How the Feds bounced Binance
Crypto isn't "incentive compatible"
Abe and I had a piece on Binance’s guilty plea in the Wall Street Journal’s Weekend Review (ungated). It built on the argument of our book (have I mentioned Underground Empire? If you like this Substack, you’ll probably like it too - buy it from Amazon or elsewhere and I’ll love you for it).
Thanks for reading Programmable Mutter! Subscribe to receive new posts - I’m keeping this free for the foreseeable future. If you would like to support my work more practically, buy my and Abe’s book, Underground Empire: How America Weaponized the World Economy from Amazon or elsewhere and spread the good word if you like it!
Our argument is pretty straightforward. First, that the fine is probably less important than the fine print for crypto. $4.3 billion is a big, attention-grabbing number, but the really important part is the monitoring agreement. Binance will have an outside actor looking over its shoulder as it sets up an extensive internal bureaucracy to report maybe-sketchy transactions to the relevant authorities, and comply with US sanctions. Binance, and the rest of the crypto community, have historically either avoided these requirements or played ‘lets-pretend to comply.’ Now, if Binance doesn’t deliver, the monitor can squeal to the authorities.
Crypto people have argued that the Feds’ unwillingness to shut down Binance demonstrates its essential role and the Awesome Power of Cryptocurrency. Our argument suggests a somewhat different spin on that logic. If you, as the federal government, want to keep crypto under your thumb, it is quite handy to have a big central actor out there that you can control. Binance will not only help the US government monitor the money flows (the argument of our book), but plausibly act act as a regulatory super-spreader, transmitting “know your customer rules” across the ecoystem like an epidemiological contagion (the argument of an important academic article on ‘viral governance’ by Gregoire Mallard and Jin Sun).
And there’s another, complementary way of thinking about this that I kept thinking about as I wrote my part of the piece. It is how the Binance outcome - or something like it - was plausibly written into the contradictions of the cryptocurrency economy from the beginning; in particular, the contradictions between the theoretical mechanisms that it relies on, and the practical ways in which the economy operates.
To expand that out a bit: the key theoretical grounding of crypto can be found in applied economic theory - specifically, the kinds of ideas about ‘incentive compatible mechanisms’ that were super popular among micro-economists in the 1980s. But a lot of the practice of crypto is anything but incentive compatible. If you, as a crypto entrepreneur, want to maximize profits, you’re going to behave in ways that make the incentives break down. The evolving economy of crypto is large part a shifting balance between three elements - incentive compatibility, maxing out the benjamins, and getting popular buy-in.
To understand this, you ought begin from a starting point that isn’t a particularly popular one on the left. Crypto is not a crude scam. It has true believers, some of whom are very intelligent people trying to figure out solutions to genuinely important problems. This doesn’t mean that crypto is a good idea. My own working assumption is that the problems are almost certainly unsolvable within the terms that crypto is looking to solve them. But to understand why crypto has stuck around, you ought some of its ambitions seriously, and understand its beginnings on its own terms.
And it begins by applying two mathematical technologies. One is the one that everyone knows about - crypto - although the relationship between cryptocurrency and actual cryptography is both socially and technically complicated. The other is much less well known, but arguably just as important. Crypto builds on economists’ efforts to construct ‘incentive compatible mechanisms’ as an alternative to politics (this Miller and Hammond piece summarizing and responding to the 1980s literature is, imo, quite brilliant, and doesn’t take a ton of math to understand).
Very crudely simplifying, an incentive-compatible mechanism is a carefully calibrated means to channel actors’ selfish interests so that they reliably produce some pro-social outcome (honesty; provision of a collective good). Economists hoped that this would provide a way to reliably provide public goods, without government or any of that second-best-at-best nonsense of democracy.
It didn’t work. Still, the crucial grounding technology of crypto - the blockchain - was invented as just such an incentive compatible mechanism. It is designed to solve a problem of dishonest reporting (the so-called ‘Byzantine Generals’ problem) that had arguably undermined previous efforts to create a decentralized currency. Blockchain supposedly does this by creating a system in which actors have an incentive to report honestly and collectively keep the books straight on who has given money to whom, without any central authority.
The blockchain, on its own terms, is quite inventive, whatever one may think of the by-products (vast amounts of wasted energy and computing power in ‘proof-of-work’ blockchains like Bitcoin’s). And much of the consequent development of crypto techniques (e.g. ZK-SNARKs) etc are designed to serve the same basic goals. They are a family of economic/cryptographic technologies that are designed to provide the building blocks of a decentralized economy, in which (a) we don’t have to trust each other to interact, and (b) we don’t have to turn to some centralized enforcer to provide a substitute for trust by imposing its will.
There is a lot of ingenious work building on top of this basic foundation. Most of it, frankly, is far too mathematically sophisticated for the likes of me to do more than appreciate from a very considerable distance. But even skeptics should acknowledge there is (a) a coherent political goal motivating the technology, and (b) a lot of hard work that has been put in, often by quite sincerely committed people, to achieve it. And if you distrust Leviathan (and - in many respects - you ought to) there is something admirable in the motivation of this project, whatever about the implementation. And truefans of political economy who don’t buy any of this should still pay attention to the work, the ideas, and the consequences (there is a lot about comparative political economy that you might be able to figure out e.g. by systematically comparing the different governance structures and consequences of DAOs, or the cultures of different blockchain currencies).
