Last week, Congresswoman and "Squad" member Rashida Tlaib (D-Mich.) sent Crypto Twitter into a tizzy with the following proposal:
The bill's academic/think tank proponents followed up with posts such as this:
There's a lot to unpack here and a lot of crossed wires, mostly due to (I suspect) the fact that the proponents of the bill are MMT theorists and not engineers. While they may have fairly elaborate theories about what function cryptocurrency serves (and in particular how it has the potential to undermine their macro strategy of money printer go brr), they may have a somewhat looser grip on how cryptocurrency actually works.
1. What the bill does
I preface this essay by saying that stablecoin issuers should be licensed. What sort of license is anybody's guess. Currently I should think a money transmitter license would be the thing but there's no reason in principle why an issuer shouldn't go get a bank license as well.
The STABLE Act does way more than that, and appears to require any blockchain that runs stablecoin code to be licensed, among other things. For example:
- The bill outlaws the issuance of a stablecoin otherwise than by "an insured depository instiution that is a member of the Federal Reserve System," i.e., a bank.
- The bill bans the issuance of stablecoins, provision of "stablecoin-related" services, or "otherwise engaging in any stablecoin-related commercial activity,” including activity involving stablecoins issued by other persons, without obtaining written approval in advance from the appropriate federal banking agency.
- The bill creates a requirement for preapproval, among other things, for "otherwise engaging in any stablecoin-related commercial activity."
It's a swing and a miss:
- First, the largest stablecoins available in the marketplace – which shall remain nameless for the purposes of this blog post – have lists of compliance issues a mile long already. Adding another requirement doesn't answer the question of how we get non-compliant stablecoins to adhere to the rules that currently exist.
- Second, one of the stated purposes of this bill is to protect underserved communities from being discriminated against by stablecoin issuers. To this I would reply that any stablecoin issuer worth doing business with will operate in New York State and need to comply with the provisions of the N.Y. Human Rights Law, which prohibits discrimination. (For the disabled, I note also that the Second Circuit federal court thinks that under Title III of the Americans With Disabilities Act there is no requirement for a "public accommodation" to have a physical location, so that aspect of equal access might also be covered by New York-based stablecoin providers.) Additionally, given the regulatory problems with some existing stablecoins, and in particular their role as dollar liquidity providers for offshore exchanges with lax know-your-customer enforcement that can't get banking access, it is likely those who would access stablecoin markets don't need to be protected from denial of access to stablecoins, but rather they need to be protected from most of the stablecoins they are likely to encounter in the wild.
- Third, the plain text of the bill presents the bizarre possibility, one which is apparently intended by the drafters, that node operation on any unlicensed chain that supported any stablecoin contracts would be unlawful and, pursuant to 12 U.S. Code § 1833a, would be subject to fines of up to $1,000,000. Criminal penalties might also be possible. The rest of this post deals with this point.
2. Introducing the Ethereum Rule of Statutory Construction
Lawyers have these little critters called "canons of statutory construction" we use to interpret laws. For example, in England they have something called "the golden rule," which basically means that when trying to understand what a law calls for, you give the statute its plain and ordinary meaning unless doing so would render the statute absurd. In the alternative there is an approach called the "purposive approach," which is generally used to interpret indirectly effective European Union law, where interpretation of the rule is driven by the purpose for which the statute is drafted.
In America, by contrast, you may have heard of "textualism," "originalism" or the "living Constitution" approach in recent U.S. Supreme Court hearings. It's the same game, choosing which rules we use to understand language.
I propose one for cryptocurrency. I call it the Ethereum Rule, and it holds that "A law is to be given its plain and ordinary meaning unless it would require Ethereum (as it exists in 2020) to apply for a license, in which case the law is absurd."
This bill appears to require just that. Although the definition of "stablecoin" in the act seems to exclude cryptocurrencies like ether, the issue isn't that the definition is over-broad but that the bill seeks to force anyone engaging with stablecoins to do so under the aegis of the Federal Reserve System. Just read the plain language:
This doesn't leave a lot of wiggle room: "Any" means "any," and "any stablecoin-related commercial activity" is a broad brush when we consider that any user of any smart contract blockchain will be verifying stablecoin transactions to some extent.
Lest we think that we're misreading the proposal, its own proponents publicly agree with this interpretation:
To this I respond with the Ethereum Rule of Statutory Construction. Ethereum has no central owners, forks regularly and is currently regulated as a commodity. If your law requires that kind of a system to get a bank charter, not only will the law fail to effectively control the blockchain, but the regulators tasked with enforcing it will have difficulty finding someone with standing to sign the application.
