Blockchain Governance/Token Economics- Proof of Stake

This post is an Overview and response to DFINITY economist, Yulin Liu’s, medium post “Glimpse of Blockchain Governance” It contains my own comments, criticisms, and key takeaways… I enjoyed Yulin’s article and you can find the medium post below.

A shift from off-chain to on-chain governance is underway. We are seeing this through the growth of Proof of Stake Blockchains with on chain governance.

On-Chain governance is more aligned with the “decentralized ethos” as everyone in network has a vote… with off-Chain governance however, the core developers and full node providers have much more power than ordinary users.

So rather then endless debates on community forums to have something upgraded or implemented with an off-chain governance. In a PoS blockchain, the voting and rules are predefined and made automatically.

Architecture of On-Chain Governance

Proposals require a collateral stake to prevent spamming. High quality proposals are rewarded and spam proposals penalized.

The more tokens that get behind a proposal, the quicker it will be addressed.

Create a weighted voting system that incentivizes members to stake their tokens. Give more voting weight to members staking long term (e.g: a person with 10 tokens locked for a year would have more voting right then someone with 50 tokens but not staked) This gives chances for all community members to have their voice heard, rather then just a few large whales and also incentivizes members to stake there tokens, and in my opinion would create a higher demand and ultimately push price upwards. This price movement could attract more speculators and enthusiasts and we can see Network Effects takeover.

Vote Delegation : Sometimes a particular area is just not your area of expertise and you may have a thought leader you trust. You can delegate your votes to the person of your choice.

Random voting body : rather then having entire network voting on each issue. Then take a random segment of the network to vote on the issue. Randomness is CRUCIAL here. My assumption is that this is where threshold relay will come into play as well. True randomness in the network for low costs.

Zero Knowledge Proofs : Anyone that knows me, knows I’m a huge proponent of ZKP. When I saw Jens Groth joined the DFINITY team, I was beyond excited. So, with zero knowledge proofs, you can have registered users that gain more voting power. But thanks to ZKP, the user’s private data is ensured safe, preserved, and secure. I am greatly behind this implementation in the networks.


Obviously in the beginning a central organization needs to bootstrap the network and get it kicked off. But how do you sufficiently decentralize it over time. How do you prevent an organization/entity to keeping just the right number of tokens so that they will always have the necessary voting power to make whatever upgrades they choose? How do you prevent a decentralized entity from only giving the illusion of being decentralized?


Great overview , Andrew! A lot of food for thought. Really thought provoking criticisms too. Definitely discussion worthy.


Thanks Andrew, it seems that I have a lot of DFINITY related stuff to read this night :smiley:


great thoughts andrew! enjoyed reading them

1 Like

I enjoyed your thoughts Andrew. The question if you can start centralized and become decentralized over time seems to be an ongoing debate. Some think it can never be truly decentralized if it starts off centralized in certain data centers or people/organizations with a lot of capital. I guess time will tell if decentralization over time can become a reality.


I’m also so very excited to see more conversations develop around governance for the Internet Computer and the open internet services that people build!


Nice write up Andrew! I remember when there was talk about creating a sort of constitution for dfinity which could be updated by the then called blockchain nervous system. I wonder what happened to that idea.


I really like the idea of stake being worth n times more with registered ID. If this is combined with ZKP it seems like an excellent solution. My only worry is that we don’t know what the multiplier should be yet. Rather than making an unknown coefficient permanent, maybe it should be variable while the network is reaching stasis, then locked down. Future changes to the multiplier could then be voted on.

Also proof-of-stake seems like the best choice for sybil prevention. It can’t be avoided at this point until the problem of decentralized identity among autonomous actors (who could be programs) is solved.


Welcome to the forums jimbo. Nice thinking, a variable parameter could be voted on or even automated (set a threshold on number of IDs and incentivise more if we drop below it).


hi Jimbo… really great points and welcome to forum


thanks! good to be here.


Are things like whether dfinity will use quadratic voting or one-token-one-vote schemes still up to debate and not yet fully determined? Is dfinity really considering to rely on government issued ids? I know his article does not represent the Dfinity foundation’s view so… is there some place I can read the foundations view about this?


Foundation didn’t come out with any specifics so yes everything still yet to be determined… but the ID’s aren’t required and folks can still vote albeit with lesser influence… i would imagine zero knowledge proofs would come into play for validating certain information on ID without sharing all information

1 Like

If you look back through the history of dfinity (there are articles, website revisions, tweets, ect.) you can see that the vision has been going through an evolution as more and more of the tech is actually built.

However from the beginning dfinity believed in the proponent for starting more centralized and then decentralizing as the network becomes more stable. However as Andrew stated the specifics are still to be determined by the foundation.