Education Level and Income Disparities: Implications for Financial Inclusion through Mobile Money Adoption in South Africa

Nyoka, Charles Article Education Level and Income Disparities: Implications for Financial Inclusion through Mobile Money Adoption in South Africa Comparative Economic Research. Central and Eastern Europe Provided in Cooperation with: Institute of Economics, University of Łódź Suggested Citation: Nyoka, Charles (2019) : Education Level and Income Disparities: Implications for Financial Inclusion through Mobile Money Adoption in South Africa, Comparative Economic Research. Central and Eastern Europe, ISSN 2082-6737, De Gruyter, Warsaw, Vol. 22, Iss. 4, pp. 129-142, https://doi.org/10.2478/cer-2019-0036 This Version is available at: https://hdl.handle.net/10419/259221 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0 Comparative Economic Research. Central and Eastern Europe Volume 22, Number 4, 2019 http://doi.org/10.2478/cer‑2019‑0036 Education Level and Income Disparities: Implications for Financial Inclusion through Mobile Money Adoption in South Africa Charles Nyoka Ph.D.; Senior Lecturer; Department of Finance Risk Management and Banking University of South Africa, Pretoria, South Africa e‑mail: ny[email protected] or [email protected] Abstract Financial inclusion has recently become an issue of concern the world over for gov‑ ernments, policymakers, non‑governmental organizations (NGOs), and financial and non‑financial institutions alike. McKinnon (1973) and Shaw (1973), in seminal pres‑ entations, brought the world’s attention to the importance of an effective financial system for economic development. In recent years, there has been growing theoretical and empirical works showing the strong linkages between financial development with economic growth and poverty alleviation. After conducting statistical analysis using Stata version 14 for Windows with a mul‑ tivariate binary logistic regression modeling technique applied, this paper tested and concluded that there is a statically significant relationship between educational levels on the one‑hand and income levels on the other on the probability of one having a mobile banking account in South Africa. From a policy perspective, this information will assist policymakers in making more informed decisions with respect to education, and from the banking fraternity point of view it will help, them in the developments of products that are more in line with the population’s education and income levels. Keywords: financial inclusion, level of education, level of income, mobile banking JEL: G21, N3, N27 130 Charles Nyoka Introduction Financial inclusion has recently become anissue ofconcern the world over for gov‑ ernments, policymakers, non‑governmental organizations (NGOs), and financial and non‑financial institutions (Quaye etal.2014). According toUNDESA and UNCDF (2006), the concept offinancial inclusion encom‑ passes two primary dimensions: (1) that financial inclusion refers toacustomer having ac‑ cess toarange offormal financial services, from simple credit and savings services tomore complex ones such asinsurance and pensions; and (2) that financial inclusion implies that customers have access tomore than one financial services provider, which ensures avarie‑ ty ofcompetitive options. Understanding and identifying the factors which affect the level offinancial inclusion isimportant topolicymakers and the academic field aswell. Research onissues regarding economic growth was pioneered byearly economic theorists, who focused onshortages ofreal factors such asland and capital (e.g., ma‑ chinery), but not the finance and financial markets asbeing constraints oneconomic growth. Inthe early 1990s, theorists like Schumpeter (1991) brought tothe fore the im‑ portance offinancial intermediary services for innovation and economic growth. McKinnon (1973) and Shaw (1973), inseminal presentations, brought the world’s atten‑ tion tothe importance ofan effective financial system for economic development. Inrecent years, there has been growing theoretical and empirical works showing the strong linkages between financial development with economic growth and pov‑ erty alleviation. Banerjee and Newman (1993) reported onthe critical role that access tofinance played inenabling people toexit poverty byenhancing productivity. Bin‑ swanger and Khandker (1995) and Eastwood and Kohli (1999) investigated the impact