Cardano’s “Governance Game Problem” and How to Solve It

The Voltaire phase of Cardano is now in full swing, and roughly a year and a half has passed since the DRep system became operational.
The governance system introduced to Cardano is based on “Liquid Democracy,” taking a different approach from traditional democracy, designed to achieve greater decentralization and prevent centralization.
However, through observing governance over this past year, I have come to see the structural limitations of the current governance system through the behavioral patterns of DReps and the delegation behaviors of holders. If left unchecked, Cardano’s governance risks losing its balance, potentially leading to an accelerating decline of the ecosystem.

In this article, I will outline the problems with Cardano’s governance and propose solutions, drawing on my experience as an active DRep since September 2024.

Part 1: What “Centralization” Means in Cardano Governance

Before diving into theoretical analysis, let’s first look at the data to see how decentralized the current Cardano DRep system actually is. This article combines multiple data sources (cexplorer, adastat, tempo.vote) to evaluate “decentralization” in Cardano governance from multiple angles (as of late April 2026).

Analyzing Delegation Concentration

How is delegation distributed among the approximately 5.82B ADA actually delegated to DReps? Let’s look at the top DRep rankings using adastat data.

Top DRep Delegation Amounts (adastat, April 2026)

Rank	DRep Name	Delegation Amount	Number of Delegators
1	Yoroi Wallet	695.84M ADA	19,725
2	YUTA	484.38M ADA	1,496
3	Blockdaemon	304.05M ADA	93
4	EMURGO	297.39M ADA	429
5	Eternl DRep Committee	247.86M ADA	9,348
6	CryptoCrow	206.92M ADA	2,630
7	Sebastien Guillemot	181.85M ADA	1,381
8	Everstake	140.16M ADA	7,340
9	Cardano Foundation	134.86M ADA	594
10	JAZZ	101.83M ADA	331

Top Concentration Ratio in Cardano

In economics, the “CR5 (Concentration Ratio 5, top-5 concentration ratio)” is a standard metric used to measure market competitiveness. Expressed in the form “CRn,” it indicates the oligopoly share of the top n firms in a given market.

Top 5 Accounts Concentration Ratio for DReps

CR5 = Total delegation of top 5 / Total delegation to DReps
    = (695.84 + 484.38 + 304.05 + 297.39 + 247.86)M / 5,820M
    = 2,029.52M / 5,820M
    = 34.9%

The concentration ratio of the top 5 DReps is CR5 = 34.9%

The Top 10 Accounts Concentration Ratio for DReps

CR10 = Total delegation of top 10 / Total delegation to DReps
     = 2,795.14M / 5,820M
     = 48.0%

The concentration ratio of the top 10 DReps is CR10 = 48.0%

In other words, in Cardano’s current governance, “the top 10 hold nearly half of all stake.”

Centralization Is Accelerating

Compared to January 2026, concentration has worsened further in just three months.

Metric	January 2026	April 2026	Change
Total DReps	1,551	1,016	-34.5%
CR5	32.5%	34.9%	+2.4pt
CR10	46.0%	48.0%	+2.0pt

What’s particularly significant is that concentration has worsened despite Yoroi announcing it would stop accepting delegations. This suggests that delegation leaving Yoroi has flowed to other top DReps. Despite the Target 15 campaign’s call for greater decentralization, the market has moved in the opposite direction.

Out of 1,016 DRep Accounts, the Top 100 Dominate

Using the same calculation method, the concentration ratio of the top 100 reaches over 90% (exact figures fluctuate as delegation amounts update).
This means that just 9.8% (the top 100 out of 1,016 accounts) controls over 90% of all DRep delegation.
To summarize:

• Top 5 (0.49%) → CR5 = 34.9% (71x concentration)
• Top 10 (0.98%) → CR10 = 48.0% (49x concentration)
• Top 100 (9.8%) → CR100 ≈ over 90% (approximately 9x concentration)
• Remaining ~900 (90%) → less than 10% of the total

These are the numbers that describe the current situation.

DReps May Appear Relatively Decentralized?

Many people believe that while DRep concentration is high, it is not critically severe. This likely stems from viewing DReps through the lens of “market concentration.” Compared to markets like smartphones, cloud services, and cryptocurrency exchanges, the concentration may not look as extreme.

• Smartphones: CR5=67%, CR10=93%
• Cloud Services: CR5=71%, CR10=80%
• Cryptocurrency Exchanges: CR5=67%, CR10=91%
• Cardano DReps: CR5=34.9%, CR10=48.0%

In a market economy, “34.9% concentration” means “you can still freely choose from the remaining 65%.” From this comparison, one might conclude that “significant centralization has not occurred.” As a result, many participants tend to believe that “Cardano’s governance is relatively decentralized.”

From a Corporate Governance Perspective, Extremely Concentrated

However, this is merely a “delegator’s perspective.” Cardano’s governance has fundamentally different characteristics from a typical market economy.
In a market economy, while monopolistic firms may exist, the ultimate decision-making power remains with consumers. Companies provide products but do not participate in setting market rules. In Cardano’s governance, however, DReps have the authority to change the rules of the “Cardano market” itself. Therefore, judging Cardano’s governance by simple market concentration metrics does not reflect reality.

A closer analogy is corporate governance — specifically, decision-making at shareholder meetings. Voting “for,” “against,” or “abstaining” on proposals, with influence determined by shareholding (analogous to delegated stake), makes shareholder meetings a very close model.

Influence Thresholds at Shareholder Meetings

At shareholder meetings of enterprise companies, shareholder influence is evaluated by ownership ratio as follows:

• 33.4% or more: Blocking power over special resolutions (de facto controlling shareholder)
• 25-30%: De facto control when other shareholders are dispersed
• 10-20%: Strong pressure on management, influence over board appointments
• 5% or more: Major shareholder with disclosure obligations, individually addressed by the IR department

Applying this to Cardano DReps (calculated with a base of 5.82B ADA total delegation):

• Yoroi (11.96%): The largest shareholder most closely watched by management
• YUTA (8.32%): A major shareholder with influence over board appointments
• Blockdaemon (5.22%): A major shareholder with legally required disclosure obligations
• EMURGO (5.11%): A major shareholder with legally required disclosure obligations
• Eternl DRep Committee (4.26%): Near-major shareholder level

And looking at overall concentration:

• CR5 = 34.9%: At the level of a de facto controlling shareholder group. They hold blocking power over special resolutions and decisive influence over management direction
• CR10 = 48.0%: Can sway nearly all decisions. If other shareholders are dispersed, this constitutes de facto control
• CR100 ≈ over 90%: 9.8% of 1,016 accounts hold dominant influence

Concentration at the “Listed Subsidiary” Level

From a corporate governance code perspective, this structure would be evaluated as follows:

• No single controlling shareholder exists, but a shareholder group with controlling influence does
• Flagged as a “minority shareholder control risk”
• Concerns over independence and democratic integrity

This is comparable to a “listed subsidiary” in Japan where a parent company holds 30-40%. In Japan, such entities face special disclosure requirements due to “questions about independence.” In the United States, institutional investors would raise alarms about “collusion risk between management and major shareholders” and demand reforms.

In other words, while Cardano may appear “somewhat concentrated” by market economy metrics, from a corporate governance perspective, it is judged as “an extremely unhealthy structure where de facto control by a few is possible.”

