PoS Governance evolved: the urgent need for a Reddit-ized governance UI linked to staked tokens
The current Proof-of-stake governance system (validators, Twitter/Youtube influencers and delegators) creates echo chambers, not robust governance. Reform is needed.
Problem statement: validators are infrastructure specialists whose ability & incentive to contribute to governance is limited.
Solution: PoS chains should allocate [10%] percentage of staking rewards towards “governance commentator” roles, with a Reddit-like upvoting system linked to staked tokens to upvote quality.
The core concept behind “delegated-proof-of-stake” (DPoS) blockchain governance is that validators provide the necessary computer hardware to run a network, while delegator-stakers, in turn, stake capital against validators’ infrastructure to insure against dishonest network activity. Both parties jointly conduct the network’s governance. The network pays for these resources in the form of token emissions, and the market ultimately decides if these investments justify themselves over time.
The historical track record of DPoS leads to 3 logical takeaways.
DPoS does a fantastic job at network security: frontal-assault network hijack attempts on DPoS chains of any scale are basically unheard of.
DPoS does a great job (too good of a job?) of incentivizing infrastructure: DPoS chains are awash in infrastructure which dwarfs their networks’ compute requirements.
While DPoS effectively solves formal network dishonesty, it totally fails to guard against informal network dishonesty, in which social consensus is hijacked by actors engaging in political collusion.
The third issue gets virtually no attention from blockchain engineers relative to the first two. Which is strange, because the third issue lies at the root of the vast majority of DPoS network failures, as well as DAO failures.
90% of professional crypto investors you talk to will say that “blockchain technology, as long as it’s not terrible, doesn’t really matter. It’s about the narrative.” Narrative attracts users to a chain. Sticky users attract devs to a chain. Sticky devs keep sticky users on a chain and start bringing in new users at accelerating rates.
What is “narrative”? It’s a story of how a blockchain fits into the world, and what value it will add. A story which is created, and enhanced, via robust public debate within a community.
However, “robust public debate” has its own potential drawbacks. If the community engages in debates all the time, it never gets anything done. High-quality contributions to governance debate tend to be poorly rewarded relative to crypto influencers, especially when they’re contrarian views.
Professional validators aren’t dedicated governors
In the Cosmos chain-of-chains vision, different digital tribes were supposed to spring up with their own validator sets and appchains around different use cases.
In practice, running a validator is its own small beast of technical specialization and execution risks. This has led to a small elite of professional validator teams running most of the major validators across many Cosmos chains, leveraging their experience to programmatically automate DPoS network security at 99.99%+ success rates.
However, while these validators do a great job of supplying compute resources and securing the network (validating transactions with virtually 0.00% rates of double-signed or missed blocks), they lack the bandwidth to really perform the other core function expected of them: actually governing the network.
Elite professional validators, instead, view validating as an asset management business. They compete for delegations based on their operational excellence as well as governance participation, but the best risk/reward for them is to passively reflect governance consensus 90%+ of the time to maximize delegations. Thus, they collect 5-10% of delegators’ staking rewards (1-3% same-token yield on total delegated assets per year).
On top of this centralization bias, most DPoS chains have highly centralized ownership (10-40% of total voting power). In these chains, governance is a formality, and “decentralization” remains aspirational in perpetuity.
As a result, governance malpractice is rampant on DPoS chains, at least in my experience. Small teams of insiders tend to pay themselves heavily, indefinitely, to produce weak results. Professional validators lack the necessary incentives to check corrupt governance behavior, and lack bandwidth to care about governance on any other than the largest of the 3-10 chains that they validate on.
Proposed alterations to governance mechanisms have generally only tinkered at the edges of this badly flawed system. Given how poor the track record is for decentralized governance, the underinvestment in tooling here is strange.
Governance as the true network attack vector
99.5%+ of altcoin chains are regarded as failures.
Technically, DPoS failure occurs when a network no longer has multiple validators, or the minimum number of validators, validating newly produced blocks: thus, the chain stops producing blocks. The dynamic ledger becomes frozen in time, a static ledger of historical transactions.
