CBDC is a black and white issue
I was reading J.P. Koning’s excellent paper [PDF] on Central Bank Digital Currency (CDBC) for Brazil and came across his reference in passing to Narayana Kocherlakota, former CEO of the Federal Reserve Bank of Minneapolis, who wrote (in 2016) that economists do not know very much about the topic of anonymity and "calls for the profession to model it more systematically”. I think this is a really critical point, because the decision about where to set the anonymity dial for a cash replacement product is an important one, and not one that should be left to technologists.
This decision is discussed in the context of implementing a digital fiat currency of one form or another. The paper explores three ways to implement a CBDC for Brazil.
MoedaElectronico (Electronic Cash): this is the most cash-like of the three CBDCs. It pays neither positive interest nor docks negative interest and is anonymous. Like cash, it is a bearer token.
ContaBCB (BCBAccounts): this is the most account-based of the three templates. Ac- counts are non-anonymous and pay interest, like a normal bank account.
MoedaHíbrida (Hybridcoins): provides a mix of cash and account-like features, including the ability to pay a varying positive and negative interest rate, while offering users the choice between anonymity or not.
Now, the first two are well-known and well-understood. I wrote about them again last month (I’ve discussed “BritCoin” and “BritPESA” several times before), in a comment on Christine Lagarde’s speech [15Mb: Central banks, tokens and privacy] and I don’t propose to look at them further here. It’s that last example that interests me.
Let’s go back to that point about anonymity. In the paper J.P. says that the case can also be made for a permanently negative interest rate on anonymous CBDC. Why? Well, since we all understand that criminality and tax evasion impose costs on society, it may be worthwhile to design anonymous payments systems in a way that recoups some of the costs these activities impose.
In other words, construct a cash replacement in which anonymous transactions cost more than non-anonymous transactions. One way to do this, which is referenced by J.P. in his paper, was the “Crime Pays System” or CPS as conceived by the artist Austin Houldsworth. Austin is most well-known for designing the cover of my book "Before Babylon, Beyond Bitcoin" of course, but he also ran the Future of Money Design Award for Consult Hyperion’s annual Tomorrow’s Transactions Forum for many years. Oh, and he was awarded a Ph.D by the Royal College of Art (RCA). It was his idea to have me present CPS at the British Computer Society (BCS). We had my alter ego set out the new payment system to an unsuspecting audience who, I have to say, were excellent sports about the whole thing! It turned out to be an entertaining and enlightening experience (you can read more and see the video here).
In CPS, digital payments would be either “light” or “dark”. The default transaction type would be light and free to the end users. All transaction histories would be uploaded to a public space (we were, of course, thinking about the Bitcoin blockchain here) which would allow anybody anywhere to view the transaction details. The alternative transaction type would be dark. With this option advanced cryptographic techniques would make the payment completely invisible with a small levy in the region of 10% to 20% would be paid per transaction.
The system would therefore offer privacy for your finances at a reasonable price. The revenue generated from the use of this system would be taken by the government to substitute for the loss of taxes in the dark economy.
What a cool idea.
Now, at the time it was just a concept. We didn’t spend much time thinking about how it would actually work (I was basing the pretend implementation for the BCS presentation on Chaumian blinding a la Digicash, hence this gratuitous picture of me influencing David in Vegas.)
That was then. In the meantime, however, along came ZCash and the mechanism of shielded and unshielded transactions that J.P. has used as the basis for MoedaHíbrida's two different modes. If the user decides to hold shielded (ie, dark) MoedaHíbrida tokens, then all transactions made with those tokens are completely anonymous and untrackable. The user can decide to unshield his or her MoedaHíbrida tokens so that all transactions can be seen (ie, light).
Offering users the choice of anonymity but making them pay for is a radical solution but I’m with J.P. in thinking that it deserves attention. What I think is very clever about using negative interest rates (which had never occurred to me) is that it allows for anonymous transactions without imposing a transaction friction, thus providing the cash substitute in the marketplace, but it penalises the stashing of anonymous cash. The negative interest rate means that dark tokens will be subject to a negative interest rate of, say -5% per annum, while light tokens will receive a competitive SELIC-linked interest rate.
Whether or not this is the way forward I or not, it is a line of thought that deserves serious examination in the context of CBDC design. If it is considered important to society to provide anonymous means of exchange, then the “tax” on the anonymous store of value seems a reasonable way to distribute the costs and benefits for society as whole.