Incentivizing Governance On-chain: A Case Study of Steem (Steemit)
As the saying goes, don't reinvent the wheel.
Photo by Matthew TenBruggencate on Unsplash.
Summary
The Steemit blogging platform built on the Steem blockchain is one of the earliest social-fi projects in the space (still up and running at the time of writing). It was created in 2016 - long before the boom of Dapp during the dubbed Defi Summer in 2020. Basically, the users could (1) post articles, (2) post commentaries, and (3) upvote content to earn the STEEM tokens. The payout of their interactions is dependent on their staked STEEM tokens as STEEM Power (SP). The glaring similarities of aligning token emissions to active interactions on Steemit depending on their staked position have made it an interesting specimen to be investigated. In retrospect, one might connect the dot to discover that the newly-proposed ve(3,3) concept is tasked to achieve the same intention. Thus, the model of Steemit (or Hive Blog on its hard-forked blockchain of Hive from Steem in 2020) is explored here. On a cautious note, however, this opinion piece would not discuss the hostile takeover of Steem by Tron - which ultimately led to the hard-fork of Hive.
The Founders
Ned Scott and Daniel Larimer founded the Steem Blockchain and created Steemit. Ned was the CEO at STEEM and has now retired from the scene. Daniel is the creator of Delegated-Proof-of-Stake (DPoS), and he previously worked at BitShares and EOS.
How does Steem work?
Steem adopts a Delegated-Proof-of-Stake (DPoS) consensus model where 20 ‘witnesses’ will serve as Block Producers and validate the transactions across the network. It has an inflationary supply model at a rate of 9.5% per annum. The inflation rate would decrease roughly 0.5% per year until it reaches a final inflation rate of 0.95%, which would take approximately 20.5 years. The reasoning of the inflation rate is as follow:
It is often said that a coin with an inflationary model is not sustainable. Still, we know from countless real-world examples that the quantity of money does not have a direct and immediate impact on its value, though it certainly plays a role.
From August 2008 through January 2009, the U.S. money supply grew from $871B to $1,737B, a rate of over 100% per year, and then continued to grow at about 20% per year for the next 6 years. All told, the money supply in the U.S. has increased by 4.59x over less than 7 years. During that same time, the dollar value relative to goods and services has fallen less than 10%, according to the government’s price index. This real-world example demonstrates that supply is only one component of price.
For the first 2 years of Bitcoin’s life, the network sustained an annual inflation rate of over 100%. For the first 5 years, it was around 30%, and for the first 8 years, it was over 10%. All told, the total “spending” Steem does to fund content, curation, and block production amounts to less than 10% APR.
Distribution of Tokens
Disclaimer, the analytic piece on the DPoS concept here does not tackle the debacle between PoS and DPoS on the blockchain consensus level. Instead, DPoS is discussed here as we view the ‘block production in DPoS mechanism’ on the blockchain as the equivalence of ‘emissions of new tokens’ by a DeFi project. The new STEEM tokens mined are distributed to the whole community, which consists of 3 main entities:
75% goes towards the rewards pool or the community for engagement.
15% to vested token holders (SP holders).
10% goes to the ‘witnesses’ or the block producers.
Note: This slightly changed after Justin Sun acquired Steemit from Ned in 2020. The updated distribution is as follows:
65% goes towards the rewards pool or the community for engagement.
15% to vested token holders (SP holders).
10% goes to the ‘witnesses’ or the block producers.
10% goes to the Steem Proposal System - essentially a benevolent holding (check the concept of BDFL) which votes for the betterment of the whole ecosystem, can be assumed as the team’s share.
As Steem operates on the DPoS consensus model, there is a need to create a token to represent the staked/ vested share - dubbed the Steem Power (SP). So, it is evident that Steem has a multi-currency system. The liquid token of the system is Steem tokens (STEEM) which can be traded on both the DEX and CEX. If one wishes to gain voting power, they would need to stake their STEEM through a process of ‘powering-up’ to SP. 1 SP would hold the same value as 1 STEEM. Powering-down SP to STEEM is possible, but it is linearly released over 13 payments for 13 weeks - constraining the mobility of the liquidity. Lastly, a Steem Blockchain Dollars (SBD) is designed to be the Steem ecosystem’s stable currency pegged to 1 USD. The interoperability of the 3 currencies in Steem is depicted below:
While ve-tokenomics is celebrated in the space as one of the main innovations brought by Curve finance, it is clear that the ‘powering-up’ process of STEEM to SP highly resembles that of locking up CRV as veCRV. A similar flywheel structure is observed for the case of Steem as well. First, the whole economic system is inflationary; hence, the STEEM needed to acquire SP will increase over time - check the attached diagram at the first part of Distribution of Tokens. Secondly, the users are incentivized to actively partake in the social interactions on the platform via upvoting or downvoting. The following chapter, Promotion of Active Participation, will describe the said actions in detail. Lastly, the rewards payouts in SBD are intended to alleviate a certain selling pressure of STEEM - it is working, albeit it is not trading at 1 USD in reality. The design of SBD is poor due to the blockchain’s prioritization to protect STEEM price instead of the SBD peg, more reading on the issue here. Nonetheless, it showcases the plausibility of segregating the payout rewards.
