The Case for Inflation on EOS
There are so many things that make EOS exciting: industrial-level scaling onchain, seamless interblockchain communication, the vast amounts of electricity saved compared to other blockchains based on proof-of-work, the possibility to resolve disputes between members (a.k.a. “The Governed Blockchain”), and so on. And most of it was made possible by the consensus mechanism — DPoS.
DPoS stands for Delegated Proof of Stake, where block producers are determined by the community through popular vote. Ideally, this should lead to the best candidates being elected, as users would vote for the ones that will push the chain forward. In other words, it’s the concept of representative democracy applied to blockchain.
Or at least, in theory. In practice, DPoS is not a democracy — it’s not based on “one person one vote” but on the amount of stake, or “one token one vote”. This means that the more financial decentralization exists in the network, the more democratic it really becomes.
Building a better democracy
Ideally, we want to give more voting weight to the people who contribute more. However, since value is subjective, there is no objective way to determine the value that was created. So what we do in DPoS is that we go by incentives instead. The assumption is that the more tokens someone has, the more skin in the game they have, and the more interested they would be in creating value for the network (or at least protecting its value). Therefore, voting weight is determined by the amount of stake.
A lot of thought has been put into making sure that EOS tokens are as widely distributed as possible to ensure that they are sufficiently decentralized. The token sale that lasted a full year has been a great help, as it gave a lot of people the chance to buy tokens. However, the Pareto principle eventually kicks in, and wealth starts becoming more and more centralized.
Let’s imagine that markets are perfect and reward participants fairly according to the value they create. I will be rewarded with a fair amount for something that I create today. If I create something of similar value tomorrow, I will again be rewarded a fair amount according to tomorrow’s value of the currency. However, with a deflationary policy, what markets would reward me today will be worth more than what a fair reward would be tomorrow. In other words, a deflationary policy will value past work more than future work, whereas an inflationary policy will value future work more than past work. This is a great philosophical debate, and there are merits to both sides, but in the context of DPoS it has a deeper implication — rewarding past value leads to greater centralization of wealth, making the whole network less secure.
More importantly, if we want to have something that is closer to democracy and farther away from oligarchy, we need to make sure that wealth and power are distributed as widely as possible. In the context of DPoS the monetary policy is not just about wealth — it’s about making collective decisions based on the amount of stake everyone has. That is why giving more voting weight to people who create value today than to people who created value in the past makes sense. Of course, getting stuck living in the past is always an option.
But wait…
Just saying “we need inflation” is not enough, though. The more important part is how this inflation is distributed. If it gets distributed top-down, we don’t accomplish anything — the “trickle down” mechanism has been tried, and it doesn’t seem to work. A simple distribution by stake does not change the voting weight distribution — on the contrary, it probably centralizes it even more, due to the higher percentage that the poorest holders need to spend each month, just to pay the bills. Just giving it to the poorest is also not a good option because it can be abused.
This is a hard problem to crack, considering that any system we create should be robust enough to withstand all the ways it will be attacked. In my opinion, the best way to create a robust system is to make it antifragile. We should avoid a central point of failure and use as many different methods to distribute inflation as we can, as long as we try to avoid the top-down approach: Worker Proposals, UBI, paying block producers (yay, we already do this!), maybe others?
How much is too much?
Finally, we should not go overboard with this. Determining how much inflation is needed is hard, and math won’t save us. It has to be done through trial and error. My gut feeling says that 5% is pretty good, but we can start somewhere, measure results, and make more informed decisions in the future.
