DAO Governance Models Compared: Token Voting vs Delegation

DAO governance models fall into a few broad categories — one-token-one-vote, quadratic voting, and delegation — each attempting to balance the same underlying tension: giving stakeholders proportional influence based on their commitment to the protocol, without letting a small number of large token holders dominate every decision regardless of the broader community's views.

Understanding these tradeoffs matters both for evaluating a specific DAO before participating and for recognizing why identical-looking governance tokens can produce very different real-world decision-making dynamics depending on the voting mechanism layered on top.

One-token-one-vote (token-weighted voting)

The simplest and most common model gives voting power directly proportional to token holdings — one token equals one vote, and a thousand tokens carries a thousand times the influence of a single token. This is straightforward to implement and mirrors traditional shareholder voting, but it produces an inherently plutocratic outcome: large holders (whales, early investors, or the founding team if they retain a large token allocation) can outvote a much larger number of smaller holders combined, regardless of how the broader community feels about a proposal.

Quadratic voting

Quadratic voting attempts to soften this concentration effect by making the cost of additional votes on a single proposal increase quadratically rather than linearly — casting one vote might cost one token-equivalent "voice credit," but casting four votes on the same proposal costs sixteen, not four. The practical effect is that a single large holder gets diminishing marginal influence on any one proposal compared to the same total token value spread across many smaller, independent voters, better reflecting breadth of support rather than pure capital concentration. The tradeoff is added complexity (voters need to understand and correctly use a less intuitive voting cost structure) and continued vulnerability to sybil attacks — someone splitting their tokens across many wallets to simulate broader support — unless the DAO has some identity or reputation layer to limit that.

Delegation (liquid democracy)

Delegated models let token holders assign their voting power to another member — a delegate — who votes on their behalf, without giving up ownership of the underlying tokens. This addresses the low-participation problem common in DAOs (see our DAO explainer for why voter turnout is often low) by letting less engaged holders default to a trusted, more actively engaged delegate rather than simply not voting at all. Delegation can be revoked or reassigned at any time in most implementations, and delegates' voting histories are typically public, letting holders choose based on track record.

Model	Influence basis	Strength	Main weakness
One-token-one-vote	Raw token holdings	Simple, easy to implement	Plutocratic, whale-dominated
Quadratic voting	Diminishing returns per additional vote on one proposal	Better reflects breadth of support	Sybil-vulnerable, more complex
Delegation	Voting power assigned to trusted representatives	Raises effective participation	Still concentrates power in popular delegates

Hybrid approaches

Many active DAOs combine elements of these models rather than picking one purely — for example, using token-weighted voting as the base mechanism but layering delegation on top to address participation, or applying quadratic funding (a related but distinct mechanism, see our quadratic funding explainer) specifically for grant allocation decisions while using simple token voting for core protocol parameter changes. The right combination depends heavily on a DAO's specific goals — a protocol DAO managing highly technical parameter changes may reasonably favor informed, larger stakeholders having more say, while a grants DAO distributing funds to community projects may prioritize broad-based input more heavily.

What this means for evaluating a DAO

Before assuming a DAO's governance is meaningfully decentralized, check its actual voting mechanism and the real distribution of its governance token — a DAO using simple one-token-one-vote where a handful of wallets hold the majority of tokens offers governance in name only, regardless of how democratic the marketing sounds. Public governance forums and voting platforms usually let you check both the mechanism and the actual historical vote outcomes and participation rates directly.

Conviction voting and time-weighted models

A less common but notable additional model is conviction voting, where a vote's weight increases the longer a voter maintains their position on a specific proposal without changing it, rather than counting all votes equally regardless of how long they've been held. This is designed to favor sustained, considered preference over rapid, reactive voting, and has been used particularly in continuous funding decisions where proposals aren't decided at a single fixed moment but rather approved once accumulated conviction crosses a threshold over time.

Practical due diligence before trusting any governance claim

Given the range of models available, a reasonable due-diligence step before participating in or evaluating a specific DAO is checking three things directly: what voting mechanism the DAO's smart contracts or governance platform actually implements, how concentrated the top holders of its governance token currently are (many block explorers and analytics platforms make this checkable), and whether historical proposals show genuine, substantive back-and-forth discussion or largely unanimous, low-turnout rubber-stamping. Together these give a far more reliable read on a DAO's real governance character than its stated philosophy or marketing materials alone.

Bottom line

No single DAO governance model fully solves the tension between broad participation and informed, engaged decision-making — one-token-one-vote is simple but plutocratic, quadratic voting better reflects breadth of support at the cost of complexity and sybil risk, and delegation raises effective participation without eliminating power concentration among popular delegates. Evaluate a specific DAO's actual token distribution and voting mechanism directly rather than assuming "DAO" implies any particular level of genuine decentralization.