Informal feature proposal: quantified proxying of post voting power

This is an informal proposal for a new feature following a discussion on Slack with users @clayop, @steemed, @dan, @arhag, @pharesim, @nextgencrypto and @talerecursion (myself). Links to the original discussion and original proposal on slack.

PREMISES

This proposal attempts to address several concerns that were raised on slack:

• The fact that current distribution of stake is centralized among a little group of people with similar cultural, social and ethnic background which, it is felt, is going to create an important imbalance between the amount of voting power that will be exercised in "subs" for which the original group of stakeholders have higher affinity, and other "subs" less relevant to this group's interests or simply unreachable due to the language barrier. This risks affecting negatively diversity on Steem, and prevent Steem from expanding in key markets such as China or catering to all age groups, genders, cultural groups or political sensibilities.
• The arbitrariness of assuming that stake ownership and social contribution to the project have to map 1:1. Several examples have been raised that show that this assumption cannot reasonably hold in all situations. It is a common practice for a single entity wishing to express as a single voice to hold equity under multiple accounts so a 1:N relationship has to be assumed between social contributions and equity holding accounts: reasons range from operational (cold storage) to organizatinal (legal entity formed of pooled stake contributions from founders, mergers etc.) to patrimonial (one account per family member, but only one curator, accounts created in view of transmitting estate to several beneficiaries etc.).
• The fact that people with higher stake may not necessarily be the most apt communicators, the most equanimous arbitrators, the most knowledgeable experts or the most available contributors which leads to the logical conclusion that it would be beneficial both to large stakeholders and the community if stakeholders were allowed to nominate competent people to handle the moderation on their behalf in all the "subs" that they would be wishing but don't have the time, knowledge or linguistic skills to support.
• The systemic risk that large stakeholders' direct contribution to moderation with all their stake weigh could encourage herding behavior behind perceived preferences and voting habits of said stakeholders and distort the neutrality of the platform that would increasingly become the reflection of large stakeholders preferences and personality.
• The systemic risk that heavy weight voting decisions will be perceived as arbitrary and biased and affect users image of the platform as a fair and neutral venue for free speech.

PROPOSAL

The common point of all the aforementioned problems is that they can all be attributed to the existence of an artificial 1:1 mapping between stake ownership and the duty of social contribution, and can be addressed by lifting this constraint and allowing 1 estate account to map to multiple posting account (1:N) or multiple estate accounts to map to a single posting account (N:1). This can be achieved by using quantified proxying of voting power as explained below.

General principle

Quantified proxying of voting power is a mechanism similar in essence to witness voting. Under this mechanism, stake holding accounts will be allowed to nominate any number of accounts to contribute socially to the plateform on their behalf. The amount of power granted to a proxy is quantified as an amount of VESTS. Only post voting power of vests is to be transmitted. Ownership of the stake and other rights attached such as witness voting remains attached to the stake. A stakeholder may be allowed to delegate the post voting power of all the VEST that isn't pending withdrawal. Inversely, a stakeholder may only be allowed to withdraw the portion of VEST which post voting power isn't currently being proxied to any third party.

Economic incentive

To incentivize both stakeholders and social contributors to use proxying, the voting reward generated using the proxied amount of stake voting power would be split between the the stakeholder and the proxy based on a rate that is negotiable between the two, and that will be a function of the perceived benefit by both parties of the proxying relationship. By relying on qualified proxies with proven expertise of their field, the stakeholder stands to benefit most from saving time and effort spent daily on reading posts and voting as well as increasing the width, depth, relevance and diversification of his portfolio of votes and therefore the expectation of return no to mention improved quality all across the board on the Steem network and therefore the appreciation of this stake. The proxy stands to benefit most by increasing his influence on votes and leveraging the effect of additional stake on his returns. Negotiation is not covered by the protocol and will occur on a discretionary basis between the stakeholder and his nominees.

