A Better World Cannot Depend on Everyone Becoming Wise

I would prefer a world in which more people understood systems, followed evidence, recognized manipulation and thought past the next headline. I would prefer politicians who did not want the status, money, attention or power attached to the office.

But that is not a governing model. That is a wish.

A functioning civilization cannot assume that every citizen will become informed, every candidate will become decent, every expert will become neutral or every institution will correct itself. The system has to survive ignorance, vanity, fear, tribalism, propaganda, greed and ordinary human error because those things are not temporary bugs in the species.

Modern electoral government makes a basic category error. It treats the ability to acquire power as evidence that a person should exercise power. It confuses campaign skill with judgment, attention with competence, confidence with understanding and popularity with legitimacy.

The answer is not to find one perfect leader. It is to build an architecture in which no leader can become permanent.

I call the larger principle High-Entropy Governance. I call the operating mechanism Qualified Civic Rotation. The national-scale implementation is the 90-Day State.

“Entropy” is an institutional analogy here, not a claim about thermodynamics. A high-entropy political system keeps the identity of future decision-makers uncertain, breaks repeat relationships, rotates authority rapidly and prevents control from staying attached to the same human nodes. The laws and the operating institutions remain stable. The people holding political authorization do not.

Elections Do Not Select “the Best.” They Select the People Best Adapted to Elections.

The ballot is the final stage of a much longer filter.

Before a voter sees a candidate, somebody had to want the job. The candidate had to tolerate public exposure, raise money, build alliances, survive party selection, compress complex questions into slogans, repeat themselves for months, attack competitors and persuade strangers that they should possess authority.

Some capable and ethical people can do that. The important point is that none of those activities is the same as governing well.

A 2025 European Journal of Political Research study provides one empirical reason to take this selection problem seriously. Across Canada, Denmark, Israel, the Netherlands and Switzerland, the researchers found that higher honesty-humility was consistently associated with lower political ambition and recruitment. They explicitly cautioned that the study was associational, not causal. It still tells us that “who steps forward” is not a neutral sample of the population.^[1]

The title of this article—no one who wants power should have it—is therefore a design rule, not a claim that every person who seeks office is evil. Wanting power should not be the admission ticket to power.

Back to top

High-Entropy Governance, Operated Through Qualified Civic Rotation

The framework separates four things that modern politics keeps collapsing into the same office:

1. Representation: who is present when choices are made.
2. Competence: whether those people can understand the process, evidence and consequences.
3. Authorization: who can make a binding public decision.
4. Operations: who has the technical knowledge to execute that decision over time.

Elections try to bundle all four into a candidate. The candidate claims to represent millions of people, understand every domain, possess authority and supervise the permanent state. That is too much abstraction attached to one person.

Qualified Civic Rotation unbundles the machine.

This is neither pure Athenian sortition nor ordinary technocracy. It is a civic lottery after a broadly accessible competency floor, connected to professional operations and wrapped in public audit.

Back to top

Nine Non-Negotiable Principles

1. No self-nomination for sovereign office. Nobody files to run. Nobody asks the public to give them power.
2. Competence before chance. Random selection occurs only inside a large pool that has demonstrated a public minimum capability.
3. The competency floor must remain broadly reachable. It certifies readiness to learn and deliberate; it does not create a priesthood of degrees, IQ scores or approved opinions.
4. Representation is engineered, not assumed. The draw uses transparent stratification so the assembly resembles the population across relevant dimensions.
5. Service is temporary, paid and presumptively mandatory. Civic duty cannot be restricted to people wealthy enough to volunteer.
6. Power never repeats. One completed public-authority term removes a person permanently from future governing draws.
7. All influence leaves a record. Lobbying moves from private access to a public evidence docket.
8. Memory and authority stay separate. Professionals preserve capability; citizens retain the final mandate.
9. The system is allowed to fail small before it fails large. Pilots, sunset clauses, outcome metrics and automatic review come before sovereign replacement.

