The free-rider problem arises when a good or service has private costs and non-excludable public benefits. A classic example is a firework show. Fireworks can be enjoyed by anyone in a particular area regardless of whether they pay for it (it’s hard to stop people from looking at the sky if they don’t have a ticket).
Lots of folks may want the firework show, but there’s not enough economic incentive to go through the effort and cost of arranging it if you only reap the private benefit (the enjoyment of getting to watch the fireworks yourself) and don’t reap some of the public benefit (everyone else’s enjoyment of the fireworks).
DAO Governance free riders
In DAOs, all token holders benefit from good governance whether they participate in it (create proposals, vote on proposals) or not.
Ultimately, this doesn’t matter much as long as good governance takes place, but I am not one to rely on the goodwill of some to do work for an entire group--anyone who has suffered through a group project in school can attest that this does not always lead to desirable outcomes.
Furthermore, the current state of affairs can have a centralizing effect--you’re incentivized to participate if your private benefit is a significant chunk of the group benefit, i.e. in the case of DAOs, you own a large percentage of the tokens.
Incentivizing governance participation
What if we privatize some of the benefits of governance participation?
I say we pay people to participate in governance. Those who participate should be rewarded with a greater stake in the protocol. From another perspective, we should be punishing those that don’t participate in governance with dilution of their stake in the protocol.
There are two primary activities in on-chain governance--creating proposals and voting on proposals.
Creating proposals is hard work, and should be rewarded through bounties, one-off grants, streaming contributor grants, and one-off payments baked into proposals. Proposals are not created equally in terms of their difficulty and impact (a parameter update is much simpler and typically less impactful than rewriting a contract).
While I think it’s important to incentivize the creation of good proposals, I don’t think there should be a one size fits all solution.
Voting, on the other hand, merits a more formalized approach. Among all the DAOs I’ve seen, voting is not explicitly incentivized, and so non-voters free-ride on the diligence of voters. I think voting should come with an explicit reward of more governance tokens in proportion to one’s voting balance.
I discuss some potential incentive misalignments with voting rewards and how to address them below.
Casting uninformed votes
One way to game voting incentives would be to simply vote (either always “yes,” always “no,” or randomly) on every proposal to earn rewards regardless of the proposal. This isn’t even hypothetical at this point as, @transmissions11 implemented a lovely contract that only allows you to vote no on every proposal.
My solution here is to provide substantially higher rewards for votes that end up winning. This encourages token holders to vote the best option and incentives folks to try to persuade others to their viewpoint as well.
Voting with the majority near the end of the voting period
Another type of free rider voter could simply wait until the voting period is near complete and vote at the end once the outcome seems assured.
My solution here is to have a reward decay schedule. The voting reward should be highest at the beginning of the vote period and decay to zero at the end. This encourages token holders to stay on top of upcoming votes and diligence the proposal before the voting period begins.
Spamming nonsense proposals
Another potential way to game voting rewards is to create useless proposals and vote on them.
However, many DeFi protocols, including Uniswap and Compound, have some degree of spam prevention built in--the proposer must meet a minimum threshold of delegated voting power. Although incomplete, one solution to this problem is to exclude the proposer (and thus their delegated voting power) from voting rewards.
If it becomes an issue, the vote threshold can be raised to discourage this type of behavior.
A Renaissance in Delegation
Direct voting is not for everyone. In my opinion, it’s difficult to be informed enough about more than 2 or 3 protocols to be an active participant, and that’s assuming that you have the ability to dedicate a significant amount of time to DeFi governance (and I’ve been told that most people, unlike me, have a life outside of crypto).
For many DAO token holders, delegation to an active community member makes the most sense. The beauty of voting incentives is that it makes being a delegate not just a labor of love, but a potential occupation.
By default voting rewards would accrue to the address that voted, i.e. delegates would get rewarded not only for their own voting power but the voting power of everyone who delegated to them.
While this is great for the delegate, it doesn’t leave much incentive to delegate your tokens. It's easy to imagine delegation contracts that allow delegates to return a percentage of rewards back to the delegators. Delegates would compete along two dimensions:
- how good of a delegate they are (high vote participation, good reputation, sound judgment, etc.)
- what percent of rewards they return to delegators
Concluding thoughts
DAOs (and web3 in general) offer a rich design space to create incentive schemes that allow us to build better applications. The current DAO voting model, however, is uninspired from an incentive perspective.
Voting rewards can better incentivize good governance and create a competitive market for delegation.