0
MOBILE BANKING IN
SUB-SAHARAN AFRICA: DIFFERENCES
IN ADOPTION AND USE
Anthony Oweke Msc Political Economy University of Amsterdam Student Number: 12088773
Source: Global Findex Database and Jupiter
1
LIST OF TABLES AND FIGURES ... 3
1. INTRODUCTION ... 4
2. LITERATURE REVIEW ... 7
2.1 INNOVATION ...7
2.1.1 Supply-Side Factors ...7
2.1.2 Demand Side Factors ...8
2.1.3 Structural Factors ...9
2.2 MOBILE BANKING AS AN INNOVATION ... 10
2.2.1 Delivery ... 11
2.2.2 Environmental/Structural Factors ... 11
2.2.3 Impacts of Mobile Banking Innovation ... 12
2.3 CONCLUSION ... 15
3. METHODOLOGY ... 17
3.1 RESEARCH DESIGN:COMPARATIVE CASE STUDY ... 17
3.2 COUNTRIES ... 17
3.3 VARIABLES ... 18
3.3.1 Dependent Variables: Mobile Banking Adoption and Mobile Banking Usage... 19
3.3.2 Independent Variables: Supply, Demand, and Structural Factors ... 20
3.4 CONCLUSION ... 22
4. RESULTS CHAPTER: REGRESSION ANALYSIS ... 24
4.1 DESCRIPTIVE STATISTICS ... 24
4.2 MOBILE BANKING ADOPTION ... 29
4.2.1 Model 1: Mobile Subscriptions and Mobile Banking Adoption ... 31
4.2.2 Model 2: The Rule of Law and Mobile Banking Adoption ... 31
4.2.3 Model 3: Formal Account and Mobile Banking Adoption ... 32
4.2.4 Model 4: All Independent Variables and Mobile Banking Adoption ... 33
4.3 MOBILE BANKING USE ... 35
4.3.1 Model 1: Mobile Subscriptions and Mobile Banking Use ... 36
4.3.2 Model 2: The Rule of Law and Mobile Banking Use ... 37
4.3.3 Model 3: Formal Account and Mobile Banking Use ... 38
4.3.4 Model 4: All independent Variables and Mobile Banking Use ... 39
5. CONCLUSION ... 41
BIBLIOGRAPHY ... 46
2
APPENDIX 2: ORIGINAL SAMPLE OF COUNTRIES ... 51
APPENDIX 3: FINAL SAMPLE OF COUNTRIES ... 52
APPENDIX 4: CORRELATION COEFFICIENTS FOR DEPENDENT VARIABLES TO FACILITATE MERGING... 53
APPENDIX 5: TESTS FOR MULTICOLLINEARITY FOR THE INDEPENDENT VARIABLES... 54
APPENDIX 6: REGRESSION RESULTS FOR MOBILE BANKING ADOPTION ... 55
3
List of Tables and Figures
Table 1: Overview of the theoretical model ... 16
Table 2: Operationalization of the variables within the theoretical framework to answer the research question(s). ... 23
Table 3: Descriptive statistics for all variables ... 25
Figure 1: Country scores for the variable Mobile Banking Adoption ... 25
Figure 2: Country scores for the variable Mobile Banking Use ... 26
Figure 3: Country scores for the percentage of the population with a mobile subscription .. 27
Figure 4: Country scores for the Rule of Law estimate... 28
Figure 5: Country scores for the percentage of the population with an account at a formal institution ... 28
Table 4: Regression coefficients for independent variables to Mobile Banking Adoption ... 30
Figure 6: Scatter plot for Mobile Subscriptions and Mobile Banking Adoption ... 31
Figure 7: Scatter plot for The Rule of Law and Mobile Banking Adoption ... 32
Figure 8: Scatter plot for Formal Account and Mobile Banking Adoption ... 33
Table 5: Regression coefficients for independent variables to Mobile Banking Adoption ... 36
Figure 9: Scatter plot for Mobile Subscriptions and Mobile Banking Use ... 37
Figure 10: Scatter plot for The Rule of Law and Mobile Banking Use ... 38
Figure 11: Scatter plot for The Rule of Law and Mobile Banking Adoption ... 39
Table 6: Regression coefficients for independent variables to Mobile Banking Adoption ... 42
4
1. Introduction
Mobile banking or e-money initiatives have proliferated within the continent as an alternative to traditional banking services. In contrast to banks, these initiatives can reach the rural segments of the population, who lack access to traditional banks for various reasons. In some countries such as Kenya, mobile banking has been extremely successful and evolved from a simple peer-2-peer exchange (p2p) mechanism to provide a broad range of services such as paying for utility bills, goods, and services, as a storage mechanism and providing international transfers. Essentially, mobile banking in Kenya has surpassed traditional banks as a financial instrument for the average individual and accounts for 85% of Kenya's GDP as of 2016 (Lashitew et al., 2019, pg.1201). In other countries within the region, mobile banking initiatives have been markedly less successful. Although the scholarship on mobile banking in Sub-Saharan Africa has grown over the last decade, scholars have advocated for research aimed at explaining the differences in the success of mobile banking operations across the continent (Ibid., pg.1202).
In attempting to understand this disparity in the adoption and success of mobile banking initiatives, this thesis seeks to employ a quantitative research design. Specifically, the
research question asks: how can the differing rates of use and adoption of Mobile banking innovations across sub-Saharan Africa be explained? The factors to be examined are conditioned by the literature on innovation and are split into three: supply, demand, and structural factors. In order to test the statistical relationship between the variables mentioned above, 13 countries in Sub-Saharan Africa comprise the sample. These include Burkina Faso, Chad, Cameroon, Ghana, Kenya, Madagascar, Mozambique, Nigeria, Rwanda, Togo, Uganda, Zambia, and Zimbabwe. This introduction chapter commences with a brief introduction to the topic at hand and concludes with an overview of the chapters ahead.
The development of so-called "Least developed countries" or "Third World' or similar euphemisms utilized to describe the countries inhabiting regions as diverse as Latin America, Asia, and Africa, has preoccupied the global socio-political and economic agenda for the past 40 years. In this sense, there has been a distinctive effort on behalf of countries from the West and global institutions to "aid" countries form the regions mentioned above in their development based on a particular ideology. This ideology can broadly be defined as
5 neoliberalism and, in terms of reform, is operationalized as the Washington consensus. Oversimplified, this refers to the notion that the state should play a negligible role in the affairs of the market, rather seek to create opportunities to enact policies that stimulate innovation and, therefore, growth. In this spirit, enormous amounts of aid have been provided to countries in Sub-Saharan Africa but has arguably been ineffective. Amongst the various obstacles to stimulating economic growth in Sub-Saharan Africa, the issue of
financial inclusion has been ever-present. The World Bank argues in a recent report that financial inclusion is key to facilitating economic growth through providing a household with more economic opportunities, thereby reducing poverty and economic inequality (Riley and Kulathunga, 2017, pg.XI). Financial exclusion generally refers to the inaccessibility of
financial services related to geographic constraints (Kablana and Chhikara, 2013, pg. 104). This may include inaccessibility to banks and ATMs. Only about 46% of the world's
population or 4.3 billion people are estimated to have access to financial services (Riley and Kulathunga, 2017, pg. 23). In sub-Saharan Africa, only 20% or 295 million people have access to financial services (Ibid).