10 People Around a Table Decide the Allocation of 350 Million ADA

Cardano’s governance has the authority to determine the allocation of massive funds through treasury withdrawal actions and similar mechanisms. The 2026 NCL (Net Change Limit) is set at 350 million ADA. This is equivalent to $140 million (at $0.4 per ADA), or roughly 21 billion yen in budget.
This budget scale is comparable to:

In corporate terms:

• Annual R&D expenditure of a mid-sized listed company
• Annual investment budget of a regional bank

In public budget terms:

• Total annual budget of a small municipality
• Annual operating budget of a single national university

Per day:

• Approximately 960,000 ADA
• $380,000 USD equivalent
• Approximately 58 million yen

This massive budget allocation is structurally possible for the top 10 to decide. A group small enough to fit around a restaurant table can determine the allocation of $140 million (21 billion yen).
Currently, there are 1,016 DReps registered on the network, but just 10 of them — 0.98% of stakeholders — can decide budget allocations of this magnitude. Per individual, this is equivalent to overseeing approximately $14 million (2.1 billion yen) in budget — the kind of sum entrusted to professional venture capitalists or institutional fund managers.
Yet under Cardano’s rules, these individuals lack the following checks that professionals handling such amounts would normally be subject to:

• ✗ No legal obligation of fiduciary duty
• ✗ No investor protection regulations
• ✗ No auditing
• ✗ No conflict of interest disclosure requirements

To put this in corporate terms: “A company with 1,016 employees, tens of thousands of users, and a 21-billion-yen budget — where the budget is decided by a board meeting of just 10 executives.” This is because it is extremely easy for 10 people (48%) to communicate and build a coalition, while it is mathematically almost impossible for the remaining 1,006 (52%) to uniformly coordinate.
Note that the current Cardano Constitution requires a 67% ratification threshold, so the top 10 alone cannot pass proposals. However, a CR10 of 48% is more than sufficient to function as a de facto veto power. The structural risk is that a minority’s “veto” can distort the direction of governance.

The List of Problems in Our Governance

Why has this centralization occurred? Based on my experience participating in governance as a Constitutional Committee member and then as a DRep, I believe the current Cardano governance structure has the following problems:

[Structural Problems]
1. Delegation to DReps Becomes Entrenched
Once delegated to a DRep, delegation tends to persist. While mechanisms exist for delegators to periodically review DRep performance, awareness of them is low. As a result, delegation is maintained even when a DRep’s activity declines. Additionally, many holders report that governance topics are too complex for them to evaluate DRep decisions.

2. A “War of All Against All” Among DReps
Unlike SPO block production, DReps are ideally supposed to cooperate, deepen discussions, and arrive at better governance decisions. In practice, however, they are trapped in a “competitive” structure fighting over delegation. Due to this structural problem, DReps are forced into governance activity in a state closely resembling Thomas Hobbes‘s “war of all against all.”

3. The “Asset-ification” of Voting Power (VP)
Voting Power (VP) gained through delegation has become the definitive metric determining a DRep’s value. As a result, high-VP DReps’ statements are given priority, and they often receive preferential treatment at events, while the quality of their statements tends to be ignored. VP has effectively become “wealth” for DReps, and “acquiring and maintaining as much Voting Power as possible” has become the goal itself.

These three elements create a “DRep delegation competition game” that produces the following governance-degrading problems:

[Resulting Bigger Problems]
A: Prioritizing “Competition Over Discussion”
As a result of 1, 2, and 3, DReps focus not on “collaborative discussion before voting” or “group position statements,” but rather on “individual promotion after voting.” In practice, many DReps make independent voting decisions and then engage in marketing: “I voted this way. Here’s why. If you agree, delegate to me.” This is the product of a structure where “differentiation is more advantageous than cooperation for gaining delegation.” DReps tend to prioritize actions that maximize their own incentives over deeper discussion.

B: Obstruction of Cooperation Among DReps
When VP is treated as “wealth,” joint statements and joint proposals among DReps become extremely rare. Cooperating with other DReps makes “everyone look the same,” ultimately causing significant damage to a DRep’s “asset building.” This structure undermines the very cooperation that should enhance governance quality.

C: “Marketing Ability” Trumps “Quality of Discussion”
The most serious problem is the tendency for “good-looking posts” to become more important than “sound logic” in governance. Deep technical analysis, time-consuming contributions to discussions, and expressing opposition to proposals from large organizations do not directly translate into delegation gains. Rather, frequent social media posts, simple messages, catchy appeals, and showing support for large organizations — these “marketing activities” and “going along with the crowd” contribute far more to a DRep’s “asset building.”
As a result, an “adverse selection” occurs where DReps skilled at marketing accumulate delegation over those with technical expertise. This tendency is especially strong for technically complex proposals. For example, discussing governance mechanisms and protocol design requires enormous time and effort, but that effort is invisible to delegators. Meanwhile, posting brief opinions on political proposals or budgets on social media is far more likely to be evaluated as “visible activity.”

Unlike SPO’s block production, which has the relatively simple goal of “maintaining a secure blockchain,” governance deals with far more complex issues. Therefore, rather than simply enhancing their own prominence, DReps should deepen discussions and examine issues from multiple perspectives to elevate governance quality overall. However, the current structure obstructs this.

These problems have not been quantitatively proven, but I believe they are prevalent tendencies observable in ecosystem communications.
Rather than focusing on quantitative verification, this article aims to examine problems in Cardano’s governance structure and their solutions on a theoretical basis.

Part 2: A Theoretical Analysis of Cardano’s Governance Failure

So why does this “marketing-dominated” structure emerge?
This is less about the beliefs or abilities of individual DReps (though that plays a role) and more about “the governance system itself producing distorted incentives that run counter to its intended purpose.”
From here, I will delve deeper into the structural flaws of governance using game theory and economics.

“Asset-ification of Voting Power” and “Zero-Sum Game”

To understand the problems in Cardano’s governance, we must first grasp two structural premises.

Premise 1: The “Asset-ification” of Voting Power

Cardano’s governance is built on a “Delegation Game” where each DRep gains influence by collecting delegation from ADA holders across the ecosystem.
Delegation volume is called “Voting Power (VP)” and should ideally serve as “a tool for improving governance quality.” In practice, however, VP functions as a DRep’s “asset.”

On-chain = Influence in governance voting
Off-chain = Greater voice, better treatment at events and projects
↓
VP increase = Increase in value as a DRep
VP decrease = Decrease in value as a DRep

Specifically, DReps with large VP:

• Are regarded with respect by governance action proposers
• Earn high regard within the Cardano community as “trusted by delegators”
• Receive priority event invitations and more opportunities for direct contact with developers
• Get more engagement on social media posts, amplifying their influence
• Can enjoy financial benefits if they also serve as an SPO

In other words, for DReps, voting power has become “an asset to acquire as much of as possible and maintain.”

Premise 2: The Zero-Sum Game Structure

Furthermore, voting power — the asset in this “delegation game” — has a “zero-sum game” structure. Simply put, a “zero-sum game” describes the following situation:

"Participants' gains" + "Participants' losses" = 0 (zero)

In other words, “if one person gains 10, others lose 10 (gains + losses = 0).”
In Cardano, the total supply of ADA at any given point is fixed, meaning the total amount of delegatable ADA is fixed. Therefore, “when one DRep’s Voting Power (VP) increases by 100 ADA, other DReps lose that 100 ADA of voting power.” One person’s gain is necessarily another’s loss.

These two structural premises — “the asset-ification of voting power” and the “zero-sum game” — cause Cardano’s governance to become a “survival game of every DRep against all other DReps.” This is remarkably close to what Thomas Hobbes called the “war of all against all.”