ofIndian’s rural banks’ expansion program and found that rural poverty reduced and non‑agricultural employment increased. Burgess and Pande (2005) also echoed asimilar positive impact onpoverty reduc‑ tion because ofthe bank branching regulations ofIndia (between the 1970s and the 1990s) which required banks toopen four branches inunbanked locations for every new branch opened inan urban area. The study provides direct evidence that this expansion ofthe bank branch network had apositive impact onfinancial inclusion and, thereby, itled toaconsiderable de‑ cline inrural poverty (World Bank 2014). Beck etal. (2009) showed how well developed financial markets and accessible fi‑ nancial services all reduce information and transaction costs, and influence savings rates, investment decisions, technological innovations, and long‑run growth rates. According toFrost and Sullivan (2009), banking services are being viewed increas‑ ingly asapublic good that needs tobe made available tothe entire population with‑ out discrimination. Itis against this background that this study seeks toexamine the impact ofincome levels and education level onfinancial inclusion with respect tomobile banking and other modern financial instruments and gadgets. 131 Education Level and Income Disparities: Implications for Financial Inclusion… Inthe context ofSouth Africa, with its history ofapartheid, anunderstanding ofthe causes offinancial exclusion onboth its economy and population ismerited. Literature review The current drive towards financial inclusion isan initiative that started 20 to30 years ago (Asian Development Bank Institute 2014; Consultative Group toAssist the Poor (CGAP) 2010; Demirgüç‑Kunt and Klapper 2012a). Patel and Graham (2012) maintain that South Africa has along history ofeconomic exclusion ofthe majority ofthe pop‑ ulation through colonialism and apartheid. Preisendoerfer, Bitz, and Bezuidenhout (2014) argued that the black population ofSouth Africa has alow participation rate inentrepreneurial activities and alow level ofentrepreneurial ambitions due tothe apartheid regime. The challenges ofaccess tofinance and poor knowledge ofavailable formal financial institutions minimized the rate ofblack‑owned small business topar‑ ticipate inthe formal sector (Cant, Erdis and Sephapo 2014). Gertler and Rose (1991) argue that economic growth and financial sector devel‑ opment are mutually dependent. Other authors have emphasized that financial sec‑ tor policy can affect the pace ofeconomic development. King and Levine (1993) em‑ pirically investigated the dynamic link between financial innovation and economic development. They argue that financial institutions lower the social cost ofinvesting inintangible capital through the evaluation, monitoring, and provision offinancing services. Montiel (1994) also noted that economic growth could bespurred byinnova‑ tion infinancial development that improves the efficiency ofintermediation, thereby increasing the marginal product ofcapital and raising the savings rate. Levine etal. (2000) found that the exogenous components offinancial intermediary development were positively associated with economic growth. Inthe context ofSouth Africa, the inability ofblack‑owned businesses toaccess fi‑ nance has animpact onthe economic growth ofthe country asthe black population accounts for 90% inthe small and medium enterprise (SME) sector compared tothe larger sector (SEDA 2017). The OECD (2015) states that itis primarily the responsi‑ bility ofgovernments the world over toensure that small enterprises dowell, asgov‑ ernments carry the burden ofassessing the extent ofsmall enterprises financing needs and eliminating the gaps with relevant stakeholders, such asbanks and financial insti‑ tutions. According toFreeman (2008), one ofthe main issues that remain ofconcern isthat the apartheid government used legislation asone ofits main tools toprovide white South Africans with abundant resources and economic opportunities. Inspite ofthe racist division ofresources inpre‑ and post‑apartheid South Africa, the coun‑ try’s economy and societal well‑being have been drastically affected bythe lack ofac‑ cess tofinance bymost black‑owned small businesses. Anumber offactors can beattributed tothis demise, among them, education lev‑ els and income disparities, which are the subject ofexamination inthis paper. 