Various initiatives, externally and internally driven, have attempted to improve financial inclusion on the continent with varied success. Taking into consideration that about 60% of Africa's population resides in rural dwellings that are inaccessible to banks (Lashitew et al., 2019, pg.1203), the need for innovative strategies such as e-money is apparent. Concomitantly, the spread of mobile phone usage and access on the continent over the last two decades has been astounding. From 2000 to 2010, mobile phone
subscriptions jumped from around 11 million to 246 million (Subramaniam, 2013, pg. 6). This spread has corresponded with the use of mobile banking innovations, aimed at addressing the high costs and inaccessibility associated with formal banks (Myovella et al., 2020, pg.2). While 81% of countries in sub-Saharan Africa have adopted some form of mobile money, the success of such programs has varied (Lashitew et al., 2019, pg.1203). In Kenya, for instance, the inception of Mpesa, an e-money initiative by Safaricom, has had a profound impact on economic performance. Various studies have illustrated that Mpesa: comprises up to 85% of Kenya’s GDP (Ibid), has led to an increase in GDP per capita in the country (Ibid), has led 2% of Kenya’s population out of extreme poverty (ibid., 1203), and has increased domestic savings by 32% (Subramaniam, 2013, pg.3), amongst other
6 economic impacts. In other countries, such as in South Africa, the introduction of mobile banking innovations has markedly been less successful (Baer et al., 2018, pg.520). Innovations such as mobile banking have the potential to expand opportunities for individuals and firms, increasing investment, saving, and consumption.
The proliferation of mobile technologies on the continent, such as mobile banking, has coincided with increased scholarly contributions to this effect. Understandably given its nascent stage, the academic scholarship on mobile banking suffers from several obstacles. For one, the majority of the studies tend to focus solely on Kenya as a case study, which given its relative success to other initiatives, makes Kenya a standpoint upon which to examine other cases. Concomitantly, data on mobile banking in Kenya is generally accessible through the network operator Safaricom on their Mpesa service, which is not always the case with other countries. This has not prevented other scholars from assessing other countries, but some such as Duncombe (2012, pg.386), who researched Uganda, find the inaccessibility of data as prohibitive. Likewise, the empirical literature on the continent tends to focus on micro-scale usage or socio-economic impacts solely within Kenya. There is a paucity of studies that seek to explain differences in mobile banking innovations across countries in the continent.
This thesis attempts to contribute to the existing literature on mobile banking; specifically, this thesis seeks to explicate for the differences in the use and adoption of Mobile Banking initiatives in 13 sub-Saharan African countries. The relationship, if any, between mobile banking and the supply, demand, and structural factors, will be assessed utilizing regression analysis. To this effect, the subsequent chapter will provide a review of the relevant literature for this study. Chapter 3 will provide an overview of the methodology employed in the study, including data to be collected, an explanation of the variables
chosen, countries that form the sample, and the method of analysis. The results of the multivariate regression analysis on the relation between mobile banking and supply-side factors and mobile banking and demand-side factors will be provided in chapter 4. Finally, a synopsis of the findings of this study and its implications will be provided in Chapter 5.
7
2. Literature Review
In order to answer the research questions, several concepts and themes need to be unpacked. For one, innovation and the sources thereof are pertinent to understanding the drivers of mobile banking in sub-Saharan Africa. A definition to be utilized for the study will be selected, and the three drivers of innovation in the literature will be reviewed. Secondly, the literature on mobile banking is assessed to identify under what conditions this
innovation occurs in order to identify gaps that this thesis hopes to contribute to. To this end, section 2.1 will discuss innovation, section 2.2 will provide an overview of the literature on mobile banking, and lastly, a synopsis which highlights the chosen aspects of the
literature reviewed that are pertinent for the study will be provided in section 2.3.
2.1 Innovation
Innovation has long been considered as an essential component in achieving economic growth and development. Due to its multi-faceted nature, there is a lack of consensus within the literature regarding a definition. Moreover, as a concept, it is difficult to
conceptualize and operationalize within empirical studies (Kalcheva et al., 2018, pg.441). For this study, innovation is defined according to Schumpeter (1939) as “any ‘doing things differently in the realm of economic life” (Ibid). Scholars also differ concerning the factors they attribute as responsible for stimulating innovation. The literature can be divided into three camps: scholars who focus on supply-side factors, scholars who focus on the demand side, and scholars who propose structural factors as stimuli for innovation.
2.1.1 Supply-Side Factors
Scholars who focus on supply-side factors as drivers for innovation emphasize that firms and entrepreneurs play key roles in driving innovation. In this sense, firms introduce
changes to an established process, which in turn leads to changes in output or cost.
Audretsch and Thurik (2001, pg.277), for instance, argue that firms, through investment into research and development, promote "technical advance" and, therefore, innovation within a particular industry. Weitzman (1998, pg.359), through his contribution, argues that the stock of knowledge and specifically that the limits to economic growth lie in a firm's ability
8 to transform ideas into usable forms. In other words, new ideas and novel ways of
transforming old ideas are critical to long term economic growth. Essentially, this refers to improvements in technology and ways of utilizing technology that, in turn, cultivate long term economic growth.
2.1.2 Demand Side Factors
As regards the demand side, scholars argue that innovation is a factor of innovators and entrepreneurs responding to the demands of the market. Scherer and Schmookler (1982, pg.226) outlining the earlier seminal contribution of Schmookler (1966), postulates for this position. Through an analysis of patents, his results indicated that innovation might occur as a result of changes to demand (Ibid). In this sense, the market stimulates firms or
entrepreneurs to change or produce new forms or products to respond to demand. Priem, R. L. et al. (2012) similarly propose increased scholarly attention on the efficacy of demand-pull explanations for stimulating innovation based on several articles reviewed. They argue that technology-push explanations of innovation, tend to assume customer or user needs are static, whereas demand-pull approaches perceive these needs as diverse and the market as encompassing heterogeneity (Priem et al., 2012, pg. 351). Innovation in this context can be driven by changes in demand for products by consumers resulting in
decisions by managers or a result of innovations produced by users themselves or ‘users as innovators’ (Priem et al., 2012, pg. 349). In some instances, thus, demand-pull innovation works "along with, or instead of, technology push” (Priem et al., 2012, pg. 364), in
stimulating innovation.