These principles are intentionally harder than “randomly pick a parliament.” Sortition is a selection method. It is not a complete constitution.

Back to top

The Qualification Gate Is Necessary—and It Is the Part Most Likely to Become Tyrannical

If the lottery includes every adult without preparation, a selected body can be overwhelmed by basic process, manipulated by experts or forced to rely on permanent staff for every interpretation. If the qualification threshold is controlled by elites, the system becomes an epistocracy: a credentialed class deciding who is allowed near power.

That is the central design tension.

The United States has already demonstrated the danger. Literacy and knowledge tests were used to exclude Black citizens and poor citizens from political participation. The Voting Rights Act of 1965 outlawed those discriminatory practices.^[2] Any proposal that contains the phrase “minimum criteria” has to confront that history directly.

The answer is not to hide the gate. It is to make the gate a public commons with constitutional restraints.

The certification should test six trainable capabilities

• Rights and process: the constitutional floor, due process, equal protection, separation of functions and limits on public authority.
• Quantitative scale: percentages, rates, orders of magnitude, base rates, budgets, debt, uncertainty and the difference between stocks and flows.
• Evidence reasoning: source quality, correlation versus causation, model assumptions, conflicts, missing data and falsifiability.
• Systems reasoning: incentives, feedback loops, second-order effects, bottlenecks, time lags and unintended consequences.
• Deliberative procedure: listening, questioning, steelmanning, recording dissent and separating factual disagreement from value disagreement.
• Ethics and recusal: recognizing personal conflicts, gifts, indirect interests, coercion, confidential information and the duty to disclose.

It should not test party identity, religious belief, ideology, wealth, formal education, occupation, cultural vocabulary or agreement with the existing government.

The gate needs its own anti-capture constitution

• The full curriculum, scoring rubric and sample bank are public.
• Preparation is free, multilingual, available online and in person, and designed for disability access.
• People may retake the certification without a lifetime penalty.
• Multiple equivalent demonstrations of competence are available; one language style or school background cannot be the only route.
• No sitting assembly, party, ministry or private contractor can change the standard alone.
• Rule changes require approval by an independent court, a randomly selected qualification-review jury and a public notice period.
• Pass rates are audited by geography, income, race, sex, disability, language and education. Unexplained disparate effects trigger automatic suspension and review.
• Every failed applicant receives the scored reason, the underlying rule and a fast appeal.
• The target is a large qualified public—eventually a clear majority of adults—not a narrow elite.

The certification is not a claim that qualified people will always be right. It is a floor beneath a structured learning process. The Evidence Service still has to teach each issue because no general test can make someone an expert in agriculture, semiconductors, epidemiology, water systems, defense procurement and monetary policy.

Back to top

The Draw Must Be More Transparent Than an Election Count

Random selection is trustworthy only if nobody can quietly shape the pool, the strata, the substitutions or the random number.

The jury system already gives us a familiar institutional analogy. Federal courts randomly select qualified citizens from source lists and expand those sources when necessary to pursue a representative cross-section.^[9] Modern public-randomness systems add a cryptographic layer. NIST describes signed, time-stamped random pulses, public access to past outputs, chained verification and methods for combining multiple independent beacons. NIST explicitly identifies publicly verifiable random sampling as a design goal.^[8]

A valid civic draw needs seven public objects

1. The eligible population count. The public can see how many people exist in each relevant stratum without exposing individual private data.
2. The qualification rule. No invisible exclusions or discretionary invitations.
3. The stratification plan. Geography, age, sex, income, education, work status, disability, household structure and urban/rural status can be balanced as relevant to the jurisdiction.
4. The precommitted algorithm. Code, version hash, substitution rules and draw date are frozen before the random seed exists.
5. Independent entropy. Several public beacons and physical public randomness sources are combined so no single operator controls the result.
6. The live draw record. Observers can reproduce the draw from the public inputs and verify the output.
7. The replacement log. Every deferral, disqualification, recusal and replacement is recorded with a reason code and audited for patterns.

