Colony - Governance
Case study
4.1 What Colony Does
Mannan, Morshed:
"Colony is a platform that provides the infrastructure for creating an ecosystem of self-organising companies (i.e. ‘colonies’),134 by lowering the costs of a diverse group of people coordinating their efforts and resources to realise shared goals, even when they do not necessarily know or trust each other. The ambition of Colony is that this coordination will occur in the organisation created through its platform in a meritocratic manner through the dynamic allocation of reputation. Reputation is a number that is associated with a person, reflecting the value of their recent contributions to a colony. It may be earned by bootstrapping colonies, successfully completing tasks and constructively resolving disputes.135 This figure affects the extent of a person’s control rights in the organisation as well as their share of rewards. Significantly, unlike currencies or securities, reputation cannot be transferred and is non-negotiable in crypto-capital markets.136 Colony is still at an early stage of development and much of what is described below is based on its white paper, setting out the features the development team expects the layers of Colony to have. The development team have been building the Colony Network and Colony JS, a software library that enables independent developers to develop applications (dApps) that can interact with the underlying smart contracts. These colonies may be established to create software but also for tangible goods, such as jewellery. As one of the founders of Colony, Jack du Rose, began developing the platform as a way of solving problems he encountered while coordinating persons in a global, high-end jewellery supply chain,137 the illustrative examples in the following subsection draw from the jewellery industry.
4.2
The Governance of Colony To understand the governance of the Colony platform, it is necessary to consider the Colony Network, the Meta Colony and individual colony layers separately. The Colony protocol138 is built on the Colony Network, a collection of smart contracts deployed on the Ethereum blockchain by the Colony development team. These contracts provide the broad parameters in which colonies may be created, such as the fees charged to use the Network, upgrades of its functionality and the reputation mining mechanism.139 Management of the Colony Network will be gradually ceded to a Meta Colony, the first, parent colony to be created on the Network.140 When this has occurred, tokens in the Meta Colony (CLNY) will have been distributed and reputation can be earned in the Meta Colony through the completion of tasks, such as making updates to individual colony smart contracts. CLNY and reputation holders get to vote on the fundamental parameters of the Network (control rights) and receive a portion of the fee charged by the Network when individuals are paid.141 Moreover, CLNY holders act as reputation miners, calculating reputation scores off-chain and updating reputation scores on-chain, for which new CLNY tokens and reputation are conferred as rewards. The functionality of CLNY tokens will be set initially by the Colony development team and the Ethereum community but eventually by the Meta Colony. Individual colonies may be created to achieve a single goal or multiple goals, over a short or long time frame. They are entities with discrete purposes but act within the broad parameters set by the Colony Network. Regardless of the goal, they will substantially share the membership and governance rules described below due to the underlying smart contract code. As these rules are embodied in code, when they are being used they are much harder to skirt than institutional and social rules in a worker cooperative, where they may be underenforced.143 When a colony is created, it will generate its own native token that will primarily have financial value.144 To achieve its goal(s), the work needed can be broken down into tasks and (sub) domains (e.g. assembly) in which tasks can be clustered. This is analogous to departments in an organisation. Domains can also be nested within wider domains, with the widest domain being the colony itself. Along with allocating a task to a domain, tasks will be tagged with relevant skills needed for its completion (e.g. #casting, #soldering). This may be a specific skill within a broader skill set (e.g. #design). Thus, there is an organisational tree and a skills tree, with participants able to earn and lose reputation in both. To create and define a task, a person with sufficient reputation must deposit (‘stake’) colony tokens proportionate to the amount of reputation in the domain.145 Reputation and colony tokens may be initially assigned as control rights and working capital at the time a colony is created to allow certain persons to set up tasks.146 Otherwise, usually, a task initiator will submit a funding proposal from the pot (wallet) of a parent domain.147 The proposal will specify the amount of funds needed and can be denominated in the colony’s own currency or in Ether. If there is only one funding proposal for a task, there are sufficient funds in the pot and there are no objections, the smart contract will begin to release funds to the pot of the task. This materialises Colony’s emphasis on completing work efficiently rather than voting on every decision. Once the funds needed for payment are in place (the bounty), the manager will have to enter into a tentative agreement with a worker who has the necessary skill set and reputation. When joining the Colony platform, workers would have tagged their skill sets and managers can use this to search for one who is most appropriate for a task. After an agreement is reached, a task may be specified to them along with working guidelines, a due date and payment terms (for the worker, evaluator and manager).148 While the manager may also act in the capacity of evaluator, this role can be delegated to a separate person as well.
