Advancing Financial Inclusion in Developing Countries with CBDCs
Monica Singer (MS): Dear Jean-Michel, in your interview published by “OECD On The Level” on 11th January 2021, you said in conclusion: “CBDCs have very strong potential to foster financial inclusion in developing countries where CBDCs could replace mobile money”. Could you please elaborate?
Jean Michel Goddefroy (JMG): A major financial revolution is currently under way in the developing world, but it has been relatively ignored by western economists; this probably explains why the link between CBDCs, mobile money and financial inclusion is rarely made. The catalyst for this revolution is the mobile phone, which is now widely available everywhere, including in the poorest countries. In Guinea for example, there are more mobile phones in circulation than there are people in the country. In most developing countries, the mobile phone penetration rate in the adult population is higher than 80%.
However, in many developing countries, few people have a traditional bank account. Until the end of the 2000-decade, access of the general population to financial services (which we call “financial inclusion”) progressed very slowly because banks considered that it was unprofitable for them to open accounts for the poor, or for those located in rural areas. This situation changed in 2007, when Safaricom, a Kenyan subsidiary of Vodafone, launched M-Pesa, a completely new way of making electronic payments, based on the use of mobile phones. In this new business model, telecom companies open electronic wallets to mobile phones owners and allow them to exchange banknotes and coins for a digital form of money deposited on the wallets. This new form of money is called mobile money; for payments specialists, mobile money is a form of electronic money issued by telecom companies. The new system was immediately successful and other telecom companies, such as Orange and MTN started similar projects covering most countries where the banking sector was too small to cover the population’s needs.
MS: What is the difference between “mobile money” and “mobile payments”?
JMG: Mobile payments are payments where mobile phones are used as access devices to the payment machinery. In developed countries, mobile payments typically transfer bank money, i.e. funds held in bank accounts, using “instant payment schemes” for example. Mobile money does not refer to the access device, but to the asset used as currency, i.e. funds managed by mobile phone companies. But, of course, payments made with mobile money are a form of mobile payments. However, most transactions made with mobile money are not payments.
MS: Could you please elaborate? Why do you say that most mobile money transactions are not payments
JMG: I do not want to be pedantic. I just want to show that access to mobile money is just a first step in the financial inclusion process. A payment can be defined as the transfer of a sum of money, in exchange for ownership of goods or access to services. In mature economies funds transfers are mostly payments. But in developing countries, mobile money is mostly used for funds transfers which are not payments. By far the most extensive use of mobile money is to process cash-in/cash-out transactions, in order to avoid storing banknotes that can easily be lost or stolen. The second use case for mobile money is remittances, i.e. funds transfers between different locations in the country, typically from the capital city to family members in rural areas. To make digital money easily exchangeable against notes and coins, mobile operators have licensed individuals or small shop owners as their local “agents”. In many African capitals, you see lots of small wooden booths on the pavement of the main roads with the logo of a telecom company very visible; you may also see the telecom companies’ logos at the entrance of small shops. These booths and these shops are the physical interfaces between individuals and telecom companies offering access to mobile money and funds transfer services.
What matters is that mobile money is a way to introduce digital money in a population used to cash only. You may be surprised to know that sometimes people go to a shop to withdraw cash from their mobile wallets and then use this cash to purchase goods in the same shop. But sooner or later the shopkeeper and his/her customer realize that funds transfers can become payments. In other words, mobile money is progressively used to make mobile payments. And once customers are used to buy goods without using notes and coins, they start buying services, financial services in particular such as insurance services, savings services or credit services. For instance, mobile money facilitates the granting of micro-credits which have allowed many persons (mostly ladies) in rural areas to run their own business and feed their families. Financial inclusion is a virtuous circle and, so far, mobile money has been the start of it.
MS: You have explained that mobile money schemes are very much praised for their ability to foster financial inclusion. But it is not clear why you are in favour of introducing CBDCs to change a business model that works well?
Introducing CBDCs in developing countries would have no relevance if the benefits of present mobile money schemes were threatened. Any new payment solution should improve mobile money schemes, not replace them. My proposal is therefore: keep the infrastructure, including the mobile phones, the telecom networks, the local agents and the wallets, and just replace mobile money with central bank money.
This is not purely theoretical. In the Indian subcontinent, financial inclusion has followed another route which is called “branchless banking”. The infrastructure is the same: telecom companies, local agents, wallets and, mostly, mobile phones. But the money exchanged is bank money, not electronic money issued by telecom companies. So, if the same infrastructure works for bank money and for mobile money, then I see no reason why it would not work for central bank money.
MS: But why do you want to replace mobile money with central bank money? Is it because mobile money is not safe?
I would not say that mobile money is not safe. In most countries, mobile money is entirely backed by short term bank deposits and, therefore, in principle, mobile money is as good as bank money which itself should be as good as banknotes and coins (i.e., central bank money). However, if the telecom companies have problems, or if the custodian banks (i.e., the banks which safekeep the funds used as collateral by the telecom companies) have problems, then mobile money owners would be in trouble. This is not likely to happen very often, but the risk exists. And because mobile money is not covered by deposit insurance, it is safer for depositors, especially the poorest ones, to hold public money rather than private money.
MS: But are you sure that people trust governments more than telcos (or other issuers of electronic money)?
JMG: There is no easy answer to this, especially because the situation may vary from country to country. Psychology may also play a role. Nevertheless, banknotes are rather popular in all developing countries, including in those which extensively use mobile money.
