A better design for defi grant programs
Grant programs in the defi and blockchain space should take some lessons from existing real-world competitive and entrepreneurial grant programs.
By Darcy W. E. Allen and Chris Berg, originally posted here.
The blockchain and defi sector should understand more about how real world grant giving bodies function. Nowhere is this clearer than in the recent debate about UniSwap and its new $20 million Defi Education Fund.
In the real world, grant giving is a lot like venture finance. It is an entrepreneurial activity involving the discovery of new information, new opportunities, and new ideas. It helps realise those opportunities and ideas and is rewarded for doing so.
The fact that grants are done with a for-purpose goal while venture finance is done with a for-profit goal only makes a difference at the margin. The best grant giving bodies in the world work very hard to ensure that the custodians of funds have incentives tightly aligned to the overall objectives of the body. Some even use external independent auditors to see whether grants align to objectives, and penalise the program’s management if they do not. These rules bind the grant makers, allowing the grant seekers to innovate and discover how best to achieve the programs objectives.
Admittedly, it can be sometimes hard to see the entrepreneurial and discovery nature of grant programs. Academic research grants tend to be highly bureaucratic processes with layers of committees and appointed experts collating and judging grant proposals at arms-length from the funders.
But ultimately this bureaucracy has a purpose. Those systems of rules might seem inefficient, but they have been designed to align the dispersal of funds with the objectives of the fund. In the case of the Australian Research Council, all those committees are intended to fulfil the objectives of the Department of Education’s scientific mandate through discovery and investment. (Let’s not get hung up about how effective these government programs are.)
At the other end of the spectrum is Tyler Cowen’s Emergent Ventures grant program, where almost all decision-making is Cowen’s judgement. But this too is a structure designed to align objectives with fund dispersal. The objectives of the fund are to allow Cowen to use his knowledge to support “high-risk, high-reward ideas that advance prosperity, opportunity, and wellbeing” — and by all accounts the program is an incredible success.
Two approaches to defi grants
Right now we broadly have two models of grant giving in the defi space. The first is small centralised grant committees. These tend to be small groups of authoritative community leaders with near absolute control of large treasuries assessing and granting funds to desirable projects. These leaders may be elected or appointed, but either way they are using their authority in the community to legitimate their decisions. They may have a deep understanding of their ecosystem and its funding needs. An obvious problem with this is the risk that committee leaders opportunistically fund projects based on personal relationships, rather than ecosystem value.
The alternative model — and the most common one — is putting all grant proposals up to a vote of all relevant stakeholders, that is, holders of a governance token. Designing structures for effective collective decision-making is one of the hardest problems in political science. It is no surprise that some decision-making in the nascent blockchain governance world have been controversial.
But there’s a fundamental problem with this democratic model to grant making: it makes very little sense to believe that a full distributed democratic community can make the sort of entrepreneurial decisions that we expect from both venture finance and grant giving bodies themselves. Why would we expect a diverse, pseudonymous community of governance token holders to coordinate around extremely uncertain entrepreneurial decisions?
Throwing every proposal to a mass vote is the worst of all worlds. First, every proposal ultimately becomes a public vote about the objectives of the program itself. Should the treasury’s funds be used for marketing, or research, or to build new infrastructure? Grant recipients, and the ecosystem that relies on them, are left with inconsistency and unpredictability.
Second, there is little reason to believe that a mass vote will reveal the best investments. Highly decentralised voting may protect against opportunism, but it isn’t likely to surface information about entrepreneurial investment opportunities — exactly what is needed for successful grant-giving. This precise information-revelation problem is the motivation and intuition between mechanisms such as quadratic funding, futarchy, and commitment voting.
A better grant program design
This is a solvable problem. Treasuries should give budgets to individual ‘philanthropists’. Those philanthropists should then make entrepreneurial investments and be rewards on the basis of the success of their invested projects.
In this approach, the full set of tokenholders sets the objective of the grant program, or an individual round. These objectives would shift as a given ecosystem and the broader industry develops — for instance from funding oracle feeds, to bridging infrastructure, to policy change. Grants are broken into funding rounds. The length of those rounds, say a year or two, must be long enough that there are observable outcomes from grant projects. Rounds could be sequential or overlap.
Each round, a set of philanthropists (say, five) are chosen (elected or appointed) and given discrete budgets. The number of philanthropists for a given round could also be decided by all tokenholders.
Once the funds are dispersed to each philanthropist, they run separate and independent grant programs. They must have credible autonomy: with their own rules, their own application processes, and their own interpretation of the objectives of the overall grant program.
At the end of the round, the full set of tokenholders rank each of the five philanthropists according to how successful (how much value was added, how closely they aligned to objectives) their grants were. The philanthropists are compensated for their work based on that ranking, with the top-ranked getting the most reward.
In this way the grants program is designed to both fund projects, and to incentivise decision-making philanthropists to do a good job.
Our proposal drives the same sort of competitive, entrepreneurial energy that we see in venture finance into defi grant distribution.
Through grant program design we can encourage effective decision-making through feedback loops, while maintaining decentralisation (the risk that philanthropists will behave badly is limited to the length of a grant round) and giving philanthropists a personal stake in the success of the grants that they have distributed (encouraging them to support and shepherd them to fruition).
Grant program design matters a lot
It might be easy to dismiss grant program design as a sideshow in the blockchain industry, marginally interesting but ultimately not a central part of the success of any particular protocol. It would be wrong to do so.
Analogies in blockchain are difficult. But if DAOs are like corporations, then grant programs are how they do internal capital allocation — and as Alfred D. Chandler Jr. has shown, internal capital allocation has determined the shape of global capitalism. Alternatively, if blockchain ecosystems are like countries with governments, then when we talk about grant programs we’re talking about public finance — they are how we pay for public goods and deploy scarce resources in a democratic context.
Ultimately, the sustainability and robustness blockchain ecosystems require effective use of resources. The success of grant programs will form a critical part of the success of blockchain and dapp protocols. They should seek to harness the same entrepreneurial energy and effort that has driven the rest of the blockchain industry.
Darcy W.E. Allen and Chris Berg are with the RMIT Blockchain Innovation Hub.