Financial institutions concentration and financial inclusion penetration in Nigeria: a comparative analysis

ABSTRACT

This paper examines financial inclusion penetration in two southwestern states (Lagos and Ekiti) in Nigeria. Lagos has a high concentration of financial Institutions, while Ekiti has few financial institutions. This paper uses survey research design and logit regression analysis to show evidence of a significant difference in financial inclusion penetration in the two states and its impact on the business performance and well-being of the citizens of the states. The study reveals that penetration is higher in Lagos at 81% and lower in Ekiti at 60%. Irregular income/job loss, unknown/hidden charges, long queues in the bank, and high maintenance fees constitute a top threat to 80% financial inclusion achieved in the Southwest zone. The study, therefore, recommends intervention policy that considers state-level characteristics. Also, government policy on employment generation should target low-income earners who are worse-off during an economic downturn.

KEYWORDS:

• Financial institution concentration
• financial inclusion penetration
• logit regression
• well-being

Introduction

A large number of the people living in developing economies are low-income earners who have a sizeable amount of money in numerous small portions idle in various hiding places. The world’s unbanked population has been estimated at 2.5 billion, half of whom live in Sub-Saharan Africa (Demirgüç-Kunt and Klapper 2012). There is consensus among development economists on the importance of a financial system in an economy. A functional and efficient financial system accelerates financial intermediation, ensures efficient allocation of resources for productive use, enhances the payment system, and facilitates access to credit. These, in turn, stimulate economic activities, promote equality, improve opportunities, and enhance growth and development in an economy. Therefore, building a functional and inclusive financial system must be a priority for all governments at all levels that want to achieve accelerated economic growth.

Better access to financial services implies making financial services accessible to everyone irrespective of their social status or class in the society, thereby enhancing equality and wealth redistribution, which will ultimately reduce poverty and improve national prosperity. Access to essential financial services implies strengthening the quality and reach of savings products, credit facility, payments system, transfers, remittances, and insurance financial services for sustained productivity and growth.

The global financial inclusion report published in 2015 shows that the percentage of adults without a formal bank account fell by 11% between 2011 and 2014. The total number of unbanked people remained over 2 billion, with more than half living in developing countries. Of the adults in developed economies, 89% had a bank account, while only 24% of those living in developing economies had a bank account. The report also shows that women, rural dwellers, and those with an unstable income were less likely to own a formal bank account (Demirguc-Kunt et al. 2015).

In 2012, the Nigerian Government made a definite commitment to enhancing financial inclusion in the country through its national financial inclusion strategic plan. The main objective of the plan is to reduce the number of adults that can not access financial service in the country from 46.3% to 20% and increase the formal sector’s share of financial inclusion from 30% to 70% by the year 2020. This objective will be achieved through a simplified risk base tiered framework that allows easy means of identification at the bank, implements the agent banking regulation framework, and implements national financial literacy framework in order to increase the level of awareness and knowledge of financial products. Also, it aims to put in place a comprehensive consumer protection framework to boost the populace’s confidence in the financial sector. Besides, measures have been put in place to pursue a mobile payment system and other cashless policies to reduce the cost and ease financial service penetration, and empower micro, small and medium-sized enterprises by implementing credit enhancement schemes and programmes. The purpose of these measures is to remove the barriers to accessing financial services and ensure that adult Nigerians can access essential financial services irrespective of their status in society.

The number of adult Nigerians excluded from the financial system, which was 36.9% in 2014, grew to 40.1% in 2016, as the contribution of microfinance banks declined substantially due to the economic recession recorded in the country in 2016. The number stood at 36.8% in 2018 as the economy recovered. The formal sector’s share of financial inclusion grew from 30% in 2010 to 36.3% in 2014, then to 38.3 percent in 2016, and then to 39.6% in 2018 as deposit money banks remained steadfast in their financial inclusion drive throughout the recession period (CBN 2018; EFINA 2018).

Financial inclusion has always been part of financial industry reforms since the 1970s. In 1977, a rural banking programme was introduced that forced all the commercial banks in Nigeria to establish rural branches. In 1989, the People’s Bank was introduced to bring banking closer to people; in 1990, the community banking system was established, and these banks were to be owned by community members with support from the Government to function effectively. Each community was to float its bank to enable members of the community to have access to financial services, but many of the banks could not function due to a lack of capital. The community banks transformed into microfinance banks in 2005 but adopted a more regulated supervisory framework. The effect of microfinance is still continuously debated in the literature. In 2012, the monetary authorities embarked on a massive financial literacy campaign, introduced a cashless policy and an electronic payment system, and between 2014–2016 approved an agent banking scheme, developed a framework for mobile banking, and put in place a financial service consumer protection strategy. All of these efforts yielded a massive reduction in financially excluded adults in Nigeria, from 76.4 in 2008 to 41.6 in 2016.

