EOS Is Not a DemocracyBy Anthony Di Franco
Blockchain technology is supposed to be a force for decentralization and democratization. But is it really? Transactions in most systems are verified by proof of work, also known as mining, which relies on supercomputer-scale hashing farms deliberately wasting enough electricity to power a medium-sized country. Theoretically, the enormous expenses associated with mining make it hard for any one party to sneak malicious changes into the transaction records kept by the network.
But it also means that powerful miners with access to inexpensive electricity, for example in China, can wield disproportionate influence. The enormous expense of mining has led transaction verification and its rewards to be centralized in small coalitions of large, well-financed operations.
To address these problems, many blockchain projects are now experimenting with voting systems. Each node’s ability to verify blockchain transactions is not based on how much energy they’ve wasted, but on how many votes they collect. That way, securing the blockchain ledger doesn’t come at a tremendous real-world cost. These savings will mean a significant competitive advantage for any system that makes it work.
But blockchain-based voting systems are now running into new problems. Take EOS, a high-profile blockchain project that recently implemented its own voting system. In EOS, a small group of elected delegates or “block producers” validate blocks for the EOS blockchain, and these delegates earn rewards for their work. This is a lucrative activity that currently yields profits at a rate of about US$400 million a year.
Some people think this is a big problem. “Because the delegate rewards in EOS are now so high [...] the competition on who gets to run nodes has essentially become yet another frontier of US-China geopolitical economic warfare,” wrote Vitalik Buterin, creator of Ethereum, in a blog post entitled “Plutocracy is Still Bad.” In other words, EOS network participants are highly incentivized to collude.
In an analysis of EOS’s voting system, LongHash’s data team found that votes cast for block producers on the EOS chain show a remarkable degree of consolidation of both voting power and the winners in its high-stakes elections.
According to LongHash research, at the time of EOS’s launch, several EOS block producers received a large percentage of votes from a small group of voters. In some cases, the top ten voters made up 50% of the votes for an elected delegate.
This poses a serious risk of bad governance for the system. How many votes one EOS participant can allocate to each candidate node depends on how many EOS tokens it “stakes” in the network, which holds the tokens in escrow for a short time. Hence voting power is directly tied to wealth held in the system.
This process has amplified the risk of wealth concentration and market manipulation. Cryptocurrency exchanges, which often hold a large portion of altcoins such as EOS, are particularly well-positioned to subvert the system. According to EOS’s voting rules, network participants can appoint proxies to cast their votes for them. This feature is most likely to be used by crypto exchanges to vote on their users’ behalf, which raises concerns of collusion between exchanges and block producers, as well as of exchanges dominating block validation. Exchanges are already well-positioned to manipulate the crypto markets by handling their order books in ways that subvert the markets, and this kind of voting system would remove one of the few remaining checks on their power to meddle in the markets.
Voting patterns for EOS are also suspect. By examining some of the top voters that staked many EOS tokens in the last election, LongHash’s data team discovered some interesting behavior: one super voter (who staked US$72.9 million worth of EOS tokens) cast votes for 10 nodes, none of them belonging to candidates in China ― even though Chinese delegate nodes make up 38% of EOS’s nodes (8 out of 21). Another super voter (US$98 million staked on the EOS network) only voted for two candidate nodes, both run by exchanges in China.
Low voter turnout
The election process has just begun to muddle through its first real test. And it has already deviated from its plans: instead of the 21 block producers normally involved in stewarding the state of the system, only a single block producer ran the first election.
To make matters worse, there was remarkably low voter turnout, possibly the result of a lack of an informed electorate. The rules of how to participate in the election process are obscure. The main source of information about how to vote for block producers is maintained in the form of short blog posts and videos by EOS Canada, itself a candidate block producer and an entity distinct from Block.one, the main architects and stewards of EOS.
The only means of voting that are generally available are two command-line tools. This has evidently excluded a large proportion of EOS participants from casting their votes: as of 13th June, about two weeks after the system came online, only 6.45% of participants had cast a vote.
In defense of EOS’s governance structure, specifically in response to Buterin's claims that it amounts to “an Oligopoly, or a Plutocracy,” Thomas Cox, vice president of product at Block.one during the EOS presale period and mainnet launch, said that: “[...] Having more tokens might give you more power — it also puts you at greater risk. To own more tokens is to be more exposed to negative outcomes on that blockchain.”
While it might be true that those who have staked the most EOS tokens on the blockchain are at greater risk if the system as a whole fails, the rewards of subverting the system may outweigh the drawbacks. In any case, the viability of EOS’s voting system will remain in question until it can demonstrate that it provides effective governance. To paraphrase the saying, it may yet prove to be more profitable to be king of the smoldering ruins than just another well-off citizen in a prosperous city.
Good governance will decide the viability of EOS itself ― whether it becomes a vehicle for even more aggressive collusion and speculation, or a thriving economic ecosystem with a profound technical and economic advantage over the current generation of blockchain systems.
Anthony Di Franco is a San Francsico-based writer and machine intelligence tool builder.