Uptake of financial technologies: the role of the legal system origin
Abstract: The research studies the implications of a legal system origin for the uptake of
financial technologies. The sample consists of three developing country regions; Africa, Asia and Latin America and three groups of legal systems; common law, civil law and mixed law. The main region of interest is Africa, where I show through series of regressions that the legal system employed does affect the uptake of financial technologies, with common law countries through more flexible regulation approach outperform both civil and mixed law countries. Additional inquiry through instrumental variable approach finds a positive relationship between financial development and the uptake of financial technologies.
Key words: financial technologies, legal origin, developing countries, Africa, financial
development
Name: Katarína Bičanová Study programme: MSc Finance
Student number: S3513076 Supervisor: prof. C.K.D. Adjasi
1. Introduction
The Fintech revolution is becoming a global phenomenon. Western countries are leading the way by developing new modes of alternative financial services. Greatly due to the world-wide spread of mobile networks, internet coverage and most recently the blockchain technology advancements, academic literature, corporations as well as Fintech start-ups, are becoming more interested in bridging the gap between financial innovations and financial inclusion in developing countries.
The motivation for this research is sparked by a recent plenary paper investigating whether there is a notable difference in adaptation of financial technologies among the sub-Saharan African countries, based on a local legal system employed (Yermack, 2018). By comparing the mean adaption rates for selected financial technologies, henceforth the fintech, the author shows that there is a considerable gap between common law countries, which show higher adaption rates, and civil law countries, which are lagging behind. These preliminary results reveal a substantial disparity, however they should be looked at as a base for further investigation into the specific characteristics of a given legal system which may have an impact on the adaptation of financial technologies and more broadly on financial development.
The relationship between the legal origin of a country and financial development has been studied by a number of researchers, most notably La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1997) conduct an extensive research on 49 developed as well as developing countries. Empirical research confirms that stronger investor protection laws, attributable to common law countries, are linked to stronger capital markets in these countries. An influential paper by La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998), where the original premise for the relationship between law and finance was established, noted that French civil law countries suffer from the weakest investor protection, followed by German and Scandinavian civil law systems and the most favourable investor environment in terms of legal protection was displayed in common law countries. The research predates the current Fintech revolution, however through its extensive analysis serves as a base for subsequent researchers investigating the different aspects of legal environment and financial sector advancements.
The primary question of this research is:
Is the legal system of a country a determinant for the adoption of financial technologies?
The main research objective of this study is to determine whether there is a relationship between the inherited legal structure of a country, as well as other characteristics of the legal system, and the uptake of financial technologies, often associated with financial inclusion, primarily in the geographical region of Africa. Subsequent complementary analysis is done on a sample of developing countries from two other regions; Asia and Latin America. To formulate a relation, firstly the older and hence relatively more established financial technologies are studied, namely the automated teller machines (ATMs).
Hypothesis 1a: The legal origin of a country has an impact on the adoption of established financial technologies.
In order to gain insights about the current financial innovations, the uptake of more recently developed financial technologies, such as mobile money and electronic payments, is studied separately to see whether the relationship from the first hypothesis (1a) carries on or whether there are any notable differences.
Hypothesis 1b: The legal origin of a country has an impact on the adoption of new financial technologies.
To attain a deeper understanding into the issue, the relationship between both established and newer financial technologies is examined with respect to financial development of a given country. Primary interest is again on Africa, however regional differences will be studied for a more comprehensive overview.
Hypothesis 2: Uptake of financial technologies has an impact on the financial development of a country.
2. Literature Review
2.1. The Fintech Revolution
Financial innovations in the era before the still quite recent 2008 financial crisis were regarded as creating uncertainty and were associated with risky investments, however since the aftermath of the crisis, innovations in the area of financial technology that are disrupting the traditional financial sector, are now associated with a more positive outlook (Gomber, Kauffman, Parker, and Weber, 2018). Overall the financial sector is relatively more innovative than manufacturing or other services sectors, more prone to regulation scrutiny and more sensitive to changes in regulations with regards to information technology (IT) advancements (Mention and Torkkeli, 2012).
Financial innovation is, as stated by Lerner and Tufano (2011), pp6: “the act of creating and then popularising new financial instruments, as well as new financial technologies, institutions and markets”. The research of Fontin and Lin (2019) sheds a light on the determinants of financial innovation specifically in low-income countries. New inventions or transfer of existing technologies to developing countries is highly dependent on local context. To address the heterogeneity amongst developing countries, the authors collect data from 168 banks from low-income countries across the continents to find the determinants of the technological gap, using stochastic meta-frontier approach. The authors find a positive relationship between financial innovation and increased competition in banking, financial inclusion, banking access, and financial depth. Furthermore, Poghosyan (2013) finds that a barrier to adoption of financial innovation in developing countries stems from the higher financial intermediation costs faced by banks stemming from riskier credit portfolios among other things. The implication is the need to encourage greater competition amongst local banks and strengthen institutional frameworks.
approvals due to regulatory issues. Financial innovations have the potential to increase financial inclusion by mitigating the risks otherwise faced by the banking sector in serving the people that would else be unbanked (Agoba, Sare, and Bugri-Anarfo, 2017). Gomber, Kauffman, Parker, and Weber (2018) further distinguish three main determinants of the fintech revolution: process disruption, transformation of services and technology innovation. Assessing these factors on a regular basis can deepen the understanding between fintech and financial as well as economic development. Based on the United Nations report Fintech, Green Finance and Developing Countries (2017), if grasped and allowed to flourish, fintech can provide many solutions for empowering local communities with easier access to finance, whether it is through increased mobile network penetration, digital payments or more complex blockchain solutions Fintech has the expectation of providing more outreaching financial inclusion through reducing costs and improving delivery services.
A notable and often cited example of a successful financial technology innovation is the story of Safaricom launching M-PESA in Kenya, a common law country, which has revolutionised the facilitation of money transfers and later savings services through mobile phones, as described by Dittus and Klein (2011) among others. In only four years since launch, over half of the adult population was signed onto the program. Although the benefits M-PESA brought about are clear now, the launch could have been halted if not for the fast and flexible regulations that made it possible. The potential of financial inclusion and poverty alleviation through access to these payment mechanisms sparked initiatives for mobile-money programs around the developing world. Dittus and Klein (2011) further argue for the regulatory environment of countries needing to be permissive for the formation of new business models of financial technologies that can work in individual countries. As mobile phones get more common and affordable, they present the most efficient instrument for financial innovations in recent years, with Africa leading the way in annual growth in mobile phones.
2.2. The Legal Origin and Financial Development
The African continent presents an intriguingly divided region in terms of the legal origins (Yermack, 2018) and the inherent approach to regulations that can be examined with respect to implications for financial and economic advancements. The most widely accepted taxonomy of legal traditions is to distinguish two main separate historical families of legal origin; the common law and the civil law, and their subsequent slight modifications depending on the country where they have developed and where they are transplanted.
tempering of the ruling royalty by strengthening the protection of contracts and personal property. With rulings based on precedence, common law today is often characterized by stronger shareholder protection laws (La Porta, Lopez-de-Silanes, and Shleifer, 2008). On the other hand, the civil law and its subsequent variations of German, French and Scandinavian civil law originate from the ancient Roman law. The civil law formations are more prevailing around the world, in modern history most recently after the Russian Federation and the former soviet satellite states converted back to the German civil law in the aftermath of the fall of the USSR. Civil law is highly reliant on established codes and orderings, in result granting weaker shareholder protection and more power to the centralized government (La Porta, Lopez-de-Silanes, and Shleifer, 2008).