Beate (2005) explains that the concentration of financial institution in one location leads to improved information flow, greater liquidity, higher efficiency in the organised market and centralisation of support services. The concentration of banking firms in a location engenders economies of scale and lock-in effects, which attract high-quality professionals and support services into the environment such as accounting, actuarial, legal services, security services, telecommunication infrastructure and office accommodation. The concentration creates a hub which engenders economies of scale and lower cost of production. Reduced cost of production brings about affordable prices of financial services product, quality product design and availability of varieties of product.

The early work of Allen, Demirguc-Kunt, Levine and Haubrich (2004) and Beck, Demirgüç-Kunt and Maksimovic (2004) explain bank concentration and competition impact on bank performance and firm access to finance. The potential impact of financial market structure on access to finance was emphasised in the work of Beck Demirgüç-Kunt, and Maksimovic (2004). The traditional school on market power argues that competition in the banking market reduces the cost of finance and increases the availability of financial services (e.g., Berger & Hannan 1998). Other schools of thought believe that bank competition harms credit availability because it may interact with the level of asymmetric information in the credit market. There is no consensus on the impact of bank concentration and competition on access to finance in the credit market and availability of affordable financial service (see Cetorelli & Paretto 2000). This study, therefore, attempts to fill this observed gap in the literature on financial institutions concentration in Nigeria.

Also, financial sector outreach to low-income groups is essential for several reasons. Borrowing may be the only available option for poor households to cope with emergencies and to access basic amenities such as energy, water, education, and health services (Peachey and Roe 2006). Affordable savings product help to smoothen consumption during the off-season period; therefore, savings and credit products are welfare-enhancing products. The financial market imperfections, which often translate to informational asymmetries and high transaction costs, result to financial constraints for poor household due to a lack of credit history, collateral, and connections (Beck Degryse, and Kneer 2014; Chauvet and Jacolin 2017; Owen and Pereira 2018). The lack of data at the aggregate household level often exacerbate the gap on characteristics of who has access to which financial product and the barriers to broad access. This study, therefore, fills the observed gaps in the literature in three distinct ways: i) examining financial institutions concentration induced competition in driving financial inclusion in rich and poor sub-nationals; ii) examining the influence of financial inclusion in the sub-nationals in driving firm performance and individual well-being; and iii) examining whether there are significant barriers to broad access to financial inclusion among low-income earners in different sub-nationals with a significant difference in financial institution concentration.

This paper adds to the evolving literature on financial inclusion in general and the influence of bank concentration and competition in ‘endowed’ and ‘less endowed’ subnational in financial outreach for low-income households in particular. Efforts to understand how formal financial systems affect the poor household remain inadequate as studies on financial inclusion are still evolving. The country-level research does not provide sufficient information on location-specific characteristic as countries, and different location response to financial inclusion differs significantly. The significance of income level, gender, literacy, age and location as the determinant of financial inclusion suggest this (Allen et al. 2016; Beck et al. 2014a). This implies that there is a socioeconomic limitation in determining inclusion and exclusion. Proffering a one-size-fits-all solution to problems of financial inclusion may be grossly inadequate despite the high cost of implementing such study. Ozil (2020) and Sarma and Pais (2008) show that financial inclusion is influenced by poverty level, state of economic development, financial sector stability, financial literacy, consumer protection framework, regulatory and supervisory framework and cultural orientation which differs significantly from country to country, and region to region even within the same country. This study uses location peculiarities to understand financial inclusion, the impact of inclusion and barriers to inclusion. To our knowledge, this is the first study in the study area to do so.

This paper is divided into five sections. After the study background, section two consists of an empirical literature review, the methodology employed is discussed in section three, the result is presented in section four, while the summary and conclusion are discussed in section five.

Literature review

Conceptual framework

Financial development is one of the major drivers of economic growth and development (Levine 1997). The provision of cost-effective, relevant, sustainable and reliable financial services to the low-income groups in the society is referred to as financial inclusion. A variety of financial services should be made available in the society without discrimination or social status to as many as needs it. An individual member of the society can own a formal bank account to save and plan to spend, for remittances and transfer, to insure products against adverse economic conditions, and access credit for personal development or business expansion. The World Bank (2014) suggests that financial inclusion consists of sustainable, low-cost delivery of financial services to low-income groups in society (Sahay et al. 2015). One of the importance of financial inclusion is its ability to foster economic growth through investment in education and easy access to entrepreneurial finance. Individuals are also empowered to invest in their health and well-being, manage shocks, and consumption effectively (World Bank Group 2014).

