A DAO is a bureaucrat
The following post was written prior to a session at the Smart Contract Research Forum on the topic of ‘DAO theory’. Questions we were asked to consider included ‘what is governance?’, ‘what is a DAO?’, ‘how do we design for resilience?’, and ‘are plutocracies ok?’ I recommend watching the session to hear other viewpoints from Michael Zargham, Joshua Tan, Quinn DuPont, and moderators Eugene Leventhal and Richard Brown.
People think of DAOs as code instantiations of democratic political process. A better way to think of them is as a competent civil service.
What is Governance?
To govern is to structure the field of action that is possible to others (Foucault, 1982). Groups develop governance practices to deal with risks, which involves setting and enforcing rules as well as shaping outcomes through services (such as health and education). Negotiation and general ‘collective puzzling’ can be part of the process of governing (Colebatch, 2014).
What is a DAO?
DAOs (Decentralised Autonomous Organisations) are typically defined and assessed in legal terms and as a bundle of rules. In Wright’s (2021, p. 155) description, DAOs consist of a ‘network of hard to change rules that establish the standards and procedures of anyone interacting with, or taking part in, a DAO’. I am less interested in the rules than the interaction aspect of this definition — how the technology shapes actions within a field and what new capabilities or harms this produces for individuals or groups.
A DAO is a software actor in these processes of governing, designed to carry out the tasks prescribed to it by a group (the members or founders) or other software actors (using information received from oracles, for instance). For the purposes of ethnographic observation and analysis, making a distinction between the software actor and the human governors enables a more accurate insight into how the technology is altering human capabilities. It also suggests that some things will remain as problems for the human actors to grapple with as they have throughout history, such as moderating deliberation or encouraging voter participation.
When seen this way, the closest thing to a DAO is not a rule book, or even a judge who is called on when disputes arise. Rather, a DAO is the bureaucrat who works from contracts to produce services for constituents on an ongoing basis. It changes how those constituents coordinate with each other and with external parties and stops some events from happening.
Social scientists have observed and compared bureaucracies and debated how to design them for maximum social benefit. A recurring theme in this body of work is the importance of separating the administration of the state from the executive to avoid corruption.
How do we design for resilience?
Fukuyama (2013) provides a framework for evaluating what he calls the infrastructural power of bureaucracies, meaning the performance of agents carrying out the wishes of a political principal. To understand infrastructural power, we need to look at two qualities: capacity and autonomy. Capacity includes merit, including professionalisation and extraction (the ability to tax). Autonomy is the extent to which the bureaucracy has an appropriate degree of scope to implement programs without direction from the political principal. Fukuyama makes the point that capacity and autonomy need to be considered together, in that you don’t want an incompetent bureaucracy to have too much autonomy.
If a DAO is a bureaucrat, then capacity relates to the functions that are performed by the DAO (what it can do), and the extent to which it is susceptible to exploits. Autonomy would encompass the benefits of automating particular processes — leaving the ‘doing’ to software rather than having token holders micromanage every task through lengthy deliberation processes.
The ‘sweet spot’ is a DAO that has merit in its design (i.e., competent at carrying out tasks and services) and that will not accommodate collusion by the human agents who make decisions, possibly by automating tasks. At the other end of the spectrum is what some refer to as Governance Extractible Value.
DAO capacity may be improved by what Max Weber described in his ideal bureaucracy as a ‘fixed official hierarchy’ of offices (Weber 1978, 347). Nested DAOs, sometimes called minion or baby DAOs, are generally owned by another DAO (which has a majority decision-making stake), creating a structure like a branch office inside a larger government department (see also Pat Rawson’s thoughts on this).
While Fukuyama’s framework provides a good starting point for thinking about the performance of DAOs, turning rules into programs is typically a multi-stakeholder process and rarely straightforward. Tess Lea, in her book Wild Policy (2020, loc.387), writes:
“[W]e need to remember that at every level, even when policies are formulated in their most black-letter artifactual form, policy intentions keep company with compromises, struggles, and ad hockery. Policy translations are capricious, and to make them work requires creative interpretations and clever evasions.”
In DAOs, the messiness of governing occurs in forums like Discord or Discourse. These aspects of governance relate to the human actors rather than the software actor in a DAO, but the nature of the discussions, who participates (or defects) and the extent to which they manifest in outcomes will be influenced by the boundaries set out in the code.
From plutocracy to democracy
The rule makers in the field I am describing are those who choose to create a DAO and those who are accepted into it (when there is some level of participation). Bureaucracies don’t require democracy to exist; monarchies had bureaucracies too.
A DAO that is a managed fund for investment, in which a small group makes decisions about the portfolio, is more like a monarchy than a democracy. Even in this case there are benefits to an automated bureaucrat that cannot be manipulated by its despotic fund manager. As a constituent you will still want assurances on the capacity of the DAO, including its ability to not be corrupted by the decision-maker or security vulnerabilities. A DAO such as this may benefit from a high degree of automation, in that prices determined by interaction with other machine actors (oracles, AMMs etc), are a way to limit the powers of the fund management team.
A DAO with a plutocratic power structure (one token one vote) resembles corporate shareholder structures. If minority shareholders needs are met through automated functions, they may still be happy to use the services. The more systemic danger, discussed by Pat Rawson (2021), is that we may export the risks of global financial capitalism to the blockchain economy if too many of our decentralised infrastructures and services end up controlled by a few powerful investors. The same group of majority shareholders could also conspire to make changes to a DAO in order to benefit another DAO where they have more to gain.
The political economy of inter-DAO relations may therefore inspire structures that are more like democracy than plutocracy. In this scenario, DAOs are a new iteration in monitory democracy, whereby a variety of organisations and agencies develop to monitor the powerful and hold them to account — downwards and sideways through the blockchain ecosystem. As political scientist John Keane (2009, loc108) writes, “[d]emocracy recognised that although people were not angels or gods or goddesses, they were at least good enough to prevent some humans from thinking they were”.
About me: I am a Professor at RMIT, working across the RMIT Blockchain Innovation Hub and the Digital Ethnography Research Centre. I am also an Associate Investigator (AI 🤖) of the ARC Centre of Excellence for Automated Decision-Making and Society. I acknowledge the support of the Australian Research Council, FT19010372.
I use Medium for work-in-progress communication and may change my views and analysis as the work progresses. I welcome constructive and respectful feedback from blockchain practitioners. DM me on Twitter @elinorrennie or email ellie.rennie@rmit.edu.au
References
Colebatch, H. K. (2014). Making sense of governance. Policy and Society, 33(4), 307–316.
Fukuyama, F. (2013). What is governance? Governance, 26(3), 347–368.
Foucault, M. (1982). The subject and power. Critical Inquiry, 8(4), 777–795.
Keane, J. (2009). The Life and Death of Democracy. Simon and Schuster. [Kindle edition]
Lea, T. (2020). Wild Policy. Stanford University Press. [Kindle edition]
Lee, L., & Kladges-Mundt, A. (2021). Our Network: Deep Dive #2. https://ournetwork.substack.com/p/our-network-deep-dive-2
Rawson, P. (2021). Ownership in Cryptonetworks. Medium. https://blog.curvelabs.eu/ownership-in-cryptonetworks-96f13f4a113e
Weber, M (1922/2019). Economy and Society: A New Translation. Harvard University Press.
