The content of this post was originally presented by @davehrlichman as a talk at ETHDenver 2023. Published with many thanks to @spengrah and @nintnick for their contributions and valuable feedback that greatly improved this talk.

DAOs (decentralized autonomous organizations) could be amazing.

But they’re falling short of their potential.

Why? One of the biggest issues I see is that DAOs, like all non-hierarchical organizations, don’t have a good way to delegate roles, responsibilities and authorities to get things done.

There’s been a lot of talk in this space about voting delegation and its importance in DAO governance -- giving your voting power to a delegate who will vote in DAO governance on your behalf. I’m talking here about operational delegation -- giving people and small groups revocable authority so they can act on behalf of the organization and get things done efficiently.

Currently only hierarchies are good at this. And so many DAOs either end up larping as a decentralized network but really are controlled by a small group, or they wallow in the tyranny of structurelessness, unable to get things done with any level of coherence and consistency.

I believe this is one of the most important problems in the space that we need to address. This post presents a framework for how we can think about and solve this issue.

The good and bad news is this problem is not unique to DAOs. It’s following the same pattern seen in decentralized organizations, cooperatives, and networks of all kinds, which I wrote about in the book Impact Networks. Over and over again I’ve seen collaborative efforts that have so much promise, but something is getting in the way of them really fulfilling their potential: the tragedy of the responsibility commons.

What is a responsibility commons? Whenever groups of people coordinate to do something together, things will come up that need to get done. Often, as things come up, it’s unclear who is responsible to take specific things on, and those things then fall into a responsibility commons.

In hierarchies, people have very clear roles and are delegated authority to cover the necessary responsibilities, so the responsibility commons stays small and manageable. But in networks of agents (like DAOs), where nobody has explicit power over anybody else, there is no mechanism to assign – or ensure that the right individuals are claiming – responsibility.

As a result, many of the things that need to get done either go unclaimed or undelegated, or they are claimed by the wrong person, or they are claimed by someone who doesn’t actually follow through. This is the tragedy of the responsibility commons, and it's a huge reason why DAOs today are largely unable to get things done at the same level of efficiency as traditional organizations.

To understand how we can overcome the tragedy of the responsibility commons, we have to look closely at how we’re organizing, and what’s getting in the way.

Hierarchies have been a dominant organizing structure for thousands of years. They are great at getting things done efficiently in complicated domains where we know what needs to be done and we can plan things out in advance. People in hierarchical organizations are generally pretty clear on their responsibilities, or they can be delegated responsibility by someone higher in the hierarchy. Hierarchical structures are also predictable and stable, so they can scale effectively.

However, hierarchical structures are often limited in their ability to address complex and multifaceted issues that don’t have a clear solution. In the face of complexity, the rigid structure creates bottlenecks. And while the structure may be reliable, its inflexibility also means it is slow to adapt.

Meanwhile, the high concentration of power within hierarchies means that certain groups at the top have enduring structural advantages and are rewarded far more compared to those lower in the organization, perpetuating inequities and discouraging those at the bottom from fully engaging. In other words, they are “captured”. The system’s distribution of power is out of balance. Certain groups and individuals resemble a black hole, capturing a disproportionate amount of resources from the system compared to the value they are providing.

For these reasons, we are now seeing a widespread shift away from entrenched power hierarchies, and a rise of cooperatives, purpose trusts, and DAOs around the world, each of which are taking a network approach to coordination.

Networks are actually quite good at engaging complex issues, because they are able to combine diverse perspectives to discover new possibilities, they are resilient and adaptive, and they’re much more capture resistant than hierarchies. Overall, networks are far better suited than hierarchies at governing and stewarding shared resources and public goods, which is why DAOs have evolved to take a network approach to coordination and favor collective decision-making over unilateral control.

But as they grow larger and more diffuse, DAOs, like all networks, struggle to achieve clarity on roles and responsibilities, hold people accountable to their commitments, and explicitly delegate and revoke authorities.

All this leads to… the tragedy of the responsibility commons

The solution to this tragedy is the ability to explicitly delegate roles and responsibilities in a nuanced way down to the level of the individual, provide (and revoke if needed) the necessary authorities and permissions to fulfill those roles and responsibilities, and create accountability mechanisms so people follow through on their commitments.

As stated above, currently only hierarchical organizations are good at this. Therefore, to solve the tragedy of the responsibility commons in non-hierarchical organizations, we need a means of delegating or revoking roles and authorities in a way that still reinforces decentralization of power.

We need structure that doesn’t result in capture. Structure where the ultimate authority is held not by an individual, but by the collective as a whole. Structure that gives people power-to without creating entrenched power-over structures across the whole organization.

What’s missing from DAOs is the ability to create finite hierarchical structures within a collectively-governed whole.

We can blend the best of hierarchies and networks so that we get the resilience and capture resistance of networks with the implementation chops of hierarchies.

What could it look like to create fit-for purpose hierarchical structures within a collectively governed whole? Let’s take at two examples.

Let’s say a DAO wanted to add the capability to run a grants program, which will require some structure. The DAO could give the authority to a grants committee to make decisions about how grants are selected.