Theoretical framework

The theoretical framework for the study is base on two strands of the literature; on the one hand, is bank concentration and competition theory and Consumer Choice theory. The bank concentration literature examines how competition in banking market (bank concentration, market share) influences the variety of banking products, the cost-effectiveness of banking services, and firm performance (Beck et al. 2014a; Owen and Pereira 2018). The traditional market power view, bank competition can affect the efficiency of the financial sector, the quality of the products, the degree of innovation and drive inclusiveness (Chauvet and Jacolin 2017). Owen and Pereira (2018), and Inessa and Martınez-Perıa (2015) find that increase competition due to bank concentration intensifies firms inclusion and access to credit. Claessens (2006) affirms that bank competition only favours the sectors of the economy that rely more on bank financing. Market structure effects are more pronounced in developing and emerging countries where poor regulation, low efficiency and the high cost of credit amplifies the gains probable from international best practices (Hermes and Lensink 2003). Although a strand of the literature also stressed that bank competition might not necessarily favour financial inclusion or result in the efficiency of the credit market, especially in developing economy. This strand of the literature argues that bank concentration may decrease access to credit as banks struggle to overcome information asymmetric and invest more in assessing the creditworthiness of opaque borrowers (Owen and Pereira 2018).

Consumer choice theory

The neoclassical economic theory explains how consumer spends their money based on individual preferences and budget constraints. The theory is built around an understanding of how individuals’ tastes and incomes influence demand and supply of goods and services, and the impact of this decision on the economy. A rational consumer will maximise consumption by spending all their income on products and services that gives the highest satisfaction. The consumer theory is based on three assumptions, i. utility maximisation ii. Nonsatiation and iii. decreasing marginal utility. Consumer makes economic choices that give the highest satisfaction subject to their budget. Consumers are never satisfied at first consumption; they are satisfied after repeated use and satisfaction decline with every additional consumption. The theory emphasises the optimal usage of available resources to achieve maximum benefit which brings about optimal satisfaction. This implies that it is not enough to own a bank account; it is the consistent usage of the bank account that gives satisfaction. The behavioural economists are quick to point out that consumers of financial product are not always rational, and some decisions are difficult to make when consumers are not familiar with the product. Campbell (2006) drew the attention of the research community to household finance when he commented that the effective functioning of the financial market depends on households financial decision and behaviour.

Shefrin and Thaler (1988) emphasise the role of self-control, mental accounting, and framing in financial decision making. They also documented evidence of inefficient consumer financial decisions and factors that affect suboptimal choices in various aspect of the financial markets such as credit and mortgage market, invest and retirement saving market and on consumption and savings generally. This they reported may be as a result of limited cognitive ability and financial literacy, psychological biases and social network, and peer pressure in financial decision making. Since this study, there has been increasing academic research on household financial and consumption decision (see Daniel, Hirshleifer & Subranmayam 1998; Daniel & Hirshleifer 2015). Before this, early researchers have propounded theories on consumer financial decision (see Agarwal, Chomsisengphet, & Lim, 2017). Many years of research has led to the discovery of limitations in cognitive ability, and behavioural biases household faces in financial management decision which led to the incorporation of financial literacy and consumer protection framework in financial inclusion drive by the Alliance for Financial Inclusion(AFI). The decision is not only to enhance household optimal financial decision capability but also to protect the financial market from eminence collapse resulting from aggregate bad financial decision making by consumers of financial products (Agarwal, et al., 2017). This came in focus shortly after the global financial crisis in 2008 resulting from the housing market meltdown. The contention is, what is the role of consumer financial decisions and financial intermediaries in financial market instability?

Empirical review

The debate has been on over the last two decades on the effect of greater access to finance on inequality and poverty reduction. Extant literature has shown that access to finance through a well-developed financial system can reduce income inequality and poverty, as well as enhance economic growth (Kappel 2010; Uddin et al. 2014; Abosedra, Shahbaz, & Nawaz 2016). In contrast, other studies reveal that financial development may fail due to reducing income, income inequality and poverty. Adeleye, Osabuohien, Bowale, Matthew and Oduntan (2018) suggest rapid financial development with strong institutions to enhance inclusive growth. Migap, Okwanya and Ojeka (2015) found that the Nigerian financial inclusion indicator is still shallow compared to other emerging economies both within and outside Africa. The study suggested that active participation of the media and educational institutions should be encouraged to promote financial literacy in Nigeria. Nkwede (2015) findings revealed a negative relationship between financial inclusion and economic growth in Nigeria, which the researcher attributed to the exclusion of a high percentage of adult Nigerians from financial services. Since the well-being of the population depends on attributes, such as income, health, and housing, access to financial services can also be regarded as an essential ingredient of human well-being. It is, therefore, necessary to design an appropriate policy for financial inclusion (Chakravarty & Pal, 2013).

Zins, and Weill (2016) found that male gender, higher income level, older age group and a higher level of education are determinants of financial inclusion in 37 African countries. Chikalipah (2017) found that illiteracy is the main barrier to financial inclusion in Sub-Saharan Africa. Ali (2019) shows that women that are disadvantage face barriers hindering their access to Islamic financial services in Comoros. He also shows that ‘no money’ and financial illiteracy are the reasons for perpetual poverty in the country. Mitchell and Scott (2019) conclude that formal financial inclusion, cashless policy and increased use of cards for transaction increased public tax revenue in Argentina. Ghosh and Bhattacharya (2019) attest that financial innovation, such as ‘SureCash’ increased financial inclusion in Bangladesh as more women and poor adults are reached through the product. Susilowati and Leonnard (2019) find individual income level as the primary factor that influences ownership and usage of financial services in Indonesia. Income level accelerates access to loan because of collateral. Other factors include age, gender, level of education, and type of job.