The branch of civil law transplanted into Africa is in majority the French civil law, brought on in the early 19th century during the Napoleon era colonisations. According to La Porta, Lopez-de-Silanes and Shleifer (2008), Napoleon increased the power of the state at the expense of the judicial system and arguably due to the prevalent inflexibility of a powerful centralized government, the results of La Porta, Lopez-de-Silanes, Shleifer and Vishny (1998) show that specifically the French civil law countries perform the worst in terms of investor protection laws, subsequently undermining financial development. As La Porta, Lopez-de-Silanes and Shleifer (2008) also point out, it can be problematic to investigate the results of adoption of a new technology due to the uncertainty of whether to attribute the uptake to the necessity and positives of the technology itself, or to the environment that makes the adaptation feasible.
Beck, Levine and, Demirgüç-Kunt (2002) conduct a research on 115 countries to deepen the understanding between legal origin and financial development. They conclude the effect is carried through two channels; political and adaptability. The consistent results show that through both channels, common law countries are better equipped to address and absorb socioeconomic changes than, especially, the French civil law countries. The legal origin is thus an important determinant in explaining cross-country differences in financial development.
Economic growth and development are to be striven for to further enhance and ameliorate living conditions within individual countries and on a global scale. The importance of economic development is preliminary to achieving many of the Sustainable Development Goals outlined by 2030, as stressed in the UN World Economic Situation and Prospects (2018). Negligible growth is predicted within the next year in parts of Southern, Western and Central Africa, with the UN emphasizing the need for long-term oriented structural improvements of institutions and infrastructure in order to eradicate extreme poverty and hone future economic as well as financial development. The report also highlights the global need to re-orient towards environmental responsibility and in this aspect the now relatively less developed parts of Africa present an opportunity to establish environmentally sustainable growth through socially responsible policies and investment.
Along the lines of law and finance, Liang and Renneboog (2017) conduct an empirical research on Corporate Social Responsibility (CSR) ratings of 23,000 companies across 114 countries and conclude that countries with civil law formations score significantly higher on CSR ratings, in terms of environmental, social and governance issues. Simply put, unlike common law countries that prioritize stronger shareholder rights, civil law countries give scope for a higher possibility of satisfying the needs of a wider group of stakeholders. On a similar note, Asongu, Nwachukwu, and Pyke (2019) conducted an interesting research on 44 Sub-Saharan countries which concludes that advancements in information and communication technologies can lead to lessening of environmental deterioration. The authors find that common law countries foster higher mobile phone and internet penetration rates and hence have a higher potential to decrease environmental harm.
3. Methodology and Hypotheses Development
3.1 Data collection
The data employed to perform the subsequent ordinary least squares (OLS) panel regressions in majority come from the various World Bank publically available databases. The analysis is conducted over a time horizon of 14 years, from 2004 to 2017. The reason for this particular period is the availability of specific data. To ensure comparability of observations, monetary indicators such as GDP and Inflation are measured in constant US dollars.
The main focus is on the region of Africa with 51 countries included, constituting of both Northern and Southern (sub-Saharan) Africa. However, to infer whether the phenomenon of the influence of legal origin is widespread or limited to Africa, I conduct supplementary analyses on developing countries from two additional broad regions, namely Asia and Latin America with 33 and 34 countries included respectively. To form a reliable sample of the developing world for the purpose of this paper, I follow the country classifications as outlined by the UN’s “World Economic Situation and Prospects” (UN, 2014). Based on their common characteristics, I only include countries classified as developing and omit economies in transition; Eastern Europe and parts of western Asia. I also omit the developing countries of the Middle East, due to the particularities of their legal and financial systems that incorporate religious norms (Alzahrani, 2019), adding strain on the relative comparability of the developing countries otherwise included. Countries that on top of common or civil legal system encompass some features of the Muslim Law, such as Indonesia or Sri Lanka, are included in the sample and characterised as pertaining to a mixed law system. Final interference with the sample is the exclusion of countries which do not possess any data for any of the dependent variables, although might show data for certain independent variables; these are mostly individual small islands in the Caribbean and DPR Korea, which offers no data at all. The prevailing legal system of specific countries is determined through individual research on the countries and legal summaries obtained through the GlobaLex website (2019).The list of all the countries included and their respective legal systems is enclosed in Appendix 1.
3.2 Variables
3.2.1 Hypothesis 1a
The legal origin of a country has an impact on the adoption of established financial technologies.
ATMi,t = α0 + β1 x Legal i,t + β2 x Governance i,t + β3 x Internet i,t + β4 x Mobilesub i,t + β3 x
Ln_GDP i,t + ε i,t (1)
To analyse the proposed hypothesis, I start with the static unbalanced panel model, using OLS regression. I perform the analysis for Africa, Asia, Latin America and the overall developing world separately to detect any differences, however I only report results for Africa in the main body of the paper and the results of the developing world in the Appendix section. Due to the nature and limited amount of the data and the non-consecutive reporting of the fintech variables, a dynamic panel estimation is not feasible. A GMM estimator can only be used with ATM as the dependent variable, hence I do perform this estimation for a robustness test. I perform a Housman test to determine whether to use a fixed or random effects model. Due to the likelihood of heteroskedasticity present in the data, I re-run the regressions with robust standard errors. Finally, to check whether there truly are statistically significant differences between the studied outputs for different legal systems, I perform the Wald test.
3.2.1.1 Dependent variable
Firstly, the density of ATMs with regards to population is studied to proxy for the use of an established medium of financial technology. The dependent variable, ATM, stands for automated teller machines per 100,000 adults as recorded by the World Bank (2004-2017). Although primary interest lies in more evolved uses of financial technologies, the number of ATMs provides a good base in establishing a relationship between the legal origin and fintech. ATMs are still relevant for newer technologies, as for instance users of schemes such as E-wallet and M-PESA in a number of African countries can now withdraw cash from an ATM through the use of a mobile phone and an identification number (Mothobi and Grzybowski, 2017). As a more and longer established financial technology, the measure has been documented for a longer period and hence provides a more bespoke indicator, used as a proxy for financial innovation by for instance Sichei and Kamau, 2012.
previous state, for instance Tufano (2003) estimates that the introduction of ATMs has a potential to dramatically decrease transaction costs, up to 10 times. Muthinja and Chipeta (2018) point out that it is especially due to transaction costs rising for banks on a global scale that they will continue to increase the number of ATMs in place.
Saloner and Shepard (1995) examine the network effect of banks mounting ATMs. The authors confirm that the more ATMs are in place, the higher is their value for the account-holding clients. In the time of this empirical study, the value to clients was measured mainly by the dispersion of the ATMs of the bank where they held an account, however nowadays arguably most ATMs, even of a specific bank, allow operations to clients of other banks, hence increasing the value in terms of convenience for all clients. As mentioned by Della Peruta (2018), ATMs as a part of banking infrastructure, among other mechanisms, can help to decrease financial exclusion.
