Bermuda has its “sand dollar”; China is well into its
experimentation phase, as is Sweden; the European Central Bank is
seriously analyzing the idea with President Christine Lagarde indicating
a digital Euro could be in use by 2025; and virtually every major
central bank is extensively analyzing the possibility of issuance.
Unfortunately, there is a danger that a number of central banks may
make one or more of six serious policy mistakes, described in this
paper. Our concern is based on extensive discussions with senior central
bankers and other relevant officials, as well as experts outside of the
official sector. In essence, these errors all have their roots in
bureaucratic and political constraints, combined with the novelty of
this instrument, making them more difficult to avoid.
Raising these concerns is not a statement of opposition to CBDCs.
They are a powerful policy tool that can be used, in conjunction with
other policy tools, to achieve a number of objectives. It is up to each
central bank, and other relevant policymakers, to weigh the pros and
cons and to choose the optimal design and implementation for any CBDC
they may issue. We are simply urging the avoidance of some critical
errors.
One note on nomenclature: this discussion focuses on “retail CBDCs,”
meaning ones that are designed for wide use in the economy and to be
held by individuals and businesses broadly, as opposed to “wholesale
CBDCs,” primarily used between financial institutions. For convenience,
we will use “CBDC” in this paper, rather than repeating “retail” each
time.
The six deadly sins
- Siloed decision-making. CBDCs do not exist in isolation; policy
regarding CBDCs, stablecoins, other digital assets, and the Payments
ecosystem must fit together coherently
- Putting off the hard choices about policy objectives
- Ignoring other policy tools to reach these objectives
- Creating the "Swiss Army Knife" of digital currencies
- Downplaying political constraints and forces
- Muddled communications
Siloed decision-making
CBDCs do not exist in isolation and they are only one policy tool of
many to achieve the various policy objectives for which they might be
used. CBDCs will be one part of a complex Payments ecosystem which also
overlaps with digital assets, such as Bitcoin, that are not primarily
payments vehicles. Thus, policy choices need to be coherent across at
least the following areas:
- Modernization of Payments systems
- CBDCs themselves
- Stablecoins
- Digital assets that are not tied to conventional currencies
Few, if any, countries have one authority responsible for all four of
these categories. Even if they do, the different parts of that
authority may not coordinate as well as they should. The US and the
European Union (EU) may be particularly prone to silo problems in this
area, as noted below.
US and EU considerations
The problem in the US is that authority over financial policy is
fragmented. The Federal Reserve (Fed) would be responsible for a digital
dollar. Regulation of stablecoins and digital assets is split between
at least the Securities and Exchange Commission (SEC), the Commodities
Future Trading Commission (CFTC), the Federal Trade Commission (FTC),
and various state authorities. The Office of the Comptroller of the
Currency (OCC) and Federal Deposit Insurance Corporation (FDIC) have
important regulatory authorities over other aspects of the digital
assets ecosystem. Payments policy is set by a mix of the Fed and state
authorities and even at the Fed there is some division between authority
at the Federal Reserve Board and the twelve district banks. And, of
course, the Administration has an important role, including the
authority of the Treasury Secretary as Chair of the Financial Stability
Oversight Council, which has a coordinating role on financial regulation
and related issues. One can easily imagine a number of ways in which
policy across all these bodies might not mesh well.
The EU has some analogous issues with split responsibilities across
different types of authorities. For example, there is currently a
difference of opinion between the ECB and the European Commission and
other Brussels bodies about the ECB’s legal authority to proceed on its
own in regard to CBDCs. Further, the EU has additional issues resulting
from the complexities of the European institutions and their
interactions with national authorities.
There are a number of practical reasons why it is important to
consider the bigger picture, of which CBDC is just a part. First, the
existing banking and wider financial systems are built around the
current roles of the central bank and commercial banks, including the
importance of deposits as a funding source for banks, much of which is
tied to payments activities that could be affected by CBDC. Every
central bank is aware that a CBDC could cause disruptions, potentially
severe ones, to the existing banking system. All the major central banks
have a stated goal of minimizing this disruption, to the extent
consistent with their core reasons for creating a CBDC. The devil,
however, will be in the details, as discussed below.