Research methods and data

The survey research design was employed for this study. Survey research facilitates data collection across many strata and allows interaction with respondents to obtain detailed information. It also allows the collection of sufficient data through a questionnaire (Osuala, 2005). This study was both exploratory and explanatory. The exploratory aspect helped to determine the extent to which different individual characteristics and state-level variables, as well as policies, are associated with financial inclusion. It also helps to ascertain the influence of the inclusion on firm performance and individual well-being among target groups such as low-income earners, rural areas dwellers, women and young adult. The explanatory aspect helped to establish causal relationships between the dependent (ownership, usage, & business performance) and the independent (determinants of financial inclusion) variables and the interactions between them (Saunders, Lewis & Thornhill 2009). This study also employed quantitative techniques, which capture effects and impacts in comparison with the existing theories (Creswell 2014). The study used primary data to explain the character behind published secondary data, which, according to Blaikie (2010) and Fielding and Fielding (1986), provides a better perspective for a proper understanding of a subject of this nature. It is appropriate to ascertain the perception and opinions of account holders regarding why they decided to own a bank account.

The population for the study includes all low-income self-employed and salary earners in Lagos and Ekiti States, South-West Nigeria. The South-West geopolitical zone has a high banking sector presence because of Lagos, which is the commercial nerve centre of the country. Apart from deposit money banks (DMBs), it has the highest concentration of other financial institutions, such as microfinance, mortgage, and development banks, finance houses, discount houses, pension managers, insurance companies and bureaux de change. Lagos has 1,478 DMBs branches, while Ekiti has just 76 (CBN 2018). A sample of 500 low-income earners was drawn from the population of the two states; 185 returned questionnaires from Lagos and 102 from Ekiti were found useable. The sample was drawn using the simple random sampling technique. A structured questionnaire was used as an instrument of data collection.

A reliability test was conducted on the instrument through a pilot test, forty (40) copies of the questionnaire were sent to respondents to ascertain the reliability of the research instrument. The internal consistency of the questionnaire items was determined using the Cronbach alpha statistic; the result obtained was 0.812, which indicated that the instrument was reliable and consistent with the intended measurement parameter. Structural Equation Modelling (SEM) was used to perform confirmatory factor analysis (CFA) to assess the scale validity and the fit of the research instrument. The convergent phase was adopted for the validation of the research items. The study employed three conditions to assess the convergent validity: 1) the loading indicator that all the scale and measurement items are significant and exceed the minimum value criterion of 0.70; 2) each construct of composite reliability (CR) exceeds 0.80, and 3) each construct’s average variance extracted (AVE) estimate exceeds 0.50. The results obtained indicate that conditions for validity were met, and items used in the study are suitable.

Method of analysis

The evaluation of the relationship between the dependent and the independent variables was performed using logistic regression. Previous studies (see Allen et al. 2016; Zin & Weill 2016) used a similar technique. Babajide, Adegboye and Omankhanlen (2015), and Sahay et al. (2015) all applied regression analysis with the ordinary least square method. When a dependent variable is categorical, the ordinary least square (OLS) method can no longer produce the best linear unbiased estimator; logistic regression is designed to fit the categorical dependent variable, which adopts the maximum likelihood estimate method or weighted least squares estimation method. It also assumes probability distribution functions; logit models use the standard logistic probability distribution. The first step to fit the models consists of defining the variables of interest. In this study, we used account ownership, the frequency of account usage, the type of account (savings), business profit and income, well-being, and location as proxies for the dependent variable, while we used individual characteristics, financial inclusion, state-level characteristics and perceived barriers to financial inclusion as the independent variables.

The general form for the binary logistic model is: (1) $Prob(Y=1/x)=\mathrm{\Lambda }\left(x\text{β}\right)=\frac{Exp\left(x\text{β}\right)}{1+Exp\left(x\text{β}\right)}$(1) where Λ indicates a link function. The stepwise selection option is applied; the likelihood ratio test is performed to test the significance of the model coefficient. The fitted model is plotted, and predictions are generated from the specified model. Usually, residuals are identified and modelled.

In a specific form, equation (1)(1) $Prob(Y=1/x)=\mathrm{\Lambda }\left(x\text{β}\right)=\frac{Exp\left(x\text{β}\right)}{1+Exp\left(x\text{β}\right)}$(1) transforms into (2) $Pri=(1+Exp(-{\lambda }_1){)}^{-1}$(2) where ʎ is linearly dependent on the variable hypothesised to affect the probability $\lambda 1=\alpha +\text{β}{X}_1$; the probability thus varies from 0 to 1 $(\lambda =\pm \mathrm{\infty })$, and the model is simplified by rearranging into a log of the odds, (3) $\ln \left(P_1/\right(1-P_1\left)\right)-\alpha +\text{β}{X}_1$(3) which consists of the individual outcome and can be estimated with maximum likelihood. An interpretation can then be made by reverting to the probabilities interacted with the financial inclusion variables regarding the likelihood of opening a bank account as a probable function of many covariates, as specified below. (4) $Pr(FI=1)=f({\text{β}}_0+{\text{β}}_{1}X1+{\text{β}}_2{X}_2+\dots .+\text{β}nXn+\frac{y}{z})$(4) The dependent variable is a measure of account ownership, account type use (savings), and frequency of usage (number of withdrawals in a month).

