- Choices in token distribution strategy are early signals of how different kinds of contribution will be valued and how contributors are likely to be treated. Given its impacts on sustaining key contributor support, fairness is a quality worth consideration; rightly or wrongly, it will be a quality by which a project is judged.
- However, “fairness” is a contested concept with different meanings that all can be, and are, invoked. Three primary grounds have emerged in broader literature of fairness; all are relevant to a distribution strategy and should be understood by token issuers.
- Choices in token design, particularly those concerning governance and contribution functions, can pose special challenges in trying to achieve fair distributions.
- The design considerations involved in distributing tokens are numerous and complex, and projects would do well to think through their objectives, limitations, and options before finalizing a strategy.
Why Be Concerned With Fairness?
Much figurative ink has been spilt debating the respective fairness of a project’s launch or mechanisms for token distribution. Given the inherent difficulties in supporting strong claims about fairness, a project considering its own launch may be forgiven for asking: why should I care about something as indeterminate as fairness in token distribution? Several reasons illustrate why fairness, if not the value to shape a project’s approach to distribution (financial, security, and governance objectives being others), nevertheless remains a quality worthy of consideration in choice of launch strategy:
- Attracting entrepreneurs to build on an open source protocol is typically a major objective for open-source projects and protocols. Such entrepreneurs typically care about being treated fairly: developers care whether contributions will be adequately compensated; potential businesses care whether they will compete with incumbents and rivals on a level field; investors care whether risks taken in sustaining a project’s early vision will be commensurately rewarded upon its success; communities care whether their values or area expertise will be respected in decision making. A project’s approach to token distribution is an early signal of what to expect from the relationship—perceived unfairness in distribution influences critical actors’ future expectations.
- Many arguments for decentralized projects frame the project as ‘solving’ industry-specific ‘problems’ arising from centralization. Such problems often fall under the purview of ‘fairness’; DEXs, for example, are sometimes promoted as ‘fairer’ alternatives to centralized exchanges on grounds of their transparency. Token distribution practices perceived as unfair provide fodder for tu quoque style criticisms of these projects, which can prompt cynicism towards a project’s motivations. Perception of the authenticity of a project’s motivation can influence whether developers and other early value-adding backers chose to support an endeavor, even if the average end-user’s choice of platform is unlikely to be influenced.
- Some projects attempt to realize network security or functional distributed governance through initial distribution targets, aiming to either distribute tokens so to optimize decentralization or to target individuals with certain qualifications. For example, proof of stake protocols should endeavor to distribute tokens so that a sufficient number of different actors can stake tokens and run nodes. Charges of ‘unfairness’ in distribution can influence perception of such network’s security or governance’s legitimacy. Likewise, charges of unfairness can undermine project’s efforts to create healthy secondary markets, as concentrated holdings or insider trading may reduce the ability for newcomers to participate in fair markets.
- Investors, developers, and other contributors contribute to a platform’s success in different ways. Distribution strategy, along with token mechanism design, can signal that a project values investment contributions disproportionately to labor, and is concerned with creating value for early investors over other potential contributors. The perception that developer engagement primarily functions to enrich early investors can make it difficult for a project to attract needed widespread developer support.
Thus, while acknowledging the conceptual and empirical hurdles to definitely answering questions about a token distribution’s fairness, considering whether available strategies for token distribution will be received as more or less fair warrants a project’s time.
This report seeks to arm readers with conceptual tools to think more discerningly about the many respects with which initial token distributions give rise to issues of fairness, allowing readers to better support their opinions on launch’s merits and anticipate the concerns inherent to particular distribution mechanisms. It concentrates on the issues surrounding the attribution of fairness to initial token distribution events—cases where, prior to network launch, all or some of a project’s total token supply are allocated to particular groups or individuals, with or without an understood basis. The report clarifies considerations that bear on characterizing as “fair” or “unfair” a project’s uses of currently available distribution mechanisms, largely by drawing from and applying as appropriate frameworks developed more extensively in the philosophical literature on distributive and procedural justice—while there may always be disputes on what is or is not fair, a common conceptual framework facilitates discussions that go beyond surface-level disagreement. The report aims not to be an exhaustive treatment of the subject, but rather raise awareness of some industry particular factors influencing a launch’s fairness.
