Aim of the Project
The project has two interrelated aims. The first, descriptive aim consists in mapping, through time, the decisions, actions and outcomes following the foundation, growth, consol-idation and eventual decline of selected so-called Decentralized Autonomous Organizations (DAOs). The second aim is to explain how different governance mechanisms, in particular reputation based governance and consensus voting and the social processes they trigger can account for variations in DAOs’ capacity to sustain value creation.
Background
Decentralized Autonomous Organizations (DAOs) were first proposed in 2015. They are vir-tual hybrid organizational arrangements that use blockchain technology to foster joint pro-duction. Though not a prerequisite for DAOs, many DAO’s are collectively owned and run by a committee whose members formulated (and agreed to follow) a set of rules in order to realize joint outcomes. Some of these rules are part of so-called “smart contracts”, which means they are implemented in algorithms that automatically get activated once specific pre-defined contractual conditions are met.
Depending on the kind of governance implemented, DAOs may have several ad-vantages when compared to traditional organizations (Choi et al, 2022; Ghavi et al., 2022). First, blockchain governance allows, in principle, providing direct control over organizational operations to all members of the organization, without interference of third parties like managers or banks. Second, they increase transparency, since all actions are recorded in the blockchain and are accessible to every member of the organization. Third, DAOs may bring considerable efficiency gains when it comes to founding and running them. In fact, one of the platforms hosting DAOs advertises with the words “Go from zero to DAO in 90 seconds” (Colony.io). Fourth, DAOs enable and encourage widespread participation and democratic decision making, which may improve decision quality. Finally, unless hackers use flaws in the code (as was the case in The DAO), DAOs foster trustworthiness about the protection of its funds, because self-executing contracts cannot violate the pre-defined rules they carry out.
Not surprisingly therefore, interest in DAOs is booming. According to a recent sur-vey, the number of DAOs increased from about 700 in May 2021 to about 6,000 as of June 2022 (Ghavi et al., 2022; see also Faqir-Rhazoui et al, 2021). During this period, the total value of crypto funds held in DAO treasuries exploded from $400 million to $16 billion, and the number of holders of interests in DAOs rose from just 13,000 to 1.6 million (Ghavi et al., 2022). In fact, some hail DAOs as the Future of Organizations. But this claim may be premature. For one, so far there is not really much empirical evidence to substantiate it. If anything, there is a large gap between the number of DAOs and the number of active DAOs. For example, as a comparative analysis of the three most important DAO platforms has revealed, of the 2000+ DAOs, only about 10% qualify as “active”, i.e. it “performed an action” during a month (Faqir-Rhazoui et al, 2021). Furthermore, DAOs, like any other or-ganizational form, come with their own risks (Morrison et al., 2020). Among these are crypto-related risks (for example flaws in the blockchain software or security breaches), legal risks (most DAOs lack a legal status, which formally turns them into general partner-ships, with the consequence participants bearing unlimited legal liability), and environmen-tal risks (related to the immense energy use of blockchain technology). Another class of problems are related to governance (Rikken et al., 2019). First, DAOs may fall prey to the iron law of oligarchy, i.e. despite its decentralized intentions de facto power may come to be concentrated in the hands of a small subgroup. This may also be the result of delibera-tive “governance attacks”, during which (groups of) participants may instrumentalize the rules of the DAO for their own ends. Second, deliberative democratic decision making in-volving the whole community may become too cumbersome and time consuming, in partic-ular with increasing problem complexity. Finally, the need to reprogram self-executing smart contracts that have become obsolete or dysfunctional inhibits quick adaptation to changing circumstances.
Features of DAOs meanwhile have been well-described (Bellavitis et al., 2022; San-tana & Albareda, 2022). In fact, the academic literature on DAOs so far excels in conceptu-al, theoretical, and programmatic analyses that try to come to grips with the essence of a DAO, how it differs from other cooperative arrangements, and how it may impact economy and society in the mid and long run. In contrast, with some rare exceptions (e.g. Du Pont, 2018; Schirrmacher et al., 2021) we hardly have any first hand descriptions or studies of what is actually going on inside a DAO. What motivates their founders to create a DAO, what might explain variations in choices for different governance structures and their even-tual change? What kind of cooperative interactions take place, how productive are they? What kind of tensions emerge and how are they handled? But the question that probably most entices practitioners and academics alike is what it takes to build a sustainable DAO. This project aims to answer this question.
Theory
The Theory of Governance Traps (Wittek, 2022) is used as a point of departure and extend-ed to the context of Blockchain Governance in DAOs. Several theoretical analyses have pointed to the endogenous downward spirals challenging the viability of DAOs and related organizational forms. For example, some scholars argue that DAOs face the same “paradox of flexibility and structure” that threatens the viability of what has been labeled Fluid Or-ganizations (Schirmacher et al., 2021; Schreyögg & Sydow, 2010). Similarly, analyses of algorithmic decision making and control point to the contested nature of the related practic-es (Kellogg et al., 2022) and highlight the inherent problems of ambiguity intolerance and pressures on social decision making practices (Herzog, 2021).