Three Governance Problems Created by the Struggle for Delegation

Building on these two premises, let’s examine how this survival game at the individual DRep level affects governance as a whole. The three problems (A, B, C) presented below are all logically derived from the premises of “VP asset-ification” and the “zero-sum game.”

A: Self-Promotion Increases Voting Power More Than Discussion

In the current governance landscape, DReps’ outreach to delegators is primarily conducted through social media (mainly X). Combined with the characteristics of social media, simple individual appeals are more effective at “increasing asset value” than deep technical discussions.
Given this dynamic, the incentive to vote first and then appeal to the community outweighs the incentive for deep pre-vote discussion, resulting in reduced motivation for thorough deliberation.
Let’s look at a specific logical model. DRep “Alice” is considering her approach to an important proposal X.

Pattern 1: Engaging in deep technical discussion

Publishing detailed analysis on social media
↓ Takes time to understand (high burden on readers)
Delegator reaction:
"Too difficult to understand. I'll delegate to a DRep who explains things more simply."
↓
Result: Delegation is hard to gain relative to effort invested

Pattern 2: Emphasizing a simple original perspective

"Here's what I think" — a simple message
↓ Easy to understand, clear differentiation
Delegator reaction:
"Alice has a unique and interesting perspective. I'll delegate to her."
↓
Result: Easy to gain delegation with less effort

Clearly, “Pattern 2” is more rational for Alice. This can be understood as a matter of ROI (Return on Investment) for DReps.

Strategy	Investment	Return	ROI
Deep technical discussion	Enormous time (days to weeks)	Hard for delegators to understand	Low
Simple individual appeal	Little time (a few hours)	Easy to understand, clear differentiation	High

The zero-sum game structure makes this choice even clearer. Since total delegation is fixed, DReps must constantly “stand out more than others.” Deep technical discussions involve specialized terminology in long-form content that only a fraction of followers can engage with. Meanwhile, simple, approachable content posted frequently provides clear differentiation.
As a result, the “delegation game” ends up promoting the wrong kind of behavior among DReps.

Deep discussion "before" voting:
→ Low ROI → Limited VP increase

Promoting one's own views "after" voting:
→ High ROI → VP increase can be expected

The ideal for Cardano’s governance is that technically important proposals receive the deepest discussion. In reality, however, “producing eye-catching voting appeals” has become a more rational choice than “promoting accurate decision-making through deep discussion.”

B: The Prisoner’s Dilemma Makes Cooperation Costly

Better decision-making requires DReps to cooperate on discussion and analysis, issuing joint statements and working as a team. However, the “war of all against all” in governance has made cooperation among DReps uncommon.
This too is a logical consequence of “VP asset-ification” and the “zero-sum game.” While section A addressed the problem that “deep discussion has low ROI,” section B focuses on “cooperation itself being structurally disadvantageous for DReps.” This can be understood through game theory’s “Prisoner’s Dilemma.”

The basic structure of the Prisoner’s Dilemma

Imagine two suspects being interrogated separately.

Both stay silent (cooperate):
→ Insufficient evidence, both receive light sentences (global optimum)

One confesses, the other stays silent:
→ The confessor goes free, the silent one receives a heavy sentence

Both confess (betray):
→ Both receive moderate sentences (worst collective outcome)

The rational calculus:

If the other stays silent → Better to confess
If the other confesses → Must also confess to avoid the worst outcome

In either case, confessing is rational

As a result, both confess, leading to the worst collective outcome.

The Prisoner’s Dilemma in Cardano DRep Governance

Suppose DReps Alice and Bob are considering whether to cooperate on an important proposal X.

Scenario 1: Both cooperate

Alice and Bob jointly publish a deep analysis
↓ Delegator reaction: "Alice and Bob are saying the same thing. Either one is fine."
↓ Impact on delegation: Delegation splits between Alice and Bob (50:50)
Impact on asset (VP): Both have similar chances of growth

Scenario 2: Only Alice defects

Alice: Emphasizes her unique perspective A (effort: 50 hours)
Bob: Publishes a cooperative, detailed analysis (effort: 100 hours)
↓ Delegator reaction: "Alice has a unique perspective. Bob is diligent but bland."
↓ Impact on delegation: Alice takes delegation from Bob (70:30)
Impact on asset (VP): Alice gains significantly

Scenario 3: Both defect

Alice: Emphasizes perspective A
Bob: Emphasizes perspective B
↓ Both get lost among other DReps
Impact on asset (VP): No worse than cooperating, but no new perspectives emerge

To summarize:

If the other cooperates → Better to defect
If the other defects → Must also defect to avoid losing

In either case, defecting is rational

As a result, both choose to defect (compete), and cooperation fails to materialize.

Three Structural Disadvantages of Cooperation

As long as VP is an “asset,” cooperation among DReps carries the following risks:

1. Loss of Differentiation = Decrease in Asset Value

DRep Alice publishes an independent opinion
→ "Alice has good insight" → Alice's reputation rises

DRep Alice and Bob issue a joint statement
→ "Alice and Bob share the same view" → Reputation is split between them

2. Dilution of Results = Lost Opportunities for Asset Growth

In joint activities, it becomes unclear “whose achievement this is,” diluting the delegation acquisition effect. This leads to the mindset that thinking independently and publishing under one’s own name is more beneficial.

3. Risk of Betrayal = One-Sided Loss

Because of the uncertainty that the other party might defect, DReps conclude that it’s safer to compete from the start than to cooperate. Due to these structures, Cardano’s governance is governed by the behavioral principle that it is rational for DReps not to cooperate.
The difficulty DReps face in building cooperative relationships is not because they lack morality or a collaborative spirit. The cause is that in Cardano’s delegation game, where “voting power = asset,” the incentive structure is designed so that cooperation means losing. The cooperation that should enhance governance quality is structurally unsustainable.

C: Marketing Determines the Winner, Not Quality of Discussion

In sections A (zero-sum game) and B (disadvantage of cooperation), we established that “competition and differentiation are rational.” In section C, we present an even more serious problem: the way differentiation itself is fundamentally distorted.
The ideal for Cardano’s governance is for DRep delegation to be differentiated by “quality of governance judgment.” This would enable governance that is more transparent and corruption-resistant than traditional democracy.
In reality, however, the structure favors differentiation through “marketing ability.”

Information Asymmetry: Delegators Cannot Evaluate “DRep Quality”

Economics has a concept called information asymmetry — a situation where the “seller” and the “buyer” possess different levels of information. Economist George Akerlof used the used car market as an example to develop the theory of adverse selection (the Lemons problem).
In the used car market, cars may look similar, but their actual quality varies greatly. Put simply, even if two cars have equally clean exteriors, the internal quality — engine deterioration, component wear — can differ enormously.
However, buyers without specialized knowledge cannot assess internal quality and must purchase based on appearance and price. Meanwhile, dealers have engineers with the expertise to know the car’s internal details.
As a result, the following occurs in the used car market:

True value of a good car (carefully maintained): $10,000
True value of a bad car (poorly maintained): $5,000
Price the buyer offers: $7,500 ("I can't tell which it is, so I'll offer somewhere in between")

Good car seller: "It's worth $10,000 but they only offer $7,500?" → Doesn't sell
Bad car seller: "It's only worth $5,000 but they're offering $7,500?" → Sells

→ High-quality cars disappear from the market, leaving only low-quality ones

The Same Happens in the DRep Market

DReps share the characteristic that “quality cannot be judged from the outside alone.”