Party affiliation should not be a quota. That would preserve the parties inside the machine. But the selection process does need to monitor broad attitudinal balance without creating a permanent ideological database. One method is a blinded, one-time policy-values instrument stored separately from identity and used only to prevent a cohort from becoming an obvious attitudinal outlier.

Ireland’s citizens’ assemblies demonstrate a real-world version of stratified random selection. Its current process aims to mirror the public across gender, age, geography, socioeconomic status and nationality, while generally excluding elected politicians and registered lobbyists.^[4] The 90-Day State takes that selection logic much further, but it does not invent the basic mechanism from nothing.

Back to top

Short Enough to Break Capture. Staggered Enough to Preserve Continuity.

Ninety days is not a number handed down by history. It is a working design choice built around a trade-off.

If the term is too long, the citizen becomes a politician. Donor networks, personal staff, factional identities, press relationships and future-career incentives begin to form. If the term is too short, members never get beyond orientation and become dependent on the permanent bureaucracy.

The solution is not simply “90 days.” It is 90 days plus staggered cohorts plus 30 days of paid onboarding before authority begins.

National working specification

• 600 Assembly members. Large enough to reduce sampling error and distribute work.
• Three cohorts of 200. One cohort enters and one exits every 30 days.
• 30-day paid onboarding. Selected citizens learn procedure, rights, budget architecture, security obligations and the current public docket before they vote.
• 90 days of voting authority. The term cannot be extended by the Assembly, an emergency declaration or a court.
• One public-authority term per lifetime. After service, the person is permanently removed from every governing draw.
• No future political conversion. Service cannot become a credential for another governing office, a party nomination or a repeat appointment.

At any moment, one third of the Assembly is in its first month, one third in its second and one third in its third. Two thirds have more experience than the incoming cohort. Nobody has enough time to become permanent.

Staggering also changes the economics of capture. An outside actor cannot buy one election and work with the same office for years. It would have to identify an unpredictable population, influence hundreds of people under public-contact rules and start again every month.

That does not make bribery impossible. It raises its cost, reduces the value of any single relationship and shortens the useful life of a compromised node.

Back to top

No President, No Permanent Legislature, No Single Chokepoint

The 90-Day State is not a legislature pasted on top of the current executive. If one person still controls appointments, emergency orders, security access, procurement and the public narrative, the system still has a throne.

1. The Rotating Civic Assembly

The 600-person Assembly controls statutes, taxation, the budget, national objectives, treaty approval, war authorization, senior oversight and the legal limits of executive action. Most work occurs in randomly assigned issue panels of 45 to 75 members. Panel recommendations return to the full Assembly for binding votes.

2. The Civic Executive Council

There is no single president. A 15-person Executive Council is drawn by lot from the sitting Assembly for 30-day internal shifts. It coordinates lawful execution, receives operational briefings and resolves cross-agency conflicts. It cannot legislate, alter appropriations, extend its own authority or issue secret standing law.

3. The Independent Evidence Service

Every major docket receives a public evidence record built by competing teams:

• a primary analytical team,
• an adversarial red team,
• a distributional-impact team,
• and an uncertainty and forecast audit.

Sources, models, assumptions, confidence ranges, conflicts and data gaps are public by default. Experts can advise but cannot vote. Artificial intelligence may organize evidence, trace claims, translate material and run simulations, but it cannot decide policy or conceal the provenance of an answer.

4. The Professional Operations Service

Roads, benefits, air traffic, disease surveillance, defense logistics, energy regulation, tax systems and procurement cannot relearn their jobs every 90 days. Operations therefore remain in a professional, merit-based civil service.

That continuity is a feature, but it is also a capture risk. GAO has repeatedly emphasized institutional knowledge, expertise and continuity in public operations.^[13] The Operations Service preserves those assets under three restraints: it cannot create an independent political mandate; every material rule must trace to an enacted statute and public delegation; and audit juries can examine whether administrative implementation changed the policy in practice.