Some scholars have further provided evidence that consumers themselves may be innovators as they adapt current products to fulfill their otherwise unmet demands. Van Der Boor et al. (2014, pg.1595), for instance, argue that in the Philippines, consumers utilized mobile phones to transfer airtime to each other, which had not been previously facilitated by the network operator. The network operator, upon discovering this, provided a service that allowed for airtime transfer (Ibid). The demand for this otherwise unprovided service stimulates the network provider to provide a service. In other words, demand drove innovation in this case.
9 In recent decades it has become apparent within the academic scholarship, that in fact, supply-side factors prevail over demand side in driving innovation. Demand factors instead play a complementary role (Di Stefano et al., 2012, pg.1284).
2.1.3 Structural Factors
In recent years, scholars have been discussing a possible third perspective to explicating innovation, namely, structural factors. The rationale being that political and economic outcomes are produced by the interaction between formal and informal mechanisms of power within a society. In this sense, politicians can play a constructive or destructive role in innovation based on the effect on their political/economic rents and power. Economic innovation and technological progress then are intrinsically linked to the power
configurations that underlie a society or state. This perspective is argued by Acemoglu and Robinson (2000, pg.127), who find through their study that innovation or lack thereof can be explained through an examination of political structures. This perspective has opened up new research avenues for scholars and provided for the use of political economy discourse in working with innovation. Tyce (2020, pg.2), for instance, utilized a framework that combines the political settlements framework by Khan (2010) with concepts from political economy literature to assess the success of innovation of mobile phone banking in Kenya. While several studies attempt to explain innovation through either of the three factors, there are few that attempt to account for all three. Examining innovation from all three perspectives at the onset may appear complicated due to conceptualization. While the demand and supply-side arguments can be defined into measurable concepts such as research and development investment whereas structural factors may not be so easily operationalized. For instance, If one takes the rule of law as a measure of institutional quality, this may describe the extent to which the rule of law is applied in relevant situations, but does not provide a detailed account of how productive outcomes arise in situations whereby the rule of law is weak. The rule of law in itself can be an outcome of formal and informal power relations within a society, which dependent on the context at hand, may be more formal and thus legal or more informal depending on how the particular scenario impacts the power relations of elites. The rule of law then is not a rigid measure but flexible and fluid according to the preferences of elites for any given situation. Despite
10 its flaws, the Rule of Law index provides the best option for assessing a conglomeration of structural factors in the sense that it encompasses numerous dimensions including: "the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence" (Rule of Law: Estimate, 2019).
This thesis attempts to encompass all three factors that stimulate innovation in order to provide for a wholesome analysis. Therefore, this thesis will analyze Mobile Banking as an innovation and the relationship between supply, demand, and external factors that drive innovation. The aim again is to provide an answer to the question of how different rates of mobile banking use and adoption across sub-Saharan Africa be explained. The theoretical framework guiding this research revolves around the three factors driving innovation, as outlined in the literature on innovation above.
2.2 Mobile Banking As An Innovation
As aforementioned, mobile banking is a phenomenon that has had profound impacts on the lives of individuals, particularly in developing countries. As of 2019, sub-Saharan Africa accounts for 48% of the global mobile money accounts and 64% of the global transaction volume using mobile money accounts (GSMA, 2019, pg.8). The innovation has changed the way individuals and firms engage within the economy, as it fosters inclusion in successful initiatives. Scholars and practitioners alike have attempted to provide an account for the establishment and impact of this mobile banking innovation. The majority of this literature tends to be qualitative in nature, focuses on Africa, and has been produced within the last decade (Kim et al., 2018, pg.5). The literature on mobile banking within this review can be classified according to the clustering method outlined by Kim et al. (2018). This method entails dividing articles into delivery, environmental factors, and impact (ibid.,6). This division within the literature on Mobile Banking mirrors the division within the literature on innovation outlined above. As in the innovation literature, scholars tend to focus on drivers of mobile banking from a supply, demand, or environmental/ structural perspective. Additionally, some scholars focus on the impact of mobile banking as an innovation.
11 2.2.1 Delivery
Delivery can be split into two aspects, namely supply and demand. Supply refers to those articles that focus on the provision of mobile services and entails an analysis of network operators and agent-client networks, for instance (Ibid). An example of this can be seen in the noteworthy contribution by Jack et al. (2010,pg.92)), which, amongst other aspects, examined the structure of Safaricom's mobile banking initiative Mpesa. Demand-side refers to how individuals and firms engage with mobile banking services and entails such studies that focus on perceptions and usage (Kim et al., 2018, pg.7). Numerous studies have highlighted the usage patterns and perceptions of individuals, which can be seen as pertinent indicators for the efficacy of an initiative and evolution in the initiative itself. For instance, Mbiti (2011,pg.252-255) documented changes in how money was sent, received, and the use of Mpesa in Kenya based on surveys conducted in 2006 and 2009. He found that the imposition of Mpesa as a money transfer service had a significant impact on the use of money transfer methods, quickly becoming the preferred method at the detriment of previously used methods (Ibid.,254). In addition, they documented the various ways that people used the Mpesa system. At the time of their study in 2009, they found that the majority of the people using the system sought to buy airtime credit (just over 40%)
(Ibid.,255). According to Riley and Kulathanga (2017, pg.69), today, the majority of the users utilizing Mpesa do so for financial services, which includes paying bills, paying for goods and services, and transferring money.
Thus studies focused on either supply, or demand-side factors can display an evolution in the use of mobile, but when accounting for adoption and success, some scholars have opted to include both. Jack and Suri (2011) adopted this approach in their study focused on the economic impacts of Mpesa, which both outlined the network structure and the perceptions and usage of the service by customers.
2.2.2 Environmental/Structural Factors
Environmental factors refer to the studies that predominantly focus on the contextual and structural factors that may enable or constrain the delivery of mobile banking
initiatives. These include intuitional, cultural, and demographic factors, according to Kim et al. (2018, pg.7). Within the context of sub-Saharan Africa, structural factors are particularly
12 pertinent as neopatrimonialism has been identified as either constraining or enabling
innovations like mobile banking (Johnson and Williams, 2016, pg.726). The way scholars have approached analyzing the impact of environmental factors has differed. Some scholars, such as Lashitew et al. (2019), have accounted for environmental factors by including
related variables in their empirical research. They included an index for Rule of Law
produced by the World Governance Indicators and an index that measures telecom sector regulation as part of their "macro-level factors" Lashitew et al., 2019, pg.1203). The Rule of Law broadly refers to the level by which people respect and abide by the rules of society (Ibid., 1205).
Their results indicate that The Rule of Law has a negative and significant effect on the rates of sending and receiving money but has a negligible impact on adoption (Ibid.,
pg.1206). Thus the Rule of Law does affect mobile banking innovations. They also found that countries that have a more efficient and transparent telecoms regulation mechanism tend to display higher adoption rates of mobile banking technology (Ibid). Alternatively, some scholars like Tyce(2020) utilize a qualitative approach to assess the impact of environmental factors on mobile banking initiatives. He sought an explanation for the success of Mpesa as an innovation through his framework rooted in political science theories. He found that MPesa's success can be explained by Its ability to centralize rents, and its close connection with the state has ensured its dominance in the telecoms market (Tyce, 2020, pg.12). The two studies highlighted above point to the impact of structural factors on mobile banking initiatives and the need to incorporate structural factors in attempting to understand the conditions which facilitate the establishment and use of the innovation.
