
bl
e

Fi
na
nc
ia
li
nc
lu
si
on

in
de
x
(u
sa
ge
)

(1
)

(2
)

(3
)

(4
)

FD
I

0.
00

76
*

0.
00

89
*

0.
00

90
0.
01

05
*

(0
.0
05

)
(0
.0
05

)
(0
.0
06

)
(0
.0
06

)

L
aw

0.
14

89
**

*

(0
.0
39

)

G
ov
er
nm

en
t

0.
19

55
**

*

(0
.0
68

)

R
eg
ul
at
or
y
qu

al
ity

0.
10

10
**

(0
.0
50

)

C
or
ru
pt
io
n

0.
11

09
**

*

(0
.0
39

)

Pr
iv
at
e
cr
ed
it

−0
.0
00

4
−0

.0
00

7
−0

.0
00

1
−0

.0
00

4

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

N
at
ur
al
lo
g
re
al
G
D
P
pe
r
ca
pi
ta

0.
12

13
**

*
0.
10

34
*

0.
16

12
**

*
0.
15

05
**

*

(0
.0
45

)
(0
.0
59

)
(0
.0
51

)
(0
.0
45

)

A
du

lt
po

pu
la
tio

n
−0

.0
08

6*
−0

.0
10

6*
*

−0
.0
12

2*
*

−0
.0
09

6*

(0
.0
05

)
(0
.0
05

)
(0
.0
05

)
(0
.0
05

)

IC
T

−0
.0
00

4
−0

.0
00

4
−0

.0
00

7
−0

.0
00

6

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

123



Foreign direct investment and inclusive finance...

Ta
bl
e
20

co
nt
in
ue
d

V
ar
ia
bl
e

Fi
na
nc
ia
li
nc
lu
si
on

in
de
x
(u
sa
ge
)

(1
)

(2
)

(3
)

(4
)

Se
co
nd

ar
y
ed
uc
at
io
n

0.
00

10
0.
00

10
0.
00

16
0.
00

13

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

Te
rt
ia
ry

ed
uc
at
io
n

0.
00

16
0.
00

14
0.
00

11
0.
00

14

(0
.0
01

)
(0
.0
01

)
(0
.0
01

)
(0
.0
01

)

F
ir
st
-s
ta
ge

re
gr
es
si
on

B
IT
s

3.
56

1*
**

(0
.9
43

)
3.
48

2*
**

(0
.9
30

)
3.
36

1*
**

(0
.9
41

)
3.
46

7*
**

(0
.9
21

)

A
dj
us
te
d
R
2

0.
20

0.
20

0.
20

0.
20

D
ur
bi
n

χ
2
-t
es
t

4.
51

**
6.
10

**
5.
59

**
7.
86

**
*

W
u–
H
au
sm

an
F
-t
es
t

4.
24

**
5.
84

**
5.
32

**
7.
66

**
*

1s
t-
st
ag
e
F

5.
36

**
5.
33

**
4.
79

**
6.
23

**

R
2

0.
72

0.
70

0.
67

0.
65

N
o.

of
co
un

tr
ie
s

96
96

96
96

R
ob
us
ts
ta
nd
ar
d
er
ro
rs
in

th
e
pa
re
nt
he
si
s.
*,

**
an
d
**
*
de
no
te
st
at
is
tic
al
si
gn
ifi
ca
nc
e
at
th
e
10
,5

an
d
1%

le
ve
ls
,r
es
pe
ct
iv
el
y

123



J. Y. Abor et al.

Fig. 5 Mean financial inclusion (access) by country. Note: Country labels adjusted with 2 intervals. Based
on 142 countries

123



Foreign direct investment and inclusive finance...

Fig. 6 Mean financial inclusion (usage) by country. Note: Country labels adjusted with 2 intervals. Based
on 123 countries

while also increasing the usage of these financial services by those who otherwise
were excluded.

