What the demand for “one of our own” reveals about political power
Abstract
This is the first article in my newly started Awate column, The Hidden Premise. I examine the conception of power beneath debates about representation: why should a community need “one of its own” in office to remain safe? I distinguish recognition, participation, a share of authority, and enforceable protection. These distinctions suggest why we can repeatedly discuss the same political questions while leaving their underlying assumptions untouched. Representation matters, but citizenship requires protection that people can invoke regardless of who governs them, including against those who claim to represent them
Among the many questions Eritreans are trying to think through about a future beyond the PFDJ, representation is one of the most persistent. Who should participate in shaping the transition? How should political offices be distributed? How can Muslims, Christians, ethnic communities, regions, women, refugees, former fighters, and younger Eritreans be assured that the next government will not reproduce earlier patterns of exclusion?
These are not abstract concerns. Eritrean history gives different communities legitimate reasons to worry about who will define the nation, control its institutions, distribute its resources, and decide which grievances deserve recognition. Any discussion of a future political order must therefore take representation seriously.
The discussion on Awate reflects this concern. The Eritrean Covenant1 documented Muslim exclusion and demanded a fairer distribution of political and economic power. Amanuel Hidrat2 has connected domination to concentrated government and proposed decentralization and institutional representation for Eritrea’s social groups. Ismael Ibraheem Al-Mukhtar3 has shown how numerical representation can be deceptive when apparent representatives serve themselves, occupy nominal positions, or remain subordinate to the real centers of power. Younis Omer4 has questioned elites who invoke constituencies they may have no mandate to represent.
While following these discussions, I began to wonder whether the word representation was being asked to carry several different political expectations. Are we asking for recognition, participation, a fair share of authority, or protection?
These aims can overlap, but they are not identical. The difference becomes especially important when we say that a community needs “one of its own” inside the state. That demand does not merely say that government should resemble the country it governs. It suggests that the safety of people outside the room may depend on the identity of those inside it.
This raises a question that precedes the debate over how representation should be arranged: What kind of state makes the identity of an officeholder feel necessary for a community’s survival?
Imagine Eritreans gathered to design a government after the PFDJ. The composition of the room would immediately become political. How many Muslims are present? How many Christians? Which regions and ethnic communities are missing? Who speaks for the Lowlands? Who represents refugees, women, former fighters, and the younger generation?
These are reasonable concerns. After a history of domination and exclusion, no community should be asked to accept another government whose national character exists only in its name. A visibly narrow government would reproduce old suspicions before writing its first law.
But suppose we solve the problem perfectly. Every significant community is present, and the major offices have been distributed with great care. Yet Muslims still depend on Muslim officials for protection, Christians on Christian officials, and ethnic communities on their respective representatives. Have we established democratic government, or merely negotiated a system through which each community can protect itself from the government?
My argument is not simply that representatives can betray those they supposedly represent. That problem is already well understood. I am asking what the protective demand for representation reveals about our underlying conception of power.
The Assumption Beneath the Demand
Suppose Muslims insist that Muslims must occupy important state offices. The demand becomes protective only if the identity of the officeholder is expected to influence how he uses authority. He is expected to recognize injuries that a Christian might overlook, defend interests another official might neglect, or resist decisions directed against people he considers his own. Christians, ethnic minorities, regions, and other populations can reason in exactly the same way.
But notice what has already been assumed. We expect an official to possess enough discretion to favor, neglect, protect, or injure. Because we cannot reliably control that discretion, we search for someone whose sympathies might incline him toward us.
Where that is the only protection available, shared identity becomes a substitute for political control.
This is not really a question of trusting the right person. A democratic state should not require me to trust the person holding an office, including someone from my own community. I should be able to distrust him and still remain protected because his authority is limited, his decisions require reasons, his conduct is visible, and what he does can be challenged.
When those mechanisms are absent, resemblance begins to feel like protection. Someone like me may use arbitrary power more gently against me.
At that point, we are no longer asking only for recognition or participation. We are bargaining for survival.
Dividing Power Without Transforming It
We often criticize dictatorship while leaving its conception of power largely untouched. We reject the dictator but continue to imagine the state as a commanding position from which whoever reaches the top distributes resources, recognizes communities, and determines whose grievances matter.
The political struggle then becomes a struggle for proximity to that position. If power is going to remain kingly, every community reasonably wants someone near the king. Better still, every community wants a portion of the kingship.
Religious or ethnic power-sharing may reduce the danger of complete exclusion. During a fragile transition, that reassurance may even be necessary. But we should be precise about what it accomplishes. It distributes access to power. It does not necessarily transform the nature of that power.
The stronger argument for power-sharing is that communities need the means to restrain one another, not merely someone sympathetic in office. That is a serious argument. But it brings us back to the same question: what is being restrained? Leaders may prevent one another from monopolizing the state while leaving the citizen dependent on his own communal guardian. Can he challenge that guardian? Can he seek protection without his permission? A balance among leaders is not yet control by the people they claim to represent.