2.2.3 Impacts of Mobile Banking Innovation
Lastly, numerous studies attempt to outline the impacts of mobile banking on
various aspects. These can be divided into studies that assess impacts on a micro and macro level. As regards the former, Mbiti (2011), through his study, outlined how the introduction of the mobile banking innovation Mpesa had a drastic impact on the money transfer system within Kenya. The introduction of the service led to a decrease in the usage of competitor products such as money transfer by agencies like Western Union and alternative methods such as sending money parcels with bus drivers (Mbiti, 2011, pg.254). Jack and Suri (2011)
13 in their study found that mobile banking initiatives higher consumption by adopters as opposed to non-users (Jack and Suri, 2011, pg.17), increased the frequency of remittances as opposed to non-users (Ibid.,pg.19), and adopters who had bank accounts were more likely to use the mobile banking initiative as a saving mechanism than adopters without a bank account (Ibid). These findings from both studies indicate that the introduction of mobile initiatives can have a profound impact on micro-economic activity.
Studies that focus on the macro-economic impacts of mobile banking initiatives are faced with numerous obstacles within the literature. For one, the majority of literature tends to focus on the case study of Kenya as opposed to other countries in the region. This makes sense considering the availability of data provided by Safaricom on their Mpesa service, but excludes the experiences of other countries in mobile banking. This has not prevented other scholars from assessing other countries as case studies, but some such as Duncombe (2012, pg.386) find the lack of data to be onerous. Moreover, the empirical case studies tend to focus solely on micro-scale usage or socio-economic impacts solely within Kenya (Lashitew et al., 2019, pg.1202).
As regards to mobile banking and financial inclusion, scholars have assessed other countries with mixed success. Although scholars and practitioners alike have recognized the potential of mobile banking to foster financial inclusion, there is a paucity of literature that examines the nature of this relationship. The few studies that do attempt to have produced conflicting results. For instance, in one study by Ghosh (2016) on MENA countries finds that increased telephone usage leads to an increase in incomes and an increase in financial inclusion increase income measured as GDP (Ghosh, 2016, pg.75). Adeoye and Alenoghena (2019, pg.10) found that internet usage, with particular emphasis on mobile banking services, has a positive and significant relationship with financial inclusion. Financial inclusion, however, has a negative relationship with economic growth and that while
internet usage may increase economic growth, it does not do so via financial inclusion (Ibid), thereby contradicting the results from Ghosh. Kim et al. (2018) in their review of the
literature on mobile banking, find that there is a lack of
“.. research addressing how, and to what extent, MFS impact financial inclusion, and to what extent these services have improved the level of financial inclusion” (Kim et al., 2018, pg.9).
14 In other words, there is a need for further engagement by scholars in articulating the
relationship between mobile banking and broader macroeconomic impacts. Secondly, there has been a general lack of cross country research aimed at
understanding the differences across countries (Ibid). One noteworthy contribution is the work by Lashitew et al. (2019). They conducted mixed-method research aimed at assessing the "development and diffusion of mobile money innovations across and within countries" (Lashitew et al., 2019, pg.1202). Of particular relevance for this thesis is the quantitative approach undertaken in this study. Lashitew et al. (2019) utilize data from various sources including the World Bank, Findex (Financial Inclusion Indexes), GSMA, and the Global Financial Structure Database (GFSD) to test the relationship between various economic and institutional variables and adoption rates of mobile money across countries (Ibid., pg.1203). The variables tested against adoption rates included demand factors, supply factors, and macro-level factors. An exhaustive list of the variables tested in this study can be found in the appendix.1 The sample tested by the authors encompassed 97% of the world
population and is based on surveys conducted by the World Bank for the Financial Inclusion index or Findex (Ibid). This enabled the authors to engage in an extensive comparison across countries, regions, and continents.
The conclusions pertinent to this section include that adoption rates of mobile money is higher in countries with higher GDP growth rates, an efficiently regulated telecommunications sector, and actual usage of mobile money services increases in countries with higher GDP per capita and lower levels of the rule of law (Ibid, pg.1213). However, as noteworthy as this study by Latishew et al. (2019), they focused primarily on assessing mobile banking across the globe, while this thesis seeks to assess mobile banking within sub-Saharan Africa. By focusing on this specific region with the aforementioned political situation, this thesis expects to find contrasting results. For instance, they found a negative and significant relationship between the Rule of Law and the usage of mobile banking. Contrastingly, countries in Sub-Saharan Africa consistently score within the bottom half of the Rule of Law index percentile ranking (World Development Indicators, 2019), but yet account for 48% of the world’s active mobile money bank accounts (GSMA, 2019, pg.8).
15 Therefore it is plausible that the relationship is inverse, as the Rule of Law score decreases mobile banking usage increases. Simply, there is a need to understand the differences in mobile banking usage and adoption across countries within sub-Saharan Africa to which this thesis seeks to contribute to.
2.3 Conclusion
In conclusion, this literature review sought to identify themes and concepts that could aid the answering of the research question(s) for this thesis. The literature on innovation, mobile banking, and financial inclusion was outlined in order to position this thesis. In synopsis, innovation in this study is defined as: "any doing things differently in the realm of economic life" (Kalcheva et al., 2018, pg.441). In seeking to understand innovation, this study shall utilize the three defined underlying factors of supply, demand, and structural factors. Supply factors broadly refer to improvements or changes in the ways products are produced by firms. Demand factors refer to the heterogeneity of the market, which induces firms to innovate to meet demand. Finally, Structural factors refer to the underlying
institutional factors that shape innovation. This theoretical framework derives from both the Innovation literature and Mobile Banking literature, wherein scholars postulate that the drivers of innovation are due to supply, demand, and structural factors.
The section regarding mobile banking outlined the impact that this innovation has had on a micro and macro-economic level. In attempting to do so, most studies either focus on explicating mobile banking innovations from the perspective of supply, demand, or
environmental/structural factors While the literature is relatively new, there is a lack of substantive studies that aim to analyze cross country differences in mobile banking
innovations. This thesis seeks to provide explication for how the differences in the adoption and use of mobile banking innovations across sub-Saharan Africa through an analysis of mobile banking and supply factors, demand factors, and structural factors, respectively. As a corollary, the theoretical model to be applied comprises three independent variables, as seen in table 1.
