What is a model for direct citizen governance?
Directa begins with a question: if people must live with the consequences of public decisions, how much authority should they retain over making, examining, and correcting those decisions?
Its answer is a civic decision order in which citizens directly determine the major rules and commitments that bind their society. Administration remains necessary. Expertise remains necessary. Institutions remain necessary. What changes is the relationship between those functions and the people they serve: a professional political class no longer receives a general mandate to decide on society's behalf.
The purpose is to make public authority specific, visible, limited, and answerable to those subject to it.
This is a proposal to be tested and developed, not a claim that a complete, failure-proof democracy has already been invented. Its ambition is substantial precisely because the difficulties are substantial. A model that promises to prevent domination must examine where domination could arise inside its own design.
The word “minimum” refers to a foundation: equal political standing, fundamental rights, accessible participation, inspectable decisions, and accountable execution. “Global” means that these principles should be available to people wherever they live. It does not mean that one worldwide majority should govern every community, or that every society must adopt an identical institutional form.
The question is how people can govern their shared affairs without having to surrender the continuing ownership of those affairs.
Authority Without a Permanent Owner
Delegation creates a distance between the person who authorizes a decision and the person who exercises that authority. Their interests may align, but they may also diverge (the principal-agent problem). A public official may sincerely serve the public while also responding to career incentives, organizational loyalty, donors, professional networks, or the pressure to preserve an institution.
The problem cannot be reduced to the character of individual officeholders. Equally, it would be careless to assume that every representative is corrupt or that representation has achieved nothing. The structural question remains even when the people involved have good intentions: what happens when citizens can replace the decision-maker more easily than they can change the decision-making arrangement?
Directa would reserve major legislative choices, public priorities, and consequential resource commitments for direct public authorization. Administrators would receive defined responsibilities rather than an open-ended political mandate. Their discretion would have boundaries, their decisions would require reasons, and their performance would be subject to review.
This does not make administration apolitical. Choosing an implementation sequence, interpreting an ambiguous instruction, or deciding which delay is tolerable can distribute benefits and burdens. Calling these choices “technical” does not remove their consequences.
A credible decision order must therefore distinguish between decisions that implement an authorized policy and decisions that substantially change it. The former can belong to professional administration within published limits. The latter must return to the appropriate public decision process.
No population can vote on every operational detail. Nor should a hospital wait for a public ballot before replacing broken equipment. Direct control means retaining authority over purposes, boundaries, significant trade-offs, and correction. It does not mean turning ordinary administration into a permanent referendum.
The ambition is to prevent delegated responsibility from quietly becoming ownership of the public's future.
Who Decides What We Are Allowed to Decide?
A vote occurs near the end of a longer process. Before anyone chooses an answer, someone has defined the problem, selected the options, established the timetable, and decided which information deserves attention.
That earlier power can be as consequential as the ballot itself (agenda-setting).
If citizens may answer questions but cannot meaningfully introduce them, their sovereignty is incomplete.
Directa's public agenda, the Gündem, would begin with citizen proposals. People would be able to submit an issue, explain its relevance, obtain assistance with drafting, and seek support for its examination. They would not need a party nomination, a sponsor, or the resources to commission a professional campaign.
Support thresholds would help manage the volume of proposals, but a threshold must not become a wall that only organized wealth can climb. Its design would need public justification, periodic review, and additional routes for matters affecting small or excluded populations. A threat to a minority's basic rights must not wait until that minority becomes numerically influential.
Supporting the examination of a proposal would also be distinct from supporting its adoption. “We should discuss this” and “we should do this” are different political acts.
Proposal visibility would require rules of its own. A supposedly open platform could still privilege those who purchase attention, publish constantly, or provoke the strongest reactions. Directa's civic infrastructure should not rank public concerns by their ability to keep people scrolling. It should provide transparent discovery, opportunities for less visible proposals to be considered, and explanations when submissions are merged, delayed, or rejected.
Those procedural decisions would be appealable. An agenda office with the power to make inconvenient proposals disappear would become a legislature without admitting it.
Even the wording of a ballot deserves scrutiny. “Should we protect our children?” conceals more than it explains. “Should this specified measure be adopted, at this cost, under these conditions?” permits an actual decision.
What kind of freedom is being offered if the only available answers have already been arranged to produce the preferred result?
A Shared Basis for Disagreement
Citizens cannot make meaningful decisions without information. Yet no institution can guarantee neutrality simply by announcing that it is neutral.
Choosing which facts to include, which time horizon to use, and which effects to measure already involves judgment (framing). A proposal may appear beneficial when evaluated through total output and harmful when evaluated through how its burdens are distributed.
Directa would therefore require public decision materials that make their own construction inspectable.
