CPS-0033

DRep Voting Power Concentration

Status

Open

Abstract

Stake-weighted voting models have a structural tendency to concentrate power over time, where those with more voting power tend to gain even more over time, ¹, and empirically, this leads to low voting participation where small holders cannot justify the cost of analyzing proposals when they perceive their vote to be powerless ²³. The mitigations tend to require identity-based mechanisms that tie voting weight to verified entities rather than to stake alone ¹. Cardano governance, by design, does not rely on identity. This creates an open and unsolved question: can voting power concentration, voter apathy and principal-agent problems ⁴ be managed effectively in a stake-weighted governance system without identity? The risk of delegation being split across multiple credentials to circumvent any concentration mechanism remains a vulnerability that any proposed solution must consider in its design.

This CPS surfaces DRep voting power concentration as a priority governance challenge for Cardano and opens an investigation into whether Cardano's unique governance structure offers novel technical mechanisms for managing concentration in the absence of identity. The CPS does not propose a specific solution; it frames the problem, surveys community discussions to date, and outlines the open questions any future CIP responding to this CPS should answer. It is connected to but distinct from CPS-0020, which addresses DRep incentives. The two efforts must be coordinated so that incentive design does not work against concentration mitigation, or vice versa.

Problem

Cardano's on-chain governance model gives Delegate Representatives (DReps) voting power proportional to delegated stake. The current ledger rules place no upper bound on the amount of stake that can be delegated to a single DRep credential. Stake-weighted delegation models have known centralizing tendencies over time, and Beyond MVG's on-chain analysis of governance activity across epochs 537–609 found evidence of increasing concentration in DRep voting power, including a declining Nakamoto coefficient and an increasing Gini coefficient. ⁵⁶ While concentration remains significantly lower than that observed in many other DAOs, these measurements suggest a window of opportunity to investigate potential mitigation mechanisms before severe concentration develops:

Forum threads and polls from January through March 2025 (and continuing into 2026), governance roundtables in 2026, and off-chain sources like workshops, interviews and surveys from the Beyond Minimum Viable Governance project consistently surface community concern over top DReps holding disproportionate stake.
Beyond MVG's on-chain analysis observed a declining Nakamoto coefficient and increasing Gini coefficient for DRep voting power across epochs 537–609, indicating that effective voting control has been concentrating among fewer participants.⁵⁶
Community discussions have explored approaches such as saturation points, voting power caps, and delegation distribution mechanisms, suggesting the community recognizes the problem but has not converged on a technical solution.

The deeper problem is structural. Concentrated DRep voting power is not inherently undesirable. Delegation may legitimately accumulate around DReps who consistently demonstrate expertise, transparency, and alignment with the preferences of ada holders. In this sense, some degree of concentration can simply reflect informed stakeholder choice.

However, increasing concentration may also introduce governance risks. Governance outcomes may become increasingly dependent on a small number of representatives, reducing governance resilience if those representatives become inactive, retire, or act contrary to delegator expectations. High concentrations of delegated stake may also make it more difficult for newer or smaller DReps to gain visibility, reinforcing delegation patterns driven by existing voting power rather than informed comparison of available representatives.

The challenge is therefore not to eliminate concentration entirely, but to understand when it becomes a governance concern and whether technical mechanisms can preserve stakeholder choice while discouraging undesirable concentration. Stake-weighted voting tends to centralize over time, and common mitigation strategies such as quadratic voting must introduce identity to be effective against multi-credential circumvention ¹, tying voting weight to verified entities rather than to stake alone. Cardano governance does not use identity. The community may, in the future, reject identity-based mechanisms entirely.

This leaves an open question that has not been answered elsewhere: how, if at all, can a stake-weighted governance system manage power concentration without relying on identity and without compromising the skin-in-the-game benefit of stake-weighted voting? Waiting for identity to be implemented (and adopted) is not an adequate response, because:

There is no guarantee the community will accept identity-based mechanisms.
Concentration is beginning to occur now and may pose meaningful governance risks before any identity solution is ready.
Cardano's governance architecture, including DReps and SPOs, may offer novel technical mechanisms that have not yet been explored. Most other models only have one voting body, and many do not offer delegation.
Imperfect mechanisms may exist such as introducing costs, incentives, or frictions that discourage one entity from splitting into multiple DReps to avoid limits on single-DRep voting power, but have not yet been seriously considered.

This CPS opens an investigation into that question.

Community discussions to date have explored several broad directions, including saturation mechanisms, voting power caps, delegation distribution mechanisms, wallet UX interventions, and incentive-based approaches. No clear consensus has emerged, motivating the need for a structured investigation of the problem space.

Use Cases

A long-term ada holder wants to delegate to a DRep aligned with their values.

Today, they open a DRep explorer that defaults to sorting by voting power. The DReps with the most voting power are the most visible. Aggregation and filtering based on voting history, rationales or voting philosophy are largely absent. The holder, with limited time and no easy way to evaluate quality, picks from the top of the list. Their delegation reinforces the existing concentration. They would prefer a system that surfaces aligned DReps without defaulting to the largest, but no protocol-supported mechanism exists.

A small DRep with strong governance expertise wants to grow their delegation base.

They publish rationales, vote consistently, and engage in community discussions. Despite this, their voting power remains low, while larger DReps continue to attract a disproportionate share of new delegation. Whether this results primarily from wallet and explorer UX, delegator preferences, existing visibility, reputation effects, or some combination of these factors remains an open question. They observe that growing their delegation is difficult and consider stopping their participation.