3.2.1.1 Independent variable
The main explanatory variable used in the model is the legal origin of a country. Studies generally categorize countries based on a single legal origin (see, e.g., La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 1997; Fowowe, 2014; Yermack, 2018). This research addresses the issue by not only distinguishing countries based on a single legal origin but acknowledges that some countries do not fall under a strict category but rather adopt a mix of legal systems, be it either common, civil, customary or religious law. The distinction is made based on a comprehensive classification and country legal profiles compiled by the World Bank and GlobaLex (2019). Although countries can bear different weights of the aforementioned legal systems, only three overall categories are studied here to maintain reasonable sample sizes, namely: single-system civil or common law countries and a mixed system legal structure countries. The variable Legal is coded as 0 for common law countries, 1 for civil law and 2 for mixed law countries. Civil and mixed law dummies are thus displayed in regressions and the benchmark group comprises of common law countries.
Based on legal origin, Yermack (2018) deduces that countries under the British common law adopt new financial technologies at a significantly higher rate than civil law countries. I expect to find support for this claim in my model, however due to the gap in the literature on mixed legal systems, I expect them to tilt either way.
by the World Bank only from 2013. An alternative approach in this research paper is using selected Worldwide Governance Indicators (WGI), included as the variable Governance, collected throughout the years of early 1990s to 2017 by the World Bank. The indicators used are: Regulatory Quality, Rule of Law and Control of Corruption, and are used separately to avoid multicollinearity. The indicators are aggregated based on more than 30 sub-indicators collected through surveys and professional perceptions, with the methodology outlined by Kaufmann, Kraay, and Mastruzzi (2011). Although the WGI indicators are not directly used in the study of La Porta, Lopez-de-Silanes, and Shleifer (2008) they overlap with the outcomes of legal origins identified in their paper. The measure of the indicators ranges approximately between -2.5 and 2.5, respectively implying a weak or a strong performance of the country under the specific indicator. For clearer comparability, I transform the negative values and normalise the measures, by adding a constant to all the values before applying a log transformation. Due to the range being an approximate, the final transformation works out as ln(4.4+WGI).
Internet connectivity and infrastructure are crucial in enabling the uptake of the more advanced financial technologies as noted by Yermack (2018), Muthinja and Chipeta (2018) and Della Peruta (2018), among others. The number of secure internet servers per 1 million people,
Internet, measured yearly by the World Bank (2010-2018) is predicted to be positively related
to the adaptation of financial technologies, especially electronic and mobile payments considered in the Hypothesis 1b.It is although noted that there are financial mobile phone services that can operate solely through, for instance, text messages and hence internet access is not always crucial.
A prerequisite for the uptake of financial technologies in Africa and the rest of the developing world is the ownership of or an access to a mobile phone. To control for this, Mobilesub represents Mobile cellular subscription per 100 people as measured and collected by the World Bank, with data available through the sample time horizon (2004-2018).
Although the primary premise of this research is to determine whether the legal variables have an impact on the uptake of financial technologies, the emergence of regional research such as Fowowe (2014) suggests that the theory of law and finance might differ across regions as this author has for instance not found an evidence of legal origin being a determinant of financial development in Africa. Separate regional regressions on top of an overall developing world regression are employed and discussed although not reported, in order to study whether there are significant observable differences in the implied relationship across regions. One-way ANOVA tests on the means of selected fintech variables (Table 5) do suggest regional differences.
Andrés, Cuberes, Diouf, and Serebrisky, 2010). The premise is that the effect of legal origin on adoption of financial technologies is conditional on the level of economic development of a country; i.e.: the higher the economic development of a country, the greater is a chance that the country will be able to uptake financial innovations on a larger scale. Economic development, GDP, is measured as a log of GDP per capita in constant US dollars obtained through the World Bank database, to allow for comparability across the sample set.
3.2.2 Hypothesis 1b
The legal origin of a country has an impact on the adoption of new financial technologies. Fintechi,t = α0 + β1 x Legali,t + β2 x Governancei,t + β3 x Interneti,t + β4 x Mobilesubi,t + β3 x
Ln_GDPi,t + εi,t (2)
The estimation method employed to study Hypothesis 1b is identical to the method described for Hypothesis 1a. Due to aforementioned data restrictions, the additional dynamic model estimation performed on the ATM model cannot be used with the other fintech variables.
3.2.2.1 Dependent variables
The dependent variable under Hypothesis 1b constitutes of financial technologies that have been developed more recently, as compared to the more established technology of ATMs used in Hypothesis 1a. The variable Fintech comprises of four indicators collected by the World Bank as part of the Findex database, namely: ‘Electronic payments used to make payments’, ‘Account ownership at a financial institution or with a mobile-money service provider’, ‘Mobile phone used to pay bills’ and ‘Mobile phone used to send money’ (Demirgüç-Kunt, Klapper, Singer, Ansar, and Hess, 2018). The indicators specify the percentage of adult population aged over 15. All of these are available for most of the countries and collected for the years 2011, 2014 and 2017. It is due to the scarcity of data and slight differences in implication of their use that I choose to analyse more than one indicator of financial technologies uptake. The availability of data throughout the regions is summarised in Tables 1 and 2. Although the data is not annual, the 7-year period is a good guide on the direction in which these variables are headed; for most countries, they are gradually increasing.
Weil, 2016; Yermack, 2018). Abor, Amidu, and Issahaku (2018) show that it is the use of a mobile phone, rather than just an ownership of one, that has an effect on inclusive financial growth and lowers the chances of poverty in a study done on Ghana, hence also the variables in this research proxy the use of a mobile phone for payments and sending money.
3.2.2.2 Independent variables
The independent variables used to study hypothesis 1b are the same as in hypothesis 1a and hence follow the same rational. As noted previously, research (Yermack, 2018) suggests that countries under common law experience a larger uptake of financial technologies. Relationship with the legal environment is also expected to be confirmed through additional legal variables: Government Effectiveness, Regulatory Quality, Rule of Law, and Control of Corruption. It is anticipated that internet infrastructure and mobile phone subscription would have a stronger impact on the fintech variables as these are more reliant on secure internet connections and mobile phone usage than ATMs.
3.2.3 Hypothesis 2
Uptake of financial technologies has an impact on financial development of a country.
FDi,t = α0 + β1 x Fintechi,t + β2 x Legali,t +β3 x GDPi,t + β4 x Inflationi,t +β5 x Governancei,t
+εi,t (3)
Due to the nature of the variables studied, there is a concern for possible endogeneity issues of omitted variables or reversed causality as studies such as Ho, Huang, Shi, and Wu (2018) point to an effect of financial deepening on innovation. Financial innovation, encompassing new financial technologies, does not have a universally accepted relationship with financial development in literature yet. The data are modelled as panel data and to overcome possible endogeneity issues on the static panel, the instrumental variable approach is conducted. The instruments are treated as independent of the error term, εi,t,. The variables treated as
3.2.3.1 Dependent variable
FD refers to the measure of financial development of a country. Following the research of Ang
and Kumar (2014) the primary measure of financial development, FD, is determined by ratio of private credit to GDP of a country, aimed mainly to measure the banking sector development. There is yet to be developed a comprehensive and accepted index of financial development that would combine the dimensions of depth, access, efficiency and stability. Private sector credit to GDP represents the depth dimension of financial development. The data is sourced through the World Bank indicators (2004-2017). The data used is reported in a single currency, US dollars, to enable cross-country comparison.