Mobile money is safe, but CBDCs would be safer. This is the first reason why I see value in promoting the use of CBDC in developing markets. Another reason is the ability of central banks to strike the right balance between competition and interoperability. Mobile money has worked better in monopolistic environments where there is only one significant mobile phone company in operation; in other markets, where mobile networks compete with one another, the bigger operator typically has no interest in sharing its network of agents with its smaller competitors. In such cases, interoperability becomes a serious hurdle to ubiquity. Separating the money (under the responsibility of the central bank) from the distribution networks (under the responsibility of the telecom companies) would foster ubiquity and competition because the central bank would license only telecom companies which accept interoperability.
MS: And what about confidentiality? Central banks are public entities. Is there not a risk that governments misuse CBDC data?
JMG: This is an important question, and I would like to make two points in relation to this. First, this risk exists with any kind of electronic payment, whoever is the money issuer, and whoever is running the payments infrastructure. Payment systems should be protected against excessive curiosity from governments, but also against commercial misuse of payment data. My second point is that high-value payments will be increasingly under scrutiny because of the need to fight against money laundering and the financing of terrorism. Public authorities will not allow any new form of money, public or private, which would not be traceable, at least for high amounts
MS: But what about banknotes? What about bitcoins? These are not traceable!
JMG: Banknotes are remains of the past which will fade away progressively. I note that in countries such as France, high amounts cannot be paid with banknotes and that in other countries, such as the UK, only banknotes in small denominations are issued. As for bitcoin, it is primarily a speculative instrument, not a means of payment. Coming back to the confidentiality issue, what matters for financial inclusion is that low-value payments can remain anonymous, whoever is the issuer of the asset used as money.
MS: You have convincing arguments in favour of the adoption of CBDCs in developing countries. However, CBDCs can be implemented without using a DLT. I’d like to turn to YYY, from Consensys, to understand what advantages DLT could bring in a CBDC environment.
Matthieu Saint-Olive (MSO): First, is a difference in transaction speed. DLT-based CBDCs really work as digital surrogates of cash. CBDC tokens are exchanged directly between individuals and corporations without any help from account-holding institutions such as banks. Transactions are final immediately after being concluded, as long as the payer has enough funds in his/her wallet. Exactly like banknotes. No intermediary is used in the process and, as a result, CBDC transactions are processed more quickly with DLT rather than without.
DLT-based CBDCs will not only be processed more quickly; they will also be processed cheaply thanks to the removal of intermediaries and the elimination of reconciliation needs. With DLT, payment systems costs for society will be lower than the use of banknotes and the costs for individuals will be lower than costs associated with bank accounts.
MS: And what about confidentiality?
MSO: I believe it is important to emphasize that we would never store personal identifiable information on a blockchain. When using blockchain technology like in Ethereum, wallet owners are identified by a string of characters such as 0x52B63B0c97ce3E05B9C80a07883F45D30cdF6bf3. In a CBDC context, only the regulated entity that offers wallets to users knows the identity of the individual behind this string of characters.
However, if the ledger is widely accessible, which is the case with public blockchain networks like Ethereum, anyone could have full visibility on the wallet holdings and transaction history. This means that if, one way or the other, you identify the owner of a wallet, you would know his/her past and future transaction history. There are multiple solutions that can be used to overcome this. One of these solutions is for central banks to use permissioned networks and to allow only authorized and regulated entities to browse the ledger and see the wallet holdings and transaction history. Those entities have little reasons to illegitimately identify the wallet owner and the law could regulate what they can and cannot do.
But, as far as financial inclusion is concerned, the most promising option to preserve data privacy will be the issuance of “non-KYC wallets”, wallets which are fully anonymous, not associated with any legal identity. Of course, because of the need to fight against money laundering and the financing of terrorism, these wallets will have to be subject to rather low maximum balances, but this is not likely to be an issue in the context of financial inclusion. This approach is already used in the Bahamas, for example.
MS: What about scalability? Can blockchains be used to process millions of transactions per day?
MSO: Yes, they can. For many months Ethereum Mainnet, the public and permissionless Ethereum network, has processed 15 transactions per second on average, which corresponds to 1,300,000 transactions per day. This current transaction throughput will be increased by a factor x100 or more in the next 2 years with the transition to Ethereum 2. Moreover, it is important to note that CBDCs will most probably be based on private permissioned networks which are more scalable than permissionless networks. For example, ConsenSys Quorum already anticipates processing hundreds of transactions per second, which means tens of millions of transactions per day.
Last but not least, an Ethereum transaction can be much more than a payment transaction. In fact, it is possible to group thousands of payment transactions into a single Ethereum transaction. This technology is called “rollups”; it can handle more than 1 billion transactions per day.
MS: Are there other advantages in using DLT in a CBDC scheme that could be relevant in a developing country?
MSO: To start with, I should like to emphasize that the immutability and verifiability of the ledger could progressively increase trust in cashless payments. This is perhaps a bit theoretical, but in the long run, this should help move away from the excessive dominance of cash.
Another characteristic of DLT-based currencies is that they can be associated with smart contracts which “self-execute” themselves without any need for intermediaries or reconciliation. As an example, cross-border remittances, which are so important for some developing countries, would be facilitated if a simple payment versus payment mechanism could be put in place, using smart contracts.
To conclude, I would like to remind you all that financial inclusion is not only about lowering costs, it is also about broadening user’s opportunities. Our economy is growingly digital, which increases financial exclusion risks to individuals who are only used to holding cash. Digital currencies built on DLT could be used for remittances and other financial products such as savings and insurance at a fraction of current costs and with a better user experience than current banking solutions.