Each significant proposal would have an accessible information packet explaining its purpose, alternatives, expected costs, affected groups, implementation requirements, and possible consequences. It would distinguish established findings from forecasts, assumptions, and unresolved questions. Keeping the current arrangement would be examined as an option with consequences of its own.
The packet would present the strongest relevant arguments on different sides. This does not require giving every assertion equal weight. An unsupported claim should not acquire credibility merely because someone demands “balance.” Evidence should be assessed by its quality, while genuine uncertainty and disagreement should remain visible.
Artificial intelligence could help compare documents, locate inconsistencies, translate material, examine scenarios, and identify questions requiring investigation. Multiple systems could challenge one another's outputs. However, agreement among models would not count as independent confirmation when they share sources, assumptions, or failure patterns.
AI would serve as an analytical instrument. It would not become the official owner of reality.
Human specialists would review the evidence and methods. People directly affected by a proposal would contribute knowledge that professional categories can overlook. A transport model may estimate average travel time while missing what a route change means to someone who cannot walk to the replacement stop.
The academic “joker” function would be to expose omissions: what has not been measured, which comparison is misleading, whose experience has disappeared, and what could invalidate the forecast? Expertise would help the public see the decision more clearly without acquiring the right to make it in the public's place.
Appointments, funding, conflicts of interest, and methods of selection would be disclosed. Review roles would rotate, and dissenting assessments would remain available.
Citizens could challenge a packet's adequacy and request corrections. But a vote approving the packet would not certify its contents as true. Public acceptance cannot transform weak evidence into strong evidence. Disputes about material errors would require an independent review process, with published reasons and defined deadlines.
The aim is a common basis for examining disagreement. Agreement itself cannot be manufactured as a condition of participation.
Participation Requires More Than Access
Giving everyone an account does not give everyone an equal opportunity to participate.
A person working unpredictable shifts, a parent providing unpaid care, and a financially secure professional may possess identical voting rights while having very different amounts of time to exercise them (participatory inequality). If this difference is ignored, formal equality can conceal a concentration of influence among those most able to remain continuously involved.
Directa would require practical support: accessible formats, translation, assistance that protects privacy, offline access, predictable schedules, and time to examine consequential decisions. Paid civic leave and care support should be considered part of participation infrastructure.
Deliberation also needs space. People should be able to question a proposal, revise a view, and admit uncertainty without being punished socially for failing to defend a permanent team identity.
The OECD's 2020 study, Innovative Citizen Participation and New Democratic Institutions: Catching the Deliberative Wave, examined 289 deliberative cases. These experiences provide material for learning about how citizens can examine public questions under structured conditions. They do not demonstrate that an entire society can replace all representative institutions with Directa.
A useful component would be temporary citizen panels selected through a transparent process designed to include different parts of the population. Such panels could hear evidence, question specialists, and publish competing recommendations. Their selection, non-participation, and limitations would be disclosed. They would inform the electorate rather than replace it.
Voting frequency should respect human attention. Stable decision windows, clear priorities, and appropriate local responsibility would reduce unnecessary demands. Urgency would require reasons rather than become a permanent excuse to shorten scrutiny.
Citizens would remain free to abstain. Abstention could reflect uncertainty, limited time, satisfaction with a process, or rejection of the options. It should not automatically be interpreted as consent.
Quorum and approval rules would be established in advance, with attention to their own vulnerabilities. A high threshold can protect against narrow decisions while also rewarding strategic boycotts. A low threshold can enable action while allowing a small mobilized group to bind everyone else. No single percentage resolves that tension for every kind of decision.
A democracy should not make exhaustion the price of remaining politically equal.
Privacy Must Exist Beyond the Screen
A digital identity can help establish who is participating. It cannot establish whether that person is free to choose.
Someone voting at home may be watched by a partner, pressured by an employer, or instructed by a person on whom they depend. A biometric match can authenticate a body while revealing nothing about the conditions under which the choice was made.
Directa must therefore abandon the assumption that home-based biometric voting automatically eliminates coercion.
Registration, eligibility verification, ballot secrecy, accurate counting, accessibility, and resistance to intimidation are separate requirements. Each needs its own safeguards.
Identity verification should be separated from the recorded choice. The system should minimize retained personal information and provide alternatives for people who cannot reliably use a particular authentication method. A failed biometric reading must not become the loss of a political right.
Protected voting spaces would remain necessary, including accessible arrangements for people unable to travel. Private assistance must be available without turning the assistant into the decision-maker.
Digital systems may support participation, but binding voting arrangements would require independent security evaluation, meaningful auditability, and recovery procedures. No participant should have to trust a vendor's assurance that an invisible process worked correctly.
