Reading Group - Schneider, et al, 2020. Modular Politics

[ebeth]

Welcome to the reading group post about Modular Politics: Toward a Governance Layer for Online Communities. The SCRF reading group aims to help new researchers gain knowledge and be a resource for new community members. Readings are solicited from the community in Discord and voted on by the reading group. Once finished with the work, one or more members can summarize the work and publish it on the forum for general discussion.

Our first piece will be Schneider, N., et al., (2020). Modular Politics: Toward a Governance Layer for Online Communities. Computers and Society, 5, 16-42. Retrieved from: https://arxiv.org/abs/2005.13701

Please put your questions or discussion points about the reading in this thread by Friday the 27th. We can follow up with discussion and answers after then. We can also meet up for a conversation, but we’d like to keep it as async as possible.

[Fizzymidas]

So, how exactly does this reading group work. I don’t think I get it.

For books, it is a bit clearer to me, usually we do like one chapter per week and then meet weekly to discuss that chapter.

So, how does it work for this paper?

[zube.paul]

For a paper, I think we’re all going to take one read through of the paper and then put questions and observations on this thread. Then we can start digging into those questions as responses in the thread. If people are interested, there could possibly be a call that gets set up too.

[tukwila]

So I spent some time reading the paper this weekend, and took a couple of margin notes. A few quotes and my brief thoughts as I go through them:

The concept of a modular governance system really spoke to me.

Such mechanisms as elections, boards, term limits, and transparent decision-making are norms for any incorporated entity in robust legal regimes, but to the extent that they do occur in online networks, they are costly to build and maintain, and they typically must be implemented by means extraneous to the feature-set of the network’s platform itself.

Governance is important, but secondary to most project’s core competencies

Portability: Governance tools developed for one platform should be portable to another
platform for reuse and adaptation.

It’s not a governance system, it’s a governance protocol. It doesn’t judge/value the content but provides a framework to process it as a community.

Independent units of these platforms may be seen to adopt many types of governance styles, but considered as populations of populations, they can be characterized by the freedom users have to exit any one community for any other—that is, they favor “exit” over “voice”

We don’t have good systems for keeping oppositional parties involved in communities right now, so they become monopolized by one governance style or oligopolic leadership

Self-organization is also evident in the ability of communities to develop and sustain widely held informal norm systems

It’s often the emergent rules that are most powerful and important

a tension captured by “Ostrom’s Law,” which asserts that “a resource arrangement that works in practice can work in theory”

This commitment to the experiential component of design was important to the Ostroms, who emphasized the craft of institution design and likened its practitioners to artisans

By instituting governance as a protocol with modular component architecture, you can iterate more freely and also make the process of getting it right more technical and, ideally, changing it less emotionally charged.

Through an SDK with an intuitive interface, Platform operators and participants alike might browse existing Modules or create their own; they might also adapt, customize, and debug these Modules before importing them into an Instance.

If implemented as an open standard, rather than as a centralized platform, Modular Politics could
contribute to furthering the vision of what has been called “Web 3.0”—an open and interoperable
network of online services that resist the tendencies of centralization that have become so persistent
on the Internet today. In the context of governance, this means that a particular community could rely on a variety of integrated services in order to achieve a particular governance structure that
spans across multiple platforms.

I think what this paper is describing is the next level of the types of governance we are seeing in DeFi right now. For instance, Compound’s Governor Bravo, along with OpenZeppelin, are offering open source governance solutions that basically are this SDK. These systems have been widely adopted while being slightly tweaked, and there is even a live vote on Uniswap to upgrade to Governor Bravo from Governor Alpha.

What we haven’t seen yet, however, are interoperable governance systems that work across platforms. What we’re left with so far is basically three options that can loosely be termed Meta Governance.

In option 1, a community like Index Coop votes on what to vote on the underlying proposal. Since there are 48.2m UNI held by DPI, then INDEX holders vote on what to vote in UNI governance.

Option 2 is a pass through. I haven’t really seen it, it’s complex and costly.

Option 3 is what we are working on at xToken (disclosure: I am on the team). We’ve developed mandates that indicate how the fund will interpret proposals. Over time as more proposals are voted on the body of precedence increases, and the community can also give us feedback on how the mandates are applied.

But as DeFi grows, so too will the types of stakeholders and ways of interacting. Due to the inherent composability of DeFi, when one protocol changes course, often other protocols are impacted. Sometimes they are collaborators, and sometimes competitors, but they are also stakeholders!

I think the direct governance questions are being addressed right now, but the need to have a more flexible cross-project interoperable governance layer is only beginning to be addressed.