3.2.3.2 Independent variables
The GDP variable in hypothesis 2 is the GDP growth of the given country, obtained through the World Bank Databases (2004-2011). Financial development has been proven to have a positive relationship with economic growth (Bijlsma, Kool, and Non, 2018). On the reversed side, economic growth can be seen as promoting financial development (Voghouei, Azali, and Jamali, 2011). GDP per capita growth is used in Hypotheses 2 to control for the level of economic development. It is noted, from the data collected, that GDP per capita growth experienced a greater variability during the global financial crisis of 2008.
A variable of inflation is included as it is believed that inflation has negative implications on financial development (see, e.g., Huybens and Smith, 1999; Asongu, 2013). Inflation as an annual percentage rate, relating to changes in consumer prices is obtained through the World Bank Databases (2004-2017).
The legal origin dummy, is reintroduced in this hypothesis based on the premise of the original law and finance theory presented by La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1997) as these document an effect of the legal tradition on financial development. Beck, Levine and, Demirgüç-Kunt (2002) study the legal origin as a determinant of financial development and argue that through two channels, political and adaptability, the countries under the British common law are more suited to foster financial development. According to the empirical research by Sayılır, Doğan, and Soud (2018), there is a significant positive relationship between a country’s governance, institutions and financial development. Hence the Worldwide Governance Indicators, Governance, are reintroduced in this hypothesis as well.
3.3 Descriptive statistics
Tables 1-3 summarise the distribution of observations across years, regions and legal systems.
Table 1. Number of countries with dependent variables data in a particular year
Year ATM Account Electronic Mobile bills Mobile send
2004 69 2005 74 2006 79 2007 86 2008 98 2009 99 2010 108 2011 114 75 77 68 71 2012 114 2013 112 2014 110 74 75 63 52 2015 109 2016 99 2017 82 77 77 74 63
Table 1 summarises the number of observations per dependent fintech variable for each year (2004-2017). The sample includes countries from Africa, Asia and Latin America.
Table 1 shows the number of countries with available data for the dependant variables of Hypotheses 1a and 1b, used to proxy for financial technologies. ATM data have been systematically collected by the World Bank for a longer period of time and hence are present for the majority of countries throughout the studied time-horizon. The remaining four dependent variables have only started to be collected recently through the Findex database on a triannual basis. The lowest numbers of country data collected are for the mobile phone used to send money.
Table 2. Number of countries with data available for dependent variables
Region ATM Account Electronic Mobile bills Mobile send
Africa 50 46 45 45 44
Asia 33 19 19 19 16
Latin America 33 22 23 23 22
Total 116 87 87 87 82
Table 2 summarises the number of observations per fintech dependent variable across the regions of Africa, Asia and Latin America. Period covers years from 2004-2017 for ATM and 2011-2017 for the rest of the fintech.
Table 2 shows the distribution of observations available for the aforementioned dependent variables based on the country region. There are more country data available for the ATM measures, 116, than the remaining four, with mobile phone used to send money being the variable with the lowest number of responding countries, 82. The number of countries with available data per indicator in Africa ranges between 44 and 50, whereas the range for Asia is between 16 and 33 and for Latin America between 22 and 33.
Table 3. Number of countries per legal system
Region Common Law Civil Law Mixed Law Total
Africa 15 17 19 51
Asia 12 7 14 33
Latin America 10 21 3 34
Total 37 45 36 118
Table 3 summarises the number of countries in each region based on their legal origin and the total number of countries per region included. The distribution of legal systems in Africa is relatively equal. In Asia the majority of countries included are either following a common or a mixed legal system whereas in Latin America the civil law is prevailing. Overall, civil law has the highest representation with 45 countries, followed by 37 common law countries and 36 from mixed law.
Table 4. Summary statistics, developing world
Observations Mean Std. Dev. Min Max
ATM 1351 25.66 29.92 0.01 313.15 Electronic payment 229 19.74 19.04 0.04 85.44 Mobile bills 205 3.37 4.69 0.03 37.1 Mobile send 186 7.26 10.14 0.06 60.48 Account ownership 226 39.16 23.72 1.52 98.22 Internet servers 917 479.06 5644.77 0.02 155191.3 Mobile subscriptions 1607 71.64 48.38 0.19 321.8 Ln_GDP 1631 7.87 1.28 4.85 11.45 Regulatory quality 1585 1.37 0.2 0.56 1.9 Rule of Law 1589 1.37 0.2 0.58 1.83 Control of corruption 1586 1.38 0.19 0.93 1.91 Private credit/GDP 1522 37.51 33.16 0.02 233.21 GDP growth 1616 2.58 5.97 -62.23 122.97 Inflation 1495 6.66 14.56 -18.11 379.85 Internet Users 1521 21.27 21.01 0.02 93.54
Table 4 presents the developing world summary statistics for key variables: number of observations, mean, standard deviation, minimum and maximum. Data are provided for period 2004-2017.
Table 5 summarises the mean values of the dependent variables used across the sample, categorised according to the geographical region and legal origin of countries. No direct conclusions should be drawn, however the summary does show a pattern that is worth exploring in a more rigorous setting in further regression analysis.
Solely for the ATM measure, overall civil law countries seem to perform better. With the exception of the established financial technology medium, the ATM, all of the other measures of fintech show higher means, and thus higher uptake for countries following a common law system when looking at the total values for the developing world. Africa is the only region where this pattern of common law countries displaying higher average measure prevails across all of the fintech indicators. Asia and Latin America vary, as for example in the mobile phone used to send money, civil law countries seem to outperform the other two categories. Important note is that the mixed system in Latin America for other than ATM is represented by data available from only one country, Puerto Rico, hence skewing the mean results. It is interesting to note that unlike for other indicators, Africa seems to outperform the other two regions in the use of mobile technology as shown by higher averages in both, mobile phone used to pay bills and to send money. A phenomenon noted by researchers is that throughout the developing world, Africa reports higher average use of mobile money with as much as half of these users being unbanked in other formal ways (Demirgüç-Kunt and Klapper, 2013).
