On Signaling and Social Pressure

How the Ethereum community could begin making meaningful signaling-based decisions

People often either over- or underestimate the power of signaling and social pressures influencing group decisions. A good example of overestimating the power of signaling is the notion that Vitalik is a benevolent dictator due to his influence in the community. Or that Ethereum's governance boils down to a cartel of developers, despite neither party being able to make any direct changes to the Ethereum blockchain. On the flip-side token holders tend to underestimate their signaling clout. Token holders (for any project, including Aragon and Ethereum) can signal support for arbitrary proposals with relative ease due to tokens being easy to quantify, publicly visible, and resistant to censorship.

CarbonVotes can be used by ETH holders to weight group decisions. Anyone can deploy an instance of CarbonVote or a similar tool and create their own polls. And if people participate in the process and produce a clear signal it could have a massive impact on the actions of other stakeholder groups, including core protocol developers. In practice CarbonVotes haven't been very effective so far, but that is almost certainly to do with the specific implementation and application rather than an issue with signaling in general.

The issue is that it is incredibly difficult to coordinate large, diverse groups in order to produce compelling signals. In the absence of clear aggregate signals only the most prestigious voices or the voices which are able to most effectively astro-turf social media are heard.

This is an important distinction, because the solution to a lack of clear signals is often presented as a need to formalize the process so that a decision can be reached regardless of signal quality. While that strategy allows decisions to be reached quickly, it doesn't ensure decisions are representative of a community and does not address the inherent coordination challenge of aggregating preferences from a diverse set of stakeholders.

Regardless of an organization's formal governance process, improving the ability of the organization's stakeholders to produce clear signals will improve governance outcomes. This is why in most democratic countries the right to free speech is so valuable, and in autocratic regimes communication is typically tightly controlled. In many cases the actual governance process is less important than the ability for groups to coordinate through strong signals. Which is why many corrupt governments adopt formal processes that appear to be fair, but in practice produce social outcomes that benefit only a handful of powerful individuals.

Creating Impactful Signals

As groups grow they tend to encompass many overlapping groups of stakeholders, and collective decisions may impact stakeholder groups differently. In order for a signal to be impactful, it needs to be clear that the signal is representative of a stakeholder group and that the stakeholder group's opinion is relevant to the decisions being made.

Being representative of a stakeholder group requires a mechanism to quantify the preferences of a specific group in such a way that it is obvious that the result closely aligns with the aggregate interests of that group. In an ideal case this would mean that every member of the group participates in the creation of the signal. However, in practice signals can be representative even with low levels of participation so long as they are applied consistently and are viewed as legitimate by the stakeholder group they represent.

Being relevant to a decision being made is important because even if a signal is representative of a group, if that group's opinion is irrelevant to the decision being made it won't have a significant impact on the outcome. In practice this could mean that a group which is too small may never be able to produce relevant signals, or that a group which is too diverse may not be able to consistently produce relevant signals.

When signals are both representative and relevant, regardless of who is making decisions, the decision makers are forced to take strong signals into account. In the context of decentralized blockchain networks which are easily forked, the social pressure on decision makers is amplified due the relative ease with which neglected stakeholders can coordinate an exit.

Empowering Stakeholders with Signaling DAOs

In the Ethereum ecosystem it has been relatively difficult for stakeholders to collectively organize into clear, quantifiable groups in order to coordinate the creation of representative and relevant signals. The lack of tools for creating and processing signals was one of the most pressing concerns that emerged as a result of the EIP0 Summit earlier this year.

With the recent Mainnet release of our Survey app, I would like to propose the idea of Signaling DAOs as a way for stakeholder groups to self organize and produce clear signals. A Signaling DAO could be created by anyone, but in order for it to be impactful it should attempt to be representative of a relevant stakeholder group. Tokens that are used to signal could be assigned based on arbitrary processes, a simple example might look like this:

1. A group of stakeholders is defined (e.g Exchanges, Core Devs, Dapp Devs, Reddit users by Karma, etc.) and non-transferable tokens are airdropped/assigned to relevant organizations and individuals.

2. The signaling organization can designate a committee to manage and adjust token balances as new members are added to the organization, or balances are updated to reflect changes in stakeholders weightings.

3. When an issue is relevant to a specific stakeholder group a survey can be created to gauge sentiment within that group in a transparent and easily quantifiable way.

This type of signaling could easily be combined with node operators if they wish to create an automatic way decide on a hard-fork based on the on-chain signals of key organizations. Dan Finlay's Strange Loop governance framework proposal for Ethereum.

If the signaling organization is administered in a legitimate way, the signals will be meaningful. If there are obvious deficiencies, then the signals will be less impactful. If a malicious actor tries to manipulate a signaling DAO, the rest of the community can simply ignore or discount that particular signal.

When an issue comes up that is controversial, rather than relying mostly on unstructured signals on platforms like Twitter and Reddit, the community can look at signaling organizations to understand what specific and relevant stakeholder groups have strong opinions and how they are impacted by the decision in order to come to an independent conclusion.