The National Academies' 2018 report, Securing the Vote: Protecting American Democracy, emphasized human-readable paper ballots and warned against internet transmission of marked ballots under the conditions it assessed. Its relevance here is methodological: convenience does not establish security, and a digital record needs a trustworthy basis for verification.
Directa should retain independently auditable physical evidence and usable fallback procedures for binding votes. Any proposed remote-voting method would carry the burden of demonstrating how it addresses coercion, compromised devices, secrecy, and verification.
A voter should be able to check that participation was properly recorded without receiving a transferable proof that enables someone else to purchase or demand a particular vote.
Disputed results, outages, and suspected interference would follow rules published before the vote. A malfunction must not allow whichever institution controls the machinery to invent the remedy afterward.
The ballot is where citizens should become hardest to identify and the counting process should become easiest to scrutinize.
Influence, Manipulation, and the Right to Disagree
Public reasoning necessarily influences people. A moving testimony, a persuasive argument, and a warning about foreseeable harm can all change a person's decision without violating their freedom.
An anti-manipulation framework that prohibits influence itself would prohibit much of democratic life.
Directa would instead target identifiable abuses: intimidation, vote buying, impersonation, concealed sponsorship, fabricated official communications, and covert operations that misrepresent who is speaking or how support was obtained.
Restrictions should attach to conduct and evidence. They should not depend on whether an opinion is inconvenient, emotional, unpopular, or opposed to Directa.
Claims of deliberate deception would require careful examination of intent, material harm, and the difference between false statements of fact and disputed interpretations. Errors should be correctable without automatically becoming crimes. Allegations would receive independent review and due process; the same body that publishes an official explanation should not possess unchecked power to punish those who challenge it.
Official civic information would face demanding standards of accuracy and disclosure. Citizens, journalists, researchers, and advocacy groups would retain the freedom to criticize that information and campaign openly for different outcomes.
Directa would give political parties no reserved role as gateways to public decisions or as holders of a general governing mandate. Yet people would remain free to associate around shared convictions. Suppressing association would contradict the freedom the model seeks to protect. Organized groups would instead face transparent, enforceable rules concerning funding, access, conflicts of interest, and attempts to dominate civic infrastructure.
A wealthy organization may call itself a foundation rather than a party. A charismatic network may exercise discipline without formal membership. Regulation must examine how influence operates, not merely what an organization calls itself.
The relevant question is not whether citizens have been influenced. It is whether they can recognize, examine, and resist that influence without losing their rights or their livelihood.
The Majority Does Not Own the Minority
A decision does not become just solely because many people support it.
Directa requires a foundation of equal citizenship, bodily security, freedom of conscience and expression, due process, and protection against discrimination. Participation itself must remain protected. A temporary majority must not be able to remove another group's standing as an equal participant.
These protections cannot be implemented by a slogan about “untouchable rights.” Rights can conflict, and applying them requires interpretation.
Independent adjudication would therefore remain necessary. Its mandate would be to explain and enforce the constitutional boundaries of public power. Review bodies would themselves require transparent appointment procedures, conflict rules, reasoned decisions, and meaningful avenues of appeal.
This gives them real authority. Directa should acknowledge that authority and constrain it rather than hide it behind the language of technical administration.
Constitutional development would remain possible through a more demanding process than ordinary policy change, including extended deliberation and rights review. But the capacity to amend the rules must not become a convenient route to stripping people of equal legal standing.
Accommodations and exemptions could sometimes protect different ways of living. They would need to respect the rights of everyone affected, including vulnerable people within the requesting group. A community's wish to govern itself cannot automatically authorize it to coerce its own dissenters.
Consider a public infrastructure project that benefits most residents while displacing a small neighborhood. Counting votes answers which option attracts more support. It does not answer whether displacement is necessary, whether a less harmful alternative exists, or what obligations are owed to those carrying the cost.
Those questions must enter the decision before the majority celebrates its preference.
The ethical test of collective authority is partly whether those who lose a vote can still recognize themselves as equal members of the society that held it.
Execution, Oversight, and the Power to Interrupt
Every approved policy would need an execution mandate: its purpose, budget, responsible bodies, permitted discretion, reporting duties, review dates, and conditions for reconsideration.
Public authorization should be specific enough to constrain abuse without pretending that every future circumstance can be predicted.
Administrators encountering a material change would be required to disclose it. They could not quietly redefine a policy while continuing to invoke the original vote as permission.
Independent audit bodies would inspect performance, spending, compliance, and conflicts of interest. Citizens would have routes to request investigation, while safeguards would discourage harassment and repetitive complaints designed to paralyze ordinary work.
Auditors may need power to request a temporary pause when serious harm or misconduct is credibly suspected. That is an exercise of authority, even when an automated rule carries it out.
