On Blockchain Scaling and Collective Action Problems
Are psychiatrists happy because their job is to treat depression? Do only thin people offer weight-loss advice? Or how about a harder question: is bitcoin (the network) a working market simply because bitcoin (the token) can be used to create working markets?
Satoshi Nakamoto never asked this question, which may be why most people in the blockchain space ignore it. People assume that markets will form inside the consensus mechanism to pay for whatever is needed to keep the network running, but if you ask them how the market will work — who will pay for what and why exactly you’ll quickly discover that their reasoning boils down to “someone has to because otherwise the network won’t work.”
For a practical example of this, consider Craig Wright’s claims that miners will pay for the network because otherwise they won’t get paid. There is a logical fallacy: we know that miners won’t earn transaction fees unless the network exists, but that does not mean that they will subsidize a network that others can use too. There is an entire class of market failures called collective action problems that cover issues exactly like this. In many markets participants are better off not paying for public goods even if their neighbours also refuse to pay.
There is already empirical evidence that these problems exist in bitcoin. Read this blog post by Craig Wright that complains about miners spending their money buying ASICs. CSW tries to argue that miners don’t understand the incentive structure of bitcoin, but surely they are just responding to the incentives created by a network that pays exclusively for hashing. For another example, consider Vitalik Buterin’s recent post on collusion that overlooked the inevitability of transaction-sourcing companies like Infura partnering with block producers to monetize their transaction flows. And forget about Casper’s balancing mechanics once that happens, because every single economic assumption underpinning the Ethereum security model will vanish overnight once Infura starts deciding who gets to be profitable.
At conference after conference, speakers enthuse over how blockchains will fix problems in healthcare, insurance, and gaming. But how many are even aware that the permanent ledger is the best example of the tragedy of the commons problem in the tech space: miners are incentivized to add data (sheep) to the blockchain (pasture) even though doing so destroys the sustainability of the blockchain over time. And this is just one example — there are several other incentive problems that are eating away at every single proof-of-work and proof-of-stake network in existence.
The irony of the situation is palpable, because how can fixing problems in external markets be more important than fixing them in the consensus mechanisms that are supposed to undergird the solution? Why are blockchain people so excited about the former and indifferent to the latter?
I suspect the problem is a lack of awareness about the economic nature of the underlying problems and a tendency to view as a “technical” challenge instead of an incentivization problem. Treating scaling as a technical issue feels productive because it encourages everyone to spend time and money on complicated approaches (like sharding!) that tend to hide their trade-offs. But incentive problems aren’t fixed by technical workarounds: the only way to fix them is to correct the underlying incentive problems so that the free market can coordinate a solution by itself.
This is a much harder problem than it seems — and one of the big challenges is just getting people to think about the problem from the incentive angle. For those interested in how Saito actually solves the tragedy of the commons problem, we have just put together this video explaining the solution. Until we figure out how to offer a simplified explanation of the method Saito uses to prevent other problems like free-riding and transaction hoarding, anyone interested in knowing how Saito works is really encouraged to one of the various conferences at which Richard and I will be presenting the solution in the coming months. Or drop by Beijing and grab a drink with us. We also have the solution described in our whitepaper, although it requires a deep familiarity with blockchain fundamentals to understand since the core problem that needs fixing changes from how to make block production difficult to simply ensuring it always remains prohibitively expensive.
Either way, supply curves and demand curves tend to meet regardless of whether people believe in them. The blockchain community is moving in the right direction by shifting towards on-chain applications and big-data network, but if we really want a global digital currency and PKI-based internet, we need to focus on fixing the problems no-one talks about, because a sustainable blockchain needs to be a castle standing on rock, not a house of cards on a foundation of sand.