References

AcemogluD, JohnsonS,Robinson JA (2001) The colonial origins of comparative development: an empirical
investigation. Am Econ Rev 91(5):1369–1401

Ackermann K, Awaworyi Churchill S, Smyth R (2021) Mobile phone coverage and violent conflict. J Econ
Behav Organ 188:269–287

Agbloyor EK, Abor JY, Adjasi CKD,YawsonA (2014) Private capital flows and economic growth in Africa:
the role of domestic financial markets. J Int Finan Markets Inst Money 30:137–152

Alesina A, Zhuravskaya E (2011) Segregation and the quality of government in a cross section of countries.
Am Econ Rev 101(5):1872–1911

Allen F, Demirguc-Kunt A, Klapper L, Peria MSM (2016) The foundations of financial inclusion: under-
standing ownership and use of formal accounts. J Finan Intermed 27:1–30

Anarfo EB, Abor JY, Osei KA, Gyeke-Dako A (2019) Financial inclusion and financial sector development
in Sub-Saharan Africa: a panel VAR approach. Int J Manag Financ 15(4):444–463

Arellano M, Bover O (1995) Another look at the instrumental variable estimation of error-components
models. J Econom 68(1):29–51

Awaworyi Churchill S, Marisetty VB (2020) Financial inclusion and poverty: a tale of forty-five thousand
households. Appl Econ 52(16):1777–1788

123



J. Y. Abor et al.

Beck T, Demirgüç-Kunt A, Levine R (2007) Finance, inequality and the poor. J Econ Growth 12(1):27–49
Bhagwat V, Brogaard J, Julio B (2021) ABIT goes a longway: bilateral investment treaties and cross-border

mergers. J Financ Econ 140(2):514–538
Brune L,GinéX,Goldberg J, YangD (2016) Facilitating savings for agriculture: field experimental evidence

from Malawi. Econ Dev Cult Change 64(2):187–220
Ćihák M, Mare DS, Melecky M (2016) The nexus of financial inclusion and financial stability: a study

of trade-offs and synergies. World Bank Policy Research Working Paper, (7722), Washington D.C.:
World Bank

Dabla-Norris ME, Ji Y, Townsend R, Unsal MF (2015) Identifying constraints to financial inclusion and
their impact on GDP and inequality: a structural framework for policy. International Monetary Fund
Working Paper No. 2015/022

Demirgüç-Kunt A, Klapper L, Singer D (2017) Financial inclusion and inclusive growth: a review of recent
empirical evidence. World Bank Publications, Washington

Demirgüç-Kunt A, Klapper L, Singer D, Ansar S (2018) The global findex database 2017: measuring
financial inclusion and the fintech revolution. World Bank Publications, Washington

Dermirgüç-Kunt A, Klapper L (2012) Measuring financial inclusion: the global findex database. TheWorld
Bank, Washington

Dupas P, Robinson J (2013) Why don’t the poor save more? Evidence from health savings experiments.
Am Econ Rev 103(4):1138–1171

Grohmann A, Klühs T, Menkhoff L (2018) Does financial literacy improve financial inclusion? Cross
country evidence. World Dev 111:84–96

Hall R, Jones C (1999)Why do some countries produce so much more output per worker than others? Quart
J Econ 114(1):83–116

Holtz-Eakin D, Newey W, Rosen HS (1988) Estimating vector autoregressions with panel data. Economet-
rica 56(6):1371–1395

Jolliffe IT (2002) Choosing a subset of principal components or variables. Principal Comp Anal 2002:111–
149

Kaiser HF (1974) An index of factorial simplicity. Psychometrika 39(1):31–36
Karlan D, Morduch J (2010) Access to Finance. In: Dani R, Mark R (eds) Handbook of Development

Economics. Springer, USA, pp 4703–4784
Kaufmann D, Kraay A, Mastruzzi M (2010) The worldwide governance indicators: methodology and

analytical issues. World Bank Policy Research Working Paper No. 5430, The World Bank
Kennedy P (2008) A guide to econometrics, 6th edn. Blackwell Publishing, Oxford
Klapper L, Lusardi A, van Oudheusden P (2015) Financial literacy around the literacy survey. Available at:

https://www.finlit.mhfi.com
Knack S, Keefer P (1995) Institutions and economic performance: cross-country tests using alternative

institutional measures. Econ Politics 7(3):207–227
Lewbel A (2012) Using heteroscedasticity to identify and estimate mismeasured and endogenous regressor

models. J Bus Econ Stat 30(1):67–80
Liu Z, Spiegel MM, Zhang J (2023) Capital flows and income inequality. Federal Reserve Bank of San