Thanks for the opportunity to read and discuss!

[zube.paul]

I’m working my way through the paper today and will be adding my margin notes ideas soon, but I wanted to take a moment to thank you for such a great post with your questions and observations.

[tukwila]

Thank you @zube.paul. It’s a topic I am very passionate about. Just found this forum recently, looking forward to learning from you all and participating more.

[zube.paul]

This was an exciting paper to read and I’m thoroughly looking forward to discussing it. The margin notes/questions I had were:

There are some real behavioral and technological barriers to implementation. As Modular Politics is developed, what are the steps to reduce or address these barriers? For example, what might some incentives be to help Platform operators?

Platform operators will have to implement strategies for defining and circumscribing the role of Modular Politics in the context of the host system, restricting the scope to particular participants and spheres of influence. (p. 12)

I’m going to be following the citation bread crumbs regarding IAD. The paper provides a quality summary, but as Modular Politics has IAD as a foundational framework, I want to come to our discussion with a better understanding of its strengths and weaknesses.

There are some interesting intersections here with what @jasonanastas has been posting about regarding data ethics. I would love to have him join this discussion at some point. Previously we had somewhat discussed if and where a balancing point was regarding privacy as a desirable data ethic. I think the following quote re-opens this thought for me.

User data has the potential to inform an ongoing conversation with existing social and political theories. (p. 18)

The core question here (and in the governance space overall) is about the balance between design and users. I think this can be a great discussion point in this thread. Without getting too much into esoteric philosophy, I am interested in the “how much of the tooling problem needs to be solved in order to evaluate the human coordination problem” type of question. Claims such as:

Online communities could become powerful laboratories testing new ideas and learning from the failures and successes of other communities, potentially spurring a renaissance in democratic culture. (p. 19)

really opens this question up for me as being foundational to projects interested in governance, agency, and human coordination. In particular, it seems there is an underlying assumption here that if barriers are removed we would likely see more coordination/participation in governance. So the claim that:

Prior research suggests that although sophisticated and democratic governance is possible in online communities, it is difficult and rare. (p.5)

assumes that the reason behind the difficult and rarity is fundamentally a tooling problem as opposed to a how humans tend to behavior observation. For discussion purposes is there evidence for that assumption?

I did want to add that I would like us to discuss if this belongs in our Notable Works in Governance.

[ebeth]

I’m excited to dig into this piece with you all.

Some things I’ve been thinking about reading this piece:

1. What features of blockchain governance make it different than online governance in web 2 that would make modular politics specific to its needs?

2. Once question that’s been plaguing me - how can communities identify their needs in order to know which module’s to pick in order to select the best governance modules for themselves?

3. I think one of @zube.paul 's questions is vital - there are real behavioral barriers for people to participating in governance - how would this model alleviate those barriers? When does a community need to enforce other barriers in order to address bad actors in the space?

4. Why does an interdisciplinary approach like using IAD matter in decentralized spaces particularly?

5. In what ways does IAD as a framework serve for decentralized governance? In what was does it not? The authors write that IAD is “game-like” that allows for creativity. How would a more traditional framework for political science or social science complement or this model?

[ntnsndr]

So nice to see this discussion here, and I’m really enjoying what folks are raising so far. One thing worth adding, at the outset, is that we are actively working on implementing this paper in two spaces:

Metagov prototype: https://prototype.metagov.org
Modpol game mod: Media Enterprise Design Lab / Modpol · GitLab

On the question of the barriers to participating, it’s a super complex question. Modular Politics doesn’t really address it. And, I would argue, participation as such isn’t necessarily an unmitigated good. Maybe we might want to design a system to reduce the need for participation, so that people can spend that time doing better things—but enabling participation when it really, actually matters. One of the things I hope we can explore with our prototypes is how people in the wild end up behaving around the tools, how they use and don’t use them. In a world where we have more “governable spaces,” we are going to have to be much more attentive to the interface and experience designs for governance, as well as the culture we build around it.

On the design front, CommunityRule is another project I’m leading to experiment with interface design for governance practices. Would love your feedback on the current version, as we may explore doing a next stage of development in partnership with SCRF!

[zube.paul]

I greatly appreciate you taking the time to jump into this discussion. Great point that participation is not inherently good. I think we see this engagement/participation burnout in a variety of online and offline communities and governance systems. Keeping that engagement energy available for key moments is something that a well-designed system might provide. I suppose my slight pushback on that might be when/where we make that decision? I suspect that there will be some really interesting observations as this gets prototyped and used in the wild.