The grants committee then determines that it needs specific individuals, who they call grants champions, to be accountable for following the progress of each grant, and supporting the grantee along the way. So the committee delegates authority to the champion, who oversees the work of the grantee and makes sure the grantee has access to the relevant documents and resources they need to do their best work.

As you can see here, there is structure, but it is fit for purpose: it arises when needed, and will dissipate when it is no longer needed. You can also see here the conditions for solving the tragedy of the responsibility commons: there is a clear delegation of responsibility and authority, but that authority is revocable and could even be connected to specific eligibility criteria or accountabilities, and ultimately it is the collective who holds the ultimate power — the DAO is at the root of the authority tree, not an entrenched individual.

The structure is built to get things done, while remaining entirely capture resistant.

Let’s take a second example, the case where a DAO recognizes the need to create a specific project team or working group to organize an event at EthDenver. By authority of the whole, it could spin up a project team to get that done, and give the team signing authority on a multisig to organize the event.

That’s a good start, but it’s not enough. We need to be able to delegate responsibility further, down to the level of the individual, because even teams need clear roles and accountabilities. One of the roles the team identifies is a facilitator.

So the project team elects a facilitator, to give them certain powers like authority over a discretionary budget, and the ability to cast to warpcaster on behalf of the team. The facilitator is also responsible for overseeing the work of some external contractors responsible for catering and logistics, as well as giving those contractors the permissions they need to get access to the group’s communication channel and relevant documents.

With this structure, you can see that the facilitator has certain powers to get things done, but this is not power over the members of the project team. Quite the opposite — the team members actually have the authority to revoke the powers of the facilitator, and to elect a new facilitator if they’d like. And meanwhile the DAO has the authority over the team as a whole.

So while there is a structure, created to fulfill a specific purpose, it is resistant to capture, and the structure is collectively governed by the group. Once the event is finished, the team’s authorities could be revoked by the DAO, and the structure dissolves as it’s no longer needed.

These are just two of many possible capture-resistant “structure legos” that DAOs could adopt. You can think of these structure legos as fit-for-purpose structures to unlock new capabilities for your DAO… a way to give structure its place within the self-organizing organism of a DAO.

This is what is meant by structure without capture, and this is what I believe will solve the tragedy of the responsibility commons, enabling DAOs and other networked organizations to create capture-resistant structures to get things done.

This is also precisely what we’re working on at Hats Protocol, a protocol for role and authority delegation. A protocol for structure without capture.

Learn more and get involved at hatsprotocol.xyz. We’re excited to see what structure legos you create with it!

The content of this post was originally presented by @davehrlichman as a talk at ETHDenver 2023. Published with many thanks to @spengrah and @nintnick for their contributions and valuable feedback that greatly improved this talk.

DAOs (decentralized autonomous organizations) could be amazing.

But they’re falling short of their potential.

Why? One of the biggest issues I see is that DAOs, like all non-hierarchical organizations, don’t have a good way to delegate roles, responsibilities and authorities to get things done.

There’s been a lot of talk in this space about voting delegation and its importance in DAO governance -- giving your voting power to a delegate who will vote in DAO governance on your behalf. I’m talking here about operational delegation -- giving people and small groups revocable authority so they can act on behalf of the organization and get things done efficiently.

Currently only hierarchies are good at this. And so many DAOs either end up larping as a decentralized network but really are controlled by a small group, or they wallow in the tyranny of structurelessness, unable to get things done with any level of coherence and consistency.

I believe this is one of the most important problems in the space that we need to address. This post presents a framework for how we can think about and solve this issue.

The good and bad news is this problem is not unique to DAOs. It’s following the same pattern seen in decentralized organizations, cooperatives, and networks of all kinds, which I wrote about in the book Impact Networks. Over and over again I’ve seen collaborative efforts that have so much promise, but something is getting in the way of them really fulfilling their potential: the tragedy of the responsibility commons.

What is a responsibility commons? Whenever groups of people coordinate to do something together, things will come up that need to get done. Often, as things come up, it’s unclear who is responsible to take specific things on, and those things then fall into a responsibility commons.

In hierarchies, people have very clear roles and are delegated authority to cover the necessary responsibilities, so the responsibility commons stays small and manageable. But in networks of agents (like DAOs), where nobody has explicit power over anybody else, there is no mechanism to assign – or ensure that the right individuals are claiming – responsibility.

As a result, many of the things that need to get done either go unclaimed or undelegated, or they are claimed by the wrong person, or they are claimed by someone who doesn’t actually follow through. This is the tragedy of the responsibility commons, and it's a huge reason why DAOs today are largely unable to get things done at the same level of efficiency as traditional organizations.

To understand how we can overcome the tragedy of the responsibility commons, we have to look closely at how we’re organizing, and what’s getting in the way.

Hierarchies have been a dominant organizing structure for thousands of years. They are great at getting things done efficiently in complicated domains where we know what needs to be done and we can plan things out in advance. People in hierarchical organizations are generally pretty clear on their responsibilities, or they can be delegated responsibility by someone higher in the hierarchy. Hierarchical structures are also predictable and stable, so they can scale effectively.