DRep side (the informed party):

• Knows their own technical abilities
• Knows their level of understanding of proposals
• Knows their contribution to discussions

Delegator side (the less-informed party):

• Cannot accurately evaluate a DRep’s technical abilities
• Cannot judge the quality of analysis
• Can only evaluate based on “visible activities” like social media

Metrics that are easy for delegators to evaluate:

• Social media presence (posting frequency, impressions, etc.)
• Clarity and approachability of messaging
• Quantitative metrics (follower count, delegation amount, etc.)
• Known brands (Yoroi, EMURGO, etc.)

Metrics that are hard for delegators to evaluate:

• Depth of technical analysis
• Track record of protocol development contributions
• Contributions to governance discussions

As a result, the same adverse selection that plagues the used car market occurs in governance.

DRep Type	Evaluation Metric	Outcome
Technical DRep	Technical accuracy, depth of discussion	Delegation stagnates → Reduces activity (exits the market)
Marketing-oriented DRep	Follower count, posting frequency	Gains delegation → Continues activity
AI-powered DRep	Post volume, content presentation	Efficiently gains delegation → Continues activity

Among DReps, those who are “visually appealing and easy to understand” attract delegation, while those who “post infrequently and attend fewer events” lose delegation. In this process of natural selection, the “actual ability of DReps” that is invisible to delegators is completely ignored.
In the current “delegation game,” no matter how knowledgeable or high-contributing a DRep may be, those with weak marketing skills will gradually lose delegation and be forced to exit the governance market.

The “Asset-ification” of VP Accelerates Adverse Selection

VP functioning as an “asset” accelerates this adverse selection as follows:

DReps successful at marketing:

VP increases → Asset value rises → More visibility at events
→ Further recognition → Further VP increase
→ "The rich get richer" cycle is complete

Technically excellent DReps:

Delegation stagnates → VP remains low → Recognition doesn't grow
→ Delegation stagnates further
→ A vicious cycle where effort doesn't translate to delegation

In the used car market, sellers of good cars “exit the market.” Similarly, in the DRep market, “once a DRep falls behind, they become even more disadvantaged,” ultimately being pushed toward the decision to “quit being a DRep.”

The System Is Generating the Wrong Competition

This is not a problem with individual delegators making poor decisions. It is the result of the system forcing this “delegation game.”

Values that Cardano’s “Liquid Democracy” should promote:

• Technical depth
• Quality of discussion
• Long-term contributions

However, the values actually being promoted:

• Marketing ability
• Short-term visual appeal
• Brand power

This tendency is especially pronounced for technically complex proposals, which are hardest for delegators to understand. The delegation system was originally designed to make governance participation accessible to people without specialized knowledge. However, the very premise that delegators lack specialized knowledge has resulted in a paradoxical reversal where metrics that should not be prioritized end up dominating.

The Structural Problems in the Delegation Game

So far, we have analyzed three structural problems in Cardano’s governance from the perspectives of game theory and economics.

Problem	Mechanism	Result
A: Competition prioritized over discussion	The zero-sum game trap	Competition is more rational than cooperation (Prisoner’s Dilemma)
B: Obstruction of cooperation	Cooperation erodes differentiation and asset value	Individual activity is more rational than collaboration
C: Marketing dominance	Information asymmetry and adverse selection	Marketing-oriented DReps have the advantage (Lemons problem)

The crucial point is that none of these are problems of individual morality or ability.

• Technical DReps are undervalued not because their quality is low
• Marketing-oriented DReps gain delegation not because they are doing something dishonest
• Delegators judge by “visible activities” not out of laziness

Everything is a product of the incentive structure created by the system. What makes this even more serious is that these problems reinforce each other.

1. The zero-sum game (A) generates competition
2. VP asset-ification (B) obstructs cooperation
3. Information asymmetry (C) promotes marketing focus
4. Marketing winners increase their VP → Further asset value growth
5. Technical DReps exit → Quality evaluation becomes even harder
6. A vicious cycle of governance degradation accelerates

The rules of the “delegation game” intertwine in complex ways, and individual rational decisions drive centralization.

Centralization Is a Result of the “Fallacy of Composition”

When rational decisions at the micro level (individual level) produce irrational outcomes at the macro level, economists call this the fallacy of composition.
Left unaddressed, this structure will continue to self-reinforce. Ultimately, a handful of marketing-oriented DReps will dominate the market, technical discussion will all but disappear, and governance will become a hollow formality. This would mean the collapse of Cardano’s foundational principle of “decentralization.”

The Governance Must Be Reformed

What the theoretical analysis so far has made clear is that the system itself needs to change.
In game theory, human behavior is shaped by “game design.” Appealing to individuals’ effort or morality will not solve the problem as long as the game’s rules and incentive structures remain unchanged.
In the next section, I will present concrete proposals for solving these structural problems.

Part 3: From Centralization to Decentralization

From here, I will present concrete proposals to address these structural problems. I propose a solution combining “delegation caps & delegation resets.”
First, I will examine why “DRep compensation” — a frequently discussed alternative — falls short as a solution, before diving into the details of my proposal.

Examining “DRep Compensation”

One of the most commonly discussed solutions in Cardano governance is a compensation system for DReps. The argument that “if we pay DReps, higher-quality activity will follow” appears attractive at first glance. However, my conclusion is that introducing a compensation system under the current governance structure would more likely worsen the situation rather than improve it.

Theoretical Merits of DRep Compensation

Theoretically, a DRep compensation system offers the following merits:

1. Increased resource investment in governance activity With compensation, DReps can devote more time and resources to governance — hiring specialists, using analytical tools, and conducting deeper research on technical proposals.
2. Incentivizing higher-quality activity If compensation can be linked to “quality,” DReps would be motivated to raise quality in pursuit of self-interest.
3. Lowering barriers to entry Currently, with no compensation, only those who can afford it participate as DReps. Compensation could broaden participation to include more talented individuals and people from diverse backgrounds.
4. Enhanced accountability Receiving compensation legitimizes demands for accountability from DReps.

These are all “theoretically correct” and appealing merits.

Two Problems on DRep Compensation

DRep compensation is a frequently debated topic in the Cardano community. The argument that “if we pay DReps, higher-quality activity will follow” seems attractive. However, I believe this proposal has problems from two perspectives: “Effectivity” and “Capability (implementation feasibility).”

Problem 1: Effectivity — Assumptions and Results May Not Align

The first concern about “effectivity” is the perspective that “even if you pay, there is no guarantee it will lead to better governance.”
Research in psychology and behavioral economics has repeatedly shown that the naive assumption “monetary rewards increase motivation” can empirically work in the opposite direction. This is known as Motivation Crowding Out, a theory demonstrated through multiple classic studies since the 1970s.

Titmuss’s blood donation study (1970) When monetary compensation was introduced for blood donation, willingness to donate actually declined. What had been “social contribution” was transformed into a “business transaction.”
Children’s drawing experiment (Lepper et al., 1973) Children promised a reward for drawing lost interest in drawing after receiving the reward.
Daycare late-pickup fine experiment (Gneezy & Rustichini, 2000) Introducing fines for parents who picked up their children late actually increased late pickups. What had been a “moral obligation” was transformed into “I paid, so it’s fine.”
What these studies demonstrate is that monetary compensation can fundamentally alter human motivation. Applying this to Cardano DReps:

Current DRep motivation:
  - Motivation: Contributing to governance, sense of responsibility to the ecosystem
  - Behavior: Careful voting, scrutiny of proposals, long-term perspective

After compensation introduction:
  - Motivation: Maximizing compensation
  - Behavior: Gaming the system, multiple DRep strategies, short-term profit seeking

Economics assumes that “higher rewards elicit better behavior.” However, actual human behavior follows the mechanism: “higher rewards → motivation shifts → quality declines.”
Monetary rewards replace intrinsic motivation — “I want to,” “because it’s right” — with extrinsic motivation — “because I get paid.” This causes the brain to reinterpret “why am I doing this?” and arrive at “for the money.” Academic research has repeatedly shown cases where monetary rewards extinguish the original motivation entirely.

