Financial Inclusion and Income Inequality in Nigeria

: Studies show that the low level of economic development in countries occasioned by high income inequality levels may be addressed by financial inclusion policies via providing increased access and availability of formal financial services to the financially excluded segment of the economy. This paper investigates the extent to which this is applicable in Nigeria using the Auto regressive distributed lag estimation technique on time series data for the period 1985 to 2019. The results, consistent with the special agent theory of financial inclusion, show that financial inclusion significantly reduce the income inequality gap in Nigeria. Furthermore, the results also suggest that education and economic growth lowers the gap while population widens the gap. Based on this, the study recommends that policy makers in Nigeria put in more efforts to achieve the set target rate of financial inclusion as well put in place population reducing polices to further reduce the gap while promoting policies that will actualise the potentials of the other indices.


Introduction
One of the goals of the United Nations Sustainable Development Goals (UN-SDGs) that has gained recent and significant attention is Goal 10 which has the objective of reducing inequality within and among countries.This primarily stems from literature that shows that countries with high level income inequality gap tend to have low investment in infrastructural development, education, and technology thus, slowing down economic development in such countries (Altunbaş, Thornton, 2019;Demir et al., 2022;Doyle, Stiglitz, 2014).Additionally, it is observed that countries have a distinct force that shape the level and impact of inequality making it necessary for each country to require an independent assessment (Doyle, Stiglitz, 2014).Differences that arise from this assessment show that inequality is determined by economic forces, polices and politics of individual countries.Consequently, a number of countries introduced measures / polices that have the potential to reduce inequality and promote development as well as social benefits (Easterly, 2007).One of such measure is the introduction of polices to reduce financially excluded economic units as well as increase the ratio of people having access to formal financial services (Kling et al., 2022;Nandelenga, Oduor, 2020).This in turn is expected to increase the economic potential of citizens that will empower them to earn more and reduce income inequality.
A cursory observation of the income inequality measure (proxied with the Gini coefficient) for Nigeria in Figure 1 shows a decline over time.Despite an increase from 41.9% in 1985 to 42.8% in 1996, subsequent years showed a gradual decrease to 39.6% in 2019.Igwegbe and Amaka (2021) recently noted that income inequality in Nigeria may be observed from the daily struggles of majority of the population despite the accumulation of obscene wealth by few individuals.This position affirms Oxfam's 2017 report that the richest man in Nigeria will expend all his wealth in 42 years by spending 1 million Naira per day despite the existence of over 100 million people experiencing poverty.In addition, the money the richest man in Nigeria earns annually is sufficient to uplift two million people from poverty for a year (Deinne, 2023).These assertions clearly show the extent of income inequality in Nigeria and of more serious concerns, are the adverse social and political effects that may arise due to rising inequality.