What Does it Mean for a Token Distribution to be Fair?
When a project’s initial token distribution event is characterized as “fair” or “unfair”, what does that mean? Broadly speaking, there are three types of grounds generally appealed to support attributions of fairness or unfairness:
1. DISTRIBUTIVE GROUNDS
Was the outcome of an allocation of goods and burdens fair? To claim that the initial token distribution event was fair is to claim that its outcome—that is, who receives what amount of tokens—is fair. Distributive principles, accounts of what makes an outcome (e.g. ‘distribution’ in the literature’s parlance) fair, are numerous and vary in answer to key questions, such as what distributed (wealth, opportunity, etc) is relevant to fairness, how recipients should be considered (individuals, communities, etc), and whether equality, need, desert, free transaction, utility maximization, etc should determine how goods are distributed.
According to one of the earliest accounts of distributive justice, a distribution of goods is just insofar as individuals receive goods in proportion equal to their claim to those goods; an individual does not receive more or less goods than she has claim to. From this perspective, fairness requires that buyers who pay the same price at the same time receive the same quantity of tokens, presuming no difference in claim. To put another way, differences in tokens received must be justified by appeal to differences in claim—for example, perhaps on the grounds that some purchasers made early project contributions or took greater risks (financial, regulatory, technical, etc) that warrant a discounted price.
Historically, fairness in distribution has been understood through several compelling perspectives, such that proposed distributive principles differ on key question :
- The distribution of what is relevant to the distribution’s fairness?(Wealth, opportunities, income, jobs, burdens, etc.)
- The distribution should be made on what basis?(Proportionality, equality, utility maximization, need, according to free transactions, etc.)
- How should the recipients be considered? (Individual persons, communities, etc.)
Of these, the nature of the goods and burdens distributed in initial token distribution events is particularly worth further consideration, since access to investment opportunity can be understood as a good allocated through a distribution’s format, and because many tokens are designed to bestow holders different forms of value, like voting or contribution rights, giving rise to particular questions about their distribution. The following section examines these nuances in more detail.
How the opportunity to earn bitcoin was distributed through Bitcoin’s launch contributes to the event’s reputation for fairness. Anyone with or willing to learn the technical know-how to operate a mining-capable device--a standard computer CPU at the time--could compete for mining rewards, with hardware barriers to competitive mining comparatively less than today. In principle, anyone who was aware of Bitcoin could participate at genesis, an outcome shaped by choices in procedure.
Whether subsequent concentrations in Bitcoin holdings are fair based on distributive grounds is more debatable, in part because there are methodological obstacles to measuring distribution (wallets are not held by users on a one-to-one ratio) and in part because open questions around measures of inequality. For instance, the Gini coefficient, used as a measure of economic inequality in nations, is sometimes applied to token distributions such as Bitcoins. Yet, whether a measure of country’s wealth holdings is a suitable analog and, presuming so, what a ‘good’ Gini coefficient is, remain contentious subjects.
Even given a general interest in avoiding inequitable holdings, the extent that projects alone can influence subsequent concentrations of wealth may be limited. Some of the differences in concentration are the results of differences in contribution: miners who chose to provide network-securing mining services in exchange for tokens (potential wealth) take on risks and add value that arguably justifies the difference in holdings compared to those who do not. Yet, vast amounts of tokens are also held by addresses belonging to Satoshi, Bitcoins’s anonymous creator, as well as a few other unclaimed or anonymous addresses, and changes in mining technology have affected who has the opportunity to accrue tokens through mining. One could argue that a general purpose currency substantially owned by a small number of early adopters is unfair based on distributive grounds and general arguments against economic inequality, even if mining was conducted on a generally level playing field. Other differences in holdings (likely the bulk of present bitcoin concentrations) reflect more market dynamics—who was historically interested in buying and selling—than distribution strategies, initial or ongoing. One could argue that this in fact is a testament to the distribution strategy itself, in that has allowed market dynamics to shape distribution beyond up-front design choices.