The evaluator may be unknown to the worker, as they may only be identifiable by their public key. Following the completion and evaluation of the task, there will be three days to raise objections and disputes regarding the quality of the task performed. When there are no objections, the worker gets paid in the colony’s native token or another approved cryptocurrency.149 If paid in native tokens, the workers’ reputation in their domain increases, as well as all the wider domains of which it is part, including the colony itself (i.e. the toplevel domain). Simultaneously, their reputation for performing the tagged skill increases, as well as any wider, parent skills of which the skill is a part.150 The sum of their top-level domain and top-level skills reputations determines their influence on decisions that affect the individual colony. To avoid disproportionate gains in reputation following the completion of a task, the bounty initially set should be consistent.151 If there is an objection, an objector must be able to defend his/her objection.
Its content should not only specify why a task is inadequate and what could be done better, but also suggestions as to the ‘reputations’ (i.e. Colony members with a certain level of reputation) that should vote if a dispute arises and reasoning for why these reputations should vote. This allows objections to be scaled to a larger group of peers, whether at a domain, colony or Meta Colony level. This objection can only be made if an objector has a certain reputation score and stakes some of their own tokens.152 If no one makes a counter-stake to object to the objection, then the objection will pass and the worker will receive less/no pay. If someone does sufficiently counter-stake within three days, then a dispute will arise. The staking of tokens is needed not only to avoid frivolous objections but also to compensate the persons involved in settling a dispute through voting. The weight of their votes is contingent on a person’s reputation in the skill and domain in dispute.153 Being on the winning or losing side of a dispute has the corresponding effect of enhancing or diminishing reputation scores. The payment and reputational scores allotted to the worker or evaluator depends on the final score received after disputes are resolved. If the work is found to be inadequate, the worker will receive diminished payment and lose reputation in their domain and their tagged skill, as well as parent and child domains and parent and child skills. In addition to payment for completed tasks to workers, managers and evaluators, persons in the colony holding native colony tokens and reputation are entitled to rewards from the revenue earned by the colony.154 This means that a worker in a colony, waiting for the next task to be assigned to them, can continue to earn (for a while) from the revenue they had helped generate.
4.3 Worker Cooperatives: Learning from the Colony Project
A close reading of the governance structure of colony reveals a startling resemblance to LMFs, such as worker cooperatives. Firstly, the economic activities are carried out primarily for the benefits of its participants. Secondly, most, if not all, of the capital of the organisation is held by the participants. This is indicated by the fact that tokens and reputation are issued exclusively to the participants of a new colony,155 before gaining potential investors, and as such can only be gained through various forms of work: production, evaluation and management. This is akin to the common practice in the startup technology sector of granting employees stock options,156 but in this instance it is coupled with the right to have a voice in significant strategic decisions. 200 Thirdly, as currently designed, colonies have voluntary, open membership by default. Restricted membership is not mentioned in the Colony White Paper. This is characteristic of initiatives in open source communities, where objective peer review is critical and where, instead, there are concerns about keeping participants motivated and committed.157 However, the key difference with open source communities is that colonies may not be limited to the private provision of public goods,158 for which values such as the long-term striving for excellence may come into play.159 Colonies may be used for the production of private goods as well. Fourthly, Colony has what can be broadly described as dynamic meritocratic governance, where the weight of one’s vote is dynamically adjusted according to one’s contributions to a task, domain or colony. In itself, this is not contrary to cooperative principles as there are cooperatives which weigh voting power according to, for example, production. 160 Participants still have a voice in the governance and strategic decision-making of the colony, as exemplified by the fact that anyone can set up a task for the colony to complete. Fifthly, it is clear from the White Paper that the assets of a colony are conceptually distinct from that of the participants, as they are escrowed in a smart contract and associated pots. Access to these pots is conditional on a successful funding proposal. There is also a separate revenue pot from which rewards may be distributed or working capital replenished.161 Notionally, colony smart contracts can subsist indefinitely with tokens in escrow, even after it has been abandoned, indicating that it is technologically possible for the colony to have its own capital. Moreover, the payment of Network fees, which is reinvested to maintain the Network and to do useful supportive work (e.g. build applications) is also reminiscent of the cooperative practice of building financial reserves and investing in useful services (e.g. training) to sustain the mission of the business. While taking these similarities into account, there are certain functionalities in Colony, which can potentially overcome the start-up and coordination costs that worker cooperatives often face, especially when operating across borders. Decentralised organisations prefigure ready-made governance structures that are easily accessible online and are native to globally distributed blockchains. While the governance mechanism is technically complex, as with other digital applications, once launched its use will be intuitive and user-friendly. As such, these organisations can provide capital and governance structures for digitally native worker cooperatives to adopt. In terms of financing, worker cooperatives can consider implementing a system in which financial rewards and decision-making power are generated through useful patronage, represented as separate quantified units, but with only the financial rewards being exchangeable – as they are with native tokens and reputation on the Colony platform.162 If the token gains use-value, then it can be sold or swapped for other, more widely used cryptocurrencies, which can tide over those who only