X = a vector of explanatory variables of individual-level characteristics, state-level characteristics, and perceived barriers in the location, as identified by the respondents, to financial inclusion.

X₁, X₂, X₃, … , … , … , … ,x_n are independent variables (such as age, level of education, gender etc)

b₁, b₂, b₃, … , … , … , … , bn are regression coefficients that determine the contribution of the independent variables.

e = a residual or stochastic term (which reveals the strength of b₁×₁ .. b_nx_n; if e is low, this implies that the level of unexplained factors is low, then the residual R and R² will be high, and vice versa.

Result presentation and discussion

Socioeconomic profile and account ownership of Lagos respondents

The socioeconomic profile of account ownership was interacted with the individual-level characteristics of the respondents; the result reveals that, in the Lagos area, 55% of individuals who do not have a bank account are female, while 45% are male. Among the respondents who own a bank account, 79% are male, while 21% are female. This result follows the global trend that men are more likely to own a bank account in a formal financial institution than women in both developed and developing economies because of their financial independence, literacy level, and socioeconomic status.

Socioeconomic profile and account ownership of Ekiti respondents

The results from the respondents show that account ownership is higher for men (65%) than women (35%). Among those with no formal bank account, the numbers of women (61%) and men (39%) are also higher. Regarding age groups, larger proportions of the age groups 16–19 years (55%) and above 60 years (35%) have no formal bank account, and the percentage with no regular income is higher in those age groups. For ownership of a bank account, the age group 20–40 years accounts for the highest percentage (65%), followed by the age group 41–60 years (26%), while the age groups 16–19 years and above 60 years have the lowest representation.

Table 1. Financial inclusion penetration: Lagos and Ekiti.

Formal bank account ownership	Yes	No	Chi-Square	Total
Lagos	150(81)	35(19)	7.278 (0.007)	185
Ekiti	61(60)	41(40)	4.571 (0.002)	102
Total	211	76		287

Source: Author’s computation, 2018 (% in parenthesis)

The financial inclusion penetration is higher in Lagos, with 81% than in Ekiti, with 60% of our sample, and it is statistically significant at the 1% level. This confirms the financial institution concentration theory as Lagos State that parades more institutions have higher inclusion rate than Ekiti state with fewer banks and other financial institutions. This confirms result obtained by Owen and Pereira (2018), Chauvet and Jacolin; 2017 and Beate (2005) (see Table 1).

Table 2. Effect of financial inclusion penetration in the two states.

Location		Variables in the equation
		B	S.E.	Wald	Df	Sig.	Exp(B)
Step 1^a	Lagos	.456	.289	2.497	1	.014	1.578
	Ekiti	−.692	.321	4.638	1	.031	.501
	Constant	1.048	.149	49.623	1	.000	2.852

Source: Author’s computation, 2018

^a Variable(s) entered in step 1: Lagos, Ekiti

Account ownership is 1.5 times more likely in Lagos at the 1% significance level, while formal account ownership penetration in Ekiti is 0.5 times less likely at the 5% significance level. The overall percentage is 74.2, and the chi-square is 19.298 P< 0.0005 (see Table 2).

Barriers to financial inclusion in the states of Lagos and Ekiti

Table 3 shows the results obtained regarding the barriers to financial inclusion in the local area as perceived by the respondents. The result factors with a mean difference of 3–4 are categorised as a high threat, and factors with 2–2.9 are categorised as a moderate threat, while factors with 1–1.9 are a low threat. This implies that, despite the 80% financial inclusion achieved in southwest Nigeria, the persistent increase in the high and moderate threat factors is strong enough to reverse the financial inclusion success achieved in the region.

Table 3. Barriers to financial inclusion.

	LAGOS				EKITI
	T	Sig.	Mean Difference	Rank	t	Sig.	Mean Difference	Rank
Irregular income/loss of job	59.921	.000	3.69091	1	18.665	.000	3.13725	4
High maintenance fee	35.515	.000	3.27273	4	24.339	.000	3.27451	2
Strong informal service in the area	18.010	.000	2.24545	7	14.162	.000	2.76471	7
No means of identification	18.360	.000	1.99091	8	15.613	.000	2.70588	8
Low customer care	26.353	.000	1.62727	10	18.985	.000	1.70588	11
Past bad experience with the bank	20.330	.000	1.61818	11	13.750	.000	1.94118	10
Low literacy level	19.882	.000	1.77273	9	18.513	.000	2.23529	9
No network/non-functional ATM	13.183	.000	1.41818	12	11.789	.000	1.45098
Unknown/hidden charges	60.457	.000	3.46364	2	49.324	.000	3.37255	1
Bank too far from my house	23.595	.000	2.84545	6	18.914	.000	2.80392	6
Long queue in the bank	44.126	.000	3.29091	3	26.298	.000	3.15686	3
Religious beliefs	27.587	.000	1.62727	10	20.074	.000	1.37255	12
High minimum balance	30.507	.000	1.17273	13	11.785	.000	1.17647	13
I don’t trust banks	42.306	.000	3.26364	5	20.445	.000	3.11765	5

Source: Author’s computation, 2018.