132 Charles Nyoka Kira and He(2012) agree that the main challenge preventing the small enter‑ prise sector from contributing fully toeconomic growth isalack offinance. Quaye, Abrokwah, Sarbah, and Osei (2014) share the same view asthe researchers ofastudy inGhana, who found that most small enterprises who were denied access tocredit bycommercial banks and other financial institution collapsed within the first three years ofcoming tobeing. There isagreat deal ofliterature linked tothe subject ofaccess tofinance, especially inthe small enterprise space. Beyers and Ndou’s (2016) study addressed two main is‑ sues relating tothe growth and development ofsmall enterprises inSouth Africa, that is, the lack offinancial management skills and lack ofaccess tofinance. Kasseeah and Thoplan (2012) argue that the growth ofsmall enterprises isavital source ofwealth creation for both the economy and the individual.Itis implied, there‑ fore, that denying someone access tofinancial resources for whatever reason istanta‑ mount todenying both the economy and the individual access towealth. Since banks, which spur economic growth and affect economic transformation, are key institutions inany economy (Djoumessi 2009), exploiting the advantages brought byinformation and communication technologies (ICT) iscritical for economic development inthe future. However, due tothe challenges emanating from resistance tochange, lack ofinformation for consumers, and poor levels ofeducation and income bythose that are supposed topartake innew technologies, financial inclusion may remain apipe dream for most African economies. Other common causes offinancial exclusion (price ornon‑price barriers) cited inthe literature include “geography (limiting physical ac‑ cess), regulations (lack offormal identification proof orof appropriate products for poor households), psychology (fear offinancial institution’s staff, structures, compli‑ cated financial products, etc.), information (lack ofknowledge regarding products and procedures), and low financial acumen (low income and poor financial discipline), among others” (Ramji 2009; Demirgüç‑Kunt and Klapper 2012a). Demirgüç‑Kunt etal. (2012b) observed that without inclusive financial systems, poor people have torely ontheir own limited savings toinvest intheir education orbe‑ come entrepreneurs, and small enterprises must rely ontheir limited earnings topur‑ sue promising growth opportunities. The available literature suggests that retail banks are facing huge challenges inmi‑ grating customers from the traditional ways ofconducting business totake part inmodern ways ofconducting business (Singh 2004, p.188; Brown and Molla 2005). Research conducted concluded that differences inattitude might exist between cus‑ tomers indifferent demographic groups. However, demographic factors alone are in‑ sufficient predictors ofacustomer’s attitude. Financial inclusion must thus betreated like any public good due toits importance toboth individual and economic develop‑ ment. The degree of‘publicness’ infinancial inclusion may bedifferent from atypical public good like ‘policing’ According toCreane etal. (2004), arobust financial system contributes tothe eco‑ nomic and overall functioning ofthe national economy through promoting invest‑ 133 Education Level and Income Disparities: Implications for Financial Inclusion… ment and funding good business opportunities, mobilizing savings, and enabling the trading, hedging, and diversification ofrisk. This, inturn, results inamore efficient allocation ofresources, more rapid accumulation ofphysical and human capital, and faster technological progress, and finally, itfeeds into economic growth. Honohan and Beck (2007), among others, have shown that well‑functioning, healthy and com‑ petitive financial systems are aneffective tool increating opportunities and fighting poverty byproviding people with awide range offinancial services, such assavings, credit, payment, and risk management services. Invarious studies, ithas been observed that the absence ofinclusive financial sys‑ tems contributes topersistent income inequality and slower economic growth (Demir ‑ guc‑Kunt and Levine 2009; Kempson 2006). Problem statement and research objectives Inabid tofoster financial inclusion from apolicy perspective, the South African gov‑ ernment introduced anumber ofacts, among which isthe National Credit Act (NCA); among other rights, itenforces the right tocredit byall citizens. The government went onto introduce the