have intermittent work. The relative transferability of a token compared to a partnership interest, a standard cooperative membership, or an employee share held in a trust, allows workers to diversify their risks, in the event their cooperative fails. At the same time, this allows for a certain amount of external investment to flow into the business. As (most) decision-making rights are not attached to native tokens independent of reputation, it may be acquired and held by third parties without diluting the decision-making rights of worker-members, as is the predominant concern with non-member investment.163 In terms of collective action problems, a frequent criticism of worker cooperatives is time spent on meetings to reconcile heterogeneous interests,164 and as such taking actions on the basis of tacit consent, rather than majority voting or unanimity, may in fact be preferable. Similarly, the requiring of staking of reputation and tokens in raising an objection can help avoid trivial disagreements about the quality of work. Turning to the aforementioned cross-border coordination issues, the fact that workers are drawn from different backgrounds prevents them from having a shared background in terms of politics, work and culture, which are usually associated with worker cooperatives.165 Instead, reputation-weighted governance may be especially suited for organisations seeking to coordinate a heterogeneous, pseudonymous group of actors166 who operate across a wide geographical territory with limited trust and state policing. While blockchain communities have only emerged in recent years,167 history is replete with examples of such organisations. Examples range from the Amsterdam Stock Exchange in the seventeenth century168 to modern Moroccan bazaars.169 Contemporaneous examples include Usenet newsgroups, massive multiplayer online gaming and open source software developer communities. A common theme appears to be finding counterparties with desirable qualities (e.g. a certain set of skills and experience), while at the same time coordinating these individuals to ensure contractual performance and the pursuance of the collective interest. This does not necessarily require external enforcement, through judges or regulators, but can be achieved through the threat of diminished reputation. The risk of losing reputation is sufficient motivation for performance by a party, especially when it is in their interest to have continuous transactions with a counterparty,170 on a regular171 or irregular basis.172 As such, the fear of lost reputation will ‘crowd in’ honesty in the long run.173 This is true of online communities and project-based work, particularly in creative industries.174 This, however, assumes that parties have sufficient information and knowledge of each other’s reputations. Online reputation systems are able to address these information asymmetries to an extent, as user reviews and ratings provide granular information about a potential counterparty in a digestible form. Yet, peer-to-peer systems are vulnerable to manipulation by platforms that host them and biased reviewers, raising concerns about the system’s own trustworthiness.175 decision-making rights of worker-members, as is the predominant concern with non-member investment.163 In terms of collective action problems, a frequent criticism of worker cooperatives is time spent on meetings to reconcile heterogeneous interests,164 and as such taking actions on the basis of tacit consent, rather than majority voting or unanimity, may in fact be preferable. Similarly, the requiring of staking of reputation and tokens in raising an objection can help avoid trivial disagreements about the quality of work. Turning to the aforementioned cross-border coordination issues, the fact that workers are drawn from different backgrounds prevents them from having a shared background in terms of politics, work and culture, which are usually associated with worker cooperatives.165 Instead, reputation-weighted governance may be especially suited for organisations seeking to coordinate a heterogeneous, pseudonymous group of actors166 who operate across a wide geographical territory with limited trust and state policing. While blockchain communities have only emerged in recent years, history is replete with examples of such organisations. Examples range from the Amsterdam Stock Exchange in the seventeenth century168 to modern Moroccan bazaars.169 Contemporaneous examples include Usenet newsgroups, massive multiplayer online gaming and open source software developer communities. A common theme appears to be finding counterparties with desirable qualities (e.g. a certain set of skills and experience), while at the same time coordinating these individuals to ensure contractual performance and the pursuance of the collective interest. This does not necessarily require external enforcement, through judges or regulators, but can be achieved through the threat of diminished reputation. The risk of losing reputation is sufficient motivation for performance by a party, especially when it is in their interest to have continuous transactions with a counterparty,170 on a regular171 or irregular basis.172 As such, the fear of lost. It is also less clear-cut that a token, as described herein, will constitute a security as compared to tradable shares in a worker cooperative, which generally will.
However, the manner of its deployment in the Colony protocol makes the system less prone to cronyism. Managers of tasks are incentivised to intuitively and objectively choose workers based on a quantification of their demonstrated skills and recent contributions, rather than personal characteristics, as they stake their own tokens when initiating a task. This score is not generated through ratings by (potentially) anonymous individuals with little to lose. Instead, evaluators stand to receive diminished payment and a reduced reputation score for inadequate evaluations, while contesting a task or decision through the dispute resolution mechanism requires risking tokens and reputation. A teething concern about the democratisation of reputation systems is that it will ultimately not be sustained, with its growing complexity leading to the emergence of oligarchy. One empirical study has already observed this trend with regard to peer-production projects, leading to structural changes in authority and a reorientation of organisational goals.176 A key distinguishing feature of Colony’s reputation system, however, is its degradability, which prevents early movers from resting on their laurels and incentivises the continuous, useful engagement of all members in the governance of colonies. To embed such a system in a worker cooperative, a link to a user-friendly portal that provides up-to-date individual reputation scores and accrued financial rewards may be provided in the section of the by-laws concerning membership."