5. Random Audit Juries

Audit juries are drawn separately from the Assembly. They do not write policy. They inspect the selection engine, contact logs, assets, conflicts, procurement, classified processes, implementation drift and outcome reports. Their narrow mandate makes shorter service possible and reduces dependence on a permanent political opposition.

6. Independent Courts

Courts protect the universal civic floor and enforce the boundaries between institutions. Judges are a deliberate exception to the 90-day rule because adjudication requires technical training and continuity. They would be randomly selected from a professionally qualified judicial pool for one long, nonrenewable term, with random case assignment, published ethics records and no political reappointment.

7. Local, Regional, National and Global Layers

The model scales through jurisdiction rather than one planetary super-assembly. Local decisions stay local. Regional infrastructure moves to the regional layer. National sovereignty remains national. Cross-border systems—climate, oceans, pandemics, trade standards, orbital resources—can use a global civic chamber drawn from national qualified pools, with population weighting and minimum regional floors so neither tiny states nor giant states can dominate alone.

The point is broad representation without creating one permanent global ruling class.

Back to top

A Public Decision Should Leave an Evidence Trail, a Reason Trail and an Outcome Trail

Most citizens see a political decision at the moment it becomes a headline. The 90-Day State treats a decision as a traceable pipeline.

1. Issue intake. A matter enters through a citizen petition threshold, mandatory statutory review, audit finding, budget calendar, court order, agency request or emergency trigger.
2. Scope lock. The Assembly publishes the exact question, legal authority, decision deadline and measurable objectives. Scope cannot quietly expand after evidence gathering begins.
3. Evidence build. The Evidence Service publishes the primary case, red-team case, distributional effects, alternatives, uncertainty and missing information.
4. Open submissions. Any person, company, union, university, agency or advocacy group may submit evidence and arguments to the public docket. Paid interests and beneficial owners must be disclosed.
5. Public hearings. Witnesses answer questions under recorded conditions. Members can request additional evidence. Experts appear in competing panels where material disagreement exists.
6. Structured deliberation. Members work in small randomly composed groups with rotating facilitators, equal speaking windows and recorded questions—not recorded private deliberation.
7. Preliminary vote. The result, member rationales, minority report and unresolved factual disputes are published.
8. Challenge window. The public, auditors and affected institutions receive a short period to identify factual errors, undisclosed conflicts or implementation failures.
9. Final vote. Ordinary legislation uses a majority or a defined supermajority according to scope. War, constitutional change, rights restrictions and very large long-term obligations require two independently drawn bodies and a higher threshold.
10. Execution map. The Operations Service publishes responsible agencies, milestones, budget, dependencies, forecast outcomes and sunset or review dates.
11. Outcome audit. Results are compared with the original forecast. Failure is not buried inside a new narrative.

The OECD’s review of nearly 300 representative deliberative processes identified several of the same requirements: a clearly defined task, representative selection, inclusion support, access to diverse evidence, adequate deliberation time, arm’s-length coordination, transparency and official follow-through.^[3]

The 90-Day State converts those principles from a temporary consultation into a standing decision architecture.

Back to top

If Nobody Can Be Reelected, What Makes Them Answerable?

This is the first serious objection to the framework.

Elections provide a rough accountability mechanism: perform badly and voters can remove you. In practice, incumbency, party identity, donor support, information asymmetry and safe districts weaken that mechanism. But removing reelection still removes something real. A replacement system has to do more than say, “Trust the random people.”

The 90-Day State replaces delayed popularity accountability with continuous process accountability.