Mzanzi account, anaccount with conditions for aminimum bal‑ ance and onwhich nocharges are tobe levied. All these happened inthe midst ofthe rapid introduction oftechnology bybanks. Inthe process, the government seems tohave underplayed the importance and impact oflevels ofboth education and income onfinancial inclusion. The FinMark Trust (2016) indicates that inthe Southern African Development Community (SADC) region, mobile money ownership remains low among low‑income groups and people with either noeducation orlow levels offinancial literacy because oflower education levels. Itis inlight ofthe above that this paper explores the impli‑ cations ofeducation level and income disparities onfinancial inclusion through mo‑ bile money adoption inSouth Africa. Research questions and hypothesis The paper seeks toanswer the questions ofthe effect that education levels and income disparities have onfinancial inclusion through mobile money adoption inSouth Af‑ rica. The research hypotheses developed asaresult are that: –Education levels disparities have statistically significant and disproportionate ef‑ fects onfinancial inclusion through mobile money adoption inSouth Africa. – Income disparities have statistically significant and disproportionate effects onfinancial inclusion through mobile money adoption inSouth Africa. 134 Charles Nyoka Contribution of the study This research study provides insights into the lack offinancial inclusion through mo‑ bile money adoption inSouth Africa with respect tothe existence ofdisparities ined‑ ucation and income levels among the country’s population. Itis also hoped that the research results may result inapolicy shift bygovernments and financial sector play‑ ers alike towards more consumer‑friendly policies and financial products ifthe dream offinancial inclusion isto berealized atall. Methodology and analytical technique Data description The research used data from The World Bank (WB) Microdata 2014 Financial Inclu‑ sion Survey. The data was sourced from the World Bank Microdata Library online data portal.From the total 1000 participants surveyed, 995 were valid responses relevant tothis research study, yielding a99.5% effective response rate. Mobile money account ownership status was the binary response variable (no=0; and yes=1), while educa‑ tion level and income quintile were the covariates. Education level had three catego‑ ries inthe model (completed primary orless=1; secondary=2; and completed ter‑ tiary ormore=3). Income quintile had five categories, namely poorest 20%=1; second 20%=2; middle 20%=3; fourth 20%=4; and richest 20%=5. Estimation technique Statistical analysis was conducted using Stata version 14 for Windows. Amultivariate binary logistic regression modeling technique was applied toestimate odds ratios with 95% confidence intervals based onthe function specified below: 0 1i i ii 0 1i exp( x ) ð Pr(Y 1 X x ) 1 exp( x ) aa aa + = = == ++ (1) 6. Methodology and Analytical Technique 6.1. Data description The research used data from The World Bank (WB) Microdata 2014 Financial Inclusion Survey. The data was sourced from the World Bank Microdata Library online data portal. From the total 1000 participants surveyed, 995 were valid responses relevant to this research study, yielding a 99.5% effective response rate. Mobile money account ownership status was the binary response variable (no=0; and yes=1), while education level and income quintile were the covariates. Education level had three categories in the model (completed primary or less = 1; secondary = 2; and completed tertiary or more = 3). Income quintile had five categories, namely poorest 20% = 1; second 20% = 2; middle 20% = 3; fourth 20% = 4; and richest 20% = 5. 6.2. Estimation technique Statistical analysis was conducted using Stata version 14 for Windows. A multivariate binary logistic regression modeling technique was applied to estimate odds ratios with 95% confidence intervals based on the function specified below: 0 1i i ii 0 1i exp( x ) π Pr(Y 1 X x ) 1 exp( x )        (1) i i 0 1i i π logit(π ) log x 1π            (2) where: Y represents the binary response variable (such that Yi=1 signifies the presence of mobile money account ownership, and Yi= 0 describes the absence of mobile money account ownership), and Xirepresents a vector of a set of exploratory variables which include the education level and income quintile. (2) where: Y represents the binary