Promotion of Active Participation
Now, to promote active participation on the social platform, Steem has an interesting payouts mechanism to differentiate and incentivize different levels of interactions. The second diagram at Distribution of Tokens clearly shows that the 65% share given to the reward pool would be shared by (1) content author(s) and (2) content curator(s). When the (1) content author(s) are making a post, they could choose from a drop-down menu to decide the payout:
50%/50% - Half of the payment will be in SBD and the other half in SP.
Power Up 100% - The entire payout will be in SP.
Decline Payout - No payout.
The reward pool of 65% will translate into 32.5% each distributed among the author(s) and curator(s) if the author(s) opt for the default 50/50 split. SP mentioned above would play an essential role to decide the distribution of rewards; the more the SP, the bigger the reward. Altogether, this is known as the Proof-of-Brain (PoB), where the Steem blockchain enables the independent knowledge of the individuals to determine the quality of content - then vote to represent their says based on the stake-weighted mechanism or SP. Due to the natural vested interest of SP holders to ensure the longevity of the whole ecosystem, the game-theoretical model would push them to (1) actively participate and (2) vote for those posting, which would bring benefit to the platform.
The 2 actions for active participation are the upvote and downvote functionality:
Upvote - The reward will flow from the reward pool to upvoted post.
Downvote - The reward will go back to the pool depending on the influence power of the downvoter.
Note: Upvote earns curation reward while downvote does not earn a curation reward.
In the context of payout maturity, the age of a post is considered as 7 days. During these 7 days, the post can receive upvotes (or downvotes, if any). The net amount matures at the end of 7 days and is available for claim. Of the 7 days, the first 5 mins of the post from the time of publishing in Steem blockchain is known as Reverse Auction, in which a linear discount of curation reward applies. From the diagram below, the curators are penalized if they were to vote in the first 5 mins or last 12 hours.
Several services are available on the platform to ensure the maximization of the curation rewards to allow the community members to delegate their voting power. Based on our review, there are 2 identified types of services available: (1) delegation-as-a-service such as MinnowBooster and Dlease, which creates a secondary market to trade the voting power and (2) curation trail program (an example here) by different communities which use the delegated voting power to support the posting from their communities.
This fascinating occurrence capitalizes on 2 different behaviours in active participation. Upvoting is the mean to maximize one’s earnings on the blockchain because the action will distribute rewards to the users. It is only practical for those to use the delegation-as-a-service platform to boost their earnings because the interested parties are willing to pay extra (bribe) to lobby extra voting powers. However, suppose they would like to emphasize the betterment of the ecosystem. In that case, delegating to curation trail program can be more beneficial because reputable representatives would curate and gatekeep the quality of the content. Let’s look back at the bribe system on Curve finance; it was evident that there is a pain point that we could address.
Takeaways for the Team
Congruent is a new project launched (rebranded) to leverage the concept of Governance-as-a-Service by developing transformative and disruptive products leveraging the holdings of DeFi protocol governance tokens - which ultimately optimize and lubricate the DAO-interoperability on Liquidity War.
On the Distribution of Tokens, the block productions by the DPoS consensus system resembles the token emissions using ve-tokenomics. As ve-tokenomics can initiate and sustain a ‘flywheel effect’; ergo, it is suggested that Congruent adopts a similar distribution model. The Congruent protocol can also implement the multi-currency model to allow auto-compounding of vested shares and mitigate the selling pressure by paying out in a dividend model.
As for the Promotion of Active Participation, the Proof-of-Brain (PoB) model conceptualized the emissions of tokens depending on different levels of participation. PoB rewards the brainwork contributed towards the platform with tokens in layman’s terms. This concept is similar to the so-called Proof-of-Works model conceived by the Concave team in the whitepaper of their tokenomics - the participants earn the whitelist eligibility for their contributions inside the Discord server. Therefore, we envision that Congruent could pioneer the concept of Proof-of-Governance (PoG) to align the emissions of tokens to the participants’ contributions in the governance process. For instance, the Congruent protocol could incentivize the following for the betterment of governance in the DeFi space:
The creation of a governance proposal.
Active participation in the governance forum.
Upvote or downvote a particular proposal.
As discussed in the last paragraph of the Promotion of Active Participation chapter, there are 2 different attitudes in active participation - one group would seek to increase their benefits maximally; another would thrive to enhance efficiency. This sentiment is equally echoed by Holy from the Congruent core team. Thus, Congruent protocol could quickly establish itself to solve this pain point because most thriving protocols in the space right now only tackle the maximization of benefits (see Bribe by Curve or Votium by Convex).
It is challenging to instil the true spirit of democracy, as the ‘1 token 1 vote’ connotation is flawed in design because it can be easily skewed using financial advantages. One potential solution can be to segregate the token emissions from voting strength further - the emissions of tokens are still directly proportional to the vested token amount; but, the voting strength can be formulated as ‘quadratic voting’ (as implemented by Gitcoin). Adopting such a system should address the misalignment or asymmetry in rewards distribution and voting power.
Credit:
Disclaimer:
The opinions expressed in this article are those of the author’s and do not necessarily reflect the Congruent's official policy or position. They should not form the basis for making any financial decisions either.