Cardano’s governance was originally built on the premise that delegation to DReps would flow dynamically and thus remain decentralized. However, many holders have little interest in governance, and delegation has concentrated and barely flows at all.
Policies targeting human collectives don’t always produce the intended behavior; unexpected outcomes can actually worsen the situation. Changing the rules based on the assumption that “paying money will boost DRep motivation” is extremely dangerous. Thorough research and discussion of the risks triggered by such policies is necessary.

Problem 2: Capability — Three Major Challenges

Even if the Effectivity problem were solved, Cardano currently lacks the implementation capability to operate a compensation system in a healthy manner. Specifically, three capabilities are missing.

Capability 1: The Fiscal Design Problem

In permissionless governance, determining the compensation budget is an extremely difficult problem.

• Who decides the budget size?
• How do you handle fluctuations in the number of DReps?
• How do you address ADA price volatility?
• Is it possible to set a budget that balances a reliable amount for boosting DRep motivation with treasury sustainability?

For these reasons, I believe it is extremely difficult to set an appropriate compensation amount for DReps. DRep compensation would become a permanent funding obligation with significant impact on ADA price, and considering that current network fees amount to less than 10% of the treasury budget, it must be said that balancing sustainability with motivation improvement is essentially impossible.

Capability 2: The Evaluation Problem

To distribute compensation “fairly,” an objective method of measuring DRep contribution is needed. However, judging DRep activity by metrics alone does not reflect reality. While voting rate and Rationale submission rate offer some basis for judgment, determining whether someone is “voting carelessly” or “writing empty Rationales” is extremely difficult, and at minimum cannot be done on-chain. In other words:

• If vote count alone is the criterion, DReps who “just vote on everything” receive compensation
• If delegation amount is the criterion, DReps who are “good at marketing” monopolize compensation
• If Rationale submission is the criterion, meaningless text can be submitted and still count
• Setting a standard for how much compensation matches DRep effort is fundamentally difficult

As discussed in Part 2’s section on “information asymmetry,” since no means exists to objectively evaluate DRep “quality,” a compensation system will inevitably skew toward distribution based on “visible metrics,” potentially degrading governance quality further.

Capability 3: The Implementation Problem

Networks maintain their refined quality by constantly eliminating unnecessary programs. Given the premise that on-chain governance should be kept extremely simple, adding complexity to the voting algorithm would increase network load.

• Evaluating DRep voting rates and Rationale submission rates on-chain every epoch would impose excessive load
• “Objectively evaluating Rationale quality” is, to put it simply, impossible
• Placing evaluation logic off-chain would rapidly increase centralization risk

Simply providing DReps with compensation for their activity is relatively easy to implement programmatically. However, building a fair distribution system that accounts for the above three points would impose enormous on-chain burden, and would realistically require off-chain evaluation algorithms. This poses an extremely serious risk to the core principles of on-chain governance: decentralization and transparency.

Under Current Governance, Things Would Worsen Rather Than Improve

For a compensation system to deliver its benefits, there is an essential prerequisite: “compensation must be distributed appropriately, to the right people, for the right behavior.” Recall the governance structure analysis from earlier.
The current problems with Cardano’s governance structure were:

• A: Zero-sum game “delegation competition”
• B: Cooperation reduces asset value
• C: Quality cannot be evaluated; marketing trumps quality

What would happen if compensation were introduced under this structure? Let’s examine the scenario assuming “compensation is determined based on VP (Voting Power)” under the current governance structure.

Problem 1: Risk of compensation monopoly/oligopoly
Under the current oligopolistic structure (CR10=48%), top DReps would predictably monopolize most of the compensation. Higher compensation for top players would increase their visibility, financially reinforcing the “rich get richer” structure.

Problem 2: Harmful incentivization of wrong behaviors
With information asymmetry (C) in place, marketing activities are more likely to earn compensation while deep technical analysis remains unrewarded. Compensation flows toward “visible activities” rather than “high-quality activities,” with no fundamental change.

Problem 3: Surge of compensation-seeking DReps
Without a “mechanism to evaluate quality,” DReps whose sole purpose is compensation will proliferate. DReps who vote but do not contribute to discussion will increase in number, degrading overall quality.

Problem 4: Acceleration of the “rich get richer” Matthew Effect
Analyzing cexplorer data shows that DReps who are also SPOs significantly outperform DRep-only accounts in delegation amounts.

Category	Number of Accounts	Delegation Amount	Average Delegation
DRep-only	984	4.81B ADA	4.9M ADA
SPO + DRep	40	1.20B ADA	30.0M ADA

This trend is particularly strong among accounts that were SPOs before the DRep system was introduced (accounts that became SPOs after the DRep system tend to have lower SPO delegation).
Just 40 SPO/DRep dual entities capture 20% of all DRep delegation. SPO/DRep entities already earn block production rewards, and adding a compensation system would further strengthen their advantage.
Recall that in the current “delegation game,” marketing ability is the most important factor in gaining delegation.
With large DReps earning compensation, they could use those rewards for even more marketing, potentially accelerating delegation concentration.

In sociology, this situation — where those who have already succeeded gain better rewards and environments that further increase their probability of success — is called the Matthew Effect, and this impact can be predicted in governance as well.
In other words, a compensation system is a tool meant to increase the number of good DReps, but without a structure where good DReps are properly evaluated, it becomes counterproductive.
It is critically important to first fix the “war of all against all” delegation game structure, and only then introduce a compensation system — this sequence is decisive.

Two Measures to Improve Cardano’s Governance

From here, I will present the “delegation cap & delegation reset” solution I have been proposing. While I have been advocating for these measures consistently, the analysis above gives me renewed confidence in the effectiveness of implementing both simultaneously.

I am sharing my two-layered solution for governance (ver.2):
1️⃣ Hard Cap at 87M ADA (1.5% limit)
✅Structurally limits single DRep influence
✅Expands opportunities for new/small DReps

2️⃣Annual 1-2 Delegation Resets
✅Periodically raises governance awareness
✅Reduces passive… https://t.co/43T9PvSJbr

— hix Cardano-SPO |COFFE|（カルダノSPO） (@hix_coffeepool) March 26, 2026

Change 1: An 87M ADA Hard Cap (e.g., 1.5% limit)

Mechanism: Set a hard cap of 87M ADA (approximately 1.5% of total supply) on the delegation each DRep can receive (these figures are tentative and require further discussion). By setting a ceiling on voting power, any delegation exceeding the hard cap would automatically not be counted. For example, a DRep with a measured 3% voting power would only count as 1.5% in actual votes.

Expected effects:

1. Structurally limiting any single DRep’s influence
At the time of writing, extreme concentrations exist at Yoroi (11.96%), YUTA (8.32%), and EMURGO (5.11%), but the 87M ADA cap would automatically count these accounts’ voting power as just 1.5%. This is expected to substantially improve concentration metrics like CR5 and CR10.

2. Expanded opportunities for new and small-scale DReps
As top DReps reach their saturation point, new delegation would be expected to flow to other DReps. This would open participation opportunities for small-scale and new DReps who currently face extreme difficulty in gaining delegation.