If Muslims require Muslim officials to protect them, while Christians and ethnic minorities require their own officials for the same reason, government continues to operate through competing partialities. Each community sends someone expected to watch the others and intervene on its behalf. The state does not become impartial. Its partiality is divided among several guardians.
This arrangement also creates gatekeepers. If Muslims require representation as Muslims, someone must decide who speaks for Muslims. If every ethnic community is entitled to a share, someone must define its interests, select its representatives, negotiate its portion, and manage internal disagreement.
The representative then acquires an interest in keeping the community politically consolidated. The community, in turn, is told that consolidation is essential to its security. Internal disagreement begins to look like a threat because it weakens the bloc. Citizens learn to ask whether their group has secured enough offices before asking what anyone occupying those offices is legally allowed to do.
A system designed to protect communities may therefore protect communal gatekeepers more reliably than ordinary people.
This is what I mean by a survivalist state. I do not mean merely a poor, unstable, or transitional country. I mean a political order in which communities seek safety through proximity to power because citizens lack reliable protection from power.
Such a government may visibly include every Eritrean identity and still fail to establish Eritrean citizenship.
Is This Democratic Naivety?
I can anticipate the objection. This argument may sound too idealistic. Institutions are weak, constitutions are violated, judges can be captured, and officials do not suddenly become impartial because the law tells them to. Communities that have experienced domination cannot be asked to entrust their survival to democratic promises. That, one might say, is simply not how the world works.
But I am not asking communities to trust promises or assume that officials will become virtuous. I am questioning the supposedly realistic alternative.
What assumptions does communal protection require? It assumes that a Muslim official will protect Muslims because he is Muslim, that a Christian official will protect Christians because he is Christian, and that ethnic representatives will remain accountable to the populations whose names give them authority. It assumes that communal elites can accurately define internally diverse interests, that balancing several partialities will produce security, and that representatives who acquire power through communal fear will not benefit from preserving that fear.
Why should these assumptions be considered realistic?
A shared identity does not discipline power. It does not require an official to explain his decisions, reveal how resources were distributed, submit to review, or compensate those he has harmed. It may create sympathy, but sympathy is not an enforceable political protection. A representative can neglect, exploit, or repress members of his own community while continuing to speak in its name.
The appeal to citizenship is not an appeal to goodwill. Citizenship is not a sentimental request that Eritreans forget their histories and see one another simply as human beings. Nor is it a declaration that ethnicity or religion no longer matters. It is a legal and political position from which a person can confront public authority directly.
A citizen protected by democratic institutions does not merely hope that an official will treat him fairly. He can demand the rule under which a decision was made, require the government to justify its action, examine the evidence, challenge unequal treatment, and seek a remedy from an institution that does not depend on the official’s communal loyalty.
Those mechanisms can fail. Courts can be corrupted, laws manipulated, and oversight bodies captured. But their possible failure is an argument for making them stronger, not for replacing them with identity. Communal representation can fail as well, except that when it fails, the ordinary person may have no independent claim against his own supposed representative.
The question is therefore not whether institutions are perfect. Nothing in politics is. The question is which political arrangement gives the citizen greater capacity to confront failure.
But who keeps those institutions independent? Certainly not the word “independent.” No faction should be able to appoint the judges, control their funding, and decide whether to obey them. Citizens, lawyers, journalists, opposition parties, and communal organizations need the freedom and means to expose violations and organize against them. Representation can help build that power. The point is to make the protection usable even against one’s own representative. Otherwise, we have simply moved the question of trust from one office to another.
Those who dismiss citizenship as naive should also explain their own theory of protection. Why should an ordinary Muslim be safer because a Muslim elite occupies an office he cannot control? Why should an ordinary Christian trust a Christian official whose decisions he cannot inspect? How does a community remove a representative who invokes its identity but does not answer to its members? What protects dissenters within the community when the gatekeeper claims to speak for everyone?
Calling institutional citizenship idealistic does not answer these questions. It merely treats the inherited arbitrariness of power as realism.
Democracy begins from a different commitment: public power belongs to citizens before it belongs to those who temporarily occupy office. An election authorizes officials to perform limited functions. It does not transfer ownership of the state to the winning community. Losing an election should mean losing an office, not losing one’s security.
If a community believes it must control part of the state because otherwise it may not survive the state, then the political order has not yet solved the democratic problem. It has organized a truce among groups competing over dangerous power.
What Citizenship Requires
The alternative is not to dismiss the fears behind representation. It is to translate each fear into a restriction on public power and a claim that citizens can enforce.
If the fear is that one region will capture public resources, allocations must follow published criteria that account for population, deprivation, infrastructure, and historical disadvantage. The calculations and final distributions must be open to scrutiny, and departures from the rules must be challengeable.