• Every vote is attributable. Members’ final votes, written rationales and disclosed conflicts are public.
• Every policy has a forecast. The decision record states what is expected to happen, by when, at what cost and with what uncertainty.
• Every material contact is logged. There is no private lobbying exception for friends, donors, employers, unions, companies, activists or government agencies.
• Every delegation is traceable. Administrative rules must cite the statute and the precise authority being exercised.
• Every removal has due process. A member can be suspended for bribery, undisclosed conflict, incapacity, coercion, security breach or procedural misconduct—not because a faction dislikes the member’s vote.
• Every removal is judged independently. A separately drawn review jury, advised by a court, determines whether the defined standard was violated.
• Every major law returns. Statutes carry sunset, review or outcome-audit dates unless the Assembly publicly explains why permanence is necessary.

The public still has direct constitutional functions. Any citizen can petition an issue onto the intake system after a published threshold; submit evidence; challenge conflicts; seek judicial review; and inspect the public decision record. Constitutional amendments, transfers of sovereignty and a narrow class of irreversible decisions require approval by two independent civic assemblies followed by universal public ratification.

That final ratification is a vote on a defined constitutional act, not a candidate election. It preserves a direct consent valve without rebuilding a permanent campaign industry.

Back to top

Random Selection Breaks the Campaign Market. It Does Not Break the Influence Market.

Sortition severs the direct link between money and winning office. It does not make citizen legislators immune to money after they arrive.

Recent scholarship makes the danger explicit: randomly selected lawmakers could be bombarded by sophisticated lobbying, and short terms could make them more dependent on civil servants. The same research argues that large numbers, short overlapping terms and strict outside-contact rules can raise the cost of capture, but none of that is automatic.^[11]

The framework therefore treats capture as a network problem. The attacker looks for predictable targets, private channels, repeat relationships, concentrated chokepoints and weak verification. The defense removes or exposes each one.

The minimum anti-capture stack

• Private policy lobbying is prohibited. Arguments and evidence enter through the public docket. Incidental contact is reported within 24 hours.
• Beneficial ownership follows the submission. A shell organization cannot appear as an independent citizen group while hiding the entity funding it.
• Assets and material interests are disclosed and verified. Verification matters more than collecting forms.
• Issue-specific recusals are automatic. A farmer can still serve in government, but cannot quietly vote on a measure that uniquely transfers value to a company they own.
• Members receive strong pay and complete participation support. Job protection, housing, travel, childcare, eldercare and disability accommodation are public infrastructure, not perks.
• Post-service monetization is restricted. The working specification uses a ten-year ban on paid lobbying, influence work or compensated representation before bodies affected by the member’s service.
• Unexplained wealth and conflict indicators trigger independent review. Review is evidence-based and includes due process; it is not permission for political surveillance.
• Influence networks are analyzed across cohorts. Repeated organizations, intermediaries, experts and staff connections become auditable patterns rather than isolated meetings.

The OECD’s 2026 integrity outlook shows why a law on paper is not enough. Across countries, conflict-of-interest rules met an average 85% of regulatory criteria, while actual practice met 42%—a 43-point implementation gap driven in part by weak verification and noncompliance.^[12]

A rule nobody verifies is decorative.

Back to top

War, Emergencies, Courts and Secrets Are Where Constitutions Reveal What They Really Are

Any political design looks clean under normal conditions. The real test is whether “temporary necessity” creates a permanent ruler.

Emergency power

An emergency directive requires three independent keys: a supermajority of the Civic Executive Council, approval from an emergency court panel and confirmation by a separately drawn Emergency Audit Jury. Initial authority expires after seven days. Renewal requires a new recorded showing of necessity and a vote by an Assembly cohort that did not authorize the first period.

Core due-process protections, the ban on term extension, the audit record and judicial access do not disappear during the emergency.

War and military command

The professional chain of command remains continuous. Tactical command cannot rotate every month. Political authority over objectives, geographic scope, funding and continuation does rotate.

Except in an immediate defense scenario defined by law, initiating or materially expanding armed conflict requires approval from two independently drawn civic bodies. Continuing authority expires on a short statutory clock unless renewed. Classified operational detail may remain protected, but the legal basis, objective, cost range and review date remain public.