3. A fundamental shift in competitive structure: From Voting Power to Discussion
With a maximum voting power threshold, governance influence would shift from “voting” alone to “discussion.” Since 1.5% voting power is insufficient to single-handedly achieve desired voting outcomes, DReps would rationally choose to persuade other DReps rather than accumulate more delegation.

Governance would transform from a game of “accumulating unlimited delegation” to a game of “persuading other DReps and achieving policy outcomes.”

Why a hard cap?

I also considered a soft cap (diminishing marginal returns model), but chose a hard cap for the following reasons:

• The limit is clear and easily understood by the entire community
• The specific figure of 87M ADA (1.5%) focuses the discussion
• When combined with the “annual delegation reset” described below, resistance to Sybil attacks is ensured

Change 2: Annual Delegation Reset

Mechanism: Delegation would automatically reset once or twice a year (set by epoch, with the specific period open for discussion). To continue with the same DRep, holders would need to actively re-delegate.

Expected effects:

1. Regular awareness-raising of governance
Delegators would be given periodic opportunities to consider “Is this DRep really good?” This would address the current “delegate once and forget” situation.

2. Reduction of passive delegation
According to economist Anthony Downs’s rational ignorance theory, when the expected benefit of voting falls below the cost of gathering information, remaining ignorant becomes rational. Under the current perpetual delegation system, this passivity entrenches delegation to DReps, solidifying the power base. A reset system would achieve a de facto transition to an “opt-in model” where only aware accounts participate in delegation.
Additionally, the introduction of hard caps would likely lead to multiple DRep account creation, but periodic delegation resets would make maintaining multiple accounts difficult. Having to solicit delegation for multiple accounts at each reset creates both a physical burden and a risk of ethical criticism, creating a psychological barrier to multi-account operation.

3. Accounts that do not re-delegate automatically move to Auto-Abstain
During the delegation reset period, accounts that do not re-delegate automatically move to Auto-Abstain. This stops forcing governance participation on accounts with no interest in governance.

4. Entry opportunities for new DReps
At reset time, all DReps become subject to re-evaluation. A structure where selection is based on current track record and quality — rather than existing brand power — is periodically created.

5. Turning the DRep re-delegation period into an “Annual Event”
By positioning the delegation reset period as an annual governance event, UX friction can be transformed into a positive engagement opportunity.

Specifically:

• Community-wide discussion: During the reset period, each DRep would share their past year’s track record, voting trends, and Rationale quality, with the entire community evaluating and discussing
• Debut opportunities for new DReps: New DReps could announce candidacy and platforms at reset time, creating a clear entry point
• Media and educational opportunities: Being recognized as an annual event would raise general interest in governance and create educational opportunities for new participants
• Cultural formation as a “festival”: Similar to election seasons, fostering a culture where the community periodically engages with governance as something personally relevant

This is not merely institutional design — it is an attempt to transform Cardano’s governance culture from “passive delegation” to “active participation.” An annual frequency strikes the right balance to sustain engagement as “enjoyment” rather than “obligation.”

The function of such “festivals” has been observed by literary figures like Kenzaburō Ōe and cultural anthropologists like Victor Turner. Festivals have played an important role in communities as ritualistic periods that temporarily dissolve the existing order and allow members to rebuild relationships on equal footing. An annual delegation reset period, similarly positioned as “a time to flatten and reorganize all of Cardano governance,” could become more than a mere administrative reset — it could serve as a cultural mechanism for revitalizing the entire community.

Ensuring Security During the Reset Period

During the delegation reset period, there is a risk that total governance voting power temporarily drops. As a safeguard against critical security issues during this period, the following mechanism would be established:

SPO and Constitutional Committee voting on emergency protocol upgrades

If a critical security issue arises during the delegation reset period, an “Emergency Protocol Upgrade” action would be established, where SPOs and the Constitutional Committee (CC) would temporarily exercise voting authority instead of DReps. This ensures that network security is not compromised during the reset period.
Regular governance actions (treasury withdrawals, constitutional amendments, etc.) would still require DRep votes, but this would function as a safety valve for emergencies.
Since SPOs have greater expertise in emergency protocol upgrades to begin with, omitting the DRep vote should pose no operational concerns.

Comparing DRep Compensation with Delegation Caps & Resets

Comparing the “DRep compensation system” and “delegation caps & delegation resets” discussed so far from the perspectives of Effectivity and Capability reveals clear structural differences between the two approaches.

Perspective	DRep Compensation System	Delegation Cap + Delegation Reset
Effectivity
Does it improve governance quality?	Low: Motivation Crowding Out destroys intrinsic motivation	High: Structural shift where “influence through discussion” becomes the optimal strategy
Impact on centralization	Worsens: Top DReps monopolize compensation, the rich get richer	Improves: Hard cap structurally enforces decentralization
Sybil attack resistance	Low: Compensation incentivizes multi-account operation	High: Combined with resets, sustaining multiple accounts becomes impractical
Anticipated side effects	Surge of “compensation-seeking DReps,” mass production of hollow Rationales	Temporary UX friction (mitigated through annual event approach)
Capability (Implementation Feasibility)
Fiscal design (Financial)	Difficult: Cannot balance permanent funding with motivation improvement	Not needed: No additional budget required
Evaluation design (Measurement)	Impossible: No means exist to objectively measure DRep “quality”	Not needed: Hard cap uses delegation amount, reset uses time — both mechanically determined
Implementation feasibility (Programmable)	Difficult: Requires off-chain evaluation, introducing centralization risk	Easy: Cap values and reset deadlines can be simply implemented on-chain

---

The DRep compensation system already has problems at the Effectivity (theoretical effectiveness) stage, and faces additional barriers at the Capability (implementation feasibility) stage. In other words, it is a measure burdened with problems on two fronts.
In contrast, “delegation caps & delegation resets” shows high feasibility on both Effectivity and Capability dimensions. Its decisive advantage in Capability is that “simple mechanical determination is sufficient.” It requires no complex evaluation logic and can be implemented while preserving the simplicity of on-chain governance.
This demonstrates that structural problems in governance should be solved through structural reform, not monetary incentives.

The Synergistic Effect of “Delegation Caps & Delegation Resets”

Implementing “delegation caps & delegation resets” as a package produces synergistic effects beyond what either measure achieves alone.

Effect A: Sybil Attack Resistance

If a hard cap were introduced alone, Sybil attacks through multiple accounts to circumvent the cap would be a concern. However, combined with annual resets, each account would need to continuously re-acquire delegation, raising the sustained cost of Sybil attacks to impractical levels. Together with the “ethical criticism from soliciting delegation for multiple accounts” mentioned earlier, multi-layered deterrence against Sybil attacks takes effect.

Effect B: Transforming the Game from “Competition” to “Discussion”

“Delegation caps & delegation resets” can be expected to drive this transformation from “competition to discussion-focused governance.” This is the most fundamental and positive change predicted by this proposal.

Current game: Voting power maximization competition

DRep's goal: Gain as much delegation as possible
↓
Optimal strategy: Marketing, differentiation, self-promotion
↓
Result: Zero-sum game, intensifying competition, declining discussion quality

After “delegation cap & delegation reset” implementation: Maximizing influence through discussion

DRep's goal: Achieve their desired governance outcomes
↓
Hard cap limits to 1.5%: Individual influence has a ceiling
↓
Optimal strategy: Persuade other DReps and lead consensus-building
↓
Result: Investment in discussion, cooperative behavior, quality improvement

The existence of a hard cap changes DReps’ behavioral principles themselves. Within a 1.5% cap, achieving one’s goals requires discussing and persuading other DReps. From a state where accumulating delegation had become the goal, the game’s rules shift to a state where achieving results through discussion becomes the goal.
This is closer to the ideal of “Liquid Democracy” that Cardano originally aimed for — a structure where the quality of discussion and the ability to persuade others are valued over the absolute amount of voting power.