Table 5. Mean values
Common Civil Mixed Total
ATM per 100,000 adults
Africa 16.07 7.05 9.88 10.95
Asia 22.05 50.50 22.50 28.36
Latin 47.29 41.93 26.72 42.63
Total 27.46 31.46 16.66 25.66
Account ownership at a financial institution or with a mobile-money service provider
Africa 44.14 24.33 22.25 31.02
Asia 58.16 48.67 46.81 50.51
Latin 76.54 41.49 69.74 44.25
Total 49.71 37.35 32.05 39.15
Electronic payments used to make payments
Africa 25.94 13.48 10.20 16.86
Asia 32.42 20.79 17.57 22.62
Latin 23.71 21.88 54.55 22.56
Total 27.33 19.21 13.50 19.73
Mobile phone used to pay bills
Africa 5.87 3.37 3.43 4.41
Asia 3.38 2.33 2.50 2.69
Latin 2.70 1.88 9.82 2.09
Total 5.05 2.37 3.19 3.37
Mobile phone used to send money
Africa 15.19 9.86 7.33 10.99
Asia 3.14 3.63 2.60 3.03
Latin 1.27 2.19 0.12 2.05
Total 12.27 5.10 5.61 7.26
4. Empirical results
4.1 Legal origin and fintech adoption
4.1.1 ATM
Table 6. ATM, Africa
(1) (2) (3) (4) (5) (6) Civil Law -9.812** -9.768*** -12.015*** -11.877*** -10.589*** -10.277*** (3.895) (3.720) (3.838) (3.592) (3.916) (3.679) Mixed Law -6.539* -6.443* -7.704** -7.448** -7.048* -6.791** (3.649) (3.477) (3.616) (3.380) (3.654) (3.425) Internet 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Mobile subscriptions 0.094*** 0.092*** 0.100*** 0.095*** 0.097*** 0.093*** (0.011) (0.011) (0.011) (0.011) (0.011) (0.011) Ln_GDP 0.747 1.603* 1.110 (0.885) (0.870) (0.848) Rule of Law 6.442 5.922 (4.169) (4.416) Regulatory Quality -5.395 -6.331 (3.876) (4.057) Control of Corruption 2.677 3.199 (4.413) (4.461) Constant 0.793 -3.870 16.748*** 6.588 5.850 -2.803 (6.262) (7.641) (5.921) (7.503) (6.591) (8.427) Observations 298 297 298 297 298 297 Countries 46 46 46 46 46 46
Table 6 provides results for OLS panel regression for Africa. The dependent variable is the number of ATMs per 100,000 adults. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. All models include number of secured internet servers per 1 million people and number of mobile cellular subscriptions per 100 as controls. Log of GDP per capita is included in models (2), (4) and (6). Rule of law is a control for models (1) and (2), Regulatory quality for (3) and (4) and Control of corruption for (5) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level.
The legal origin is coded so that the base group is the common law system. The dummy variables of comparison displayed in the regression tables are thus civil and mixed legal systems. Table 6 presents the output for the uptake of ATMs and its determinants. Across the 6 model specifications a clear pattern emerges. The negative statistically significant coefficients on the legal dummies suggest that the uptake of ATMs is indeed lower for both civil and mixed legal systems as compared to the common law. The results with robust errors, enclosed in Appendix confirm the statistically significant result only for the civil law countries, with negative coefficient remaining for mixed law however no longer significant.
relationship between financial innovations and bank competition. The higher adaption of ATMs in common law countries can thus be explained by greater bank competition.
The number of secure internet servers has a positive effect on the fintech uptake, as hypothesised, however the relationship is economically weaker than expected. The low in value, albeit significant coefficients can be due to missed observations prior to 2010 when the number of secured internet servers could have grown at a higher pace.
Mobile subscriptions have a significant positive effect throughout the model. The results suggest the relevance of mechanisms highlighted by Mothobi and Grzybowski (2017) of the use of mobile phones to withdraw money from an ATM even without a formal bank account. Positive effect of higher economic development is only significant when controlled for regulatory quality.
Repeating the regression with including the civil and common law dummies and omitting mixed law allows me to perform Wald test and confirm that there are significant differences in the uptake of ATM between civil and common law countries across all model specifications. Civil and mixed law systems did not prove to be significantly different based on the Wald test, hence the emphasis of the discussion is on the differences between common and civil law countries.
4.1.2 Electronic payments
Table 7. Electronic payments, Africa
(1) (2) (3) (4) (5) (6) Civil Law -11.346** -11.340** -11.150** -11.204** -12.037*** -12.175*** (4.549) (4.672) (4.449) (4.559) (4.432) (4.538) Mixed Law -15.142*** -15.891*** -14.993*** -15.780*** -15.046*** -15.826*** (3.941) (4.067) (3.868) (3.988) (3.747) (3.872) Internet 0.001 0.000 0.001 0.000 0.001 0.000 (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) Mobile subscriptions 0.191*** 0.278*** 0.183*** 0.270*** 0.204*** 0.292*** (0.053) (0.079) (0.050) (0.077) (0.052) (0.078) Ln_GDP -3.881 -3.804 -3.883 (2.754) (2.738) (2.722) Rule of Law -7.332 -6.935 (12.417) (12.929) Regulatory Quality -5.652 -5.771 (11.627) (12.140) Control of Corruption -15.730 -16.457 (12.563) (12.989) Constant 20.012 41.138* 18.264 39.579* 30.174* 52.788** (15.841) (22.941) (15.168) (22.504) (15.838) (22.936) Observations 103 102 103 102 103 102 Countries 42 41 42 41 42 41
Table 7 provides results for OLS panel regression for Africa. The dependent variable is Electronic payments used to make payments. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. All models include number of secured internet servers per 1 million people and number of mobile cellular subscriptions per 100 as controls. Log of GDP per capita is included in models (2), (4) and (6). Rule of law is a control for models (1) and (2), Regulatory quality for (3) and (4) and Control of corruption for (5) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level. Standard errors in parentheses.
Table 7 depicts the output for Africa using electronic payments as the dependent variable; first of the 4 variables of Hypothesis 1b. The legal origin dummies mirror the results of the ATM regression in direction, with the difference that mixed law countries seem to be even less adapt to uptake of electronic payments than the civil law countries. Mobile phone subscriptions have again a strong positive effect on the uptake, however internet servers do not seem to have an effect and neither do the governance indicators. The robust results displayed in the Appendix section reinforce the results with notable difference only in the size of standard errors. An additional Wald test confirms that there is a statistically significant difference between common and civil law countries but not between civil and mixed.
which helps to promote financial development. Electronic payments have been established for a shorter period of time than ATMs, hence more recent regulatory changes needed to be adapted. A greater degree of flexibility has found to be a driver of faster functional regulatory reforms that can enable the uptake of financial innovation (Bernier and Plouffe, 2019). The positive effect of mobile subscriptions is confirmed and in addition positive effect of secure internet servers is reintroduced.