16
Table 1: Overview of the theoretical model
Supply Side factors
Mobile Banking
Innovation
Demand Side Factors Structural Factors
17
3. Methodology
3.1 Research Design: Comparative Case Study
This methodology chapter seeks to outline the process by which this paper seeks to answer the research question(s). This research aims to seek an answer to how differing adoption and usage rates of mobile banking innovations sub-Saharan Africa can be
explained through an analysis of the relationship between the use and adoption of mobile banking and supply, demand, and structural factors. Consequently, the three sub-research questions are as follows:
1. What is the Impact of the supply factor on Mobile Banking Use and Adoption in Sub-Saharan Africa?
2. What is the Impact of the structural factor on Mobile Banking Use and Adoption in Sub-Saharan Africa?
3. What is the Impact of the demand factor on Mobile Banking Use and Adoption in Sub-Saharan Africa?
In order to answer the research questions, statistical analysis as regards the relationship between the variables outlined above will be provided based on data from the following databases: World Development Indicators (WDI), World Governance Indicators (WGI), Global Financial Development Database (GFDD), Financial Inclusion Index (FI). To that end, the following chapter will provide the countries for analysis in section 3.2, provide an explanation of the variables chosen to be operationalized for the study in section 3.3, and section 3.4 will provide a synopsis of the chapter.
3.2 Countries
This study aims to explain differences in mobile banking use and adoption across sub-Saharan Africa, therefore the sample of countries to be used needed to be as large as possible to facilitate analysis. This was problematic due to constraints surrounding the availability of data for several sub-Saharan African countries. The Initial sample of countries totaled 38, based on the availability of data for “Network Coverage” and “Network Quality”
18 data points produced by the GSMA in the MCI.2 All 38 countries had data for these two
points, but as the other data points from the other databases were added, the lack of countries with data for all data points became a constraint.
Countries were chosen only if they had data for each of the 11 data points needed for the study.3 Thus, out of the original 38 countries, only 13 sub-Saharan African countries had
data for all data points chosen for this study. These include Burkina Faso, Chad, Cameroon, Ghana, Kenya, Madagascar, Mozambique, Nigeria, Rwanda, Togo, Uganda, Zambia, and Zimbabwe.4 While the sample size is smaller than the ideal size, the above countries have a
total population of 487,943,723 million people, which roughly accounts for 45% of the total population of sub-Saharan Africa (Global Financial Development Database, 2019). The small number of cases prohibits exhaustive regression analysis using several predictors
(Duncombe, 1961). Thus the number of variables to be assessed has been restricted to three from the original five.
3.3 Variables
As aforementioned, the variables chosen are informed by data from various databases. These databases include WDI, WGI, GFDD, and FI. This secondary data is based on surveys conducted by the World Bank on a select number of African countries will be selected for analysis. The data which is collected by these agencies are then composed into statistics that account for a population, for instance, mobile phone subscriptions percentage of a population. The chosen statistics will then be divided into supply, structural and demand factors to comprise the independent variables. This approach is based on the framework proposed by Lashitew et al. (2019).5 However, due to the small number of cases from sub
-Saharan Africa with data, the number of predictor variables have been reduced to three from the original five chosen. Regression analysis will then be carried out to determine the relationship between the dependent variables representing Mobile Banking Adoption and Use and the independent variables representing the supply, demand, and structural factors to be outlined below.
2 See Appendix 2 for the original sample 3 See Appendix 2 for the original data points.
4 See Appendix 3 for the final sample and data points. 5 See Appendix 1
19 3.3.1 Dependent Variables: Mobile Banking Adoption and Mobile Banking Usage Similarly to Latishew et al. (2019), The dependent variables will be informed by three statistics provided by the World Bank's Financial Inclusion Index. The dependent variables can be broken into two: Mobile Banking Adoption, which is informed by the percentage of people who have a mobile account, and has used it in the last year (FI 2017) and Mobile Banking Use, which is informed by three data points. The three data points are the percentage of people who have sent or received domestic remittances in the last year, percentage of the population that have paid a utility bill using a mobile phone, and the percentage of people who have received payments for agricultural products using a mobile phone.
The rationale for selecting the datapoint for adoption is relatively clear: it indicates the percentage of people who have utilized a mobile money account in the last year—making it an effective indicator for adoption. The rationale for selecting the percentage of people who have sent or received domestic remittances through a mobile phone is based on the fact that most people in sub-Saharan Africa utilize a mobile phone to send or receive domestic remittances (Demirguc-Kunt et al., 2018, pg.51). As regards agricultural products, the FI found that in sub-Saharan Africa, over 40% of individuals receiving payment for agricultural products do so over mobile phones (Ibid.,6), indicating the importance of mobile money in this industry. Lastly, the payment of utility bills utilizing mobile phone was found to be one of the top uses of mobile banking in developing countries, particularly in sub-Saharan Africa and East Asia (Ibid.,50). Thus the three datapoints represent good indicators for Mobile Banking Use.
The number of people who have a mobile account will be conceptualized and
operationalized as the "Mobile Banking Adoption" indicator, and the three data points for remittances, agricultural products, and utility bills will inform the "Mobile Banking Use" indicator. The rationale behind aggregating the three data points lies in the sufficiently high correlation coefficients between them.6
20 3.3.2 Independent Variables: Supply, Demand, and Structural Factors
The supply factor of Mobile Subscriptions will be informed by the data point Mobile Subscription Rates derived from the World Bank’s World Development Indicators. Mobile phones and, by extension, mobile phone subscriptions provide opportunities for low-income households who would otherwise be unable to access several services, such as financial services (Mothobi & Grzybowski, 2017, pg. 71). Thus mobile banking innovations are more likely to occur in contexts where subscriptions are high, an indicator that the necessary infrastructure for network operations has been established. Indeed, In countries such as Kenya, the high level of mobile subscriptions and the extent to which the network agent structure pervades across the country have been identified as key to the
establishment and usage of Mobile banking innovations (Bosire, 2013, pg.3). As a corollary, this thesis expects this to hold true for the other 12 sub-Saharan African countries.
The structural factor of The Rule of Law will be similarly informed by statistics from the World Development indicators provided by the World Bank. The indicators to be used shall be the estimated Rule of Law score for a country which takes into consideration the:
“..Extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. Estimate gives the country's score on the aggregate indicator, in units of a standard normal distribution, i.e. ranging from approximately -2.5 to 2.5” (Rule of Law: Estimate, 2019).
Given its all-encompassing nature, the Rule Of Law index provides an expansive account for structural factors within a country. The index takes into consideration numerous structural issues such as contract enforcement, crime, amongst others, which may conceivably impact the imposition of mobile banking technology (Ibid). Arguably, mobile banking initiatives are more likely to occur in countries wherein the Rule of Law score is high, indicating respect for property rights and respect for rules of society. Lashitew et al. (2019) contrastingly found that as the rule of law score decreases, mobile banking use increases (Lashitew et al., 2019, pg.1206). However, this study included 97% of the world's population, and due to the disparities mentioned above in mobile banking adoption and usage between high-income countries, which tend to have scores on the Rule of Law index and low income developing countries which tend to have lower scores, seems logical. This study seeks to assess differences between developing countries, and thus expects that abiding by the rules of
21 society is imperative for mobile banking adopting and use. Therefore the expectation is that the Rule of Law will have a positive and significant relationship with Mobile Banking
Adoption and Use.