Second, some of the reasons put forward for CBDCs center on
weaknesses in current payments system and regulation. Therefore, it
would be important to look at the full payments ecosystem, and other
proposed changes to it, when considering issuance of a CBDC or designing
the details. We elaborate on this point further below.
A third set of reasons for CBDC issuance center on a preference by
many policymakers for public sector digital currencies over private
sector ones. The larger the role for a CBDC, the less the need for
stablecoins that effectively mimic a CBDC. However, one could easily
imagine a scenario in which a central bank chooses to create a very
limited CBDC, yet financial regulators hem in potential stablecoins with
a high degree of constraints, leaving a gap in the regulated system
that is filled by neither the public nor private sector, which could
easily encourage unregulated forms of digital currencies. Or the reverse
could occur, where central banks attempt to create an attractive CBDC,
but private sector stablecoins flourish in a lightly regulated
environment that enables them to be more attractive than the CBDC. As
one gets into the details, there are many choices which should be
harmonized between the structure of a CBDC and the regulation and
supervision of stablecoins and other digital assets.
Putting off the hard choices about policy objectives
A typical central bank which is seriously investigating the potential
for a CBDC has issued a white paper in which they describe perhaps
seven future scenarios in which they would want to have a CBDC.
Implicitly, and sometimes explicitly, these central banks do not feel a
CBDC is currently necessary, but rather that it might be in the future
and the very long lead time needed makes it imperative to investigate
now.
The future scenarios usually center on some or all of the following trends or issues:
- Severe decline in the use of cash in an economy
- The need for faster, more sustainable and more resilient payments processes
- The need to increase financial inclusion
- The potential for private sector digital currencies to crowd out central bank money
- The need to improve cross-border payments processes
- The need to decrease fraud and illicit activity
- The desire to facilitate innovation and open competition in a more digitized economy
Moving from top-down to bottom-up approaches, a number of central
banks are running, or have run, technical experiments to understand the
capabilities and constraints of the underlying technologies.
This top-down and bottom-up approach has been very reasonable, but
will soon hit its limits. In particular, it is very hard to have an
honest debate about moving forward with a CBDC, or about its detailed
design, when the objectives are so unclear. A CBDC is a very powerful
tool, but you cannot know whether it is the best tool for the task, or
specifically how to design it, without knowing its purpose. Most central
banks are at the stage of arguing for the need to be ready to create a
tool, without getting into whether that tool should be a hammer or a
screwdriver, much less whether the tasks will require a Philips-head
screwdriver or a standard one, a large screwdriver or a small one, etc.
Central banks need to move from discussing scenarios to choosing
specific policy objectives. If, as is likely, they want to achieve a
number of different objectives, then they also need to prioritize them,
at least roughly. Setting objectives is critical to the next two steps:
(1) determining how CBDCs compare with, and interact with, other policy
tools and (2) making specific structural design decisions.
Implications for structural design
There are a host of high-level structural design choices that
policymakers must decide. These are very briefly described below, as
full explanations would require much more space.
Who should have access to CBDCs?
In particular, will this be a retail CBDC, analogous to cash, or a
wholesale one, limited to use by financial institutions? For purposes of
this paper, we are confining the discussion to retail CBDC.
Will CBDC holders have a direct relationship with the central bank or will intermediaries, such as banks and Fintechs, be used?
Central banks in advanced economies are generally leaning strongly
towards an intermediary model, sometimes using “digital wallets.”
Will the system be account-based, token-based, or some combination?
Holders could be required to prove their identity in order to use
CBDC in an account-based model or could have a digital token, the
equivalent of paper money or coins, in which they only need to hold the
token in order to use CBDC – a so-called “bearer instrument” model.
Which intermediaries will be allowed?
Most central banks explicitly wish to avoid disrupting the commercial
banking system more than necessary. Does this mean only banks should be
the intermediaries or will Fintechs and Big Tech firms be included?
What capabilities should be built in?
Digital currencies can be designed to allow or facilitate the use of
“smart contracts” and “programmable money”. As part of this, a CBDC
could be designed in such a way as to automate the execution of policy,
for example, by delivering “helicopter payments” with automatic expiry
dates as a form of monetary stimulus.