The results show that irregular income/loss of job, unknown/hidden charges, high maintenance fees, and lack of trust in the banking sector constitute a high threat to financial inclusion in the Lagos area. The Government and banks’ management must work together to eradicate high threat factors to sustain the financial inclusion achieved in the region. The Government, through its policy and agencies, can reduce the unemployment rate, which is the cause of insufficient and irregular funds. Unknown/hidden charges also constitute a high threat, because bank customers’ realisation that money kept in their bank account is reducing without any justified reasons can prompt account closure. High maintenance fees and a lack of trust are also a significant threat. Like the Lagos area, similar issues also represent a high threat in the Ekiti area. The order of importance is a little different: hidden and unknown charges are the primary concern in the Ekiti area, followed by high maintenance fees, long queues in banking halls, and then irregular income. The recent distress in the banking sector has aggravated the lack of trust in the sector, particularly in the microfinance sub-sector, and waves of fraud in account holders’ bank account.

Travelling a long distance before access to financial services can result in a substantial cost for the low-income earner; technology diffusion, banking infrastructure, and agent banks can provide appropriate support, particularly for people living in the rural areas. The presence of active informal service providers is also a concern in the two states. ‘No means of identification’ and a ‘low literacy level’ are two significant factors that the National Financial Inclusion Strategy has provided adequate provisions to overcome through a risk-based first-tier identification and financial literacy campaign; this may be the reason why those issues no longer appear as a high threat in the local area. Low customer care, past bad experiences with the bank, and high minimum balances seem a low threat in the two states. These are factors that banking institutions local branches can resolve without resorting to their corporate head office or the Government. A high minimum balance used to pose a high threat; the government policy for a zero balance has gradually eliminated this in the two states. With adequate attention to the high and moderate threat factors, the financial inclusion achieved in the southwestern geopolitical zone can be sustained. The ranking differs slightly in the two states. Identification of high, moderate and low threat to financial inclusion is one of this study’s contributions to knowledge. The result obtained in this section is similar to the results of Beck et al., (2014b) and Agarwal, Chomsisengphet, and Lim (2017) on consumer choice on financial products.

Individual and state-level characteristics for financial inclusion

Table 4 shows the logistic regression results of individual and state-level characteristics interacted with account ownership in the states of Lagos and Ekiti. Column I shows the results for Lagos (a state with a high concentration of financial service providers), and Column II contains the results for Ekiti (a state with a minimal level of financial service providers).

Table 4. Regression results of financial inclusion and barriers to financial inclusion.

	Column 1				Column II
	Lagos				Ekiti
Characteristics	Coeff.	Wald	Sig.	Odds ratio	Coeff.	Wald	Sig.	Odds ratio
Sex (Male)	3.496	2.986	0.084*	32.978	1.747	3.456	0.104	4.923
Sex (Female)	0.128	1.297	0.005**	2.553	-3.456	2.116	0.002 ***	8.949
Age 20–40	.313	3.882	0.049**	1.986	0.424	1.789	0.067 *	1.234
No Edu.	-4.567	3.005	0.003 **	5.784	-0.233	0.433	0.116	2.255
Post-Sec. Edu.	0.767	1.667	0.001***	6.719	0.906	1.213	0.027**	2.475
Rural	-0.128	2.786	0.101*	1.999	-0.115	1.555	0.051**	7.565
Salary Earner	0.767	2.377	0.019***	23.308	1.752	0.982	0.191	0.967
High Internet Penetration	3.453	1.660	0.027**	20.867	0.013	1.004	0.069*	4.567
Positive State Govt. Policy	5.230	4.669	0.038**	3.619	1.916	0.769	0.718	1.419
Irregular Inc./Loss of Job	-1.547	3.471	0.062*	24.696	0.313	0.716	0.106*	14.989
Unknown/Hidden	-4.198	2.140	0.014***	46.546	-1.782	1.233	0.025**	14.985
Long Distance	-1.382 2.566	0.019***	10.223	-1.233	2.238	0.567	18.127
Omnibus Tests of Model Coefficients	Chi-square			Sig.	Chi-square			Sig.
	112.974			0.000	96.022			0.000
Pseudo R Squared	Cox and Snell			Nagelkerke	Cox and Snell			Nagelkerke
	64.2			85.6	42.9			63.3
Hosmer and Lemeshow Test	0.976				0.713
Overall Percentage	90.9				86.6
Observations	185				102

Source: Author’s computation. *** 1% significance level; ** 5% significance level; * 10% significance level

The main interest in the logit regression model is in the sign, the odds ratio (Exp(B)), and of course, the significance of the variables. As expected, the coefficients of gender (male and female) are consistent with the financial inclusion hypothesis that men are more likely to own a bank account than women in developing economies. The results show that male adults in the state of Lagos are more likely to own a bank account with a formal institution. The result for age groups shows that an individual in the age group 20–40 years is more likely to own a bank account with a formal institution in Lagos and Ekiti States. This result confirms the previous study such as (Allen et al. 2016; Zins and Weill 2016; Susilowati and Leonnard 2019).

