The world isn’t made fair, but that doesn’t mean we shouldn’t strive for a fairer world.
-Written by Darren from EOShenzhen
-Proposal by EOShenzhen, 2018/07/10

1. Building a Fairer World

The ideals of fairness and equality are two of the core tenets that society is built upon. Such values were enshrined into our beliefs, as equal rights allow for a more prosperous economy. Individuals can feel socially connected and equally participative in their community, equal opportunity to be productive, and to instill confidence that everyone is respected equally (and even the false perception of freedom). That has been the main thrust of society’s progression from single-rule to democracy, and the adoption of capitalism such that every individual feels empowered to chart his/her own destiny.

However, at times, such friends could find differences and become foes to one another. Such encouragement of fairness in meritocracy isn’t perfect, and luck gets in the way - from one’s birthright to one’s inbred personality and one’s network. Additionally, the very same effort encouraged and mindset could turn around and bite one when one doesn’t maximize their chances and slowly find themselves polarized in society to even the most basic of human rights. Does one then have the right to complain when his lack of hardwork led to a lack of food, or even the right to vote? Where then should the line be drawn between fairness and motivation? Or could it be said that resource wealth is the key indicator of who knows best?

This tussle can be constantly seen through the building of any system, even through the EOS Governance framework. Therefore, to build our referendum, the fine line between fairness and determination is examined throughout this write-up on whether either ends, or a balance, should be the ultimate objective.

2. Scales of Fairness

Justice can be thought of as two ends of the scale, and varying degrees of fairness involved within each model will be explored. The system can be as fair as a petition system, or as realistic as the divide between the MNCs and the populace.
Variables affecting fairness include monetary/voting tokens, staking and unstaking, and time. These play an important role in affecting the degree of fairness of the referendum system.
Though the models proposed seem like two faces of the same coin, involving similar elements, the reality is largely different between models.

2.1 Monetary Tokens

In nearly all democratic referendums in our society, every individual, regardless of status, is entitled to one vote. This is the fairest system we can imagine, and it’s the same ideals that decentralization was founded upon - beyond safekeeping of information, to ensure everyone now has a voice in a community well-equipped with the technology to account for every vote in any decision, without risk of fraud or wastage of copious amounts of time, and reversing the need to elect representatives who we believe we can identify with and can represent our ideals.

Yet the unfairness constantly voiced out on the American voting system for instance, lies behind real world influence that goes beyond the simplistic voting mechanism - the ability for richer individuals to influence the thoughts of the average individual, through occupying the limited time we have with propaganda for their candidates. Similar to how people are known to be the weakest link in security, we can be our own enemies against securing what we truly desire.

On the EOS referendum system, however, we can choose to make the fairness more or less blatant.
On the fairness front, every voter can be given only a binary option to each referendum that is proposed. In this instance, time does not play any difference as multiple votes can occur without compromising upon the integrity due to the lack of a shared resource between referendums.
An alternative should we choose that there is a need to pay for one’s vote is to set a fix token fee to vote. Such “payment” compromises fairness slightly, but gives those who earned such right to pay for an equal voice, similar to material purchases.
Should we choose a less fair system, every voter can be issued a similar number of tokens, similar to an auction system. For example - every voter is given 100 Tokens, and allowed to stake such tokens across a wide range of referendums. To ensure fairness, time has to play a factor from henceforth (explained below). Similarly, such tokens can be deducted from one’s earnings to provide the incentive to earn the right to vote.
At the other end is a free capitalist system, where the media influence is now blatantly translated into the right to vote instead. Anyone is free to stake whatever resources they own into voting. This encourages individuals to strive for this right at the expense of building a fair system, which could compromise the integrity of certain referendums. This has been highlighted that resourceful companies could easily stake votes in one referendum and then hop unto another.

2.2 Time

Time plays a role only when the monetary resource is shared between referendums. This potentially compromises decision-making when one has now to consider how his decision in one referendum could affect that of another. This is similar to example buying goods in the real world, where an individual has to consider how to best use his cash. This person might need to purchase a microwave oven and a bed for his home, but has only resources for one. He now is forced to choose which is more necessary to him, rather than based on the mere merits of the item itself. Forcing such a comparison could mean that an item that is equally beneficial can be neglected due to the lack of resources.

The example demonstrated above only works in symmetrical information. However, when time is introduced, there is now asymmetrical information because of the shared resource. Working upon the above example, imagine now that the consumer found a cheaper microwave oven that now allows him to buy both the bed and the oven, but because he had asymmetrical information, he believed the first oven was well-worth its’ price and bought it instead. Therefore, he was not given a fair chance to decide the optimal utility amongst the items.

This works similar to that of an auction, but the auction system only works well because it is assumed that all individuals at an auction house are wealthy individuals, and therefore money is not a ceiling, and thus the price discrimination between individuals is matched at what they exactly value that very item. Such is not the case with referendums.