What limits will be in place for users and intermediaries?
Some central banks are exploring caps on the quantity of CBDCs that
any specific individual can hold to limit financial disintermediation.
How this is monitored and enforced may have implications to design.
How will data privacy/anonymity be handled?
There are a multitude of complex considerations in this area,
revolving around which entities will have which data and how they are
permitted to use it.
How will the system discourage money laundering and financial crimes?
This ties into the data privacy question. At one extreme, digital
tokens could be as anonymous as cash, potentially making money
laundering easier. At the other extreme, tokens or accounts could be
tracked closely, reducing crime at the expense of privacy.
Optimal design choices will depend on which policy objectives are
seen as the most crucial and as being consistent with the evolving role
of the central bank. For example, an emphasis on financial inclusion
might argue for a tighter relationship between the individual and the
central bank, such as through an account at the central bank itself. In
contrast, a focus on faster payments might put a premium on the use of
intermediaries, which are already experienced at handling customer
relationships and whose role would allow the central bank to avoid
building a huge new infrastructure of accounts.
Similarly, if the goal is to mimic features of cash like its
anonymity, the choice may be to ensure clear limits to data collection
or guarantee anonymity through technology, for example, by adopting a
token-based system that does not track the identity of the payer. On the
other hand, if there is greater priority on combating illicit financial
activities, such as money laundering and financing terrorism, then a
system that collects more information on accumulation and usage is
preferred.
Ignoring other policy tools to reach these objectives
There are alternative policy tools available to achieve virtually all
of the objectives. For example, many central banks and private sector
providers are in the process of dramatically increasing the speed of
payments or have already done so. In terms of improving cross-border
payments, the Financial Stability Board and the Committee on Payments
and Market Infrastructures laid out 19 tools in their Stage 2 report on
cross-border payments. CBDC is just one of these 19. Similarly, there
are a host of potential tools to improve financial inclusion, many of
them moving forward today.
This is not to argue against CBDCs, but to emphasize that once we
know the policy objectives it will be critical to see what CBDCs are
most useful for, either instead of other tools or, more likely, as a
complement to them. There could be specific instances when a CBDC is the
ideal tool. For example, the Bahamas adopted a CBDC blockchain
structure for its sand dollar to help manage a specific set of
constraints: limited connectivity and sporadic presence of financial
institutions across a widely dispersed archipelago.
Creating the "Swiss Army Knife" of digital currencies
Unfortunately, there is a strong temptation to proceed forward with
design of a CBDC, or at least publicly floating some specific design
choices that may create strong momentum for those particular choices,
prior to accepting the political pain of narrowing down and prioritizing
the objectives. This creates a further temptation to “square the
circle” by loading the technical requirements with a wide array of
demands, so that when policymakers eventually settle on their goals, the
CBDC design will support them.
The problem is that this is likely to result in a “Swiss Army knife”
approach. The Swiss Army knife is a brilliant design, allowing the user
to access many different tools in a compact space. However, it has only a
very small share of the global tool market because it is virtually
always better to use one or two or three tools that are optimized for
the particular task. As with almost all major policy initiatives, the
implementation will be much better if the objectives are clear from the
beginning.
Some may be relying on the magic of innovative technology to rid them
of the need to make these choices now. As good as our technologists
are, it is much more likely that important design trade-offs will
remain. Further, as discussed above, some policy choices are inherently
at odds with others, such as certain trade-offs between anti-money
laundering efforts and privacy requirements. (To be clear, privacy and
AML considerations are not completely in conflict, but there are some
aspects where one has to choose.)
Downplaying political constraints and forces
Choices about CBDC issuance and design will significantly affect the
financial system and larger society. As a result, central banks will not
be able to proceed in isolation or make decisions purely on their views
about the technical merits of different approaches. Central bankers are
aware of this, but there is a danger that they downplay the importance
of politicians and other stakeholders, through a combination of a
legitimate desire to protect their independence in monetary policy and a
culture of trying to hold political considerations at arm’s length.