Figure 1. Income inequality in Nigeria (1985-2019)
Financial inclusion is defined by Allen et al. (2012) as the access to and use of formal financial services that has the objective of increasing the access of financially excluded units to formal financial services and products.Sarma and Pais (2011) further notes that it implies the ease of access and use of formal financial services and products at affordable costs.These costs arise from imperfect market conditions that give rise to information asymmetries and costly transaction fees making it expensive for economic units to avail themselves of financial services and products thereby increasing the income inequality gap (Banerjee, Newman, 1993).
The extent to which financial inclusion reduces the inequality gap is not clear cut.On one hand, some studies argue for the potential of financial inclusion in reducing the gap by empowering economic units to increase their household and firm income.This occurs when they benefit from the use of formal financial services and products (e.g., credit) at affordable costs and are able to undertake productive economic activities (Agu et al., 2023;Aslan et al., 2017;Park, Mercado Jr, 2018;Polloni-Silva et al., 2021).On the other hand, literature also show instances of financial inclusion not having any effect on income inequality and also increasing the inequality gap.For example, Kochar (2011) demonstrates the no evidence stand for India while Honohan (2008) shows the increased stand.Contributing to the inconclusiveness debate, Yang and Zhang (2020) show the time varying effect of financial inclusion on the income inequality gap.Their results reveal an increase in urban-rural inequality in the short-run and a decrease in the long-run given improvement in financial infrastructures and literacy in the rural areas.
Nigeria, as with other developing countries has over the years made concerted efforts at boosting financial inclusion to reduce the income inequality gap.In 2012, the country launched a National Financial Inclusion Strategy (NFIS) to provide a framework for an enabling environment for financial inclusion to thrive.The framework was later revised in 2019 with a target of achieving an inclusion rate of 80%.In 2023, the Central Bank of Nigeria observed that despite the progress made since the introduction of the NFIS, the target was not achieved as the inclusion rate stood at 64%.Series of measures introduced to achieve the set target ranged from institutional involvement to specific policies and programmes.These include but are not limited to the introduction of non-interest financial services and products to capture individual religious and ethical beliefs, financial literacy campaigns, streamlining of transaction costs, establishment of payment services banks, establishment of digital and agency banking etc.However, the extent to which these measures have reduced the income inequality gap in Nigeria when the policy target is at the macro-level is not clear.Studies such as Churchill et al. (2020); Ibrahim and Aliero (2020); Ibrahim et al. (2019) provide micro-level evidence of the impact of financial inclusion on income inequality but are silent on macro-level evidence.The present study is one of the first papers that addresses this concern through the use of macro-level measures of financial inclusion from the supply-side view to investigate the income inequality gap in Nigeria.The use of macro-level data to capture financial inclusion would show the extent to which borrowing costs are reduced and the efficiency of capital allocation in lowering the income disparity gap (Alshubiri, 2021;Domanski et al., 2016).Furthermore, the study provides a broader context of the financial intermediation roles of financial institutions in reducing the gap thus allowing policy makers to introduce / implement measures that would lead to income convergence in Nigeria.
The remaining part of this study is structured as; Section 2 provides the review of literature, Section 3 describes the data and methodology employed in the study, Section 4 discusses the results while Section 5 concludes with relevant policy implications.

Theoretical consideration
Theoretically, imperfections in financial markets limit access by the poorer population to formal financial services and products that could have otherwise engaged them in productive commercial activities (Aghion, Bolton, 1997;Banerjee, Newman, 1993;Galor, Zeira, 1993).The models in these studies submit that financial inclusion may lower income inequality by making available financing opportunities for entrepreneurship and provide education to the poor ultimately enhancing household / business income.Information asymmetries, rigid rules and high transaction costs that prevent this section of the population from accessing credit markets to finance viable business activities are eliminated via measures undertaken to promote financial inclusion (Nandelenga, Oduor, 2020).The special agent theory of financial inclusion emphasizes the role played by specialised financial agents, such as banks, in reducing the difficulties encountered in the use of formal financial services and products by delivering affordable and accessible financial services to excluded economic units in the society.These agents have information regarding the peculiarities of the excluded as well as possessing the structures and infrastructures needed for implementing financial inclusion measures to achieve the desired objectives (Ozili, 2020).The expectation from the theory is that increase in the level of financial inclusion would lead to excluded units having increased access to formal financial services and products.This would enable them to be involved in productive economic activities to improve their household income and thus reduce income inequality.In this regard, this paper addresses the hypothesis that financial inclusion lowers income inequality in Nigeria.