2. PROCEDURAL GROUNDS
Was the process of allocating goods and burdens conducted fairly? To claim that the initial token distribution event was fair is to claim that its process is fair. Fairness and unfairness in process is typically characterized by attributes like being (non)transparent (un)biased, (dis)honest, principled, arbitrary, etc. Some such attributes of a process are relevant to fairness in the sense that they are disposed to produce fair or unfair distributive outcomes—asymmetry in information tends to produce discrepancies in the opportunity to participate fully in token distribution that result in differences in who gets allocated how much. Other attributes are more relevant to treating parties fairly.
Questions that matter in assessing procedural fairness include:
- Are there redundancies in place, or venues allowing affected parties to appeal erroneous decisions?
- Was information communicated equally to all parties, and did parties have the same resources to resolve disputes?
Where a procedure for allocating tokens involves a sale between individuals or groups, concerns about those particular exchange’s commutative justice also become relevant for thinking of the overall fairness of process.
Beyond its distributive outcomes, Bitcoin’s distribution event gets characterized as fair due to some choices for procedure—choices that may have contributed in producing the aforementioned distributive outcomes of opportunity and tokens. For instance, the Bitcoin whitepaper was initially distributed to a popular, public cryptography mailing list, and Bitcoin had no privileged information nor token allocation to insiders.
Livepeer promoted its distribution process’s fairness on grounds of its (purported) equitable opportunity to participate and principled preference of distributing to those likely to support the platform over the long term vs mere speculators. Those claims to fairness have been disputed by commentators, such as Viktor Bunin. The fairness of Dash’s PoW mining distribution was called into question when a bug in its mining algorithm led to 10% of the supply being mineable early launch, when fewer had information about the opportunity, leading to accusations of ‘instamining.’
3. DERIVATIVE GROUNDS
Was the token distribution event productive of fairness or unfairness in other project or industry areas? A token event may be fair not so much in the sense that its distribution or process was fair, but that it was productive of fairness in other respects—the token distribution event enabled the project to be governed more fairly, or contributed to circumstances where developers were unfairly compensated.
Grin, Monero, and Bitcoin rely on donations to fund development work. Whether such structure gives rise to working conditions of questionable fairness is a subject of debate—i.e. are such projects overly beholden to donors, un-, under-, or over-pay volunteer work, provide adequate oversight and accountability of funded development work, etc. The use of an initial token distribution event to fund a centrally or decentrally managed treasury could be characterized as ‘fair’ to the extent that such infrastructure enables projects to more adequately manage such issues of fairness. Similarly, token sales that concentrate holdings to a few investors raise concerns about such holders’ ability to manipulate prices on secondary markets; initial token distribution events that contribute to such practices may be characterized as unfair.
As such plurality of grounds suggests, the simple characterization of a token distribution event as ‘fair’ or ‘unfair’ can obscure the complexity of the judgments underlying it. Likewise, many initial token distribution events are composites, not homogenous in distribution mechanism, and how, say, the airdrop portion’s lack of transparency combined with the sale’s comparatively greater access to investment opportunity contributes to an overall assessment of the token distribution event as a whole can depend on questions of scope, severity, and relevant comparison cases. Fairness also comes in spectrums and it is often more the question of whether the comparative fairness of a token distribution mechanism might be improved.
• Q2: MaidSafe sale inflates price of Mastercoin, results in illiquid currency. holdings, partly due to complex and controversial BitAngel investor agreement.
• Q3: Stellar launches, uses snapshots to distribute to Bitcoin and Ripple holders.
• Q2: The DAO fund created and subsequently hacked.
• Q2: EOS token sale begins, allows for secondary market before sale’s end. The sale lasts a year, the window offering investors notable opportunity to participate.
• Q4: Zcash launches with ongoing escrowed founders rewards.
• Q2: Numerai targets data scientists in airdrop.
• Q3: Filecoin utilizes SAFT agreement in sale.
• Q3: Tezos completes $232 million token sale that frames contributions as ‘non-refundable donations’, introduces KYC procedure after the sale.
• Q4: Telegram’s TON SAFT private sale begins, offering select investors discounted, immediately re-sellable GRAM. TON cancels public ICO, and a speculative secondary market emerges.
• Q3: Handshake proposes 70% airdropped distribution.
• Q3: FOAM token sale introduces ‘Proof-of-Use’ mechanism.
