Distributed blockchain platforms and applications can be designed in ways that make them accountable to their participants. Unlike blockchain systems that reserve some decision-making powers for executive-style committees or founders, these platforms and applications are governed by people who may be unknown to each other (often referred to as community governance, but perhaps better described as collective governance). These platforms in turn govern us, in that they shape what actions we can take within a sub-system, either through automation (algorithmic governance), or by their non-automated institutional qualities (imposing standards, enacting rules). The dynamics of distributed governance by people and machines are beginning to emerge as these platforms grow.
How do we measure and understand blockchain governance? In the first part of this series I identify some of the rules and orderings that are arising in the sphere of blockchain platforms and applications known as DeFi and sketch some research parameters. In future posts I will discuss how automation and community governance co-produce stability and security in these systems. I am working towards a framework for thinking about the outcomes of distributed technology and the relationship of outcomes to governance. This includes identifying where and how ‘pop-up economies’ are emerging, which I briefly introduce at the end of this article.
Defining DeFi governance
In the emerging area of DeFi, automation is integral to the workings of financial infrastructures, enabling decentralised trading exchanges and, in some instances, helping to produce stable assets. Many DeFi applications incentivise people to contribute to liquidity pools, which are necessary for automated market-maker (AMM) algorithms and some stable coins to work effectively. Governance rights are therefore directly attached to responsibilities and economic incentives. In essence:
1. Individuals provide liquidity to liquidity pools (they may remove as needed).
2. Liquidity enables AMMs to work for the benefit of all users (even those not providing liquidity). A good explanation of the Uniswap AMM is here.
3. In some cases, individuals earn governance tokens when they provide liquidity, which they can use to vote on the direction of the platform. Some of these governance tokens have significant value attached to them, resulting in a behaviour known as ‘yield farming’ in which individuals provide liquidity to the most profitable pools in order to earn more tokens.
These features are resulting in particular governance challenges, as well as rapid innovation in governance mechanisms. Most of these stem from the direct connections between automation, open source code, market incentives and governance.
Fair launching and governance whales
When governance tokens are offered as incentives to provide liquidity, those who can provide more liquidity have greater power. Governance ‘whales’ — those with large stores of governance tokens — can become the primary decision-makers. This resembles share-holder governance without regulatory limits, although other features can alter or weaken this likeness.
For instance, In order to overcome this feature, some blockchain applications may in the future use radical forms of voting in decision-making. One of these is quadratic voting, in which individuals signal their preferences by ‘buying votes’, but where votes get incrementally more expensive as more votes are cast. Similar mechanisms are already being used to fund research and development; with Gitcoin’s quadratic funding, a project that receives 1000 donations of $1 will receive more matching than 1 donation of $1000.
Some DeFi protocols may also depart from shareholder governance in how tokens are allocated. The ‘fair-launch movement’ aims to encourage new products with tokens that have zero value at launch, and where there are no early investor or founder pre-allocations (they may, of course, find value through markets). These are considered to be more accessible, although in reality they may be club-like due to their communication dynamics (see below). Applications that use this method have been referred to as ‘Cronje markets’ (named after Andre Cronje who invented this model with yearn.finance).
Fast-forks and Sushiswapping
Copycat DeFi applications also complicate fair launch intentions. SUSHI was an example of a ‘fast-follower fork’ that copied the code of Uniswap, did a zero-value token launch and gave 100% of its tokens to anyone who participated (beating the well-established Uniswap to its own investor-led token launch). Copycat applications are interesting to observe in terms of their durability — i.e. whether loyalty and governance emerges within new systems or whether they are short-lived, opportunistic yield farming events.
While economic incentives should theoretically incentivise governance token-holders to make products stronger, the possibility of short-term economic gains can undermine governance. The anonymous founder of the SUSHI exchange, for example, cashed-out and destroyed the market, leaving the SUSHI ‘community’ to debate whether it was worth continuing. mStable, on the other hand, has been deliberately transparent with its founders’ rewards and acted to prevent founders and investors from being able to cash-out early, thereby signalling the long-term ambitions of the platform. Other mechanisms exist to overcome short-term gain incentives, such as staking coins (locking them up in a smart contract) in order to vote with incentives dependent on staking/lock-up. An early pioneer of this method is Decred (not a DeFi app but a cryptocurrency blockchain), which has also developed co-governance for treasury spending (Politeia).
The immutability inherent in blockchain platforms creates significant risks. Kevin Werbach discusses these as commitment devices (as per Douglass North). For instance, fast-forks can appear to be copies of audited code, yet depending on when the code was copied, may have errors since resolved in the original application. How a community responds when things go wrong is revealing. When a flaw was discovered in the Yam protocol, a sufficient number of yam holders delegated their votes to re-engineer the platform (unfortunately another flaw was discovered, but the episode revealed support for the Yam protocol nonetheless).
Finally, DeFi applications exhibit club-like qualities, even though participants by definition do not have to know each other to participate. Launches are generally discovered through networks and social media among yield famers who refer to themselves as the DeFi Degenerates. Memes are a common communication device, with some tokens created as ‘meme-coins’ using emoji food icons to make them instantly recognisable. The game-like qualities of yield farming also works to attract users.
My starting point here has been to identify some of the dynamics that are emerging from DeFi by paying attention to actors and events. The next task is to begin to measure governance in these systems. Some clear lines of investigation flow from the above:
· The value of a token can be understood through its use, how long it is held for, and whether it is maintained and held following a negative event. These can be ascertained through quantitative analysis of on-chain data.
· Context-specific factors also need to be taken into account: What are people being asked to vote on? What level of automation is involved in the maintenance of the application versus human direction and decision-making? The scarcity and supply of the token may influence behaviours. Whether risk correlates with behaviours (for instance, whether an application has been audited), would be interesting to examine.
· Comparing fully community/collectively governed applications with those that are hierarchical might provide some insight into the relationship between community governance and security.
Finally, we need to understand not just how these systems are governed but also what they enable. To do this we need to go beyond acknowledging their institutional properties and consider the capabilities that emerge from them. My ‘pop-up economy’ theory, which looks at what arises through use of blockchain infrastructure, is a starting-point for analysing the outcomes of these human-machine governance systems.
You can read Part II of this series here.
*Note this article was edited on 17/9/2020 AEST to clarify that I am not aware of any DeFi projects that are currently using quadratic voting. So far I have only seen talk of this as a possibility.
*I use Medium for work-in-progress communication and may change my views and analysis as the work progresses. I welcome constructive and respectful feedback from blockchain practitioners. DM me on Twitter @elinorrennie or email email@example.com
About me: I am a Professor at RMIT, working across the RMIT Blockchain Innovation Hub and the Digital Ethnography Research Centre. I am also an Associate Investigator of the ARC Centre of Excellence for Automated Decision-Making and Society. I acknowledge the support of the Australian Research Council, FT19010372.