The demand factor of Formal Account will be informed by statistics from the Global Financial Development Database compiled by the World Bank. The data point used will be the percentage of people with an account at a formal institution (Global Financial
Development Database, 2019). Presumably, where the percentage of the population with access to a formal institution is low, there would be a demand for alternative methods of accessing financial services. Such as mobile banking innovations. In high-income countries, for instance, the percentage of the population with an account at a formal institution is high at 93%, while the percentage of people who have mobile bank accounts is negligible
(Demirguc-Kunt et al., 2018, pg.17). In some countries in Sub-Saharan Africa, in contrast, more people have mobile banking accounts than accounts at a formal financial institution due to various constraints associated with getting a formal account (Ibid., pg.20). Moreover, Beck et al. (2015) postulate that mobile banking innovations are utilized to overcome the prohibitively high costs and high risk associated with formal financial institutions (Beck et al., 2015, pg. 4). Thus within this context, mobile banking can be seen as an alternative to formal institutions, and an inverse relationship can be inferred. Demand then is reflected by the percentage of the population with an account at a formal institution.
These three independent variables mentioned above were further tested for multi-collinearity. The results of this test indicate that the chosen independent variables are independent enough (below a correlation coefficient of .9) to facilitate regression analysis.7
This thesis expects that Formal Account and Mobile Banking Adoption and Use will have a negative statistical relationship, wherein as the Formal Account variable decreases, the dependent variables increase. The hypotheses, at a minimum, expect that all the factors, supply, demand, and structural; have an impact on the dependent variables described above. The null hypotheses to be rejected in each combination of variables is that there exists no relationship between the dependent and independent variables in any given sequence. In order for the null hypothesis to be rejected, the p-value between any of the
22 three groups of independent variables and each dependent variable would have to be less than .10 or 10%. The significance levels of less than .05 or 5% and less than .01 or 1% represent highly and extremely high levels of significance, respectively. In addition to the hypotheses above, this thesis expects findings pertinent to the sub research questions. Per the first sub research question, Mobile Subscriptions as the supply factor is expected to have a positive and statistically significant impact on Mobile Banking Adoption and Use. In reference to the second sub research question, The Rule of Law is expected to have a positive and statistically significant relationship with Mobile Banking Adoption and Use. Lastly, Formal Account is expected to have a negative relationship with Mobile Banking Adoption and Use in reference to the third sub-research question.
In Excel, multiple linear regression will be conducted to ascertain whether a statistically significant relationship exists. Moreover, regression analysis will provide for explication as regards the efficacy of the model selected. This is provided for by the r squared value, which accounts for the variation in the dependent variables that can be attributed to the
independent variables.
3.4 Conclusion
The above-outlined methodology will enable a viable analysis to answer the research question. The thesis specifically asks how the differing adoption and use of mobile banking innovation can be explained across sub-Saharan Africa. Based on the literature on mobile banking and innovation, the thesis postulates that this can be explained through a statistical analysis of the relation between mobile banking adoption and use and supply, demand, and structural factors, respectively. Thirteen sub-Saharan countries that had data for all the data points in this study were chosen, including Burkina Faso, Chad, Cameroon, Ghana, Kenya, Madagascar, Mozambique, Nigeria, Rwanda, Togo, Uganda, Zambia, and Zimbabwe. The small number of cases prevents the inclusion of additional variables that would provide for more exhaustive regression analysis and is a limitation of this study.
The data points mentioned above inform the dependent and independent variables to be used in the study. An overview of the variables can be seen in table 2. Regression analysis in Excel will provide for the relationship if any between variables and the extent to which the model chosen for this study provides an account for variation in the dependent
23 variables. For the regression analysis, the significant levels for the regression coefficients to be used in this study are 10%, 5%, and 1%. The results of this analysis will provide answers to the sub-research questions for the study, which will further facilitate explication for the main research question of how the differences in the adoption and use of mobile banking innovations across sub-Saharan Africa can be explained.
Table 2: Operationalization of the variables within the theoretical framework to answer the research question(s).
Supply Side Factor
• Mobile Subscription rates (WDI)
Mobile Banking
Innovation
Mobile Banking Adoption• Percentage of people with a mobile account (FI)
Mobile Banking Use
• Percentage of people who have sent or received domestic remittances in the last year (FI)
• Percentage of people who have received payments for agricultural products on a mobile phone in the last year (FI)
• Percentage of people who have paid utility bills using a mobile phone in the last year (FI).
Demand Side Factor
• Percentage of people with an account at a formal institution(GFDD)
Structural Factor
24
4. Results Chapter: Regression Analysis
The following chapter outlines the results of the regression analysis between the independent variables that comprise the supply side, structural, and demand-side factors and the dependent variables that comprise Mobile Banking Adoption and Use. The chapter is divided as follows: Section 4.1 outlines the descriptive statistics for all of the variables utilized in the study be discussed, sections 4.2-4.3 provide the results of the regression analysis, and finally, a conclusion will discuss the findings of this chapter and the relevance of the results will be outlined in section 4.4. The regression analysis will be divided into two: section 4.2 will outline the results between the independent variables and the first
dependent variable of Mobile Banking Adoption and section 4.3 will outline the results of the multivariate regression in relation to the dependent variable of Mobile Banking Use. Therefore, the results presented in section 4.2 will address the Mobile Banking Adoption aspect of all the sub-research questions, and section 4.3 will address Mobile Banking Use. After understanding the results from both sections, this thesis can explicate how the differences in adoption and use of mobile banking innovations across sub-Saharan Africa can be explained.
4.1 Descriptive Statistics
As mentioned above, the dependent variables for this study are four variables split into two categories: Mobile Banking Adoption and Mobile Banking Use. Mobile Banking
Adoption is informed by the Mobile Money Account indicator from the FI index. This records the percentage of the population that has used a mobile money account in the last 12 months. The following indicators inform mobile Banking Use from the FI index: percentage of the population that have sent or received domestic remittances in the last year,
percentage of the population that have paid utility bills through a mobile phone, and the percentage of the population that have received payments for agricultural products through a mobile phone. The descriptive statistics for each dependent variable can be seen in table 3.
25
Table 3: Descriptive statistics for all variables Sources: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.; The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0. ; The World Bank (2019), Global Financial Development Database (GFDD), dataset, The World Bank,
https://datacatalog.worldbank.org/dataset/global-financial-development ;
The World Bank (2018). World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
All Data From: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.; The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0. ; The World Bank (2019), Global Financial Development Database (GFDD), dataset, The World Bank,
https://datacatalog.worldbank.org/dataset/global-financial-development ;
The World Bank (2018), World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
The figures in table 3 indicate a variance in the adoption and usage of mobile banking across the 13 countries surveyed for this study. As regards to Mobile Banking Adoption, the average across the countries is 30%. The standard deviation is 19%, which indicates a relatively large difference between the average and any given score. The large range of 67% is reflective of the differences in the popularity of mobile banking innovations across the region. Kenya, for instance, demonstrates the highest score for this indicator at 73%,
according to figure 1. Conversely, only 6% of the population surveyed in Nigeria have used a mobile money account in the last 12 months.