Empirical Evidence
The empirical literature on financial inclusion and income inequality is mostly concentrated in developing regions of the world due to the low level of development in these areas and the potential benefits such countries may derive from introducing financial inclusion policies.For instance, in a cross-country study of 13 Latin American countries, Polloni-Silva et al. (2021) argue that financial inclusion measured as an index of micro-level variables covering the availability, penetration, and usage dimensions reduced income inequality in the sampled countries.The results obtained from the investigation support the argument that financial inclusion is a necessary policy measure that can improve the welfare of the citizens in the 13 countries.In a similar study of Sub-Saharan African countries, Nandelenga and Oduor (2020) show that financial inclusion which improves availability and access to financial services ultimately leads to increased investment and employment opportunities.This in turn reduces the income inequality disparity in the short and long run.Omar and Inaba (2020).However, the extent to which the results may be generalised to individual countries in these studies was not accounted for given the heterogenous level of development of individual countries and the existence of distinct economic policies and forces that shape financial inclusion polices in different countries as observed by Doyle and Stiglitz (2014).
Nevertheless, not all studies show that financial inclusion lowers income inequality.Kochar (2011) for instance did not find evidence suggesting that increased access to formal services causes a reduction in household income inequality in India.In addition, Honohan (2008) show that financial inclusion when investigated with other control variables increased the income inequality gap.Extant literature such as Salazar-Cantú et al. ( 2015) and Yang and Zhang (2020) also show evidence of the time varying nature of the effect of financial inclusion on income inequality.These studies show that in the short-run, increased financial inclusion levels initially led to increased income inequality, but in the long-run, income inequality is reduced with increased levels of financial inclusion.
In the Nigerian context, existing studies used micro-level variables to investigate financial inclusion and income inequality gap.Ibrahim 2019) noted that for financial inclusion to have the desired effect on income inequality gap, financial institutions (a macro-level variable) need to play a more prominent role in making credit accessible and available to the low-income cadre of the society.However, the authors did not provide evidence to back up their observation.The present study extends the earlier Nigerian ones with the use of financial institutions data to capture financial inclusion and show the extent to which financial intermediaries in channelling resources to economic units, influence the income inequality gap in Nigeria.

Data
Following the work of Demir et al. (2022), data for income inequality are obatined from the Standardised World Income Inequality database (SWIID) and the Gini coefficient of disposable income is used as the measure for the dependent variable (income inequality).The database has the advantage of standardising the Gini coefficient as well as minimising reliance on problematic assumptions while making use of as much information as available in immediate years from the same country (Solt, 2009).The Gini coefficient ranges from 0 to 100.The main independent variable of interest (financial inclusion) is sourced from the Financial Development and Structure Database of the World Bank and is proxied using two different measures; credit to the private sector as a percentage of GDP and liquid liabilities as a percentage of GDP i .These two macro variables measure the key role of financial intermediaries in channelling of resources to economic units from the supply view as well as the availability and accessibility of financial access and depth (Afonso, Blanco-Arana, 2024; Nandelenga, Oduor, 2020; Yang, Zhang, 2020).Thus, they are expected to enhance income inequality convergence.
In line with previous studies, control variables that affect income inequality are sourced from the World Bank Development Indicators (WDI).These include population, education, growth, government expenditure, and trade.Population proxied as the annual percentage in population has been found to increase inequality (Demir et al., 2022; Lacalle-Calderon et al., 2019).In contrast, an inverse a priori expectation is hypothesised for education, growth, government expenditure and trade which has been found in literature to reduce income inequality.Education increases the awareness of citizens to profitable economic activities that increase income and reduce inequality (Bolarinwa, 2023; Lacalle-Calderon et al., 2019; Lee, Lee, 2018; Park, Mercado Jr, 2018).Trade opens up avenues for economic units to convert their produce / products to wealth thus reducing inequality (Demir et al., 2022;Furceri, Ostry, 2019).Growth proxied with GDP per capita reduces income disparity through increase in income of the poor via benefits derived from economic development (Beck et al., 2007;Demir et al., 2022).Data for all variables are from 1985 to 2019 and are based on data availability for the dependent variable.E-views software was used for the analysis in this study.Table 1 presents the variables description and data source.Data period for all variables are from 1985 to 2019, the last available date for the Gini coefficient.This is the summation of exports and imports of goods and services measured as a share of gross domestic product.

World Bank Development Indicators
Source: literature reviewed in Section 2; * See Footnote 1.

Model Specification
In determining the relationship between financial inclusion and income inequality in Nigeria, the Linear Autoregressive Distributed Lagged (ARDL) model of Pesaran et al. (2001) is used to analyse Eq. (1): Where INEQ refers to income inequality, FININCL is the measure for financial inclusion, POP is annual population growth rate, EDU is gross enrolment rate of primary education, GRW is annual GDP per capita growth rate, EXP is general government final consumption expenditure, TRD is trade openness, t is time in years and β are coefficient estimates.The variables are as described in Table 1 while Eq. ( 1) illustrates the long-run relationship of the variables.The financial inclusion variable (FININCL) is proxied by two measures -LIQLIB which is the main variable in Eq. ( 1) and PRC which is used to show the robustness of the results if a different measure is used for financial inclusion.
The short-run relationship is depicted in the error-correction mechanism in Eq.( 2): ε is the error term, γ are parameter estimates and ∆ is the first difference operator.