• Q1: Grin claims to replicate Bitcoin’s fair launch approach.
• Q1: NuCypher proposes 'WorkLock' mechanism.
• Q3: Algorand offers fungible refunds on auctioned tokens.
• Q3: Edgeware launches via lockdrop.
Token Distribution: It is About More than Just Tokens
To properly appreciate debates on the merits of particular token distribution mechanisms, one must come to appreciate how the goods and benefits distributed in an initial token distribution event extends beyond tokens.
Access to the initial distribution event is conceivably an opportunity distributed through a project’s choice of initial distribution mechanisms. On the view that justice demands equal distribution of opportunity, the fairness of certain private sales could be questioned on grounds of seemingly inequitable distribution— and be defended perhaps on grounds of the greater risks in investing in earlier stage projects and the need to secure key partnerships. Regulatory and KYC procedures likewise limit who has access to investment opportunities, with the discrepancies these give rise to defended on grounds of their protections for investors or regulatory compliance benefits. Hardware and requisite tech know-how likewise affect access to opportunity; mining today arguably distributes to fewer the opportunity for token earnings, as the increased requisite investments needed for mining competitively represents a bigger hurdle than five years ago.
Similarly, access to discounts or other favorable sales terms, is another conceivable good allocated through an initial distribution strategy. Projects that distribute discounts inequitably and who do not clearly and persuasively communicate the justification for targeted discounts risk being perceived as unfair. Even when access to discounts is in theory equitable—such as when a token sale offers discounts for earlier sale periods—poor communication, overly short length, and unthoughtful supply management will de facto influence who has access to such opportunities. The presence or absence of vesting or lockup mechanisms can bolster or undermine projects’ justification for discounts; sales with no or minimal lockup period for tokens purchased in discount periods have comparatively little holding early investors to supporting the project.
Token Functions Matters Greatly
Differences in designed token function can entail that otherwise similarly structured token distribution events may distribute goods of a different nature. This is clearest in the cases of tokens conferring governance or contribution rights.
For example, projects that utilize some form of token-based representation in project decision making essentially distribute both suffrage and wealth in allocating tokens. This can matter in cases where the justifications for conferring representation and the justifications for conferring wealth come apart; while the fairness of an exchange as a transfer of wealth might be sufficiently justified on the basis of it being a freely chosen, fully informed, accessible to all, investment of fiat for tokens, the same basis may insufficiently justify the exchange as a conference of suffrage.
Projects like DASH, Tezos, and Decred have project infrastructures such that token holders can participate in and exercise oversight over certain project decisions. In this respect, their tokens are a dual-natured good—it is as if one’s dollar holdings also conferred one voting power of decisions on monetary policy, to use a simplifying analogy. Questions of on what basis to distribute tokens of such nature can be more complicated than questions of how to distribute tokens that singularly function as currency. Peter Zeitz of 0x, a decentralized exchange that, at least initially, considered adding governance functions to its tokens, comments on the balance between a project’s need to raise capital and choice of whom to remain accountable to. As Zeitz’s discussion suggests, selling tokens may be a good mechanic to meet a project’s need for capital, yet may be imperfect in the long run as a means for allocating representation or oversight powers to those who would make well-informed decisions in the best interest of the project.
While restricting the consensual, informed, and free exchange of tokens in their function as wealth is arguably unfair, the ability to freely exchange of tokens with governance function leaves projects open to the charge of being unfairly ‘plutocratic’, affecting project’s ability to treat those building on the platform equitably (so fairly)—plutocracy is thought to undermine a project’s ability to provide ‘a level playing field.’ For instance, Melonport argues ‘If a single actor can take over control of a supposedly decentralized protocol by simply buying up tokens, then you don’t have a decentralized governance model, you have a protocol up for sale.’