However, in practice, “failure” occurs much earlier, when a community has lost confidence in its governance system to launch a successful product, and its token price enters a death spiral. A failure to launch a successful product is a failure of network governance to align good incentives around competent people, monitor returns on investment, and enforce accountability for results. In other words, the vast majority of network failures result from stakers and validators failing to put in the sweat equity required to launch successful products.
If you want responsible governance, reward good governance inputs!
Good governance in crypto has, sadly, been abstracted to “number go up.” Chains with a ripping token price are governed by geniuses who are shipping amazing products. Chains with derelict prices are run by idiots who can’t ship anything.
It sounds dumb, because it is dumb. But even worse, it’s totally backward-looking, and it misunderstands governance completely. Governance is a process, not a result. If you want a good process, create mechanisms that reward good process.
But what’s a “good governance input,” programmatically defined?
I propose a Commonwealth-like forum as a core feature of the blockchain’s code and Station UI, with a couple of changes:
Like Commonwealth, provide a space to provide detailed governance feedback in a way that supports images, charts, and rich text (bullets, numbering) necessary to break down complex ideas.
Instead of CW’s upvote/downvote by individual, upvotes are tallied based on staked tokens of that network. Only staked tokens could upvote or downvote.
A wallet is judged to “objectively add value to network governance” when its argument or comment on a governance proposal gets a proportion of upvotes that’s at a positive variance relative to the voting outcome of that vote.
Create a payout pool which rewards “positive variance” contributors.
Good governance: Implementation example
PROP. 69420 GOVERNANCE PROPOSAL: change {this old line of blockchain code} to {this new line of blockchain code}
PROP. 69420 VOTE RESULTS
YES: 70% (700 billion LUNC)
NO/NWV: 20% (200 billion LUNC)
ABSTAIN: 10% (100 billion LUNC)
4LEX_WALLET_PROP69420_VOTE: NO
4LEX_WALLET_PROP69420_ARGUMENT: “Here is my brilliant argument for why Proposition 69420 is a flawed idea.” (this would be posted within the specialized governance UI)
stLUNC upvotes on 4LEX_WALLET_PROP69420_ARGUMENT: 30bn
stLUNC downvotes + NWV votes on 4LEX_WALLET_PROP69420_ARGUMENT: 50bn
stLUNC abstain votes on 4LEX_WALLET_PROP69420_ARGUMENT: 0
4LEX_WALLET_PROP69420_ARGUMENT upvote %: 37.5% (30bn/80bn)
4LEX_WALLET_PROP69420_ARGUMENT downvote %: 62.5% (50bn/80bn)
4LEX_WALLET_PROP69420_ARGUMENT variance: since 4LEX_WALLET_ARGUMENT advocated a NO vote, his argument in favor of NO outperformed the NO vote he advocated by (17.5 percentage points) * (30 billion + 50 billion + 0 billion) = 14bn. So, he has accumulated “14 billion positive variance points.”
Total POSITIVE_VARIANCE across all debate contributors for time period: 10 trillion
4LEX_WALLET total positive variance across all debates and comments: 100 billion
4LEX_WALLET share of positive variance payout pool: 100bn / 10 trillion = 1%
POSITIVE_VARIANCE_REWARD_POOL: 10 billion LUNC
4LEX_WALLET payout for contributing positive variance to governance: 1% * 10 billion = 100 million LUNC
The idea here is that governance is not about checking boxes and voting. It’s also about debating ideas and contributing value-added variant perception to technical debates in a constructive and informative manner, while rewarding contrarianism and consensus equally. This would also provide an effective way to signal validator and staker idea preference, as opposed to the current Twitter/influencer consensus.
This would also help network validators quickly screen through a debate to find the “most positive variance” comments on governance disputes, and help communities quickly achieve efficient compromises (high positive variance ideas would tend to be incorporated into final proposals). Major validators, in aggregate, tend to have the expertise to vote wisely on any given governance issue, but are too overtaxed across chains and communities to exercise their authority with effective consistency.
In the real world, politicians (validators) alone could be trusted to honestly govern. Thus, the political economy created a second role (media anchors, commentators, etc) which specialized in professionally filtering governance information to wider society. PoS chains outside of Ethereum have repeatedly failed to balance their political economies, resulting in very poor governance outcomes.