4.1.3 Account at a financial institution or mobile-money service provider
Table 8. Account ownership, Africa
(1) (2) (3) (4) (5) (6) Civil Law -14.156** -13.948** -14.744*** -14.548*** -14.580*** -14.409** (5.706) (5.714) (5.636) (5.612) (5.629) (5.601) Mixed Law -17.646*** -17.661*** -18.170*** -18.093*** -18.048*** -18.073*** (4.989) (4.996) (4.948) (4.931) (4.764) (4.772) Internet 0.000 0.000 0.000 0.000 0.000 0.000 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Mobile subscriptions 0.298*** 0.287*** 0.307*** 0.291*** 0.303*** 0.287*** (0.059) (0.076) (0.056) (0.074) (0.057) (0.075) Ln_GDP 0.610 0.781 0.819 (2.945) (2.930) (2.921) Rule of Law 3.825 5.112 (14.231) (14.711) Regulatory Quality -1.015 0.618 (13.834) (14.269) Control of Corruption 0.871 2.204 (14.812) (15.149) Constant 13.574 8.160 19.584 12.814 17.321 10.712 (18.327) (24.951) (18.201) (25.015) (19.038) (25.422) Observations 101 100 101 100 101 100 Countries 42 41 42 41 42 41
Table 8 provides results for OLS panel regression for Africa. The dependent variable is Account ownership at a financial institution or with a mobile money service provider. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. All models include number of secured internet servers per 1 million people and number of mobile cellular subscriptions per 100 as controls. Log of GDP per capita is included in models (2), (4) and (6). Rule of law is a control for models (1) and (2), Regulatory quality for (3) and (4) and Control of corruption for (5) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level. Standard errors in parentheses.
as well. The robust model confirms the result of the regression in Table 7. Performed Wald test confirms significant differences between civil and common law countries but again not between civil and mixed.
The academic literature provides some explanations for my findings. Account ownership essentially increases access to finance. Higher degree of banking access and financial inclusion are identified as determinants of financial innovation in the study of Fontin and Lin (2019). The argument falls back on the research of La Porta, Lopez-de-Silanes, and Shleifer (2008) that links better access to finance as one of the characteristics of the common law, which stems from better investor protection.
The results for the developing world support the output for Africa, only the civil law countries perform worse than the mixed law countries, unlike in the African sample. Under model specification (1) the higher confidence recorded for the rule of law, which includes the enforcement of contracts, leads to higher subscriptions for an account. In addition higher GPD and more secure internet servers also translate to higher account ownership.
4.1.4 Mobile phone used to pay bills
Table 9. Mobile phone used to pay bills, Africa
VARIABLES (1) (2) (3) (4) (5) (6) Civil Law -3.152* -3.299* -2.685 -2.802 -2.875 -2.986 (1.898) (1.928) (1.934) (1.970) (1.899) (1.931) Mixed Law -3.652** -3.672** -3.346** -3.374** -3.195** -3.192* (1.625) (1.648) (1.671) (1.699) (1.605) (1.636) Internet -0.000 -0.000 0.000 0.000 0.000 0.000 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Mobile subscriptions 0.037* 0.047 0.029 0.039 0.034 0.044 (0.022) (0.033) (0.021) (0.033) (0.022) (0.033) Ln_GDP -0.497 -0.521 -0.515 (1.161) (1.183) (1.171) Rule of Law -8.402* -9.027* (4.908) (5.090) Regulatory Quality -5.378 -5.877 (4.794) (5.004) Control of Corruption -8.085 -8.576 (5.166) (5.336) Constant 14.588** 18.363** 10.988* 14.729 14.160** 17.851* (6.220) (9.309) (6.232) (9.433) (6.535) (9.566) Observations 90 89 90 89 90 89 Countries 42 41 42 41 42 41
Table 9 provides results for OLS panel regression for Africa. The dependent variable is Mobile phone used to pay bills. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. All models include number of secured internet servers per 1 million people and number of mobile cellular subscriptions per 100 as controls. Log of GDP per capita is included in models (2), (4) and (6). Rule of law is a control for models (1) and (2), Regulatory quality for (3) and (4) and Control of corruption for (5) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
The results for the frequency of using mobile phone to pay bills are displayed in Table 9. It is worthy to note again that the models for the use of mobile phone to pay for bills or to send money encompass fewer available data points than the models with remaining dependent variables. The effect of being under civil or mixed law country is still negative compared to common law, however the coefficients are notably smaller than for the models with previously used fintech variables. Effect of mixed law remains significant throughout the model specifications, however the negative effect of being under civil law is only significant for specifications (1) and (2). A complementary Wald test indeed confirms significant differences between common and civil law only in regressions (1) and (2). A more striking result is that under these specifications, the rule of law also becomes significant however in an opposite direction than expected. With robust errors, the relation is confirmed in model (2) however the coefficient of civil law turns insignificant, although still negative.
The rule of law encompasses the confidence in obeying the rules of society including quality of contract enforcement, aggregated from over 30 different data sources. My results thus partially contradict the research of Asongu and Nwachukwu (2016), where the institutional governance represented by the same indicator of rule of law is positively associated with mobile phone penetration is Sub-Saharan Africa. My research includes the use of a mobile phone for financial services, not just an ownership of one, which can account for the observed differences. Although outside views or government proclamations might build up higher perceptions of the rule of law, mistrust to these proclamations by the locals might be a deterrent in using a mobile phone to pay bills. Partial explanation can also be drawn from the research of Brinkerhoff, Wetterberg, and Wibbels (2018), who through the use of survey data from Afrobarometer on 17 African countries that conclude that rural citizens tend to trust the government more than the urban population. Since the share of urban population is globally declining, albeit at a slower pace in Africa, higher mistrust of the urban population in government can reflect in the negative coefficients in my results.
4.1.5 Mobile phone used to send money
Table 10. Mobile phone used to send money, Africa
(1) (2) (3) (4) (5) (6) Civil Law -6.588* -6.767* -5.540 -5.662 -7.154* -7.340* (3.920) (4.029) (3.955) (4.068) (3.762) (3.858) Mixed Law -9.066*** -9.226*** -8.193** -8.398** -8.829*** -8.915*** (3.396) (3.501) (3.441) (3.553) (3.188) (3.297) Internet -0.001 -0.001 -0.000 -0.000 -0.001 -0.001 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Mobile subscriptions 0.087* 0.108* 0.070 0.093 0.097** 0.115* (0.045) (0.061) (0.043) (0.060) (0.043) (0.060) Ln_GDP -1.148 -1.276 -1.054 (2.270) (2.299) (2.230) Rule of Law -12.062 -12.576 (10.124) (10.599) Regulatory Quality -4.229 -4.271 (9.984) (10.548) Control of Corruption -20.066* -20.702* (10.302) (10.715) Constant 24.644* 32.126* 15.041 22.610 34.511*** 41.664** (12.965) (18.745) (13.104) (19.095) (13.118) (18.647) Observations 94 93 94 93 94 93 Countries 42 41 42 41 42 41
Table 10 provides results for OLS panel regression for Africa. The dependent variable is Mobile phone used to send money. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. All models include number of secured internet servers per 1 million people and number of mobile cellular subscriptions per 100 as controls. Log of GDP per capita is included in models (2), (4) and (6). Rule of law is a control for models (1) and (2), Regulatory quality for (3) and (4) and Control of corruption for (5) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level. Standard errors in parentheses.