response variable (such that Yi=1 signifies the pres‑ ence ofmobile money account ownership, and Yi = 0 describes the absence ofmobile money account ownership), and Xirepresents avector ofaset ofexploratory variables which include the education level and income quintile. Todetermine the proportion ofoverall variation inmobile money account own‑ ership status that was accounted for byeducation level and income quintile, the Cox 135 Education Level and Income Disparities: Implications for Financial Inclusion… &Snell Pseudo R‑square and the Nagelkerke R‑square were calculated. The respective test statistics were performed based onthe functions specified below: 2 2 n null k 2 Cox & Snell Pseudo R LL 12LL æö -÷ ç÷ -ç÷ ç÷ çè =ø (3) ( ) ( ) 2 2 n null k null 2 2 n Na 2LL 2LL 1 agelkerke R which divides Cox & Snell R by i 2 ts m 1 LL m xi um æö -÷ ç÷ ç÷ ç = ÷ çèø - - = - (4) where: –2LLnull symbolizes the likelihood for the model with only anintercept; and –2LLk represents the model with the predictor. Toexamine the predictive power ofthe model, the area under the nonparametric Receiver Operating Characteristic (ROC) curve was computed. The respective curve, which isagraph ofsensitivity versus 1 minus specificity, was derived atc=0.5 prob‑ ability cutoff. Sensitivity refers tothe fraction ofobserved positive outcome cases that are correctly classified, and specificity isthe fraction ofobserved negative outcome cases that are correctly classified inthe analytical process. Results and analysis This section presents summary statistics onthe demographic profiles ofthe partici‑ pants and the mobile money account ownership status according toeducation level and income quintile. The cross‑tabulated frequencies and estimated odds results ofmobile money account ownership status for each education level group and income quintile group were provided. Table 1 above shows that from the total of995 respondents surveyed, approximate‑ ly 18% (n=183) reported they owned mobile money accounts, while 82% (n=812) ofrespondents reported that they did not have mobile money accounts. From the 18% (n=183) who had mobile money accounts, only 1% (n=13) had completed primary education orless, 12% (n=117) had completed secondary education, and 5% (n=183) had completed tertiary education ormore. The cross tabulation results presented inTable 2 above show that from the total of995 respondents surveyed, approximately 18% (n=183) reported that they owned mobile money accounts, while 82% (n=812) ofrespondents reported that they did not have mobile money accounts. From the 18% (n=183) who had mobile money ac‑ counts, 1% (n=14) were inthe poorest 20% income quintile category, 2% (n=17) were 136 Charles Nyoka inthe second 20% income quintile category, 3% (n=25) were inthe middle 20% in‑ come quintile category, 4% (n=36) were inthe fourth 20% income quintile category, and 9% (n=91) were inthe richest 20% income quintile category. Table 1. Respondent education level * Has a mobile money account Has a mobile money account Total No Yes Respondent education level completed primary or less Count 211 13 224 % of Total 21.2% 1.3% 22.5% secondary Count 539 117 656 % of Total 54.2% 11.8% 65.9% completed tertiary or more Count 62 53 115 % of Total 6.2% 5.3% 11.6% Total Count 812 183 995 % of Total 81.6% 18.4% 100.0% Source: own elaboration. Table 2. Within‑economy household income quintile * Has a mobile money account Has a mobile money account Total No Yes Within‑economy household income quintile 1 poorest 20% Count 151 14 165 % of Total 15.2% 1.4% 16.6% 2 second 20% Count 172 17 189 % of Total 17.3% 1.7% 19.0% 3 middle 20% Count 144 25 169 % of Total 14.5% 2.5% 17.0% 4 fourth 20% Count 169 36 205 % of Total 17.0% 3.6% 20.6% 5 richest 20% Count 176 91 267 % of Total 17.7% 9.1% 26.8% Total Count 812 183 995 % of Total 81.6% 18.4% 100.0% Source: own elaboration. The odds ratios (Table 3) are all statistically significant atthe 5% level and lie with‑ in the respective 95% confidence intervals. The results indicate that respondents who completed tertiary education ormore had approximately eight times the odds ofhav‑ ing amobile money account than respondents with primary education orless. Simi‑ larly, respondents who completed secondary education had approximately three times the odds ofhaving amobile money account than respondents with primary educa‑ tion orless. Concerning the within‑economy income quintile, respondents who are