On one end of fairness, referendums should be submitted by a due date within the previous month, say 20th June. Referendums are then collated and issued on 1st July, giving a month’s window to vote upon. The next month’s referendums then have to be collated by 20th July and released on 1st August. To allow for fairness, an equal number of referendums can be released simultaneously, and/or categorized into types of referendums for equal voting opportunities eg. 1 from each category or all from the same category, depending on whether the comparison is between or within categories.
On the other end, time can be a non-issue, allowing for asymmetrical information, and potentially hoping that individuals will value the referendum for what it is worth rather than comparing against what could be future opportunities.

2.3 Staking and Unstaking

This depends on the referendum system chosen, ranging from a non-necessity in non-shared resource systems (binary systems). As highlighted, a rich voter could easily hop around and stake to meet the acceptance threshold, then unstake and jump to another referendum, easily giving false positives for referendums that might not meet the widespread scrutiny and approval of the community.

This could be solved in a few ways. Firstly, the time factor could be used such that whatever is decided upon by the end of the month will be counted as the voting. Otherwise, a down period could be introduced. Currently circulating amongst the community is a down-time of 3 days. However, it can be argued that this is a non-solution if various important referendums to this block producer, or if all the referendums proposed somehow had a lead-time of 3 days or more between them.

To clarify, allowing tokens to be reused is a non-issue in voting as for example, if a poor producer produces 1 token per month while the rich individual produces 100 tokens, unstaking the 100 tokens utilized the previous month for the rich individual and 1 token for the poor producer might lead to a numerical difference, but not in percentage. So long as the threshold is a percentage issue, this is a non-concern.

One way to solve this issue other than preventing staked votes from being reused, is to collate identity, and set an internal threshold that is significantly lower than the threshold of acceptance, for example, 50% of the 15% required can only be held by any single identity.
Otherwise, another method is to ensure that on top of the threshold, a minimum of 3 individuals must make up the threshold.
Finally, these two can also be combined such that a minimum of 3 individuals, each with minimally 5%, to allow for a referendum to pass. Such measures would then allow a referendum to better represent a community.

3. Identity

Identity can be argued as an issue in fairness only if it concerns the final scenario posted above, if the argument is the prevention of double-voting or vote-laundering.

If a voter votes from multiple identities, it isn’t an issue as the voting is still illustrative of his desires. Altogether, it might be a better representation as it’s the utility of his full resources being invested upon the voting process. This is only a concern of transparency as it might create a false impression upon the community about the support, but it is staunchly a non-issue in relation to the fairness of voting if the final scenario above of creating fairness is not invested upon.

However, identity merely to create transparency for the community and affect their impressions is still an important investment that should be looked upon. If there are other threshold of acceptance, then identity is arguably a crucial factor to consider.

4. Threshold of Acceptance

Crucially, as slightly discussed above, any consecutive days for a criteria to be met could potentially reduce fraud of a rich investor jumping between referendums, but adds significant delay to the entire business process, and does not completely eliminate the issue.

Below is currently the criteria discussed. Contentious especially are the vote totals, for which there seems to be no reason to choose such arbitrary numbers.

When a new proposal is entered into a smart contract, a 120-day countdown begins and token holders can begin to vote on it: yes or no
Throughout the day, the smart contracts save account names and their vote selections to a table
At the end of each day, the smart contract takes a snap shot of the staked token balances of the saved account names
From that snapshot, count the amount of staked tokens held by each user for each vote selection
Vote totals are recorded if the following criteria are met:

total_votes > 15% of total EOS 
total_votes_yes > (total_votes_no + 10%)

If the criteria continues to be met for 30 consecutive days, the vote passes
If at any point after day 90 the vote drops or remains below either +10% or >15%, it dies, as its no longer possible to hold the two thresholds for 30 days. (Assuming it didn't already pass before day 90)
If 120-days pass before achieving the threshold for 30 consecutive days, the proposal dies

The long-standing recommendation in the constitution should be discussed as to the rationale of these numbers chosen as they neither hold simple majority or any reasonable interpretation.

5. Proposal

Conclusively, EOShenzhen believes that one should tip more towards a fairer system. Ultimately, basic human rights should be untouched as necessary, and voting should be one of them. It might be agreed upon that although the vast majority might not know what is in their best interests, it is also the vast majority who should still chart their own future.

Consequently, EOShenzhen proposes that every individual should be merely entitled to a “yes or no” vote on each and every referendum similar to that of a petition, and identities be collated, but not released, to create transparency and not influence an individual’s decision on what each individual voted for. This historic system of our established democracy ultimately has had centuries of experiment to come out triumphant, and indeed there must be a reason so.

Though there should be an incentive to collecting EOSTokens, however, voting should not be the incentive nor avenue for their battleground.