To the extent that the dual-nature of currencies with governance functions raises particular challenges with achieving fair distributive outcomes, these must be weighed against the relative gains such infrastructure provides for processes of distributing tokens fairly. Those managing an initial token distribution event make decisions bearing on the event’s fairness; what mechanisms the initial token distribution will take, how differences in goods distributed will be justified, what communities or individuals may be privileged, how information about the event will be communicated, and to what extent values like security might restrict access to sales or mining to name a few. How such decisions are made can impact an initial distribution event’s fairness as a process. Project concerned about a distribution’s procedural fairness sometimes elect to manage a portion of token allocations through some form of centralized or decentralized treasury, the intent of developing decision making procedures greater representative of broader stakeholder interests, less subject to individual bias, that are transparent and accountable for how treasury token holdings are allocated
Tokens that confer contribution rights raise similar issues. Tokens with contribution rights additionally allocate income-earning opportunities through initial token distributions, opportunities typically paired with the responsibility of bearing the cost of performing a network function—staked token holdings can determine who is responsible for maintaining network security, operating dedicated hardware, or performing the service that drives platform use. For example, PoS validators contribute hardware resources and technical expertise but can earn block rewards, and Augur REP holders must act as oracles for prediction markets but are paid for being accurate and reliable.
Contribution tokens enable projects to, through distribution, manage who (initially) has access to income earning opportunities, enabling them to justify particular distributional outcomes of such opportunities on grounds like individuals’ desire and ability to contribute to the platform's success or the ‘raise all boats’ adoption benefits significant industry partners bring—at least, to the extent that distribution mechanisms can select for such things. Likewise, contribution is an intuitive justification for fairly distributing tokens as wealth (so long as the wealth distributed is commensurate to a contribution’s value) meaning it is easier for a project to communicate such channels of allocation’s justification.
Like governance tokens, distributing goods of such a dual-valued nature can give rise to distinct challenges with fairness. Given that the value of a contribution token is intended in part to reflect the value of the opportunity to perform certain services, sales as means of distributing contribution tokens are arguably fairer when putative buyers have enough information about the network to accurately predict likely earnings. Concrete information about such likely earnings is typically more readily available after decisions on token models and fee structures are finalized, process for penalizing and arbitrating the penalization of mal-actors are disclosed, networks are functioning, and the project demonstrates progress towards adoption, all developments the sell of contribution tokens would help projects invest towards. In this respect, contribution functionality can create questions of how projects will balance their early need to raise capital through the sale of tokens with their willingness to conduct process of token distribution made fair through disclosures informative of the contribution right’s value, information more readily available later in life-cycle.
Contribution functionally can additionally introduce the need for project infrastructure sufficient for fairness. Systems managing distributed service providers that forfeit or burn stake for non-malicious behavior (i.e. an Augur reporter knowledgeably stakes REP on a non-consensus outcome that, in retrospect, turns out to be accurate), or lack arbitration mechanisms are open to charges of being unfair systems for managing labor, and systems that inappropriately price contribution’s value may be deemed unfair pricing systems. In these respects, contribution tokens’s distribution might come to be seen as supporting exploitative systems.
In their early stages, projects strongly depend on communities of investors, developers, users, entrepreneurs and other contributors to succeed. While differences in token function, consensus model, and overall vision complicate axiomatic approaches to distributing tokens fairly, as a general rule-of-thumb, projects would do well to think about the initial token distribution from the point-of-view of the communities whose support they depend on.
Projects that understand what opportunities a distribution will allocate to whom are better positioned to communicate to key constituencies the rationale and justification behind strategies employed, and to demonstrate trust by committing to stated rationales by means that go beyond mere words, such as through thoughtful integration of vesting or supporting ongoing distribution strategies. By clarifying varieties of grounds for judging distribution’s fair, this memo better prepares readers to adopt such points-of-views and more readily understand the concerns of key constituents. From a communication and community building perspective, it is important to be able to think through each ground for fairness discussed, as however a project internally prioritizes particular grounds, the community will probably judge the project in light of all.
In turn, would-be-contributors concerned with a project’s fairness are well-advised to not use a distribution’s relative fairness as the sole determinant for collaboration. Unfair mechanisms may be unintentionally implemented, and it is unreasonable to expect projects to fully anticipate a priori the actual effects of novel mechanisms. Whether a project honestly acknowledges the consequences of its choices, and what steps it takes next, can be better signals of whether a project is worthy of collaboration than simply the absence of unfairness. Likewise, circumstances can constrain project choice, with improvements in technology and infrastructures affecting the feasibility of certain distribution strategies, and experience in the emerging field of cryptoeconomic design better clarifying token design’s implications.