Classified information

Selected members can receive temporary clearances after screening and training. A randomly selected classified-audit panel inspects withheld evidence. Every classified decision produces a public summary at the time and an automatic declassification review later. “National security” cannot become a permanent blank field in the decision record.

Courts

Judges do not rotate every 90 days. They require deep legal competence, precedent continuity and insulation from the Assembly. The model uses one long, nonrenewable term drawn from a qualified judicial pool, random case assignment and independent ethics enforcement. There is no campaign and no reappointment incentive.

Public health and infrastructure shocks

Agencies operate preauthorized emergency protocols with objective triggers. Those protocols are written in normal time, audited before use and bounded by automatic expiration. Crisis action should be fast because the rails already exist—not because one person becomes law.

Back to top

How the 90-Day State Fails

A framework is not serious until it can describe the machine that destroys it.

The strongest failure may be the interaction between two individually reasonable design choices: short terms and professional continuity. Short terms reduce political capture but increase knowledge dependence. Permanent staff provide knowledge but create a stable target for capture. The answer is not to choose one side. It is to keep the interface adversarial, visible and auditable.

Back to top

What Has Been Demonstrated—and What Remains a Hypothesis

What the record supports

• Representative civic selection is operationally possible. Ireland uses stratified random recruitment for national citizens’ assemblies.^[4]
• Ordinary citizens can deliberate over complex and polarizing issues after structured learning. The OECD catalogued nearly 300 representative deliberative processes through 2020, and Stanford’s America in One Room assembled a representative, stratified sample of hundreds of voters for informed deliberation.^[3][5]
• Sortition can be institutionalized rather than used once. Ostbelgien created a permanent citizens’ council and recurring citizen assemblies by law.^[6]
• Full legislative sortition has a developed scholarly literature. Contemporary work examines lottocracy, randomly selected legislative chambers, hybrid chambers and permanent mini-publics.^[7]
• Publicly verifiable random selection is technically possible. Randomness beacons and reproducible algorithms can make manipulation harder to hide.^[8]

What the record does not prove

• That a sovereign national government can replace candidate elections without a legitimacy crisis.
• That 90 days is enough time for citizens to legislate effectively across the full national docket.
• That a competency certification can remain fair, broad and politically neutral over decades.
• That lifetime non-repeatability improves decisions more than it destroys accumulated civic expertise.
• That the anti-capture stack will outperform political parties, investigative media, electoral opposition and existing constitutional checks under real attack.
• That a rotating executive can manage war, intelligence and fast-moving crises better than a conventional executive.

There is also a direct representativeness warning. A 2025 pre-registered experiment involving 1,308 people found that mini-publics could increase perceived legitimacy, but minor composition biases cut those gains substantially and larger biases could erase them. Small samples and nonresponse are not clerical problems. They strike at legitimacy itself.^[10]

This is why the model uses a large chamber, strong participation support, presumptive service, public composition targets and staged adoption. It is also why I will not pretend the proposal is “proven.”

Back to top

Do Not Seize a Country to Test a Governance Theory

A framework meant to reduce concentrated power cannot be imposed through concentrated power. The transition has to be constitutional, voluntary, measurable and reversible at every early stage.

Phase 0 — Open design and simulation

Publish the qualification model, draw code, cohort design, evidence protocol, conflict rules and decision simulations. Run historical cases through mock assemblies. Let critics attack the system before it touches law.

Phase 1 — Local advisory trials

Cities and counties convene qualified, paid and stratified panels on defined issues. Compare their knowledge gains, recommendations, costs, participation patterns and public legitimacy with existing hearings and commissions.

Phase 2 — Binding audit and oversight

Random civic juries receive narrow binding powers over procurement review, conflict enforcement, project postmortems and agency implementation audits. Oversight is a safer first use of binding authority than general legislation.

Phase 3 — A randomly selected second chamber

A state, province or willing national jurisdiction creates a sortition chamber with defined review, delay, amendment or veto powers alongside the elected chamber. The two systems operate against the same policy environment, making comparison possible.

