19.1 DAO Architecture: Governance Tokens, Voting Mechanisms (Token-Weighted, Quadratic, Conviction Voting)

What it is

A decentralized autonomous organization (DAO) is an organization whose membership, decision rules, voting power, and authority to act are encoded in contracts and publicly auditable records. DAO architecture separates who may propose, whose preferences count, how a proposal passes, and which contract receives authority to execute the result.

How it works

A governance token or membership credential expresses voting power, but the token contract is only the source of weight. A snapshot fixes each participant’s power at a stated timepoint, preventing purchases or withdrawals during a vote from changing an already active ballot. Quorum is the minimum participation required for a result to count; delegation lets a holder authorize a delegate to cast its voting power.

A common flow separates eligibility, signaling, and execution:

    sequenceDiagram
    actor Proposer
    participant Votes as Governance token checkpoints
    participant Proposal as Proposal registry
    actor Voter
    participant Rule as Voting mechanism
    participant Decision as Quorum and outcome
    participant Executor as Authorized executor

    Proposer->>Proposal: Submit intent and encoded actions
    Proposal->>Votes: Select fixed snapshot
    Voter->>Votes: Read or delegate voting power
    Voter->>Rule: Cast choice under ballot rules
    Rule->>Decision: Add weighted or evolving signal
    Decision-->>Executor: Publish passed, defeated, or expired result
    Executor-->>Proposal: Request authorized execution

The snapshot protects a ballot from late token trading, but it does not stop an attacker from acquiring power before the snapshot. Voting delay, proposal threshold, delegation disclosure, and a cancel window address different attack windows; no one parameter removes them all.

The three principal mechanisms differ in what they measure:

Mechanism	Weight function	Passage model	Main strength	Main failure mode
Token-weighted	Usually one vote per token unit at the snapshot	Discrete ballot with quorum and choice thresholds	Simple, auditable, and compatible with transferable governance tokens	Concentrated holdings dominate, and delegated power can be captured
Quadratic	A participant’s additional votes cost the square of their count; credits can be split across choices	Discrete or streaming ballot under a per-member budget	Expresses preference intensity and reduces the marginal value of extra tokens on one issue	Splitting an address into many wallets defeats the fairness assumption
Conviction	Support grows with time and may depend on locked stake, decay, and requested resources	Continuous proposals pass when accumulated support crosses a dynamic threshold	Supports asynchronous prioritization and limited-rate treasury allocation	Parameter sensitivity, concentrated delegates, and poor explanations can make outcomes hard to predict

A quadratic voting rule assigns each participant a fixed voice-credit budget. Spending v votes on an option costs v² credits, so preferences can be spread across several choices. The fairness claim depends on enforcing a meaningful per-person budget. On a public chain, one address can own many addresses; a token-balance curve or square-root strategy limits whale influence but does not provide Sybil resistance.

Conviction voting treats support as a time-varying signal rather than a single vote. A participant places conviction behind a proposal, and the aggregate grows while remaining support decays. A proposal passes when its signal exceeds a threshold, often based partly on the share of treasury it requests. 1Hive’s implementation demonstrates the allocation model, while Moonbeam’s OpenGov uses a different conviction system based on locked voting multipliers. They are not interchangeable implementations of one standard.

A deployment should publish a decision record that the token, Governor, interfaces, tests, and monitoring services can verify:

governance:
  chain_id: 1
  governance_asset: "0x1111111111111111111111111111111111111111"
  clock: block_number
  voting_unit: token
  snapshot:
    offset_blocks: 7200
  ballot:
    mechanism: token_weighted
    choices: [against, for, abstain]
    quorum_basis: total_supply_at_snapshot
    quorum_percent: 4
  proposal:
    threshold_percent: 1
    voting_period_blocks: 50400
  delegation:
    enabled: true
    disclose_effective_delegates: true
  execution:
    delay_seconds: 172800
    cancel_window_required: true
  security:
    emergency_executor: none
    minimum_voting_delay: 1

Security review must treat governance as a security boundary. A voting delay changes the acquisition window rather than guaranteeing fair influence. Snapshot lookup can fail after a token upgrade unless the token preserves historical checkpoints. Token supply can concentrate through lending, staking wrappers, or purchase immediately before a vote. Emergency roles can restore response time but can also bypass ordinary governance, so their authority, delay, scope, and expiration must be explicit.

Tradeoffs

Decision	Gain	Cost
One token, one vote	Simple verification and liquid governance	Wealth concentration directly becomes governance power
Quadratic voting	Expressive preferences and lower marginal whale influence	Requires stronger identity assumptions and more complex accounting
Conviction voting	Encourages sustained, asynchronous attention	Outcomes depend on time, decay, thresholds, and delegate behavior
Longer voting delay	Reduces last-moment acquisition and gives delegates time to react	Slower decisions and more proposal staleness
Lower quorum	Allows a smaller active group to decide	Makes capture and nonparticipation more consequential
Emergency authority	Fast response to active incidents	Creates a privileged path that can bypass member consent

When to use

You need transferable economic alignment and can tolerate openly concentrated voting power.
You need preference intensity across several choices and can enforce a credible per-person budget.
You need continuous prioritization for treasury requests and accept a signal-based passage rule.
You need a simple proposal threshold and quorum that contracts can enforce deterministically.

Alternatives

One-member, one-vote off-chain voting — fits an identified membership and participation process, but needs a credible membership registry and off-chain result verification.
Optimistic governance — treats silence as approval for bounded, routine changes, but requires veto rights and a slower path for consequential proposals.
Council or multisig decision-making — fits small expert groups, but concentrates authority and gives members fewer direct voting rights.
Delegated liquid democracy — lets specialists cast member power, but introduces delegate competition, capture, and accountability requirements.

19.2 Treasury Management & Multi-Sig Custody