Table 9 displays the model for the final fintech dependent variable, mobile phone used to send money. Mixed law systems consistently show a negative significant effect throughout the 6 specifications, whereas the negative effect is significant under civil law in the model variations (1), (2), (5) and (6). The same models specifications also confirm positive effect of mobile phone subscriptions, which however prevail across the 6 specifications under robust errors. Wald test confirms statistically significant differences between common and civil law, except in models (4) and (5) when controlled for the regulatory quality. The robust model confirms the significant negative effect of mixed law, but not of the civil law and denies the significant effect of control of corruption.
regression with robust errors. In addition regulatory quality and GDP also display significant negative effects in models (3) and (2) respectively. The negative effect of governance should be taken with a grain of salt due to the sample size. However the explanation might simply be in the empirical research design. Sassi and Ali (2017) study the determinants of corruption in Africa. The authors show that although countries should benefit from lower corruption as a result of diffusion of financial innovations, however, this benefit is conditional on the strength of law and regulations, revealing a more complex non-linear relationship.
Across the sample of different dependant variables, the consistent result for the region of Africa is that both, civil and mixed legal systems, perform worse than a common law legal system when it comes to the uptake of financial technologies. The relation gets weaker for the last two regressions. These findings are in line with the preliminary research of Yermack (2018) and in support of the proposed Hypotheses 1a and 1b.
4.2 Fintech adoption and financial development
To employ the second hypothesis I conduct the instrumental variable approach to mitigate the endogeneity issues of the model. Due to the unclear direction of causality between financial technologies and financial development, I instrument the fintech variables with the number of internet users and mobile phone subscribers as reported by the World Bank.
Table 11. Correlation matrix
Variables (1) (2) (3) (4) (5) (6) (7) (1) Mobile subscribers 1.000 (2) Internet users 0.745 1.000 (3) ATM 0.733 0.740 1.000 (4) Electronic payment 0.519 0.668 0.555 1.000 (5) Account ownership 0.595 0.708 0.722 0.767 1.000
(6) Mobile – pay bills 0.126 0.173 0.060 0.428 0.388 1.000
(7) Mobile - send money -0.101 -0.163 -0.225 0.153 0.041 0.638 1.000
Table 10 presents the Pearson correlation matrix between the endogenous fintech variables (3-7) and instruments (1-2). Data cover the period from 2004 to 2017.
4.2.1 Financial development and ATM
Table 12. Financial development and ATM, Africa
(1) (2) (3) (4) (5) (6) ATM 0.851*** 0.855*** 0.852*** 0.804*** 0.821*** 0.812*** (0.071) (0.070) (0.070) (0.070) (0.069) (0.068) Civil law -9.991 -7.829 -8.779 -9.891 -8.053 -8.990 (8.153) (7.841) (8.084) (7.885) (7.502) (7.895) Mixed law -5.853 -4.418 -5.470 -5.592 -4.280 -5.426 (7.551) (7.287) (7.486) (7.296) (6.967) (7.308) GDP growth -0.027 -0.025 -0.026 -0.108* -0.112* -0.115* (0.034) (0.034) (0.034) (0.061) (0.061) (0.061) Inflation -0.006 -0.002 -0.004 -0.012 -0.009 -0.011 (0.016) (0.016) (0.016) (0.016) (0.017) (0.016) Rule of law 0.928 4.646 (7.343) (7.395) Regulatory quality 9.257 11.427* (6.239) (6.250) Control of corruption 7.602 9.464 (6.537) (6.573) Constant 18.865 6.691 9.604 14.576 4.439 7.858 (11.652) (10.389) (10.733) (11.577) (10.243) (10.662) Observations 446 446 446 429 429 429 Countries 44 44 44 44 44 44
Table 12 provides results for the Baltagi 2SLS error component model for Africa. The dependent variable is financial development. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. Under models (1-3) ATMs are instrumented through mobile money subscription and under models (4-6) through number of internet users. All models include GDP growth per capita and Inflation as controls. Rule of law is a control for models (1) and (4), Regulatory quality for (2) and (5) and Control of corruption for (3) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level. Standard errors in parentheses.
Table 12 presents the relationship between financial development and the uptake of ATMs instrumented by the number of mobile phone subscriptions in the first three models and the number of internet users in the remaining, obtained through the World Bank Databases. Both instruments provide a strong case for the positive effect of the uptake of ATMs on financial development. ATMs essentially capture the diffusion of financial innovation by banks. As noted by Voghouei, Azali, and Jamali, 2011, banking innovations play a stimulating role on financial and economic growth, which is in line with my results. Legal origin variables are not statistically significant, however higher regulatory quality does lead to deepening of financial development under model (5). High regulatory quality promotes private sector development and as determined by Rajan and Zingales (2003), quality of regulatory infrastructure is a crucial determinant of a developed financial system.
associated with financial development under both instruments, which is again in line with Rajan and Zingales (2003).
Table 13. Financial development and electronic payments, Africa
(1) (2) (3) (4) (5) (6) Electronic payment 0.162*** 0.208*** 0.192*** 0.140** 0.208*** 0.199*** (0.063) (0.058) (0.057) (0.069) (0.058) (0.060) Civil law -18.957 -11.508 -10.640 -15.119 -9.931 -8.973 (14.387) (16.170) (13.486) (14.883) (18.132) (13.582) Mixed law -8.373 -3.432 -4.755 -5.459 -1.417 -3.031 (12.466) (14.031) (11.646) (12.956) (15.889) (11.775) GDP growth 0.017 0.127 -0.018 0.247 0.428* 0.285 (0.242) (0.239) (0.239) (0.268) (0.243) (0.252) Inflation 0.020 0.118 0.080 -0.011 0.109 0.088 (0.155) (0.161) (0.157) (0.143) (0.143) (0.146) Rule of law -6.615 10.342 (18.385) (19.635) Regulatory quality 23.009 31.376* (19.730) (19.053) Control of corruption 35.827** 38.692*** (14.770) (14.514) Constant 39.988 -4.543 -20.217 15.311 -17.416 -25.837 (26.438) (30.562) (22.495) (28.192) (30.139) (22.479) Observations 90 90 90 73 73 73 Countries 37 37 37 37 37 37
Table 13 provides results for the Baltagi 2SLS error component model for Africa. The dependent variable is financial development. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. Under models (1-3) Electronic payments are instrumented through mobile money subscription and under models (4-6) through number of internet users. All models include GDP growth per capita and Inflation as controls. Rule of law is a control for models (1) and (4), Regulatory quality for (2) and (5) and Control of corruption for (3) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level. Standard errors in parentheses
Table 13 provides similar results with regards to the positive effect of fintech on financial development, with electronic payments being the instrumented variable. Good regulatory quality again leads to higher financial development, in line with Rajan and Zingales (2003). Under this model, high positive effect of control of corruption is observed under both instrument models. Trust in public authorities not to take advantage of their functions for private gain stimulates financial development, which is in accordance with the survey on its determinants conducted by Voghouei, Azali, and Jamali, 2011.