A large DRep is concerned about contributing to concentration.

A large DRep is concerned that continued growth in delegated voting power may contribute to concentration. However, there is currently no protocol-supported mechanism for communicating delegation preferences or influencing future delegation behavior.

A delegator wants to distribute stake across multiple DReps.

A delegator wants to spread their stake across multiple DReps to avoid contributing to any single DRep's accumulation. The ledger supports this today via multiple stake credentials per account, each delegating to a different DRep. However, no wallet standard defines a multi-DRep delegation flow, a single stake credential's voting power cannot be split fractionally, and most wallets surface a single-DRep model only.

A wallet DRep wants to surface saturation information to delegators.

A wallet wants to surface information to delegators about a DRep's current level of delegated voting power relative to a preferred or recommended threshold. There is currently no protocol-defined signal indicating whether a DRep has reached a preferred or recommended level of delegated voting power. The wallet therefore has no protocol-supported data on which to base such a UI.

Goals

The goals of this CPS, ranked by importance:

Establish DRep voting power concentration as a recognized technical problem in the Cardano protocol and standards ecosystem, with shared understanding of which technical artifacts (ledger rules, metadata standards, and wallet behavior) contribute to it and could address it.
Open investigation into protocol-level and standards-level mechanisms for managing concentration in a stake-weighted governance system that does not use identity. Mechanisms under investigation may include ledger rule changes (saturation curves, self-cap signaling at the credential level), metadata extensions (CIP-0119 successors defining optional saturation or distribution fields), and wallet and explorer behavior standards.
Coordinate with CPS-0020 (DRep incentives) so that any concentration solution and any incentive solution are designed to be compatible. The two problems are connected: poorly designed incentives can worsen concentration, and poorly designed concentration limits can undermine incentives.
Surface and address the technical vulnerabilities any solution must navigate, including the risk that any cap or saturation mechanism is circumvented by splitting delegation across multiple DRep credentials, and the risk that wallet and explorer defaults continue to undo any protocol-level mitigation.
Provide a technical reference point for future CIPs in the Ledger, Wallets, Metadata, and Meta categories that propose specific solutions to the problem framed here.

Open Questions

Any proposed CIP responding to this CPS should address the following:

On the Problem

What is the appropriate metric (or set of metrics) for "too much" concentration in Cardano DRep governance? Nakamoto coefficient? Gini? Saturation point? Some combination?
At what point does concentration become an active risk to governance integrity, rather than a long-term concern?
How do we account for the fact that some concentration may be legitimate (e.g., trusted institutional delegates and delegation that is concentrated but represents the will of the community), while other concentration may be problematic (e.g., a small number of entities controlling outcomes for their own benefit)?
Before any technical intervention is proposed, what modelling, simulations, or research should be conducted to understand the likely effects on delegation patterns, voter participation, governance quality, and power distribution?

On Ledger-Level Mechanisms

Can the ledger introduce a saturation curve on DRep voting weight, modeled on the SPO K-parameter? What is the appropriate functional form (linear taper, exponential, threshold-based)?
What is the appropriate saturation parameter value, and how should it be modeled against current and projected ADA distribution?
Should a DRep credential be able to declare an on-chain self-cap at registration or update time, and have it enforced by the ledger?
Should the ledger enforce a hard upper bound on delegation to a single DRep credential, or rely on saturation curves only?
How does any ledger-level mechanism interact with the existing delegation lifecycle (registration, update, retirement) defined in the Conway era?

On DRep Metadata (CIP-0119 and Successors)

Should CIP-0119 be extended with optional metadata fields for a self-declared delegation cap or saturation preference, even before any ledger enforcement exists?
Should CIP-0119 include fields that wallets can use to surface activity signals (rationale rate, voting record completeness, recent participation)?
Should CIP-0119 include a field for the DRep to signal that they are closed to new delegation?

On Wallet and Explorer Standards

Should a CIP define wallet UI standards for surfacing saturation, alternative DReps, or delegation distribution warnings at the point of one-click delegation?
Should a CIP define standards for DRep explorer sort and filter dimensions so that voting power is not the only or default ordering, for example surfacing participation rate, rationale quality, or consistency instead?
Should multi-DRep delegation flows be defined as a wallet standard, so that delegators can distribute stake across multiple DReps from within their wallet?

On Multi-Credential Circumvention

Without identity, any cap or saturation mechanism can be circumvented by splitting delegation across multiple DRep credentials. What economic or structural mechanisms (deposits, registration cost, on-chain proof of activity) can raise the cost of multi-credential circumvention enough to make any cap or saturation mechanism meaningful?
How can any saturation or cap mechanism avoid creating an incentive to flood the system with many small DRep credentials?

On Unintended Consequences

Could concentration mitigation reduce the quality of governance work, if smaller DReps lack the accountability, expertise, or capacity of larger ones?
Could payment for governance work (CPS-0020) create voters who do the bare minimum to qualify for compensation, reducing rationale and vote quality? How should concentration mechanisms be designed to avoid amplifying this risk?
How should concentration mechanisms and incentive mechanisms (CPS-0020) be designed so they support rather than undermine each other?
Should solutions in both areas be sequenced or developed in parallel?

References

Cardano Community Discussions

CPS-0020: Governance Stakeholder Incentivization, coordinated effort referenced throughout this CPS
Governance tagging discussion in CIP repo

Cross-Ecosystem Research

Uniswap Delegate Reward Working Group

Copyright

This CPS is licensed under CC-BY-4.0.