The results obtained for the level of education show that those without any form of formal education are less likely to own a bank account. At the same time, those with the post-secondary school are more likely to own a bank account in Lagos and Ekiti, respectively. The results for those who reside in rural areas show that they are likely to own a bank account in Lagos and Ekiti, respectively. This support the confirmation of result on location differences. The results also show that salary earner, and those residing in states with high internet penetration are more likely to own a bank account. The effect is significant at the 5% and the 1% significant level in Lagos and Ekiti, respectively.

Regarding the barriers to financial inclusion, irregular income/loss of a job, Unknown/hidden charges and long-distance are significant barriers to financial inclusion. The variables are signed correctly and statistically significant at 1 and 5% significance level.

The models’ ‘goodness of fit’ test is shown in block 1, containing the omnibus test of the model coefficient. Model 1 (Lagos) shows a chi-square of 112.974 and high significance of p<.0005, and model 2 (Ekiti) has a 96.002 significance level, P<.0005, with 5 degrees of freedom. In the model, the summary shows the Cox and Snell R square and the Nagelkerke R square, which indicate the amount of variation in the dependent variable explained by the explanatory variable. Model 1 shows that between 64.2 and 85.6% of the dependent variable is explained by the explanatory variable, and model 2 shows that the explanatory variable explains between 42.9 and 63.3 of the dependent variable. The classification table shows the overall percentage, which implies that the model correctly classified 90.9 cases in model 1 and 86.6 cases in model 2. The positive and negative predictive values for model 1 are 94.4% and 75%, respectively, while the positive and negative predictive values for model 2 are 80.5 and 77.5, respectively (see Table 4).

Table 5. Regression results for the frequency of usage, business performance, and well-being.

	Column 1				Column II
	Lagos			Exp(B)	Ekiti			Exp(B)
	Coeff.	Wald	Sig.	Odds Ratio	Coeff.	Wald	Sig .	Odds Ratio
Type of Account (Saving)	1.880	2.667	0.012**	15.666	0.823	3.889	0.073***	5.617
Type of Transaction (Receive Money)	8.051	10.775	0.029**	9.767	1.098	1.990	0.102	4.676
Age in Bus. (>5 yr)	4.675	1.780	0.005**	6.209	0.918	2.718	0.664	1.456
Acct Open (>2)	1.213	0.880	0.016 **	3.567	0.705	1.546	0.041**	0.140
Get Support	3.742	4.360	0.037**	5.374	0.373	0.783	0.222	0.676
Use Other Service	1.276	0.444	0.001*	1.273	0.613	1.212	0.177	1.775
Increase in Bus. Profit (Yes)	0.737	1.645	0.045**	2.934	1.203	4.314	0.091***	0.391
Add Value to My Life (Yes)	4.355	2.341	0.014**	3.952	0.807	2.414	0.009**	2.931
Self-Employed	0.919	0.717	0.063***	5.056	-0.119	2.717	0.114	0.877
Type of Bank (Microfinance)	0.626	0.915	0.112	1.456	0.319	0.888	0.233	2.771
Omnibus Tests of Model Coefficients	Chi-square			Sig.	Chi-square			Sig.
	81.042			0.000	72.113			0.000
Pseudo R-Squared	Cox and Snell			Nagelkerke	Cox and Snell			Nagelkerke
	54.2			75.2	48.9			60.5
Hosmer and Lemeshow Test	0.816				0.733
Overall Percentage	80.9				76.6
Observations	185				102

Source: Author’s computation (2018). *** 10% significance level; ** 5% significance level; * 1% significance level.

Table 5 shows the logistic regression results of business performance and well-being interacted with the frequent use of a bank account in the states of Lagos and Ekiti. Frequent usage is defined here as ‘I make a withdrawal at least three times in a month’ through any channel of withdrawal – ATM, PoS or transfer – since the withdrawal is self-dependent, unlike deposits, which can be initiated by another person (Allen et al. 2016). Column I shows the results for Lagos (a state with a high concentration of financial service providers), and Column II shows the results for Ekiti (a state with a minimal concentration of financial service providers).