Table 14. Financial development and account ownership, Africa (1) (2) (3) (4) (5) (6) Account 0.247** 0.357*** 0.292*** 0.265** 0.388*** 0.361*** (0.103) (0.095) (0.092) (0.107) (0.096) (0.095) Civil law -12.570 -0.704 -5.726 -8.330 1.780 -1.573 (11.395) (12.192) (11.376) (11.720) (12.277) (11.491) Mixed law -4.989 3.565 -2.289 -1.094 6.637 1.672 (9.826) (10.390) (9.747) (10.128) (10.527) (9.804) GDP growth -0.009 0.169 -0.043 0.185 0.412 0.233 (0.290) (0.292) (0.284) (0.316) (0.305) (0.304) Inflation 0.021 0.197 0.092 0.032 0.208 0.142 (0.171) (0.185) (0.170) (0.157) (0.167) (0.160) Rule of law 5.318 19.142 (19.535) (19.865) Regulatory quality 47.615** 55.893*** (20.907) (20.582) Control of corruption 37.067** 43.025*** (15.741) (15.594) Constant 16.535 -50.365 -29.706 -5.148 -64.268** -42.627* (26.255) (32.203) (23.069) (26.999) (31.774) (23.598) Observations 89 89 89 72 72 72 Countries 37 37 37 37 37 37
Table 14 provides results for the Baltagi 2SLS error component model for Africa. The dependent variable is financial development. Civil Law is a dummy variable equal to “1” if the country is under civil law. Mixed Law is a dummy variable equal to “1” if the country is under mixed law. Under models (1-3) Account ownership is instrumented through mobile money subscription and under models (4-6) through number of internet users. All models include GDP growth per capita and Inflation as controls. Rule of law is a control for models (1) and (4), Regulatory quality for (2) and (5) and Control of corruption for (3) and (6). Standard errors are in parentheses. Period covers years from 2004 to 2017.
*** Significance at 1% level, ** significance at 5% level, * significance at 10% level. Standard errors in parentheses.
Lastly, table 14 repeats the previous models with the account ownership at a financial institution or with a mobile-money service provider as the instrumented variable. Clear positive effect can again be observed on the African sample. As with electronic payments, with the same reasoning, the importance of regulatory quality and control of corruption are important determinants for financial development.
5. Conclusion
5.1 Contributions
The fintech revolution is bound to continue and the developing countries present a tempting market for new financial innovations to develop, lowering costs and potentially helping to alleviate poverty through financial inclusion.
This research is aimed to provide an initial insight into the relationship between legal systems of countries and the uptake of financial technologies. I cover a wide sample of developing countries from Africa, Asia and Latin America. The emphasis is put on Africa, due to the distribution of legal systems, the scope of academic literature and the crucial role amongst developing regions in implementing innovative financial solutions, most notably through mobile phone services. On top of the main taxonomy of legal systems as used in literature, most influentially by La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998), I distinguish three main classifications of legal system families; the common law, the civil law and the mixed law. In most cases civil and mixed law countries are not proven to be statistically significant, hence the highlighted contrasts in analysis are between civil and common law. I identify five proxies for financial technologies. The more established technologies are represented by the number of ATMs installed and more recent technologies consist of electronic payments, the use of mobile phone to either pay bills or to send money and finally the account ownership at a financial institution or with a mobile-money service provider.
Through series of OLS panel regressions I have found evidence in support of the theory that legal origin is an important determinant for the increased adaption of financial technologies, thus confirming the research proposed by Yermack (2018). Specifically in Africa, under the majority of specified models, both civil and mixed law countries underperform as compared to common law countries when it comes to the uptake of financial technologies. The broad explanation is the inherent rigidity of civil law, making it less adept to accommodate the financial innovations in terms of regulatory changes, or from another point discourages innovation due to rigidity and weaker investor protection (La Porta, Lopez-de-Silanes, and Shleifer, 2008). The relationship mostly holds for the rest of the developing world, with the exception of ATMs, where civil law countries outperform common law which is likely due to country-specific policies.
5.2 Limitations and implications
The main limitation of this research is the scarcity of data available to accurately measure the uptake of financial technologies. I believe this to be an emerging and important strand of literature in development finance. Studies examining the implications of financial technologies on financial development, financial inclusion or economic growth have been conducted so far mostly on country-level data with use of local datasets and household surveys. A larger scale study in the future could conduct a meta-analysis of these individual researches to see what drives or constrains the financial innovation in developing countries and whether the relationships hold across different country contexts.
Mixed law countries seem to perform worse than common low in adapting technologies, however the sample of mixed law countries varies in the specific attributes, hence a more local-context research into theses is needed. An implication relevant for civil and mixed law countries is to determine what which specific attributes the common law system allow for easier transfer of financial innovations or give rise to people and businesses being able to come up with and securely offer new financial technologies. Important factors seem to be the relatively greater flexibility of common law, greater investor protection and better enforcement of contracts. La Porta, Lopez-de-Silanes, and Shleifer, 2008 further note that common law countries, due to the above characteristics, are associated with greater financial development which in turn leads to higher responsiveness to investment opportunities. Large reforms can’t however happen overnight and an important first step is to strengthen the infrastructure of institutions as noted, especially for the African countries, by Ahmed (2016).
6. References
Abor, J., Amidu, M., Issahaku, H., 2018. Mobile telephony, financial inclusion and inclusive growth. Journal of African Business 19(3), 430-453.
Agoba, A., Sare, Y., Bugri-Anarfo, E., 2017. Financial inclusion and monetary policy: A review of recent studies. Ghana Journal of Development Studies 14(1), 231-254.
Ahmed, A., 2016. Integration of financial markets, financial development and growth: Is Africa different?. Journal of International Financial Markets, Institutions and Money 42, 43-59.
Alzahrani, M., 2019. Islamic corporate finance, financial markets, and institutions: An overview. Journal of Corporate Finance 55, 1-5.
Andrés, L., Cuberes, D., Diouf, M., Serebrisky, T., 2010. Diffusion of the internet: a cross-country analysis. Telecommunications Policy 34(5-6), 323–340.
Ang, J., Kumar, S., 2014. Financial development and barriers to the cross-border diffusion of financial innovation. Journal of Banking and Finance 39, 43–56.
Asongu, S., Nwachukwu, J., Pyke, C., 2019. The comparative economics of ICT, environmental degradation and inclusive human development in Sub-Saharan Africa. Social Indicators Research, 1-27.
Beck, T., Levine, R. Demirgüç-Kunt, A., 2002. Law and finance: why does legal origin matter?. The World Bank.
Beck, T., Levine, R., Loayza, N., 2000. Finance and the sources of growth. Journal of Financial Economics 58, 261–300.
Bernier, M., Plouffe, M., 2019. Financial innovation, economic growth, and the consequences of macroprudential policies. Research in Economics 73, 162-173.
Bijlsma, M., Kool, C., Non, M., 2018. The effect of financial development on economic growth: a meta-analysis. Applied Economics 50(57), 6128-6148.
Brinkerhoff, D., Wetterberg, A., Wibbels, E., 2018. Distance, services, and citizen perceptions of the state in rural Africa. Governance 31(1), 103–124.
Della Peruta, M., 2018. Adoption of mobile money and financial inclusion: a macroeconomic approach through cluster analysis. Economics of Innovation and New Technology 27(2), 154–173.
Demirgüç-Kunt, A. and Klapper, L., 2013. Measuring financial inclusion: explaining variation in use of financial services across and within countries. Brookings Papers on Economic Activity, 279–321. Demirgüç-Kunt, A., Klapper, L., Singer, D., Ansar, S., Hess, J., 2018. The global findex database 2017: measuring financial inclusion and the fintech revolution. The World Bank.