Establishment

For a proxying relationship to be established, two transactions must be posted: one by the stakeholder and one by the proxy. Both the stakeholder and the proxy will submit a proxy offer that will specify the both parties accounts, the amount of stake to be proxied, the payout rate for the stakeholder (the payout for the proxy being the complement to 1), the notice period in days and the context in which the contribution is expected to take place ("subs" path). Here is an example of what the proxy command could look like. Although the relationship is asymmetric, only one type of transaction is needed since stakeholder account and nominee account fields already encode the direction of the transfer.

proxy_setup

Revocation

The proxying relationship may be revoked at any time by either of the party at their own discretion. The revocation by only one of the parties will schedule the termination of the relationship to occur at the end of the agreed notice period. Revocation by both parties will immediately void the relationship, returning voting power to the stakeholder, and freeing the proxy from subsequent commission withholding from his voting rewards.

proxy_revoke

Reputation

Trust and reputation matters in the choice of a stakeholder or proxy are to be handled externally to the protocol. It should however be assumed that the reputational impact of defaulting on the agreement, whether it is from the proxy failing to perform agreed duties, displaying poor demeanor or integrity, or quiting on a whimps, or from the stakeholder terminating early the agreement without a good reason, will carry sufficient social stigma to effectively deter misbehavior, exactly as this is currently the case on the job market.

Effect

This is this author's conviction that the proposed mechanism will effectively allow to solve the problems introduced in premises.

• By allowing stakeholders to delegate social contributions, higher ethnic, cultural, linguistic and gender diversity can be achieved in proportion of the size of the market they represent. Although some stakeholders may carry biases that would influence their subjective evaluation of the relative importance of each social group, economic rationality is to be assumed in average as an insufficiently represented market would create an arbitrage opportunity by allowing new stake to get stronger impact per vote and therefore stronger returns.
• By allowing multiple accounts to nominate the same social contributor to handle posting, stakeholders can manage their estate more effectively by vesting stake in multiple accounts without losing the convenience of being able to intervene in a vote with their maximum weight if they so desire.
• Delegation of the duty of social contribution effectively allows to let people do what they are good at: investors can invest without having to worry that their return will be affected by the fact they don't have the time to contribute directly, founders and developers can focus on getting things done without having to feel obliged to spend time everyday reading posts and voting, other community members can also spend more time undertaking other duties like market making, future planning, marketing, building third party businesses around the platform etc. This also creates job opportunities for external talent willing to join the community, tailor made to people's particular type of expertise: someone with a good knowledge of the field will be more likely to spot the gems early and maximize returns. As a result, overall quality of post voting should increase dramatically.
• Spreading voting power among a swarm of hundreds or thousands of qualified and cherry picked moderators kills herding behavior at its root. Should some moderator become too predictable and prone to falling for vote bait, his influence can be reduced and compensated by the introduction of another moderator on the same topic.
• The more moderators, the more specific political, philosophical or cultural sensibilities can be represented an the less likely it is that some minorities would end up being left out. Again, fine control at stakeholder level together with economic rationality of occupying unexploited niche markets will lead to creation of additional positions, whether the gap is noticed by stakeholders themselves (unlikely) or by someone among said left out minorities stepping up and making a case for his community.

Risk

Rare are the proposals that have only benefits, and this proposal is no exception. Among other things, identified risks are

• Increased risk that stake would progressively migrate into the hands of profit seeking investors as opposed to purpose driven stakeholders. Proxying can be seen as a form of commoditization of the duty of social contribution, which enables Steem equity to be seen analytically as some sort of structured cash flow generating security. With proxying, stakeholders could be essentially absent from leadership and still stand to derive maximum profit. This is both a good and bad thing depending on the perspective. It could also be argued that this is unavoidable in the long term if Steem reaches the level of success of incumbent businesses such a Facebook or Twitter that are already entirely driven by profit seeking investors. Witnesses are also a guarantee that there will always be some level of leadership on board no matter how passive future stake holders may turn out to be.
• The creation of social contributor / moderator positions could lead to the avent of post voting agencies and therefore introduce some degree of centralization. The risk is however limited by the fact that, unless PoW mining, social contribution doesn't benefit much from economies of scale as a human will always require the same time to read and understand topics regardless of whether they are seating in their couch with an iPad on their lap or behind a screen at a specialized company. There is a risk that salary differentials would make post votes farming more profitable in third world economies, but in that case cultural and language barriers acts as a protection and give advantage to locality. The fact people tend to have a personal interest in participating in online discussions also indicate that for many participants the payout in emotional or intellectual terms will always outweigh the economic interest in countries where basic needs are taken as granted and personal fulfillment is considered the priority. For this reason, it is unlikely that a large differential of cost and quality would exist between amateur but personally motivated contributors from developed economies content of getting free beer money for doing their hobby and professional but emotionally detached post voting farmers in developing countries.

There are many other risks, benefits, caveats and important points that I must have overlooked, and I am probably plain wrong on some of the above.
Please review and share your views;