The diffusion of fintech, financial inclusion and income per capita

Abstract

Advances in information and communication technology (ICT) have provided a platform for the introduction and diffusion of a range of financial technologies that have transformed the financial sector. This study analyses the diffusion of financial technology (fintech) and its interaction with financial inclusion and living standards (GDP per capita). We consider the determinants and effects of technology diffusion in financial services and identify two possible transmission mechanisms from the financial sector to GDP per capita – a fintech diffusion channel and a financial inclusion channel. We specify the interactions between these two channels and their relationship with income per capita. Our empirical analysis focuses on the diffusion of two enabling fintech innovations: ATMs and associated digital networks; and mobile phones and payments systems. The relationships between fintech diffusion, financial inclusion and GDP per capita are estimated using a panel data set for up to 137 countries over the period 1991–2015 using both cross section and panel techniques, including an error correction model that distinguishes short- and long-run effects. A key finding is that fintech diffusion and financial inclusion have long-run effects on GDP per capita over and above their short-run impact and the effects of investment in fixed and human capital.

KEYWORDS:

• Fintech
• financial inclusion
• income per capita
• growth

Acknowledgements

We are grateful to the participants of the Financial Inclusion and Fintech Conference at the School of Finance and Management, SOAS University of London, on 25th-26 March 2019 and to Pasquale Scaramozzino, two anonymous referees and the editors, for valuable comments and suggestions, the usual disclaimer applies. We acknowledge financial support from the DFID-ESRC [ES/N013344/2] Research Grant on “Delivering Inclusive Financial Development and Growth” under the Growth Research Programme (DEGRP) Call 3, the ESRC-NSFC [ES/P005241/1] Research Grant on “Developing financial systems to support sustainable growth in China - The role of innovation, diversity and financial regulation”, the AXA Research Fund; Department for International Development; Economic and Social Research Council.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 See Batiz-Lazo (2009, 2018), Konheim (2016), Hall and Kahn (2003), and Scott, Van Reenen, and Zachariadis (2017) for analysis of the introduction and diffusion of ATMs and digital networks; Frame and White 2012 for an overview of financial innovations and their diffusion; and Lashitew, van Tulder, and Liasses (2019) on the diffusion of mobile phones, money and payments technology.

2 The Financial Stability Board (2019, 1) defines fintech as, ‘technology-enabled innovation in financial services that could result in new business models, applications, processes or products with an associated material effect on the provision of financial services.’ Leong and Sung (2018, 75) use a similar definition and identify three stages of fintech development – Fintech Mark 1.0 involved the development of enabling technologies that ‘breed related products of financial technology’ reflecting the ‘platform technology’ nature of ATMs, digital networks and security protocols.

3 Madsen and Ang’s study provides valuable analysis of the impact of research and development (R&D) on ideas production or invention for 20 OECD countries. However, the measures they use for the knowledge stock and R&D do not fully capture for R&D in financial services as until recently the collection of data via Business R&D surveys did not cover the financial intermediation sector and the proportion of financial innovations that is patented is low (Dal Borgo et al. 2013).

4 The distinction between invention (or ‘ideas’ production) and diffusion is captured by the ‘paradox of patents’ (Robinson 1956), i.e. that patents are designed to spur invention by slowing down the rate of diffusion (Rosenberg 1972; MacLeod 1991). Invention and innovation are also distinct; inventions only become innovations when diffusion begins, i.e. an invention becomes an innovation when it starts to be utilised in the economy by firms, consumers or government (OECD, Oslo Manual, 2005).

5 Another indicator used in the literature is the percentage of respondents who report personally using a mobile money service in the 12 months preceding the survey (see, Allen et al. 2016). We did not use this indicator due to the limited coverage of this indicator in our dataset.

6 It may be argued that survey respondents are educated people who, on average, have higher income. Thus, using the percentage of respondents who report having an account at a bank or another type of financial institution as indicator of financial inclusion might be biased. However, the Global Findex database is drawn from nationally representative samples using random selection techniques. The target population is the entire civilian, non-institutionalised population age 15 and above without targeting the level of education. Respondents are randomly selected within eligible households by following the Kish grid (eligible households are selected using random technique as well).At present, this is the best source to assess the proportion of the population with a bank account because it is based on the demand side and not collected from suppliers. In developing countries, financial inclusion measured on the supply side is biased due to duplication of the number of bank accounts (the same person has multiple bank accounts in different financial institutions).

7 For a system to be identified, the number of exogenous variables in all the equations, including the instruments (any additional variables), minus the number of exogenous variables in each equation of the system must be greater than or equal to the number of endogenous variables. In other words, there must be at least as many noncollinear exogenous variables in the remaining system as there are endogenous right-hand-side variables in an equation. When the number of exogenous variables is greater than the number of endogenous variables, the system is overidentified and we use the Hansen’s J statistic to determine the validity of the overidentifying restrictions. Tests of overidentifying restrictions test whether: (i) the instruments are uncorrelated with the error term; and (ii) the equation is mis-specified and one or more of the excluded exogenous variables should in fact be included in the structural equation. The p-value associated with the Hansen’s J statistic should be greater than 0.10 in all cases. In addition, the F-statistics calculated after the first stage of the estimation – and usually used to check whether the estimated models are globally significant – is used to test the existence of weak instruments. These F-statistics are expected to be greater than 10 (see Staiger and Stock 1997).

8 This is preferred to the Sargan test which is not robust to heteroskedasticity.

9 The maximum number of observations is 3,425, i.e. 137 countries over 25 years from 1991 to 2015, but there are missing observations for some countries and years.

10 The results are available upon request from the authors.

11 This equation does not account for country heterogeneity in parameters. Country effects are captured only through the fixed effects. This assumption can be strong, however, the results of the estimates do not invalidate our previous findings based on cross-sectional analysis.

12 The results are available upon request from the authors.

Additional information

Funding

We acknowledge financial support from the ESRC-DFID [ES/N013344/2] Research Grant on ‘Delivering Inclusive Financial Development and Growth’ under the Growth Research Programme (DEGRP) Call 3, the ESRC-NSCF [ES/P005241/1] Research Grant on Developing financial systems to support sustainable growth in China - The role of innovation, diversity and financial regulation'; Department for International Development; Economic and Social Research Council.

Notes on contributors

Désiré Kanga

Dr Désiré Kanga, Research Fellow, School of Finance and Management, SOAS University of London and Economist at the IMF.

Christine Oughton

Professor Christine Oughton, Professor of Management Economics, School of Finance and Management, SOAS University of London and Fellow of the Academy of Social Sciences.

Laurence Harris

Professor Laurence Harris, Emeritus Professor, School of Finance and Management, SOAS University of London and member of the African Economic Research Consortium.

Victor Murinde

Professor Victor Murinde, AXA Chair in Global Finance and Director of the Centre for Global Finance, School of Finance and Management, SOAS University of London.