If the fear is land dispossession, customary and communal ownership must be legally protected. Government decisions must require stated reasons, and affected people must have access to an independent institution capable of stopping unlawful action.
If the fear is discrimination in education or public employment, the criteria, procedures, and outcomes must be visible and reviewable.
If the fear is cultural disappearance, language, religion, association, and communal continuity must be protected from ordinary political majorities.
These arrangements do not ask officials to become better people. They reduce the number of decisions that depend on personal virtue, sympathy, or communal loyalty.
Nor does citizenship require the state to ignore collective injuries. Some harms are patterned across communities, and some goods are necessarily collective. Communal land, language, religious freedom, refugee return, and cultural continuity may all require explicit constitutional protection. A community should be able to identify an injury, organize around it, and compel the state to recognize its pattern.
But protecting collective goods does not require converting communities into permanent political corporations. The injured citizen must retain a claim that requires no communal intermediary: I am an Eritrean citizen. What rule permits the state to treat me this way, and where can I challenge it?
Descriptive representation still has value. Different experiences inside government can reveal neglected realities, expose exclusion, and help prevent the capture of institutions. But what does that presence allow the citizen to do? If he still needs someone from his community to intervene before the law will hear him, we have secured him a broker. We have not yet secured his citizenship.
Can Institutions Come Later?
A more practical objection is that courts, administrative law, auditing systems, and constitutional protections do not appear overnight. First, the argument goes, let us guarantee representation and prevent monopolization. We can build impersonal institutions once the transition becomes secure.
I do not think these tasks can be separated so easily.
A survival arrangement is already an institutional arrangement. It distributes offices, recognizes particular communal categories, authorizes people to speak in their names, and teaches citizens where protection comes from. These are not temporary habits existing outside the future state. They become the material from which that state is built.
Once an office is understood as a community’s protection, surrendering it later will look like surrendering the community’s security. Each group will fear abandoning its guarantee while others retain theirs. The mistrust that justified the arrangement will then prevent its removal.
Lebanon illustrates this difficulty. Its confessional allocation of offices was intended to preserve coexistence among communities that feared domination. But a political order organized around communal survival also gives communal leaders a permanent reason to preserve the divisions through which they receive authority. As Al-Mukhtar3 observed, endorsing Osman Sabbe3’s rejection of the Lebanese model for Eritrea, constitutionally distributing offices by sect may be preferable to outright domination, but it is hardly the political horizon we should choose.
“First representation, then institutions” is therefore not a neutral sequence. The first stage creates the interests and political habits from which the second stage would have to emerge. It trains communities to monitor one another rather than jointly discipline public power.
This does not mean that a dangerous transition can ignore power-sharing or corrective inclusion. Safeguards may be necessary to prevent immediate capture and reassure communities that have good reasons to fear exclusion. But limitations on power must begin at the same time as its distribution. Temporary guarantees must have defined purposes, procedures, limits, and conditions for revision. Otherwise, transitional protectors become permanent custodians.
Beyond Survival
The question is not whether representation is good or bad. It is what we expect representation to accomplish and what kind of political order that expectation creates.
If we want someone like us in office because only he can protect us from those unlike us, then we have accepted too much about the authority he will inherit. We have accepted that power may remain arbitrary and tried to survive that arbitrariness by securing a share of it.
That may be an understandable response to the state Eritrea has been. It cannot be the design of the state Eritrea is trying to become.
The democratic task is not to find a balanced collection of people whom different communities can trust with dangerous power. It is to make that power answerable to citizens regardless of who possesses it. It is to create a political order in which losing an election does not place a community’s security in question and no citizen must inspect the religion, ethnicity, or ancestry of an official before knowing whether the law will hear him.
A country is not built when every community acquires someone through whom it can survive the state. It is built when the citizen can stand before the state without communal sponsorship.
No Eritrean community should have to capture part of the state in order to be safe from the whole of it.
References
- Mejlis Ibrahim Mukhtar. “The Eritrean Covenant: Towards Sustainable Justice And Peace.” Awate, February 12, 2010 (document dated February 11, 2010).
Read on Awate - Amanuel Hidrat. “HardTalk: Group Rights And Identity Politics.” Awate, October 22, 2019.
Read on Awate - Ismael Ibraheem Al-Mukhtar. “The Challenges of Fair Representation in Eritrean Politics.” Awate, August 24, 2019. See “Deceptive representations” and “Remedies”; the latter discusses and endorses Osman Sabbe’s rejection of the Lebanese model for Eritrea.
Read on Awate - Younis Omer. “OUR NATIONAL UNITY: Why Eritrea’s Political Imagination Fails Reality.” Awate, April 4, 2026. See “VI. The Opposition’s Structural Paradox.”
Read on Awate
Comments