Results and Discussions
The pairwise correlation matrix of the variables are presented in Table 2.All variables except for government expenditure (EXP) are observed to have low values suggesting the absence of multicollinearity issues.Accordingly, EXP is dropped from the regression equation.The results reported in Table 4 indicate that the null hypothesis of no cointegration relationship is rejected due to the value of the F-statistics (4.319510) being higher than the critical value of the I(1) bound (3.79) at 5% level of significance (see Pesaran et al., 2001).This further suggests the presence of a long-run cointegrating relationship between the variables in the study.Thus, the ARDL regression technique is used to estimate both the long-run and short-run estimates of Eq. ( 1) and Eq. ( 2) respectively.Source: Author's computation Table 5 presents the LIQLIB long-run and short-run regression results from the ARDL model.The independent variable of interest, LIQLIB, which proxies financial inclusion is statistically significant at 5% level of significance with a negative coefficient ( -0.272751).This implies that increase in the financial inclusion rate in Nigeria significantly reduces income inequality.This result is not surprising given the various policy measures implemented over the years to reduce the percentage of financially excluded people in Nigeria as expounded in the introductory section.Some of these measures include the introduction of agency banking, mobile banking, the expansion of digital financial services among others.The results suggest that financial institutions while channelling resources to economic units reduce the income inequality gap by providing access to formal financial services at affordable costs.This outcome is consistent with the studies of Alshubiri Education, growth and trade in Table 5 have the expected negative coefficients but they are not statistically significant indicating that although they do not contribute to reduction in income inequality in Nigeria, they show potentials of reducing it.Thus, policy makers need to put in place more efforts / measures to realise the desired outcome of a reduction in income inequality.For instance, more funds could be allocated to the education sector to improve literacy level because over years the percentage of the national budget allocated to the education sector has not been encouraging.In addition, efforts should also be geared towards attaining a healthy and sustainable economy in order to increase GDP per capita thus reducing the income inequality gap.The promotion of trade via access to international trade opportunities and a financial system that removes impediments in international transaction and payments may also lead to enhanced welfare for economic units that would reduce income inequality.
The results from the short-run estimate showed that the error correction term (CointEq.(-1)) has the expected negative and statistically significant sign (-1.223390; p<0.05).This suggests the presence of a correcting mechanism from short-run disequilibrium to the long-run equilibrium in the regression equation.1 The Breusch-Pagan tests for serial correlation and heteroskedasticity both report non-significant pvalues at 5% level of significance.These suggests a model that is free from serial correlation and heteroskedasticity issues, thus validating the regression estimates.Similarly, the CUSUM tests to check the stability of the long-run regression estimates show the plots are within the two critical bounds suggesting stability of the model as shown in Figure 2.

Robustness (Credit to the private sector)
The results presented in Table 6 for cointegration give a similar outcome of the presence of a cointegrating relationship among the variables.The F-statistics of 3.990136 is higher than the I(1) bound of 3.79 at 5% level of significance.Consequently, the long-run relationship is estimated with the same ARDL technique.For robustness, Table 7 reports long-run and short-run parameter estimates where financial inclusion is proxied with credit to the private sector (PRC).As obtained in Table 5 for LIQLIB, PRC is seen to be negative and statistically significant at 5% level of significance (-0.513119).Similar with the results obtained with LIQLIB, using PRC as the proxy for financial inclusion suggests the reduction in income inequality in Nigeria.However, unlike LIQLIB, education (PRM) and growth (GDP) are statistically significant at 5% level of significance with negative coefficients (-5.677171; -8.641411 respectively).These suggest that the two variables also contribute to lowering of income inequality gap in Nigeria.-1)) with the expected negative and statistically significant sign (-0.121755; p<0.05) implying an adjustment process from the short-run to the long-run.The diagnostics tests also confirm the validity of the regression estimates being free from serial correlation and heteroskedasticity while stability of the regression estimates is confirmed from the CUSUM test in Figure 3.

