The DAOist invited Lawrence from Coordinape, Livia working on Praise at the Token Engineering Commons and Aaron founder of Govrn, to share their insights on rewards systems in DAOs during a panel at EthBarcelona. In a nutshell, a rewards system describes the tools and practices a community adopts to distribute financial and non-monetary rewards for member contributions.
While each organization represented on stage tackles the issues of rewards within DAOs in different, often complementary ways, moderators Elco and Angela, invited the panelists to emphasize their personal experience as long term DAO contributors, living in the rewards systems they’re building on a daily basis. Rather than shilling a single best practice on who gets what, the discussion highlighted the structural importance and challenges to keep in mind when thinking about DAO rewards systems.
Upon reflection, two major takeaways emerged:
(1) Bottom-up goal setting is hard. Rewards are here to help, and
(2) extrinsic and intrinsic rewards work best in tandem and transparency.
Bottom-up goal setting is hard. Rewards are here to help.
The core question here is: who gets to dictate the project’s mission and goals, in a decentralized, flat organization? Put differently: how can goals be set bottom-up, rather than top-down?
Aaron gave an anecdote from Govrn, where the team started by setting goals top-down but quickly came to realize that what people on the ground were working towards was in fact something completely different.
As a consequence he realized that “goal setting needs to become an emergent property of the system”.
But how can distributed online communities sense where their priorities lie at any given time?
This is where rewards come in.
Livia argues that having data on how people reward each other's contributions can provide valuable insights into understanding the lived values, mission and goals of a community. As such, rewards act as a feedback mechanism that can help to understand community values in a bottom-up manner and guide a community in adjusting its goals accordingly.
For Livia’s “transparency might be uncomfortable, but it can never be something bad”.
The argument here is that if we have more data on how people value different contributions, we’ll gain better insights into what communities value overall and be able to set more appropriate goals. This logic touches on a more salient debate which went on throughout the conference and in the wider community. It’s the old question of transparency versus privacy, now often reframed in terms of the solarpunk and lunarpunk imaginaries.
Elco points towards ZKPs as a possible way of reconciling our need for better data with the desire to protect private information and minimizing the risk of outside surveillance. In any case, confronting this tension will likely constitute a key issue for improving DAO governance, goal setting and rewards systems in the coming years.
Goodhart’s law
Another challenge to the approach, where the contributions that get the most rewards become the goal, is that it easily falls prey to Goodhart’s law.
Goodhart’s law states that: “when a measure becomes a target, it ceases to be a good measure”.
In the context of DAOs then, using quantifiable rewards for different contributions to sensemake around high level goals and values may lead communities to maximize for high reward contributions rather than activities that support the overarching goal.
Overcoming the risks of Goodhart’s law involves thinking deeply about when and how contributions get measured and quantified.
This leads us to the next point:
Extrinsic and intrinsic rewards best work in tandem and transparency.
Angela led with the provocation that extrinsic rewards, i.e. tokens, don’t only bear the risk of leading us deep into Goodhart’s forest but can also be gamed intentionally and override contributor’s intrinsic motivation. The last point especially seemed to strike a nerve.
Aaron argued that it’s “detrimental to talk too much about extrinsic versus intrinsic rewards”, instead we should be thinking more akin to Maslow's hierarchy of needs, or the MasDAO hierarchy of rewards, where extrinsic rewards such as tokens are required to put food on the table and free up capacity for intrinsically motivated contributions.
While everybody being freed from financial need and able to focus on intrinsically motivated activity sounds heavenly, it’s not easily done right. When it comes to distributing resources, the question of fairness inevitably plays a role (a question that generations of philosophers working in the field of distributive justice have been trying to grapple with). While questions around fair resource distributions in any given context remain, the discussion around rewards systems in DAOs has “given people a space to talk about money and values” , argues Lawrence - a conversation that often remains hushed and behind closed doors in many traditional contexts.
But what happens when outside conditions change and resources are scarce (read: bear markets)?
A short form, tongue in cheek answer according to Lawrence: “The system stays, the numbers just go down”. Unfortunately, for many, the reality often looks different; there seems to be a general fear that if you’re not a shadowy super coder and numbers go down, you might need to start looking for a “real” job next month. It’s a reality that all panelists aim to change.
Consequently, it becomes important to design rewards systems in a way that helps contributors take a holistic view of the organization when deciding how to allocate rewards when money is scarce.
Lawrence brings up the idea of capped rewards which could ensure that top contributors cannot make disproportionately more than others - an interesting approach I hope to see DAOs experiment with in the future. While systems design is crucial, building a community culture that helps people feel secure is key to designing resilient rewards systems.
“Rewards are only the tip of the iceberg. Being a healthy group of people requires a lot more than Coordinape circles.”, says Lawrence.
Culture determines how healthy internal competition plays out, it helps communities value participation beyond financial rewards (such as relationships, vibes and education - all of which remain abundantly available in hard times) and is essential to building the types of healthy environments that enhance members’ wellbeing beyond financial security.
Aaron suggests incorporating learning into ritual practices to strengthen a communities approach towards rewarding each other and the required culture to support it.
Overall, he argues: "there is no one way to do rewards!! The journey is the destination and getting rewards right is not a static end goal."
We look forward to this conversation continuing, an explosion of experimentation in this domain, and to facilitate translating learnings across DAOs and beyond.