There once was a man digging on the side of the road with a shovel. Soon, a merchant in a wagon passed by, offering to sell him a red shovel. “A red shovel digs ditches much better than digging with your hands,” said the merchant. “But I’m not digging with my hands,” replied the man. “I’m digging with a shovel.” The merchant continued to list the virtues of the red shovel, and the man continued to ask how the shovel being red changed anything at all. Soon the merchant threw up his hands in exasperation and drove away.
This story seems increasingly relevant for central bank digital currencies (CBDCs). From The Bahamas to Beijing, central bankers are now exploring, piloting and even implementing these novel forms of money. Their rollout almost seems inevitable.
Advocates of CBDCs often cite financial inclusion as a reason to introduce these currencies, but rarely is a CBDC’s impact on financial inclusion examined in any depth. It is frequently taken at face value that inclusion will be an inevitable byproduct of CBDCs.
But we are not convinced. CBDCs might just be shovels of a different color.
While we at CGAP don’t claim to be experts on the larger macroeconomic issues related to CBDCs, we do have a few questions on whether and how they will enable financial inclusion, particularly in emerging markets. The financial inclusion arguments in favor of CBDCs generally fall into three broad categories: improved access to digital financial services, enhanced efficiency of payments and lower cost. Let’s examine each in turn.
The access argument: CBDCs will be universally accessible and widely used
Many central banks envision a retail CBDC being accessible to every individual and business in the country, supporting efforts like social assistance payments in emergency situations. And it is true that government-to-person payments to mobile wallets have been a lifeline for many in the pandemic. But to use a fully digital currency, you must have a means of both receiving and spending that money digitally. And poor people in emerging markets often lack these options. But more than this, they have not demonstrated a desire to move from a cash economy to a fully digital one. There is little evidence that low-income customers’ behaviors and preferences for cash will change with the introduction of a different form of digital currency.
For example, mobile network operators (MNOs) in Africa have been chasing the grail of digital merchant payments for nearly a decade. In its latest State of the Industry report, GSMA reports that in December 2020 only 4% of mobile money funds in circulation were used to make such payments. In contrast, cash-in/cash-out (CICO) transactions represented 43% of total transaction value. Taken together, these figures reflect the fact that, in many cases, people and businesses with access to mobile money choose to keep using cash for retail purchases. Only a few countries like China, Norway and Sweden have achieved a level of digitization where the need for cash and CICO access points appears to be declining. Even in these countries, regulators hesitate to completely phase out cash for fear that this would adversely affect excluded segments that still use it, like rural populations and the elderly.
The bottom line is that the poor largely live in the cash economy. Reaching them with financial services requires more than just having e-money or a CBDC on a phone. It requires last-mile distribution networks and liquidity management to bridge the digital and cash economies, and an understanding of consumer behaviors to design customer-centric products that ensure adequate protections.
So here is our first question: as regulators design CBDCs for emerging markets, how do they plan to manage the interface between the cash and digital economies?
The efficiency argument: CBDCs will reduce dependence on physical cash, thereby making the distribution of money more efficient
For the reasons above, it seems unlikely that CBDCs will replace cash for low-income households anytime soon. A CBDC is more likely to be a substitute for other digital stores of value. This means that the right question to ask is whether they will be more efficient to use than a tool many poor people already have.
It is hard to see how CBDCs, as currently designed, will be a more efficient payment mechanism for the poor than other digital solutions, such as the issuance of e-money by non-banks. A retail CBDC could be as anonymous and free-moving as cash. But this is not where most regulators are headed, largely for reasons of maintaining financial integrity.
Identified accounts will still likely be required in order to ensure AML/CFT compliance. And this means that customers will be required to provide proof of identification, which is another major impediment to on-boarding unbanked customers. Today, an estimated 1 billion people around the globe and 45% of women in low-income countries do not have any official identity.
It seems to us that countries introducing a CBDC will be required to operate two and possibly three distinct money systems in the near to medium term — traditional fiat currency and e-money and now also the CBDC — while not dramatically changing the experience for customers who are using accounts today. Even if a CBDC successfully displaces e-money systems, it is unclear when and how it would displace cash, which is where the real efficiency gains would come.
Second question: how do we quantify the efficiency gains of CBDC relative to current alternatives such as e-money systems, and what is the assumed timeline for displacing cash?
The cost argument: CBDCs will be cheaper than other digital financial services
A major criticism of today’s retail financial services is that many of them carry high costs for the poor through fees, commissions and other ancillary costs. Advocates hope that CBDCs emerge as a cheaper alternative and help the poorest segments enter formal financial systems.
However, today’s CBDC systems are built on top of existing delivery channels: branches and agent networks established by banks, e-money issuers and other providers. CBDCs rely on these established channels to continue facilitating customer onboarding, distribution, and transactions. So if provider cost structures for activities like onboarding and distribution aren’t expected to change, it is fair to ask why we would expect the consumer economics of retail CBDCs to look any different than existing products.
Some have argued that a central bank could offer these accounts directly to citizens. However, the operation of retail financial services is not often a strong suit of central banks; those M-PESA agents in rural Kenya are not going to be replaced by Central Bank of Kenya agents any time soon. And the implied subsidy involved in contracting out distribution to the private sector would likely be considerable. It is hard to see how this solves the cost problem.
Our final question is therefore: once compliance and last-mile distribution costs are taken into account, how does the cost of a CBDC compare with the cost of tools like e-money? Or with cash?
There is a lot of hope that retail CBDCs will add value for the underserved. However, where financial inclusion is an objective of CBDCs, we should ask whether such a currency actually improves the value proposition of digital financial services for the excluded consumer, compared to existing alternatives like e-money issued by non-banks.
If the same market participants are engaged in the same activities with the same costs, risks, and compliance requirements, then we have to ask: is this technology moving us closer to our goal, or just changing the color of the shovel?