The Web3 Book
Chapters / Part IV
Part IV · The Components of Web3

20. Decentralized Autonomous Organizations (DAOs) and On-Chain Governance

Chapter 20 · 22 min read ·

This chapter examines one of Web3's most ambitious and most contested propositions: that an organization — a company, a club, a fund, a charity — can be run not by managers and bylaws enforced through courts, but by software that holds the group's money and executes the group's decisions automatically. The vehicle for that proposition is the decentralized autonomous organization, or DAO. The chapter explains what a DAO is and how its parts fit together, returns to the founding catastrophe of the form — "The DAO" of 2016, whose theft of roughly 3.6 million ether forced the Ethereum community into the schism examined in Chapter 13 — surveys how today's DAOs actually operate, and then confronts the hard problems that separate the movement's rhetoric of member-controlled, leaderless coordination from the concentrated, low-turnout reality that empirical study keeps finding. The DAO is, in miniature, the whole question of this book: a genuine technical innovation wrapped in claims about decentralization that deserve careful, even-handed scrutiny.

In plain terms

a DAO is a club run by a shared spreadsheet that enforces its own bylaws. Imagine a club whose treasury, membership roll, and rulebook all live in one document that everyone can see; the document is wired so that it does what the members vote for — releasing money, admitting members, changing a rule — without needing a treasurer or a president to be trusted to follow through. The "spreadsheet" is the blockchain; the wiring that makes it act on votes is the smart contract (Chapter 13, Ethereum and Smart Contracts).

thousands of strangers around the world pooled the equivalent of $150 million into a single piece of software, on the promise that the software, not any person, would manage it. It was a remarkable vote of confidence in the idea that "code is law" — and it was about to be tested in the worst way.

the very first famous DAO was emptied by a bug, and saving the investors required the human community to override the "unbreakable" code — proving, at the form's birth, that a DAO's rules are only as final as its community's willingness to leave them alone. Every claim made later in this chapter about transparency and automation has to be read against that founding lesson.

a modern DAO is a forum plus a ballot box plus a shared bank account, all wired to a blockchain. Members hold voting tokens, write proposals, debate them, and vote — sometimes directly on the blockchain (secure but costly) and sometimes by free signed messages (cheap but needing humans to enact the result). The money sits in code that pays out only when a vote says so, often guarded by a small committee holding the keys.

the pitch is that a DAO can do what a company does — pool money, make decisions, pursue a mission — but in the open, across borders, without anyone having to trust a boss to be honest or competent. The books are public, the doors are open to all, and the rules enforce themselves. If it worked as advertised, it would be a genuinely new way for humans to cooperate at scale.

in most DAOs, votes are not counted by head but by holding. The more of the governance token you own, the more votes you cast. This is not democracy ("one person, one vote"); it is plutocracy — rule by the wealthy — because whoever holds the most tokens decides, and tokens are bought with money.

quadratic schemes try to dampen the power of whales by making the second vote on an issue cost much more than the first, the third more than the second, and so on — so a rich holder pouring money into one decision faces sharply rising costs, while a broad crowd of small contributors can collectively outweigh them. In quadratic funding, the formula channels matching money toward the projects with the most contributors, not the most dollars — funding is the square of the sum of the square roots of individual contributions, written F = (Σ√cᵢ)², which rewards breadth of support over depth of pockets (Buterin, Hitzig & Weyl, 2018).

if a court decides your DAO is really just an informal partnership, then every member could be on the hook, with their own personal assets, for anything the DAO does wrong — a debt, a lawsuit, a regulatory fine. The very feature that makes a DAO feel borderless and free of paperwork could leave its members exposed to unlimited liability they never realized they had accepted by buying a token.

the technology that lets a crowd hold money and act on votes without a boss is real and works. The claim that this produces government by the many rather than the wealthy few is, so far, mostly unproven — most DAOs are run by a small number of large holders while nearly everyone else stays home. The right posture is neither dismissal nor enthusiasm but measurement: judge each DAO by who actually decides, not by what its founders say about decentralization.

What the chapter covers

  1. What a DAO is
  2. The original: "The DAO" of 2016
  3. How modern DAOs work
  4. Real examples: the range of the form
  5. The promised benefits
  6. The hard problems
  7. Synthesis

Terms explained in this chapter

The full chapter is in the book.

Buy the full book