Francisco. Working Paper 2020-14. Available at: https://doi.org/10.24148/wp2020-14
Liu F, Walheer B (2022) Financial inclusion, financial technology, and economic development: a composite

index approach. Emp Econ 63(3):1457–1487
Long C, Yang J, Zhang J (2015) Institutional impact of foreign direct investment in China. World Dev

66:31–48
Mauro P (1995) Corruption and growth. Quart J Econ 110(3):681–712
Muriu PW (2020) Role of institutional quality in promoting financial inclusion. AERC Research Paper 384,

African Economic Research Consortium, Nairobi
Naceur MSB, Barajas MA, Massara MA (2015) Can Islamic banking increase financial inclusion?. IMF

Working Paper (WP/15/31), International Monetary Fund
NeaimeS,Gaysset I (2018)Financial inclusion and stability inMENA: evidence frompoverty and inequality.

Financ Res Lett 24:230–237
Nkoa BEO, Song JS (2020) Does institutional quality affect financial inclusion in Africa? A panel data

analysis. Econ Syst 44(4):100836
North DA (1990) Institutions, institutional change and economic performance. Cambridge University Press,

New York

123

https://www.finlit.mhfi.com
https://doi.org/10.24148/wp2020-14


Foreign direct investment and inclusive finance...

Okada K (2013) The interaction effects of financial openness and institutions on international capital flows.
J Macroecon 35:131–143

Reiter SL, SteensmaHK (2010) Human development and foreign direct investment in developing countries:
the influence of FDI policy and corruption. World Dev 38(12):1678–1691

Rivera-Batiz L, Oliva M-A (2002) Political institutions, capital flows, and developing country growth: an
empirical investigation. Rev Dev Econ 6(2):248–262

Roodman D (2009) How to do xtabond2: an introduction to difference and system GMM in Stata. Stand
Genomic Sci 9(1):86–136

Sarma M (2012) Index of fnancial inclusion-a measure of fnancial sector inclusiveness. Berlin Working
Papers on Money, Finance, Trade and Development

Sharma D (2016) Nexus between financial inclusion and economic growth: evidence from the emerging
Indian economy. J Financ Econ Policy 8(1):13–36

Suri T, JackW(2016) TheLong-run poverty and gender impacts ofmobilemoney. Science 354(6317):1288–
92

Swamy, V. (2012). Bank-based financial intermediation for financial inclusion and inclusive growth. Avail-
able at SSRN: https://ssrn.com/abstract=2126834 or http://dx.doi.org/10.2139/ssrn.2126834

World Bank (2014) Financial Inclusion, Global Financial Development Report, The World Bank, Wash-
ington D.C

Zhao L (1998) The impact of foreign direct investment onwages and employment. Oxf Econ Pap 50(2):284–
301

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.

123

https://ssrn.com/abstract=2126834
http://dx.doi.org/10.2139/ssrn.2126834

	Foreign direct investment and inclusive finance: do financial markets and quality of institutions matter?
	Abstract
	1 Introduction
	2 Definitions, measures of financial inclusion and review of related studies
	2.1 Effect of financial markets development on financial inclusion
	2.2 Effect of institutions on financial inclusion 

	3 Data and sources
	3.1 Data
	3.2 Descriptive statistics

	4 Empirical methodology
	4.1 Model specification
	4.2 Identification strategy 

	5 Empirical results and discussions
	5.1 Baseline results 
	5.2 Main IV results
	5.3 Robustness checks
	5.3.1 Lewbel12 heteroscedasticity-based identification
	5.3.2 Using alternative approach: GMM estimate
	5.3.3 Examining regional differences
	5.3.4 Examining differences in income and development status of countries

	5.4 Mechanisms

	6 Conclusion and policy recommendations
	Appendix
	A.1 Correlation analysis
	A.2 Further robustness
	A.2.1 Using lags of independent variables
	A.2.2 Cross-sectional analysis: alternative measure of financial inclusion


	References