In general, CBDCs will need to be designed in close consultation with
finance ministries and financial regulators, as well as the broader
public, given their potentially large impacts on the financial system
and critical importance of payments infrastructure to the economy. In
this sense, the division of responsibilities and accountability for CBDC
design is much more akin to the situation for financial regulation than
for monetary policy. In most countries, central banks play a major role
in regulating and supervising the financial sector and are more
constrained and more closely watched by the political authorities for
these activities than for monetary policy. That said, the consultation
will need to be managed carefully to avoid the appearance or reality of
undercutting the independence of central banks in monetary policy.
Moving on to specific decisions, there are two areas in particular
where central bankers may believe they have more freedom than they will
have in practice. First, many central banks are considering placing
limitations on the amount of CBDC that can be held by any one individual
or entity, in order to keep CBDC as primarily a payments vehicle and
not a store of value that competes with bank deposits. However, if the
new CBDC works well, it is difficult to imagine that central banks will
be able to resist the high level of pressure that will arise to raise
the limit to levels high enough that virtually any transaction can be
performed with it. The ECB has floated the idea of limiting a digital
euro to 3,000 euros per holder; one can easily see demands to raise that
to 10,000 or 20,000 or higher, so that individuals and small businesses
can buy automobiles or pay for an expensive trip. This pressure will be
particularly hard to resist as the main reason for the limit is to
avoid excessive disruption of the current banking system, which can
easily be caricatured as the elites protecting the big banks – never a
popular position.
Second, central banks would prefer to pay little or no interest on
their CBDCs. This is partly to avoid disrupting the banking system too
much and partly, one presumes, to preserve the financial benefits to
central banks of being able to issue currency for free. If interest
rates in the economy are very low, as they are currently in some
countries, this may not be a major issue. However, any significant
difference between what banks pay on deposits and what CBDCs accrue is
likely to provoke anger that the “people’s bank” is treating citizens
worse than the profit-seeking banks are.
Muddled communications
Good communications will be critical, given the importance of money
and payments to the economy and therefore the wide range of
stakeholders: the public; politicians; the media; financial
institutions; tech firms; merchants; other types of businesses; etc.
This will be challenging and easy to get wrong, particularly because
there is a wide range of understanding about money across the
stakeholders. We have had conversations with otherwise intelligent and
well-informed individuals who literally thought that when a bank took a
deposit, it placed those funds in a safe deposit box. More commonly, the
great majority of the public makes no distinction between “money” they
hold in a deposit account at a bank and “money” that represents a direct
liability of a central bank. Further, we have already seen media
coverage that equates CBDCs with Bitcoin, despite the potentially huge
differences between the two types of instruments.
Good communications will be particularly important in driving
adoption of a CBDC once it has been designed. This requirement goes
beyond clarity to persuasion. Clear communications are critical, but a
persuasive message also needs to resonate.
The need for good communications is another strong reason why
policymakers need to be clear about their objectives and
prioritizations. If policymakers are not clear, how can we expect the
other stakeholders to understand clearly?
Communications choices can have important implications that may not
be clear on first glance. For instance, most central banks are likely to
go with an intermediated version of CBDC, whereby banks and some other
authorized intermediaries will have the direct relationship with the
holders of the CBDC. One reason for this choice is to take advantage of
the intermediaries’ abilities to handle AML and other duties. However,
if major money laundering problems occur, it may be difficult for
central banks to shield themselves from blame to the extent they can
today, if there has been a strong stress that this is a CENTRAL BANK
digital currency.
As a second example, there is a risk of confusion around the level of
privacy of transacting with CBDCs, as the level of privacy will depend
not only on the choices made in its design, but the country’s broader
privacy and data protection regime. While a CBDC may be designed to be
private from central banks, data and metadata may be captured by
intermediaries and combined with other data sources in ways that violate
citizen’s perceptions of the CBDC’s level of privacy.
Additionally, the introduction of CBDCs may also change expectations
from the public about monetary and fiscal policy. If a CBDC is messaged
as a way for central banks to respond more effectively to economic
challenges, then it may be more difficult to resist calls for central
bankers to engage in novel monetary stimulus measures, such as
helicopter money.
Conclusions
There is clear momentum for retail CBDCs and we believe they are
likely to be built in many jurisdictions this decade. It is important
that policymakers design and implement them well, which is why we have
emphasized six areas where the risks of an error are particularly high.