Effect C: Creating Cooperation Incentives

The “Prisoner’s Dilemma” structure demonstrated in Part 2 would transform into a cooperation game through the implementation of delegation caps & delegation resets. Let me briefly explain what this means using game theory.

How to Read a Payoff Matrix

In game theory, the combinations of actions available to two players and the resulting benefits are shown in a “payoff matrix.” Here’s how to read it:

              Other DRep cooperates  Other DRep competes
Self cooperates      3, 3               1, 4
Self competes        4, 1               2, 2

• The numbers in each cell represent “(self’s payoff, other’s payoff)”
• Higher numbers mean greater benefit
• For example, the top-left “3, 3” means “when both cooperate, both receive a payoff of 3”
• The top-right “1, 4” means “when self cooperates and the other competes, self’s payoff is 1 and the other’s is 4”

Current State: Prisoner’s Dilemma Structure

              Other DRep cooperates  Other DRep competes
Self cooperates      3, 3               1, 4
Self competes        4, 1               2, 2

Reading this matrix reveals the following:

• Top-left (3, 3): If both cooperate, both achieve a good outcome (global optimum)
• Bottom-right (2, 2): If both compete, both achieve a moderate outcome
• Bottom-left (4, 1) / Top-right (1, 4): If only one defects, the defector gains significantly

Now consider the situation two DReps face:

If the other chooses “cooperate”:

• Cooperating gives a payoff of 3
• Competing gives a payoff of 4
• → Competing is better

If the other chooses “compete”:

• Cooperating gives a payoff of 1
• Competing gives a payoff of 2
• → Competing is better

Regardless of what the other chooses, competing is the rational choice for oneself. The same logic applies to the other party.
As a result, both choose to “compete,” and the Nash equilibrium (a stable state where neither party can benefit from changing strategy) settles at the bottom-right (2, 2). However, the top-left (3, 3) would actually be the more desirable outcome for everyone (Pareto optimal).
Cooperation would be better for everyone, yet competition prevails — this is the Prisoner’s Dilemma. Cardano’s current governance is in precisely this state.

Transformation to a Cooperation Game

Implementing “delegation caps & delegation resets” changes the environment surrounding DReps, and the payoff matrix shifts accordingly.
Let’s first examine why this change occurs from the structural side.

• Effect of the hard cap: Currently, the game is simply “whoever collects more delegation wins,” but setting a ceiling creates a physical limit on individual influence. With 1.5% voting power, a DRep cannot single-handedly achieve their desired governance outcomes. As a result, DReps are compelled to shift their behavioral principle from “collecting more delegation” to “cooperating with other DReps to achieve results”
• Effect of resets: Under the current perpetual delegation, once secured, voting power tends to become “fixed as an asset.” However, under a reset system, even if one reaches the cap, delegation periodically returns to zero, undermining the very premise of treating voting power as an “accumulated asset.” The incentive to keep growing delegation diminishes, and re-acquiring delegation after each reset requires ongoing participation in governance discussions. This is a transition from “stock-based evaluation” to “flow-based evaluation”
• Combined effect: With both caps and resets in place, the era where “the size of voting power” determines a DRep’s influence ends. To achieve desired governance outcomes, DReps must develop persuasive arguments within the community and build coalitions with other DReps. “Collaborating to achieve targeted outcomes” ultimately returns benefits to the individual DRep as well (influence on discussions, community reputation, continued delegation)

In other words, “the return from defecting” decreases, while “the return from cooperating” increases. Expressed as a payoff matrix, this becomes:

              Other DRep cooperates  Other DRep competes
Self cooperates      4, 4               3, 3
Self competes        3, 3               2, 2

Reading this change:

If the other chooses “cooperate”:

• Cooperating gives a payoff of 4
• Competing gives a payoff of 3
• → Cooperating is better

If the other chooses “compete”:

• Cooperating gives a payoff of 3
• Competing gives a payoff of 2
• → Cooperating is better

Regardless of what the other chooses, cooperation is the rational choice. The Nash equilibrium shifts to the top-left (4, 4), which also coincides with the Pareto optimal outcome. This means “cooperation becomes rational for the individual as well,” triggering a paradigm shift to a cooperation game.

Concrete Behavioral Changes

Let’s see how the theoretical changes translate into actual DRep behavior.

Current typical behavioral pattern:

DRep A: Independent analysis (effort: high) → Posts unique perspective on social media → Differentiation appeal
DRep B: Independent analysis (effort: high) → Posts unique perspective on social media → Differentiation appeal
→ Duplicated work, moderate discussion quality, both competing for delegation

Behavioral pattern after delegation cap & reset implementation:

DRep A & B: Share information early → Collaborate on deep analysis (effort: medium, quality: high)
→ Publish joint statements and joint analyses
→ Both are evaluated as "having contributed to high-quality discussion"
→ Influence is exercised through discussion, not delegation amount

In this way, delegation caps & delegation resets do not merely “prevent concentration” — they transform the very relationship between DReps from “rivals” to “collaborators.” I believe this brings us closer to the “governance through dialogue and deliberation” that Cardano originally envisioned.

Effect D: Qualitative Transformation of the Entire Ecosystem

Current state: Spiral of degradation
  Technical DReps exit → Discussion quality declines → Technical DReps become even more disadvantaged

After "delegation cap & delegation reset": Upward cycle
  Quality is valued → Technical DReps increase → Discussion quality improves → Further quality focus

The hard cap reduces the ROI of excessive marketing investment, while resets demand continuous proof of quality. With “influence through discussion” becoming the primary evaluation axis, the relative position of technically specialized DReps improves.

Risk Analysis and Mitigation

Of course, the “delegation cap & delegation reset” proposal is not a perfect solution. Finally, I will address anticipated criticisms and present mitigation strategies.

Criticism 1: Increased UX Burden Leading to Delegator Fatigue

Problem: Re-delegation once or twice a year may become burdensome, leading some delegators to disengage.

Mitigation:

• DRep performance dashboard: To support the decision of “should I continue delegating to this DRep,” provide a single-screen view of the past year’s voting history, Rationale list, and participated governance actions. Features like listing DReps whose Yes/No tendencies are closest to the delegator’s preferences. When decision-making materials are readily available, the psychological burden of re-delegation decreases
• Staged reminders before the reset deadline: Progressive reminders at one month, two weeks, and one week before the deadline would prevent “I didn’t realize the reset happened” scenarios. Integration with social media and major wallet apps would ensure notifications reliably reach holders
• Transforming friction into positive engagement through “Annual Events”: As described above, positioning the reset period not as a negative “obligation” but as a community-wide “festival” transforms UX friction into participatory enthusiasm

These mechanisms are technically implementable within the existing Cardano ecosystem, and I believe the benefits outweigh the implementation costs.

Criticism 2: Validity of Auto-Abstain Transition

Problem: Objections are anticipated to accounts that fail to re-delegate being automatically moved to Auto-Abstain. The concern is that “not voting becomes permanently fixed as a statement of intent.”

Mitigation:

This concern is somewhat overestimated given the current state of governance participation.