Figure 1: Country scores for the variable Mobile Banking Adoption Source: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank. All data from: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.
33% 15% 15% 39% 73% 12% 22% 6% 31% 21% 51% 28% 49% 0% 10% 20% 30% 40% 50% 60% 70% 80% P e rce n t Countries
M OBI L E BANK I NG ADOPT I ON
Descriptive Statistics M subscriptions Rule of Law
%p with a formal account Mobile Banking Adoption Mobile Banking Use Mean 78% -0,52 33% 30% 19% Standard Error 7% 0,12 4% 5% 3% Standard Deviation 27% 0,44 13% 19% 11% Range 97% 1,22 47% 67% 43% Minimum 41% -1,10 9% 6% 3% Maximum 138% 0,12 56% 73% 46% N 13
26 The data points for the 13 countries surveyed demonstrated the general low-level usage of mobile banking innovations. The average percentage of the population was a lowly 19%, which is a reflection of the vast range of data of 43%, according to table 3. The lower average of Mobile Banking Use in comparison to Mobile Banking Adoption is reflective of the fact that while the innovation has spread across the countries in Sub-Saharan Africa, the success of such innovations has varied (Beck et al., 2015, pg. 4). As demonstrated by figure 2, Kenya and Nigeria again, posited the highest and lowest scores, at 46% and 3%,
respectively.
Figure 2: Country scores for the variable Mobile Banking Use Source: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank. All data from: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.
The population of the 13 countries that comprise the sample for this study represents 45% of Sub-Saharan Africa or 487,943,723 million people (World Development Indicators, 2019). Of this population, 82% have one or more mobile phone subscriptions (Ibid), which is indicative of the extensive spread of mobile phone usage in Sub-Saharan Africa. The Mobile Subscriptions variable was created by dividing the data point for the number of people with a mobile subscription by the data point for the total population provided by the World Bank’s WDI database. The computed variable for the percentage of the population with mobile subscriptions comprises the supply side variable for this study.
The average score for the variable Mobile Subscriptions across the 13 countries is 78%, indicative of the high level of mobile phone engagement across the selection. The high standard deviation score of 27% indicates that the data points are not very close to each other, which is a factor of the outliers, namely Ghana and Madagascar, which distort the
18% 10% 17% 29% 46% 5% 14% 3% 17% 15% 28% 22% 26% 0% 10% 20% 30% 40% 50% P e rce n t Countries
27 average distance of a given score from the mean. As demonstrated by figure 3, Madagascar had the lowest percentage of people with mobile subscriptions at 41%, and Ghana scored the highest at 138%. The latter indicating the prevalence of more than one mobile
subscription per person within the country. This large range is reflective of the reality that while mobile phone in sub-Saharan Africa is generally high, some asymmetries still exist between countries (Asongu, 2018, pg.83).
Figure 3: Country scores for the percentage of the population with a mobile subscription Source: The World Bank (2018), World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
All data from: The World Bank (2018), World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
The sole variable chosen for this study to represent the structural factor of the independent variables is the World Bank’s Rule of Law estimate. Broadly speaking, this refers to the extent to which a country abides by the rules of society and is measured on a scale from the worst score; -2.5 (worst) to the best: 2.5 (Rule of Law: Estimate, 2019). As indicated by table 3, the 13 countries chosen for this study performed poorly on this scale on average at -,52. This is reflective of the overall experience of countries in sub-Saharan Africa, wherein every-day life is characterized by behavior that often falls beyond the
boundaries of the law. Examples include corruption, political violence, despotism, to name a few. Generally, the 13 countries score negatively on the index, except for Ghana, Rwanda, and Zimbabwe, as displayed by figure 4. Chad has the worst score on the Rule of Law estimate at -1.10, indicating a widespread disregard for rules within the county. Conversely, Rwanda demonstrated the highest score at .12.
98% 73% 45% 138% 96% 41% 48% 88% 79% 78% 57% 89% 89% 0% 50% 100% 150% P e rce n t Countries
28
Figure 4: Country scores for the Rule of Law estimate Sources: The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0. All Data From: The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0
As regards the demand side variable, the descriptive statistics for Formal Account across the 13 countries indicate that the majority of the population within most countries do not have an account at a formal institution relative to countries in other regions. As displayed by table 3, the average across the 13 countries is a lowly 33%. Conversely, in high-income OECD countries, 94% of the population has an account at a formal institution (Demirguc-Kunt et al., 2018, pg.19). There is variation across the 13 countries, indicated by the range of 47%. Kenya has the highest percentage of the population with an account at a formal institution at 57%, and conversely, Chad has the lowest at 9%, as displayed in figure 5. In this context, it is reasonable to expect the viability of mobile banking initiatives as an alternative to the inability to access formal financial institutions.
Figure 5: Country scores for the percentage of the population with an account at a formal institution Sources: The World Bank (2019), Global Financial Development Database (GFDD), dataset, The World Bank,
https://datacatalog.worldbank.org/dataset/global-financial-development
All Data From: The World Bank (2019), Global Financial Development Database (GFDD), dataset, The World Bank,
https://datacatalog.worldbank.org/dataset/global-financial-development -0.45 -1.08 -1.10 0.07 -0.41 -0.81 -1.04 -0.88 0.12 -0.59 -0.29 -0.34 0.00 -1.20 -1.00 -0.80 -0.60 -0.40 -0.200.00 0.20 Sco re ( -2 .5 t o 2 .5 ) Countries R UL E OF L AW 45% 27% 9% 42% 56% 10% 33% 39% 37% 34% 33% 36% 28% 0% 10% 20% 30% 40% 50% 60% P e rce n t Countries
29 As aforementioned, this thesis has several expectations for the regression analysis between the independent variables and the dependent variables concerning the sub-questions. Firstly, a positive relationship between Mobile Subscription and Mobile Banking Adoption and Use is predicted. Whereby as Mobile Subscription increases, Mobile Banking Adoption, and Usage increases. Thus in answer to the first sub-research question, a positive impact is expected of the supply side factor on Mobile Banking Adoption and Use. Secondly, this thesis expects a positive relationship between the structural factor of the Rule of Law and dependent variables in relation to sub-question 2. Whereby as the Rule of Law increases, Mobile Banking Adoption and Usage increases. Lastly, a negative relationship between the demand factor of Formal Account and dependent variables is predicted for sub-question 3. As a low level of formal account at financial institutions is indicative of the demand for and thus the adoption and use in Mobile Banking innovations. In this sense, the relationship between the independent variable and the dependent variables is hypothesized to the negative: The dependent variables increase as the independent variables decrease.