The results obtained show that low-income earners who use a savings account, receive money regularly, operating their business for more than five years and opened their bank account more than two years ago use their bank account more frequently in both Lagos and Ekiti States but the magnitude and significance level differs significantly. Those who perceived that they receive support from their bank because of the type of account that they own and those who use other bank services such as ATMs, POS, agent banks, and digital finance are more likely to use their bank account frequently.

People who perceive that the use of a bank account increases their business profit and adds value to their well-being are more likely to use their bank account more frequently. The result is significant at the 5 and the 10% level, in Lagos and Ekiti respectively.

The models’ ‘goodness of fit’ test is shown in block 1, containing the omnibus test of the model coefficient. Model 1 (Lagos) shows a chi-square of 81.042 and is highly significant, p<.0005, and model 2 (Ekiti) has a 72.113 significance level, P<.0005, with 5 degrees of freedom. In the model, the summary shows the Cox and Snell R square and the Nagelkerke R square, which indicate the amount of variation in the dependent variable explained by the explanatory variable. Model 1 shows that between 54.2 and 75.2% of the dependent variable is explained by the explanatory variable, and model 2 shows that the explanatory variable explains between 48.9 and 60.5% of the dependent variable. The classification table shows the overall percentage, which implies that the model correctly classified 82.5 cases in model 1 and 74.5 cases in model 2. The positive and negative predictive values for model 1 are 90.6% and 75%, respectively, while the positive and negative predictive values for model 2 are 85.5 and 68.5%, respectively.

Conclusion, summary and policy implications

Recent statistics show Lagos as the wealthiest state in the country and Ekiti as one of the poorest regarding revenue generation and contribution to the GDP. Though the two states are in the same geopolitical zone, the standard of living differs significantly. This study examines how financial institution concentration influence financial inclusion penetration in the two states; highlights the effects of financial inclusion on firm performance and well being, as well as barriers to broader access considering individual and state-level characteristics. The results from this study show that competition induced by financial institution concentration drive financial inclusion penetration and differs significantly in the two subnational. Though the factors that constitute barriers to financial inclusion among low-income earners differ, yet, they are similar despite the level of wealth in the two states. The magnitude and severity of the impact of financial inclusion on firm performance and well-being vary significantly. While the results from our sample show 81% financial inclusion penetration for Lagos, Ekiti has 60%. Account ownership is 1.5% more likely in Lagos at the 1% significance level, while it is 0.5 times less likely in Ekiti; this result is significant at the 5% level. The study identified high, moderate and low threat to the already achieved financial inclusion in the Southwest region and therefore, concludes that the factors categorised as a high threat to financial inclusion in the two sub nationals should be attended to immediately to avoid reverse inclusion. Policy implementation should be state-specific, as the barriers’ impact is not the same for the two states. Government policy on employment generation should target low-income earners as they are most likely displaced during economic down-turn.

Acknowledgements

The authors thank the management of Covenant University for payment of the publication fee for this paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Abiola Ayopo Babajide is an Associate Professor of Development Finance in the Department of Banking & Finance, College of Business and Social Sciences, Covenant University, Ota. Her research interest areas are Development Finance – Microfinance, Entrepreneurial Finance and Capital Market Development. An astute researcher, she has successfully executed numerous research projects. She has published extensively in reputable local and international journals.

Adedoyin Isola Lawal is a Senior Lecturer in the Department of Accounting and Finance, Landmark University, Nigeria. He holds a Bachelor Degree in Economics from the University of Ilorin, Ilorin, a Masters in Banking and Finance from Bayero University, Kano, and a PhD in Banking and Finance from Covenant University, Ota. He has published extensively in reputable journals. Lawal reviews for a number of Journals like African Development Review (Wiley); The Quarterly Review of Economics and Finance (Elsevier); International Journal of Emerging Markets (Emerald Insight); and Cogent Social Sciences (Taylor and Francis). He is on the Editorial Board of the Binus Business Review (Binus University, Indonesia) and serves as the Editor-in-Chief of the Human and Social Science Letters.

Lanre Amodu is a Senior Lecturer in the Department of Mass Communications, Covenant University. His has a PhD in Community Relations and Conflict Resolution. His areas of research interest include public relations, corporate communication and Information and Communication Technologies. Dr Amodu has published scholarly articles in international and reputable journals.

Olabanji Olukayode Ewetan is a political economist and holds a PhD in Economic Science. He lectures in the Department of Economics and Development Studies, Covenant University, Nigeria. The title of his thesis was “Fiscal Federalism and Macroeconomic Performance in Nigeria”. His research focuses on public sector economics, public finance, monetary economics and development economics.

Susanash Limiga Esoswe is a research fellow with the Department of Business Management and Sustainable Development, School of Business, Information and Communication Technology University, USA: Cameroon Campus. Her research interests focus on development finance, poverty alleviation and financial economics.

Tochukwu Chibuzor Okafor obtained his PhD in Finance, his MSc in Finance and BSc in Banking and Finance at Covenant University, Ota, Nigeria. He is currently a faculty at Covenant University, Ota, Nigeria. He has published in reputable outlets. His areas of interest are international finance, development finance and financial economics.