• Undelegated accounts already do not participate in governance: As shown in Part 1, approximately 34% of total staking is currently undelegated, and 38% is in Always Abstain. This means roughly 72% are already effectively not participating in governance. The Auto-Abstain transition merely makes this existing situation explicit
• Auto-Abstain is a legitimate choice explicitly expressing “non-participation in governance”: Forcing holders who have no interest in governance to participate formally actually degrades governance quality. Auto-Abstain should be respected as a valid option
• Stopping forced participation is actually healthier: A state where “delegation stays with whoever was first chosen indefinitely” does not reflect the delegator’s actual intent. The transition to Auto-Abstain is a mechanism for transforming passive participation into active choice
• Setting a sufficiently long re-delegation window: For example, providing a full month ensures that interested holders are not left out. This prevents important votes from being lost due to short-term missed opportunities
• As Cardano grows, an increase in casual users who simply “use ADA for everyday purposes” is expected. Governance participation would be a significant burden for casual users and could hinder Cardano’s expansion. After all, when purchasing dollars for an overseas trip, you don’t typically think about the U.S. presidential election

In other words, Auto-Abstain does not “exclude disinterested holders from governance” — it is a mechanism for “more accurately reflecting the intent of interested holders.”

Criticism 3: Security Degradation During the Reset Period

Problem: During the reset period, the number of active governance participants may temporarily fluctuate, potentially delaying emergency responses.

Mitigation:

• Emergency protocol upgrade mechanism with SPO and CC participation: As described above, the Constitutional Committee and SPOs would temporarily exercise voting authority during emergencies. The network security safeguard functions even during the reset period
• Strategic timing of the reset period: By fixing it as an annual event, governance action proposers can adjust their proposal timing to account for this period
• Compact reset period design: While the re-delegation window should be long enough, the “actual period of DRep fluctuation” itself should be designed to be short. For example, the re-delegation acceptance period could span one month, but the activation of the new delegation structure would occur synchronously, minimizing the practical impact on governance

Through these design choices, the annual “festival” character can be maintained while managing security risks.

Conclusion: Governance For Right Direction

Throughout this article, I have discussed the structural problems facing Cardano’s governance and proposed “delegation caps & delegation resets” as a solution.
At minimum, one thing is clear: if the current situation is left unchecked, centralization will continue to advance. Both the worsening numbers shown in Part 1 and the game-theoretic structures analyzed in Part 2 point to a “self-reinforcing vicious cycle.” This is a system-level problem that cannot be solved by individual effort or moral appeals.
What is needed is an efficient, implementable concrete measure. This article has compared two options: “introducing DRep compensation” and “delegation caps & delegation resets.”

A compensation system would only introduce new risks. Motivation Crowding Out altering motivations, the difficulties of fiscal design, evaluation design, and implementation — all of these would merely pile new problems on top of structural ones rather than solving them.
In contrast, “delegation caps & delegation resets” is an approach that applies minimal patches to existing rules. It is simple, mechanically deterministic, and requires neither additional budgets nor evaluation algorithms. Yet it has the potential to transform the game’s rules from “voting power maximization competition” to “influence through discussion competition.”

Governance is not about technology — it is about the community itself. More important than technical sophistication is the nature of relationships participants build as they discuss and reach consensus. Laying that foundation is the essence of governance improvement.

However, a single mistake in rule design can send human behavior in precisely the opposite direction. Humans create rules, but once created, rules govern humans. Unless flawed rules are correctly revised, the community will march straight down the wrong path. The Cardano community stands at this crossroads.

Democracy Is Still “Incomplete”

Thousands of years have passed since democracy first emerged in ancient Greece. Democracy has been adopted to prevent the concentration of power in the hands of a few, to reflect diverse voices, and to mitigate harm to the collective from decisions made by a minority.
All of these are mechanisms to minimize conflict within communities arising from decision-making. Can we say that democracy is functioning in today’s Cardano community?

Throughout history, humanity has experimented with monarchy, dictatorship, oligarchy, theocracy, and many other forms of governance. Each time, abuse of power, corruption of privileged classes, and violent regime changes repeated themselves. “Transferring power without bloodshed” — a challenge that seems obvious yet is actually extremely difficult — and democracy represents humanity’s hard-won achievement through long trial and error.
Even so, democracy is not a completed system. Rather, it is an unfinished system that always teeters on the edge of fragility.

In the 18th century, the French philosopher Montesquieu recognized the danger of power concentrating in a single place and advocated for the “separation of powers” into legislative, executive, and judicial branches. This forms the foundation of democratic “checks and balances.” Yet even in many countries that adopted this framework, the abuse of power has not been completely prevented. Political interference in the judiciary, executive overreach, and legislative dysfunction are issues debated even in today’s advanced democracies.

Looking back at history, no country has ever “completely succeeded” at democracy. Every democratic nation is perpetually in a state of trial and error. In fact, history offers numerous examples of dictators rising through democratic elections. The dictator Adolf Hitler emerged from the Weimar Republic, the most advanced democratic system of its time.

British statesman Winston Churchill is said to have stated:

“Democracy is the worst form of Government except for all those other forms that have been tried from time to time.”

Democracy is not perfect. But it is better than every other form of governance humanity has tried. This was Churchill’s wryly ironic conclusion. And this very recognition is the starting point for improving and maintaining democracy.

Liquid Democracy Is Still Very Much Incomplete

The Liquid Democracy adopted by Cardano is an extension of this long history.
The prototype of “Liquid Democracy” is attributed to Charles Dodgson, a 19th-century British mathematician (later known as Lewis Carroll, author of “Alice’s Adventures in Wonderland”), who conceived the “Single Transferable Vote.” It is a system where votes can be delegated to others and that delegation can be changed as needed. This idea, conceived over 150 years ago, was impossible to implement with the social infrastructure of the time, when postal mail was the only means of communication. However, modern blockchain technology has finally brought it within reach of practical implementation.
In other words, liquid democracy is a system whose experiments have only just begun. Far from being perfected, it is a form of governance that no one has ever concretely tried at this level. Cardano is the cutting-edge laboratory challenging this uncharted territory. That is precisely why the structural distortions identified in this article are emerging for the first time in the world.

Democracy collapses the moment people stop thinking critically. “We should comply because the system decided it,” “We can leave it to the experts,” “Someone else will think about it even if I don’t” — the moment such thoughts take hold, the system silently begins to decay. This is the same structure that allowed Nazi Germany to establish a dictatorship with only about 30% of the popular vote.

The same is true for Cardano’s governance. The moment we think “it’s Liquid Democracy, so it must be decentralized” or “the concentration ratio was determined by the rules, so it’s fine,” we begin walking the path toward corruption. The numbers presented in Part 1 of this article clearly show the warning signs.

Keep Doubting, Keep Verifying!

In the crypto industry, the most fundamental mindset is “Trust no one.” This is not cynicism. Rather, it is the most honest attitude for building truly trustworthy systems.
Don’t blindly trust authority. Always question, always verify. Don’t take proposals at face value — think for yourself. I hope readers will approach the arguments in this article with that same attitude.
I don’t believe my proposal is the only right answer. However, I am strongly convinced in the diagnosis that “centralization will progress under the status quo” and the conclusion that “structural reform is necessary.” Concrete solutions require discussion and verification by the entire community.
Whether Cardano becomes the first successful liquid democracy in human history, or follows the same path of hollowing out and centralization as past democratic experiments — we stand at that fork in the road right now.

Keep doubting, keep verifying.

The only way to keep democracy alive is for each and every one of us to never stop thinking.

(End)