But - and this is evident - crypto is not just the ingenious math. Actually existing crypto is not just sincerely committed libertarian computer science geeks trying to construct an entire alternative economy from the foundations up. It is also people looking to make shitloads of money, through fair means or foul. That is where the skepticism should begin - even if you discount the huge amounts of outright chicanery and fraud (and - in many respects - you ought not).
The crypto political economy blends these two together. You can’t separate out the math from the money. Both the true believers and the people who are in it for the benjamins share a common desire to share the gospel as widely as possible. But their incentives clash sharply with each other.
This helps explain why crypto preaches decentralization far more than it practices it. As people like Moxie Marlinspike have observed, there is a lot of centralization across the different aspects of the crypto economy. Binance is just one important example - it has captured the lion’s share of exchange between one crypto currency and another.
Some of this is because of efficiency gains and consumer laziness, but a lot is the result of deliberate business strategy. As Peter Thiel (who has a complicated but profound relationship with electronic currencies) has observed, you need market powerto make serious, sustained profits, and you ideally want to establish a monopoly. Companies like Binance have taken such advice to heart, trying to grow as big and as powerful as possible, so that they can dominate their segments of crypto, and reap the financial rewards. My sense is that the true believers in crypto have been reasonably willing to go along with this - even those who are appalled at the out-and-out confidence artistry. They think - likely correctly - that crypto can never succeed as a financial system if there aren’t people doing their damnedest to make vast amounts of money in it.
But this - to return to our book’s argument - is why the crypto economy keeps on making itself so vulnerable to government. Crypto relies on entrepreneurs’ cupidity to grow. But inside every lean, hungry entrepreneur, there is a bloated monopolist struggling to get out. And once it escapes, its power-bloated bulk tacitly invites the government to come back in and impose control.
Why is this so? Consider. If a business becomes big and powerful, it becomes more vulnerable to government regulation. The more that a firm has have at stake, the harder it is to disappear when the regulator knocks at your door. Binance did remarkably well for years, claiming it had no headquarters, hopping from jurisdiction to jurisdiction, and setting up a purportedly independent US subsidiary to head regulators off. But sooner or later, it was bound to find itself trapped in a situation where its best response was to agree to what the feds demanded. That is what transpired last week.
Moreover, as a firm establishes market power, it also become more attractive to the government. The reason that the feds want to keep Binance around is because so much crypto activity depends on it, which now suits them very well. If they are able to get Binance to report suspicious activities, stop sanctions busting etc, and induce others behave similarly, they can keep tabs on what is happening across a very wide swathe of the crypto economy.
The result is that crypto, for all its faith in incentive compatible mechanisms, is not actually itself incentive compatible. Crypto entrepreneurs are motivated by a specific kind of self interest that will reliably lead them to behave in ways that degrade their ecology. Specifically, they will look to build monopolies, centralizing control of markets to maximize profits. And that will provide government with the hooks that it wants and needs to re-establish control
That is what is allowing government to tame crypto. Binance got big and powerful enough that government came knocking, and Binance’s managers and owners decided that it was better to render unto Caesar than be crushed by him. Others will follow in Binance’s train.
This leads then, to an interesting but highly awkward long term question. Harnessing greed is necessary if the crypto project is to spread and succeed, providing room for actors like Binance. But can crypto thrive if it genuinely knuckles under to Uncle Sam?
As I’ve noted, the blockchain is an ingenious solution to a highly particular political problem - that of creating a financial system whose core is highly decentralized and robust against manipulation. Unfortunately, it turns out that it’s hard to keep the system decentralized (it also turns out that the system was never as decentralized as it seemed).
But the blockchain’s entire purpose is political. If crypto is no longer about teh magic of decentralization, then why would anyone want to use it? Without the politics, it just becomes an incredibly cumbersome and inefficient means to reconciling ledgerbooks, which could be done much more quickly and efficiently through traditional decentralized means. You don’t need the equivalent of a mid-sized country’s energy resources to run an ordinary ledger reconciliation system, while the computing power of the entire Ethereum system … is not so large.
As Binance and other parts of the crypto ecology become subjected to state regulation, and subsumed into the regular political system, crypto is likely to lose its political point, and very quickly too. Perhaps, work on so-called “soulbound tokens” as building blocks for decentralized communities might provide an alternative - but this would likely require a sharp turn away from incentive compatible mechanisms and towards some of the fuzzier aspects of social interaction that crypto people tend to overlook. When applied to useful purpose, those might provide a more robust set of political foundations than the current dubious alliances that crypto rests on - but getting there from the crypto ecology that we have today seems to me to be extremely-hard-to-maybe-impossible.
Thanks for reading Programmable Mutter! Subscribe to receive posts - I’m keeping this free for the foreseeable future. If you want to support my work in more practical ways, buy my and Abe Newman’s new book, Underground Empire: How America Weaponized the World Economy from Amazon or elsewhere and spread the good word if you liked it!