Conclusion, Policy Implications and Direction for Future Research
In countries with a high level of income inequality, a slowdown in economic development is common which subsequently leads to a high number of poor citizens.To address this, the United Nations SDGs has one of its goals as the reduction in income inequality within and across countries and a number of countries have tried to pursue this goal.One of the ways through which this is being pursued is the introduction of financial inclusion policies to increase the availability and accessibility of formal financial services to excluded economic units in countries where the polices are introduced.With increased access to these services, it is expected that concerned units will engage in productive activities that will lead to increased income and thus a better welfare.The results from the analysis in this paper show that financial institutions in Nigeria while performing the financial intermediation role, are able to reduce the inequality gap by making available and providing access to formal services.These results are in line with the theoretical expectation of the special agent theory of financial inclusion put forward by Ozili (2020).Additionally, the findings are also consistent with studies such as Alshubiri who argued that access to affordable financial servceis and products improve household income via engagement in productive economic activities in developing economies.Policy wise, it is recommended that increased efforts be made by regulatory authorities in Nigeria towards achieving the 80% target of financial inclusion as this would help in reducing the income inequality gap.In addition, policy makers should also put in more efforts on other important indices such as education, trade, and GDP in reducing the gap because the results show that they have the potential to do so.Such efforts can include increased national budgetary allocation to education, removal of impediments in international trade transactions etc.This study highlights important limitations that future studies may build upon of which one of them is the data period.The present study is unable obtain data post 2019 for the inequality variable thereby failing to show the most current position on the financial inclusion and income inequality gap.Another limitation is the one country study of Nigeria.Future studies may improve on this by examining a larger In 22 transition economies, Le et al. (2019) also provide evidence of the existence of a negative relationship between financial inclusion index and income inequality.The authors demonstrate that accessibility and availability of low-cost formal financial services which increase financial penetration in rural areas, enhance financial responsibility and management in such areas thus improving the standard of living and leading to income inequality convergence.Similar studies in developing regions include Agyemang-Badu et al. (2018); Alshubiri (2021); Chinnakum (2023); Chinoda and Mashamba (2021); Cicchiello et al. (2021); Khan et al. (2022); and Aliero (2020) using household survey data from 2010 to 2016, show that financial inclusion in Nigeria reduces the gap.Ibrahim et al. (2019) and Churchill et al. (2020) also used micro-level variables for financial inclusion and reported similar results.However, Ibrahim et al. ( (2021); Chinoda and Mashamba (2021); Cicchiello et al. (2021); Khan et al. (2022); Omar and Inaba (2020).Furthermore, it is observed that the results do not differ from the micro-evidence studies of Churchill et al. (2020); Ibrahim and Aliero (2020); and Ibrahim et al. (2019) who show that micro-level financial inclusion measures lead to income inequality convergence.With respect to control variables, only population is observed to have a positive and statistically significant effect on income inequality (12.540326; p<0.05).This indicates that an increase in population without a corresponding growth in the wellbeing of Nigerians would lead to increase in inequality gap consistent with the findings of Demir et al. (2022) and Lacalle-Calderon et al. (2019).

Table 2 . Pairwise correlation
Source: Author's computation; Variables are as described in Table1

Table 3
reports the results of Augmented Dicky Fuller (ADF) and Phillips-Perron (PP) unit roots test for stationarity of variables.The table shows a mixed order of integration of I(0) and I(1) variables for both tests which further lends support for the use of the ARDL model.It also suggests the presence of a cointegrating relationship among the variables.Accordingly, a cointegration test is conducted using the ARDL bounds test procedure.

Table 3 . ADF Unit Root Test Results Source
: Author's computation; Variables are as described in Table1

Table 7 . Regression and Diagnostics Tests results (PRC)
Demir et al. (2022) and growth in the economic strength of the country, more people are empowered economically to engage in productive activities that increases household income and reduce the gap as argued byBolarinwa (2023)andDemir et al. (2022).The short-run results similarly show an error correction mechanism (CointEq.(