Phase 4 — Distributed executive authorization

Specific appointment, emergency-review and cross-agency authorities move from a single executive to a rotating civic council under strict sunsets. Professional operations remain unchanged.

Phase 5 — Sovereign replacement referendum

Only after multi-year evidence shows that the new system is more representative, harder to capture, operationally competent and publicly legitimate should a jurisdiction consider replacing candidate elections for the legislature and executive authorization layer. That final change requires universal ratification and a constitutional reversion mechanism.

Gradualism is not timidity here. It is how a systems proposal proves it understands failure.

Back to top

The Framework Must Be Able to Lose

A bad political theory explains every failure as sabotage and every success as proof. A usable framework precommits to the evidence that would stop it.

The trial stops, rolls back or redesigns if the gate becomes discriminatory, the bureaucracy becomes the real legislature, capture simply migrates, decision quality deteriorates or public legitimacy fails to survive actual losses.

No constitutional framework deserves immunity from its own evidence standard.

Back to top

The Unit of Analysis Is Not the Politician. It Is the Power Network.

The normal political argument stays trapped at the visible layer.

Is this candidate good?

Is that candidate bad?

Which party deserves the next turn?

The Pattern Nexus framework asks a different set of questions: What are the rails? Where are the chokepoints? Which incentives repeat? Who controls information? Where can private influence enter? What becomes more valuable as time passes? Which relationships compound?^[14]

From that perspective, the individual politician is one node inside a much larger selection and reinforcement system:

• ambition supplies candidates,
• parties filter candidates,
• capital funds access,
• media allocates attention,
• campaigns compress reality,
• elections grant a term,
• incumbency strengthens the node,
• staff and lobbyists preserve memory,
• and reelection restarts the loop.

The 90-Day State attacks the loop instead of begging the node to become virtuous.

It removes self-nomination. It removes campaigns. It removes fundraising. It removes repeat terms. It makes the target population large and unpredictable. It keeps the evidence trail public. It separates memory from mandate. It forces power to decay before a durable personal network can form.

We do not need every person to become brilliant before the world can improve.

We need institutions that stop turning ignorance, vanity and ambition into permanent leverage.

That is the entire point.

FAQ

Research and Precedents

Research cutoff: August 24, 2026. The institutional specifications in this article are the proposed High-Entropy Governance framework. Sources document evidence, precedents, critiques and implementation constraints; they do not constitute proof of the complete 90-Day State.

1. European Journal of Political Research — Too Honest and Humble to Run for Office? Citizens’ Personality Traits, Nascent Ambition, and Recruitment
2. U.S. National Archives — Voting Rights Act (1965)
3. OECD — Innovative Citizen Participation and New Democratic Institutions: Catching the Deliberative Wave
4. Ireland Citizens’ Assembly — FAQ, Eligibility and Stratified Random Selection
5. Stanford Deliberative Democracy Lab — America in One Room
6. Bertelsmann Stiftung — The Ostbelgien Model: Institutionalising Deliberative Democracy
7. Cambridge University Press — Political Lotteries and Democracy
8. NIST — Interoperable Randomness Beacons
9. United States Courts — Juror Selection Process
10. British Journal of Political Science — Mini-Publics, (Lack of) Representativeness, and Legitimacy Beliefs
11. Perspectives on Politics — Constituency Juries: Holding Elected Representatives Accountable through Sortition
12. OECD — Anti-Corruption and Integrity Outlook 2026: Conflict of Interest
13. U.S. Government Accountability Office — Civil Service: Agency Responses and Perspectives on Former Schedule F Executive Order; see also GAO guidance on institutional knowledge, expertise and leadership continuity
14. Pattern Nexus — Pattern Nexus Framework: How to See the Hidden Architecture of the Modern World; related: The Pattern Nexus Framework: How Modern Systems Actually Work