The STABLE Act says that blockchain users will be permitted to transact, if only they would first achieve the impossible. This is an absurd state of affairs and a strong indication that, as written, the STABLE Act would not make good law.
3. Would the STABLE Act actually make running a node illegal?
Of course, there is zero chance the STABLE Act is going to become law during this Congress. However, coin people – and Ethereum people in particular – have been asking the question: What if it did?
The answer is not straightforward. Peter van Valkenburgh over at Coin Center says the prohibition on "stablecoin-related commercial activity" hands down applies to node operators or anyone running the Ethereum client:
Though a reasonable conclusion, and on balance likely the correct one, it is not a forgone one, since the current language of the STABLE Act, being both overbroad and imprecise, leaves plenty of scope to poke holes in it. For example, it is not clear whether operating a node gratis (as many full nodes do) counts as "stablecoin-related commercial activity" if done on a non-commercial basis. Seeing as nodes are not ordinarily compensated it is certainly conceivable there will be situations where node operation is sub-commercial if not non-commercial. Research would be required to find the answer here.
Additionally, it is not immediately apparent to me that running a full node is "stablecoin-related commercial activity" given that many if not most cryptocurrency transactions don't have a stablecoin component. The statute's lack of specificity narrows its application. If the statute said "any commercial activity related to, or any communication which may facilitate, any stablecoin transaction," that would be one thing. But that's not what the language says. Properly understood, Ethereum is a rail, and just as we don't refer to the act of driving a car as being "jogging related" just because cars and joggers use the same roads, we shouldn't refer to the act of running a node as "stablecoin-related" just because stablecoin transactions are broadcast alongside all other transactions via devp2p. Again, more research would be needed to see whether a court would agree with that interpretation.
See also: Preston J. Byrne – Stop F**king Around With Public Token Airdrops in the United States
There is another matter: In my view, the operator of a cryptocurrency node is capable of being a provider of an interactive computer service under a legislative provision known as Section 230 of the Communications Decency Act (47 U.S. Code 230(c)(1)). This law states in relevant part that providers of interactive computer services, properly "information content providers," are not treated as the publisher or speaker of, and therefore have no liability for, content which third parties submit to their servers, subject to certain limited exceptions.
Coin Center has called, in the past, for a node operation safe harbor similar to Section 230 (which exempts social media networks from content liability). Since the blockchain is really little more than a published, cryptographically verifiable feed of transactions that have been authorized by the Bitcoin network (and other blockchains, the same for their corresponding native assets), I tend to think that it's more likely than not a blockchain application falls within the confines of Section 230.
I freely admit that whether a node operator qualifies for the exemption is an open question. The law defines an "information content provider" as a "system ... provider that provides or enables computer access by multiple users to a computer server." I'd have to do a little research to see if there are any precedents dealing with the question of what a "server" constitutes for this purpose, but at least at first glance there is an argument to be made that operating a full node on a blockchain, which in its essence is a distributed timestamp server, could qualify, at least insofar as it pertains to third-party financial communications that are being relayed by that node.
Section 230, however, only confers immunity from state criminal law and civil actions. It has no effect on federal criminal law, and there are criminal sanctions in the FDI Act (see e.g. 12 U.S. Code § 1818(g)). To figure out whether a full node could be captured within the STABLE Act the first thing to do is read the statute and try to determine whether providing peer-to-peer network access services counts as "stablecoin-related commercial activity."
If not, then node operation is not captured by the statute and the analysis ends. If so, the next questions would be (a) whether node operators were covered by Section 230(c)(1) and (b) whether the STABLE Act impliedly narrowed or repealed Section 230's application to node operators insofar as the nodes processed transactions related to stablecoins. After answering those questions the picture would be clearer.
In terms of the current federal picture, we know that providing network access services is not equivalent to money transmission, that FinCEN doesn't consider node operation to be money transmission, and that for most federal crimes accessory liability requires heightened knowledge and participation of the kind we don't usually ascribe to node operators. This is perhaps why, to the best of my knowledge, there have been no prosecutions for running a Bitcoin full node to date.
Nor should there be, now or ever, and if American leadership in the crypto arena is to continue it might be worthwhile, given how wrongheaded the STABLE Act is – not on stablecoin licensure, as I think stablecoins are properly the subject of regulation, but on blockchain node licensure – to revisit Coin Center's proposal for a blockchain node safe harbor that clearly and unambiguously accords blockchain nodes the status enjoyed by other online publishers.
Section 230's most learned interpreter, Jeff Kosseff, titled his book on the provision "The Twenty-Six Words That Created The Internet." I note for the record that Facebook, Google, Twitter and YouTube were not founded in Europe. If America is to lead the decentralized internet we would do well to look to Section 230 as an example of how to do internet regulation the right way.