4.2 Mobile Banking Adoption
The analysis for this section sought to establish a linear relationship between all the independent variables representing supply, structural, and demand factors and the dependent variable of Mobile Banking Adoption. In order to do so, linear regression was first conducted between each separate independent variable and Mobile Banking Adoption, followed by multivariate regression with all the independent variables. The latter providing the relationship between each independent variable and Mobile Banking Adoption controlled for the other independent variables. To this effect, four models of regression were conducted: Model 1 between Mobile Subscriptions and Mobile Banking Adoption, Model 2 between the Rule of Law and Mobile Banking Adoption, Model 3 between Formal Account and Mobile Banking Adoption, and Model 4 with all the independent variables and Mobile Banking Adoption. Through the results of this regression analysis, the impact of the independent variables on the Mobile Banking Adoption aspect of the three sub-research questions can be answered. Specifically, the results ascertained from Model 4 would provide for the impact of the three independent variables on Mobile Banking Adoption controlled for each other. This
30 thesis expects that Mobile Subscriptions and the Rule of Law will have a positive impact on Mobile Banking Adoption, while Formal Account will have a negative.
The variable comprising the supply side factors of the study is the percentage of the population with a mobile subscription across the 13 countries, the Rule of Law represents the structural variable, and the demand side is represented by the percentage of the population with an account at a formal institution. The dependent variable of Mobile Banking Adoption, as aforementioned, is comprised of the variable Mobile Money Account.
As a corollary of the aim of this chapter, the hypotheses that guided this research assumed a relationship between each independent or predictor variable and Mobile Banking Adoption. The null hypothesis in any given sequence of variables would then be that a relationship between the two variables does not exist or is very weak. The strength of such a relationship is demonstrated by the p-value of the intercept between the chosen independent and Mobile Banking Adoption. As indicated by table 4, the Rule of Law and Formal Account have a statistically significant relationship with Mobile Banking Adoption. No such relationship exists between Mobile Subscriptions and Mobile Banking Adoption.
Table 4: Regression coefficients for independent variables to Mobile Banking Adoption Source: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.;
The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0. ; The World Bank (2019), Global Financial Development Database (GFDD), dataset, The World Bank,
https://datacatalog.worldbank.org/dataset/global-financial-development ;
The World Bank (2018). World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
All Data from: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.;
The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0. ; The World Bank (2019), Global Financial Development Database (GFDD), dataset, The World Bank,
https://datacatalog.worldbank.org/dataset/global-financial-development ;
The World Bank (2018). World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
Mobile Banking Adoption
Variables Model 1 Model 2 Model 3 Model 4
Mobile Subscriptions 0,279
(0,195) x x
-0,298 (0,253) The Rule of Law
x 0,267 (0,102)** x 0,260 (0,121)* Formal Account x x 0,850 (0,353)** 0,885 (0,456)* R2 0,157 0,386 0,345 0,567 N 13 *p<.10; **p<.05; ***p<.01; () Standard Error
31 4.2.1 Model 1: Mobile Subscriptions and Mobile Banking Adoption
Mobile Subscription did not demonstrate any discernible impact on Mobile Banking Adoption. For one, the regression coefficient was low at (,279), and the relationship was not statistically significant at any of the confidence levels, according to table 4. Thus based on this data, a relationship between the two variables cannot be established. This may be a factor of the low number of cases tested as well as anomalies in data. As demonstrated in table 4, the standard error (,195) is relatively close to the regression coefficient (,279) indicating a wide range of data. The scatter plot in figure 6 indicates the spread of the data relative to the predicted relationship. The distance between the actual points and the predicted relationship indicates the lack of relationship between the two variables.
Figure 6: Scatter plot for Mobile Subscriptions and Mobile Banking Adoption Source: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.; The World Bank (2018). World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicatorsAll data from: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.; The World Bank (2018). World Development Indicators, dataset, The World Bank. https://datacatalog.worldbank.org/dataset/world-development-indicators
Moreover, the poor explanatory level of the model is displayed by the R2 score of (,157) or 16% as per table 4. The null hypothesis in this regression analysis cannot be rejected: A relationship between Mobile Subscriptions and Mobile Banking Adoption cannot be discerned based on the data.
4.2.2 Model 2: The Rule of Law and Mobile Banking Adoption
In contrast, the structural variable for the Rule of Law demonstrated a positive and significant relationship with Mobile Banking Adoption. As demonstrated by Table 4, Model 2, which solely included the Rule of Law as an independent variable in
conjunction with Mobile Banking Adoption, displays a regression coefficient of (,267) at
0% 50% 100% 0% 20% 40% 60% 80% 100% 120% 140% 160% M ob ile B ank ing A do pti on M subscriptions
M subscriptions Line Fit Plot
Mobile Banking Adoption Predicted Mobile Banking Adoption Linear (Predicted Mobile Banking Adoption)
32 a highly significant level of less than .05. In model 2, therefore, as the Rule of Law score increases, Mobile Banking Adoption increases. This effect, however, is relatively weak, as demonstrated by the low regression coefficient. The standard error of (,102) indicates further that there is a wide spread of data. However, upon closer inspection of the scatter plot displayed by figure 7, this is a result of one point that greatly deviates from the predicted line. This point measured at (-.41, 73%) on the x-axis refers to the data points for Kenya, which had the highest rate for Mobile Banking Adoption but retained a relatively poor Rule of Law score. The other countries, however, generally follow the trend line, indicating that in model 2, the Rule of Law has a positive and significant impact on Mobile Banking Adoption. Thus the null hypothesis of no relationship
between the two variables can be rejected. The r-squared score for this model indicates that the variable offers a decent level explication for the variation in Mobile Banking Adoption at 39%, according to table 4.
Figure 7: Scatter plot for The Rule of Law and Mobile Banking Adoption Sources: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.; The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0. All Data From: Demirguc-Kunt, A. et al. (2018) Global Findex Database 2017 : Measuring Financial Inclusion and the Fintech Revolution. [Online]. Washington, DC: World Bank.; The World Bank (2019), Rule of Law: Estimate, Data Catalog. The World Bank, November 7, 2019. https://datacatalog.worldbank.org/rule-law-estimate-0.
4.2.3 Model 3: Formal Account and Mobile Banking Adoption
Model 3 comprised of univariate regression between the demand side independent variable of Formal Account and Mobile Banking Adoption. The results, as outlined by table 4, indicated that Formal Account has a large impact on Mobile Banking Adoption (,850). Moreover, this relationship is highly statistically significant at less than ,05. Contrary to the expectation stipulated above, and an increase in Formal Account leads to an increase in Mobile Banking Adoption. This thesis expected that this relationship
0% 50% 100% -1.20 -1.00 -0.80 -0.60 -0.40 -0.20 0.00 0.20 M ob ile B ank ing A do pti on Rule of Law
Rule of Law Line Fit Plot
Mobile Banking Adoption Predicted Mobile Banking Adoption Linear (Mobile Banking Adoption)
