Digital Financial Services (DFS) is a relatively recent mobile-centric financial inclusion innovation in developing countries. Using ubiquitous mobile phones as the means of service access, DFS provides the unbanked and underserved - many of whom live in rural areas - with access to basic financial services provided by banks and nonbanks such as mobile network operators and third party DFS providers (DFSPs).
With mobile phones as the primary access mechanism for services, access to DFS is highly dependent upon the degree and quality of mobile coverage offered by mobile network operators (MNOs).
User interfaces (UIs) for access to DFS are mostly dependent on the type of mobile coverage available which in many cases in the developing world is via slower (narrowband) second generation (2G) GSM technology. Faster broadband third (3G) and fourth generation (4G) mobile coverage is available in most developing world countries, but is mostly limited to urban and peri-urban areas and along national road corridors.
Our study finds that the type of mobile coverage provided has a significant effect on the DFS UI and type of mobile phone that can be used for DFS access. Feature phones and Unstructured Supplementary Service Data (USSD) transactions continue to be the choice for the vast majority of (mobile money) users. Faster mobile broadband technologies are optimal for smartphone use – or hybrid smartphones in feature phone format - which can provide superior and more intuitive graphical user interfaces.
Indeed, the primacy of 2G coverage in developing countries where DFS is prevalent forces DFS customers to use non-intuitive, coverage-sensitive, text-based UIs -- particularly USSD -- and also STK on only 2G-centric basic and feature phones (or its limited, near equivalent on smartphones.) These UIs and resultant user experiences (UX) are sensitive though to the quality of the mobile coverage and signal, and limit the suite of potential services than can be provided to customers to primarily basic transactional ‘DFS 1.0’ type of services. Inconsistent coverage also forces users to have SIM cards and prepaid accounts for all MNOs they anticipate can provide service at particular locations. Frontline signup and cash handling services are mostly provided by agents who are similarly dependent on availability of mobile coverage of adequate quality.
Use of smartphone-type applications using a graphical menu system can provide a more intuitive UI/UX and well as being more capable of offering a larger suite of services that including easier credit provision and information as well as agent rating and liquidity mapping, but the ability to properly use them is highly dependent on the availability and quality of broadband coverage signals.
We find that the lack of broadband coverage – and fallback dependency on coverage-sensitive 2G-based UIs for DFS provision (primarily via USSD) - handicaps progress to more comprehensive service offerings as well as creating potential competition bottlenecks and system security risks in DFS provision. We define the evolution of DFS offerings as from DFS 1.0 o DFS 4.0, with DFS 1.0 being foundation, basic services such as including simple person-to-person (P2P) value transfers. The net result is that rural residents, for the most part, will have to make do with the foundational DFS services because of their reliance on 2G-based coverage and thus UIs. They will not necessarily have access to the same suite of services available to those in urban and peri-urban areas, invariably perpetuating rural-urban financial service access divides.
Growth and security of, and sustained user interest in the DFS ecosystem at a national level then is inexorably linked to the type and quality of mobile coverage available to current and potential DFS customers. Noting this crucial symbiosis between mobile coverage and the provision and evolution of DFS, this study documents the current state, limitations, and bottlenecks and potential strategies and methods for expanding mobile coverage. The role of regulators and other authorities in the coverage expansion process is also examined.
We look at why there is no impetus to national mobile broadband coverage. Findings from our research suggest that provision of mobile coverage in rural areas in developing countries does not provide an adequate return on investment for MNOs and that in many cases because of expansion costs and costs to upgrade from 2G to 3G and higher services, 2G services will be the only technology provided to rural areas by individual MNOs. Indeed, these MNOs may labor under huge debt – and some have shuttered or merged - in attempting to undertake this expansion, one of the primary reasons regulators we canvassed indicated that they have not mandated national mobile broadband provision by individual licensees. We find generally too that national upgrades of 2G infrastructure by a single MNO for provision of broadband speeds for its own customers provides a poor ROI relative to the CAPEX costs required, which must now also include expensive backhaul provision to maintain broadband quality of service levels.
Instead of ‘solo’ expansion by MNOs from 2G to 3G and higher using their own financial resources, expansion and provision of coverage – be it 2G or higher – in rural areas to achieve near-national mobile broadband provision requires either subsidization of the infrastructure installation and operation; or requires infrastructure sharing between MNOs; or is provided through national wholesale networks; or through provision of turnkey infrastructure by third party tower companies. Further, new ‘digital dividend’ spectrum resulting from a switchover to digital television and also made available by regulators through auctions and sales may significantly enhance the degree and quality of mobile coverage since fewer mobile base stations would be needed to cover wider surface areas. Evolving innovations such as provision of coverage direct to customers by drones, new low-power mobile base stations, micro-satellites and balloons also hold promise but some are far off from commercial provision. New ‘datalite’ smartphones, operating systems and apps that operate relatively efficiently in narrowband environments also hold promise in spurring national DFS 2.0 availability and adoption but are still in early stages of development and maturity.
We find that, besides limiting progression to DFS 2.0 type of services and providing a poor UX, there are also downstream competition and security related effects of not expanding or enhancing beyond the 2G-only coverage. On competition-related concerns, access to DFS via existing 2G-type UIs may be restricted by competing MNOs who control critical USSD or other gateways, with 2G-only coverage likely to be the default coverage for a number of years yet in rural areas in the developing world, competition issues are likely to manifest concurrently and will require regulatory intervention and not as it is often now, forbearance. Security concerns relate to the inherent insecurity of 2G technologies, such as USSD in use for DFS provision, such their use as a primary UI poses risks to the security of the DFS ecosystem as is currently configured. These concerns and current and potential responses from regulators are also discussed.
A stylized graphical matrix of the coverage issues, their implications as well as potential solutions is provided in Exhibit 1. To navigate the study and before beginning the reading of the paper, readers are urged to first view this stylized graphical matrix of the issues which are divided into three parts: A: infrastructure expansion; B: 2G-related issues; and C: Outcomes. We also include specific country-focus annexures on the effect of coverage on DFS in India and Uganda.