However, hierarchical structures are often limited in their ability to address complex and multifaceted issues that don’t have a clear solution. In the face of complexity, the rigid structure creates bottlenecks. And while the structure may be reliable, its inflexibility also means it is slow to adapt.

Meanwhile, the high concentration of power within hierarchies means that certain groups at the top have enduring structural advantages and are rewarded far more compared to those lower in the organization, perpetuating inequities and discouraging those at the bottom from fully engaging. In other words, they are “captured”. The system’s distribution of power is out of balance. Certain groups and individuals resemble a black hole, capturing a disproportionate amount of resources from the system compared to the value they are providing.

For these reasons, we are now seeing a widespread shift away from entrenched power hierarchies, and a rise of cooperatives, purpose trusts, and DAOs around the world, each of which are taking a network approach to coordination.

Networks are actually quite good at engaging complex issues, because they are able to combine diverse perspectives to discover new possibilities, they are resilient and adaptive, and they’re much more capture resistant than hierarchies. Overall, networks are far better suited than hierarchies at governing and stewarding shared resources and public goods, which is why DAOs have evolved to take a network approach to coordination and favor collective decision-making over unilateral control.

But as they grow larger and more diffuse, DAOs, like all networks, struggle to achieve clarity on roles and responsibilities, hold people accountable to their commitments, and explicitly delegate and revoke authorities.

All this leads to… the tragedy of the responsibility commons

The solution to this tragedy is the ability to explicitly delegate roles and responsibilities in a nuanced way down to the level of the individual, provide (and revoke if needed) the necessary authorities and permissions to fulfill those roles and responsibilities, and create accountability mechanisms so people follow through on their commitments.

As stated above, currently only hierarchical organizations are good at this. Therefore, to solve the tragedy of the responsibility commons in non-hierarchical organizations, we need a means of delegating or revoking roles and authorities in a way that still reinforces decentralization of power.

We need structure that doesn’t result in capture. Structure where the ultimate authority is held not by an individual, but by the collective as a whole. Structure that gives people power-to without creating entrenched power-over structures across the whole organization.

What’s missing from DAOs is the ability to create finite hierarchical structures within a collectively-governed whole.

We can blend the best of hierarchies and networks so that we get the resilience and capture resistance of networks with the implementation chops of hierarchies.

What could it look like to create fit-for purpose hierarchical structures within a collectively governed whole? Let’s take at two examples.

Let’s say a DAO wanted to add the capability to run a grants program, which will require some structure. The DAO could give the authority to a grants committee to make decisions about how grants are selected.

The grants committee then determines that it needs specific individuals, who they call grants champions, to be accountable for following the progress of each grant, and supporting the grantee along the way. So the committee delegates authority to the champion, who oversees the work of the grantee and makes sure the grantee has access to the relevant documents and resources they need to do their best work.

As you can see here, there is structure, but it is fit for purpose: it arises when needed, and will dissipate when it is no longer needed. You can also see here the conditions for solving the tragedy of the responsibility commons: there is a clear delegation of responsibility and authority, but that authority is revocable and could even be connected to specific eligibility criteria or accountabilities, and ultimately it is the collective who holds the ultimate power — the DAO is at the root of the authority tree, not an entrenched individual.

The structure is built to get things done, while remaining entirely capture resistant.

Let’s take a second example, the case where a DAO recognizes the need to create a specific project team or working group to organize an event at EthDenver. By authority of the whole, it could spin up a project team to get that done, and give the team signing authority on a multisig to organize the event.

That’s a good start, but it’s not enough. We need to be able to delegate responsibility further, down to the level of the individual, because even teams need clear roles and accountabilities. One of the roles the team identifies is a facilitator.

So the project team elects a facilitator, to give them certain powers like authority over a discretionary budget, and the ability to cast to warpcaster on behalf of the team. The facilitator is also responsible for overseeing the work of some external contractors responsible for catering and logistics, as well as giving those contractors the permissions they need to get access to the group’s communication channel and relevant documents.

With this structure, you can see that the facilitator has certain powers to get things done, but this is not power over the members of the project team. Quite the opposite — the team members actually have the authority to revoke the powers of the facilitator, and to elect a new facilitator if they’d like. And meanwhile the DAO has the authority over the team as a whole.

So while there is a structure, created to fulfill a specific purpose, it is resistant to capture, and the structure is collectively governed by the group. Once the event is finished, the team’s authorities could be revoked by the DAO, and the structure dissolves as it’s no longer needed.

These are just two of many possible capture-resistant “structure legos” that DAOs could adopt. You can think of these structure legos as fit-for-purpose structures to unlock new capabilities for your DAO… a way to give structure its place within the self-organizing organism of a DAO.

This is what is meant by structure without capture, and this is what I believe will solve the tragedy of the responsibility commons, enabling DAOs and other networked organizations to create capture-resistant structures to get things done.

This is also precisely what we’re working on at Hats Protocol, a protocol for role and authority delegation. A protocol for structure without capture.

Learn more and get involved at hatsprotocol.xyz. We’re excited to see what structure legos you create with it!