Such power would require a defined scope, evidence requirements, strict time limits, and prompt independent review. Essential services would need continuity arrangements. Freezing the resources of a disputed project must not accidentally cut off drinking water or emergency treatment.
The authority to raise an alarm and the authority to make the final determination should be separated. Those accused would have an opportunity to respond. Confirmed wrongdoing would have consequences; an unsupported accusation would not become a permanent punishment.
Public reporting would protect legitimate personal privacy. Transparency about institutions must not become unlimited exposure of the individuals who depend on them.
Accountability means being able to discover what happened, require an explanation, correct the harm, and impose appropriate consequences. Publishing a large quantity of documents is only one part of that task.
A Decision Must Remain Connected to Its Consequences
Democratic decisions can be mistaken. Direct participation does not remove uncertainty, prejudice, or the possibility of unintended harm.
Directa would therefore treat evaluation as part of authorization.
Before adoption, a significant proposal should state what it is trying to achieve, which effects will be monitored, who will assess them, and what would justify revision or termination. Assessments should examine the distribution of consequences, not only average results.
A policy that reduces total costs while making a vulnerable group's life substantially worse cannot be adequately described by its savings alone.
Suppose a town needs to renew its water network. Residents propose the project. The information process compares repair strategies, costs, disruption, financing, and risks. Affected neighborhoods challenge an assumption about access. The proposal changes. Citizens authorize a bounded plan. Administrators implement it. Auditors discover an unexpected procurement problem. A limited intervention protects the service while the issue is examined. Material changes return for public authorization.
Afterward, the town compares the promised improvements with actual reliability, affordability, and disruption.
This is the decision order in practice: a continuing relationship between a public purpose, a mandate, its execution, and the lives affected.
Some consequences cannot be reversed. Decisions involving permanent displacement, environmental destruction, or long-lived infrastructure therefore require stronger scrutiny before adoption. The availability of a future vote does not repair every form of harm.
A society should become better at correcting itself without treating people as expendable material for experimentation.
Beginning Without Demanding a Leap of Faith
Directa should earn wider responsibility through limited, lawful, observable trials.
A community might begin with a bounded area of participatory budgeting, a citizen-initiated policy review, or a defined local decision process. The trial should specify its authority, resources, duration, safeguards, and evaluation criteria before it begins.
Success would include more than turnout or satisfaction among enthusiastic participants. Evaluation should examine who could participate, who remained absent, whether decisions were implemented, whether burdens were fairly distributed, and whether complaints produced effective remedies.
Trials must also reveal failure. A process captured by a highly organized minority, inaccessible to caregivers, or dependent on one technology provider would supply reasons for redesign rather than excuses for expansion.
Any transition from existing institutions would require legal continuity and a clear allocation of responsibility. Public services cannot be suspended while a new theory proves itself.
Expansion should follow evidence and renewed public authorization. There should be a way to pause, revise, or reject an arrangement that performs poorly. A civic order built around public self-correction cannot declare its own adoption irreversible.
Nor should all decisions move upward. Local matters should ordinarily remain at the lowest level capable of addressing them responsibly (subsidiarity). Decisions with cross-border consequences require wider coordination and enforceable obligations. This includes situations in which those affected cannot vote in the jurisdiction creating the harm, as well as the interests of children and future generations.
No existing electorate contains everyone who may bear the consequences of its choices. Recognizing this limitation is part of responsible sovereignty.
The Decision Order Must Also Be Open to Decision
Directa would not end conflict. It would not make citizens equally informed, erase ambition, or prevent every abuse. A different distribution of formal authority does not automatically dissolve differences in wealth, status, knowledge, and social influence.
Its value would depend on whether people gain a more effective ability to shape the conditions of their lives while preserving one another's freedom.
That requires institutions strong enough to carry out shared decisions and limited enough to remain answerable. It requires expertise that can be challenged, technology that can be inspected, majorities that accept boundaries, and dissenters who retain their place in the public world.
It also requires a different relationship to losing. A decision reached through a fair process can still be painful, mistaken, or worth contesting. Accepting a result does not require surrendering the right to argue for its revision.
The political culture Directa seeks is one in which changing one's mind need not mean betraying one's identity, and questioning an institution need not mean rejecting collective life.
Perhaps the deepest change would be this: public decisions would cease to appear as events that happen somewhere above ordinary people. They would become responsibilities whose reasons, limits, and consequences people could follow.
But the final question must return to Directa itself.
If it became difficult to question, if its experts became a caste, if its technology became an authority, or if its defenders treated criticism as disloyalty, would citizens still possess the means to change it?
They must.
The population's continuing authority over its shared life is the purpose. Directa is a proposed way of serving that purpose, and it must remain subject to the very scrutiny it asks every other institution to accept.