A governance trap reflects a self-reinforcing process in which an institutional ar-rangement that is intended to elicit cooperation, also triggers behaviors that indirectly un-dermine it. An example are performance contingent incentives in organizations, like bonus-es. Whereas such incentives are powerful in eliciting the type of behavior that yields the reward, they may also lead to the neglect of other behaviors that are not rewarded, but nevertheless important for overall performance, like not taking excessive risks (Becker & Huselid, 1992). Building on insights from research on goal framing and joint production mo-tivation (Lindenberg & Foss, 2011), this theory argues that independently of its success in getting cooperation going in the short run, any governance structure also bears the seeds for its own decay in the middle and long run. This tendency towards endogenous decay has its roots in the brittle nature of human motivation when it comes to sustaining contributions to collective goods (Lindenberg, 2014). As recent experimental research has shown, main-taining a collective good is more difficult than creating a new one (Gächter et al., 2017). One implication is that governance structures geared towards keeping joint production moti-vation salient will be more successful in preventing the emergence of governance traps.
DAO platforms – the digital infrastructures that potential DAO founders can use to configure their own DAOs, like Colony or Aragon – are well aware of the many potential threats that may lead to the (early) dissolution of a DAO. This is why they equipped their platforms with a series of tools that allow founders to implement and calibrate a variety of institutional safeguards to prevent and mitigate governance failures (Baninemeh et al., 2021). Reputation and consensus systems are two particularly important elements of the broader set of governance instruments used by DAOs (e.g. Rea et al., 2020).
First, most DAOs provide the opportunity to track and reward member contributions to the collective good, like a specific project. Often, such contributions can be made visible through an individual reputation score, and thereby contribute to the reputation of the DAO member. This reputation can be expressed in the DAO’s own token, and may therefore also have monetary value for the member, or it may translate into voting or control power with-in the DAO. The opportunity to build up reputations therefore can be a powerful incentive for individuals to invest intelligent effort into joint endeavors. But reputation systems come with their own challenges. For example, how to avoid that members who have accumulated high reputation scores in the past also keep contributing in the present? DAOs therefore dif-fer with regard to their approach to reputation based governance. Particularly noteworthy is the solution that the Colony platform has developed. Here, the reputation algorithm is pro-grammed such that a member’s reputation decays through time (e.g. at an hourly rate), in order to incentivize members to keep contributing (Rea et al., 2020).
Second, most DAOs have some form of collective decision making process in place. Such processes are used to vote, for example, on budget allocations for specific projects, or on strategic issues. Also here DAOs differ in the way they design the related consensus and voting procedures. Again, the Colony platform’s approach is pioneering in its reliance on what it calls lazy consensus, i.e. “decentralized decisions without voting”. This principle is based on the idea that voting is only necessary if there is disagreement, thereby avoiding one of the potential shortcomings of participatory decision making.
A DAO is sustainable if it succeeds in eliciting and maintaining joint production ef-forts that create internal and social value - also if circumstances for this joint production deteriorate. Pre-programmed reputation decay and lazy consensus are just two of a vast array of blockchain based governance practices designed to boost the sustainability of DAOs through a radical implementation of organizational practices geared to increase accountabil-ity, objectivity and participation. But like any form of algorithmic control (Herzog, 2021; Kellogg et al., 2020), also blockchain governance creates a whole array of new challenges, some of which may actually undermine these very objectives. This project investigates un-der which conditions DAOs succeed to prevent and mitigate such governance traps.
Research Design
A mixed method approach will be used for an in-depth longitudinal comparative study of selected DAOs (for an inventory of DAOs, see for example https://daocentral.com). Digital Ethnography (Pink et al., 2017), and in particular the principles of Participatory Digital Eth-nography of Blockchain Governance as outlined by Rennie and colleagues (2022), serve as the point of departure for designing the research strategy for this project. Data collection methods include interviews with different DAO stakeholders (e.g. founders, members, ben-eficiaries), participant surveys, focus group discussions, and text analysis of communica-tions among DAO members. With DAOs being very recent phenomena that moreover con-sist to a large part of online interactions, an important task for this project will be the de-velopment of a feasible strategy of collecting and analyzing different forms of data. The technique of Ethnographic Arrays (Abramson & Dohan, 2015) will be applied for this pur-pose. The first phase of the project will consist of inventorizing DAOs that may be suitable cases for this project. This will be followed by approaching representatives of DAOs and ex-ploring opportunities for participant observation as part of a co-creation process (Rennie et al., 2022). The respective DAOs will be followed for a period of three years. Data collection will involve (computer aided) content analysis of online communication and deliberation, as well as personal interviews and focus group discussions.
Project Initiators
Rafael Wittek (UG), Francesca Giardini (UG), Lisa Herzog (UG)
Location
Sociology, University of Groningen