The Constitution of Distrust

How a founding document wrote a prisoner’s dilemma into a nation

By E. Frashie  ·  The Ethiopian Tribune

If the trap has an Ethiopian birthplace, it is not 2018. It is 1995 — the year the Constitution of the Federal Democratic Republic of Ethiopia came into force — and to see why, one must read that founding document not as a charter of rights but, in Curtis’s manner, as a model of human nature written into law. To read it so, we need first to meet the model itself, and to follow it back to the strange and paranoid room in which it was made.

The model is the suspicious actor: the isolated, rational, self-interested individual who assumes the worst of everyone around him and plans accordingly. Its formal origins lie in game theory, set out by the mathematician John von Neumann and the economist Oskar Morgenstern in Theory of Games and Economic Behavior (1944), which recast human conduct as the strategic calculation of players each pursuing advantage. At the RAND Corporation, the Pentagon’s Cold War think-tank, this became the mathematics of nuclear survival. It was there, around 1950, that Merrill Flood and Melvin Dresher devised the scenario the mathematician Albert Tucker would name the prisoner’s dilemma: two rational actors, unable to trust each other, each betraying the other and so arriving together at an outcome worse than cooperation would have produced. And it was there that the young mathematician John Nash formalised the equilibrium that would later win him a Nobel — the stable point at which suspicious players, each defending against the others, become collectively locked in place, none able to improve his position by moving alone.

Curtis makes a deliberately provocative move at this juncture, and honesty requires that it be presented as his interpretation rather than as settled fact. He links this model of man-as-suspicious-loner to Nash’s own paranoid schizophrenia, and to the paranoid atmosphere of the Cold War that funded the work — as though a picture of human beings as mutually distrustful competitors had been generalised from a mind, and a moment, in the grip of suspicion. One need not accept the biographical causation to see that the structural point stands. A specific and pessimistic picture of what a human being is was built into the machinery of the modern state at the very moment it was being sold as rationality itself.

Now hold that picture against the opening words of Ethiopia’s constitution. “We, the Nations, Nationalities and Peoples of Ethiopia…” Not we the people, singular, but we the peoples, plural — each a discrete and calculating entity, each entering the federation as a party to a contract it may one day rescind. Article 39 makes the logic explicit, and it is worth quoting exactly, since the whole reading rests upon it: “Every Nation, Nationality and People in Ethiopia has an unconditional right to self-determination, including the right to secession.” The exit is not treated as a failure of the system. It is written in as a feature of it, constitutionally guaranteed, permanently on the table.

Read through the theory we have just laid out, the design reveals itself. What Meles Zenawi and the architects of the Ethiopian People’s Revolutionary Democratic Front built, whether or not they ever used the vocabulary, was a prisoner’s dilemma given the force of law. Each people is modelled as a rational, self-interested actor. Each is invited to assume the others are pursuing their own advantage. Each is handed, in Article 39, a standing threat to walk away from the whole. In such a structure the individually rational move — as Nash’s mathematics predicts — is never to trust first, never to disarm first, always to secure one’s own group against the possibility of the others’ bad faith. The stable outcome, the equilibrium of the thing, is not cooperation. It is a low-trust standoff in which every party arms against every other, and all are left worse off than a single act of mutual trust would have made them. This is the architecture of suspicion, carried from the RAND seminar room into the founding law of an African state.

There is a scholarly bridge worth naming here, because it lets us extend Curtis with rigour rather than by loose analogy. The political scientist Barry Posen, in “The Security Dilemma and Ethnic Conflict,” published in the journal Survival in 1993, took the same game-theoretic core — the mutual, self-fulfilling fear at the heart of the prisoner’s dilemma — and applied it precisely to multi-ethnic states. His insight, drawn from the collapse of Yugoslavia, ran as follows: so long as an imperial or hegemonic power sits above the groups and forbids them to compete, the game is suppressed and the peace, however unloved, holds. But the moment that power withdraws, each group finds itself in a sudden anarchy, forced to provide for its own security, reading every neighbour’s defensive preparation as a possible offensive threat — and the dilemma, long dormant, springs shut. Curtis never made this move. Posen did. Ours is to recognise that Ethiopia ran the experiment.

The design did not fail in 2018. It was in 2018 that the design was finally, catastrophically, permitted to work.

For a quarter of a century the dilemma was held in check — not by trust, but by a hegemon. The EPRDF’s doctrine of revolutionary democracy and democratic centralism placed a single dominant party above the federation and simply forbade the peoples to play the game the constitution had written for them. Competition was suppressed; the exit clause was theoretical; the vanguard internalised the cooperative outcome and imposed it from above. The federal architecture appeared to function, but it functioned the way a coiled spring appears to be at rest. What looked like a federation of trusting nations was in truth a federation of restrained ones.

Then the restraint was removed. The transformation of 2018 — the very opening we called, in the first part, a genuine liberation — was in Posen’s exact sense the withdrawal of the hegemon. The dominant player stepped back from the board, and the game it had suppressed for twenty-seven years asserted itself with the speed of released tension. Each people, now genuinely free, did what the structure had always instructed it to do. It calculated. It looked at its neighbours, remembered its history, and moved to secure itself first.

The evidence is the last seven years of our national life. The proliferation of regional “special forces” and ethnic militias — region after region fielding its own standing army, because to be the one region without one was to accept the sucker’s payoff. The scramble for new statehood in the south, where the old Southern Nations region fractured as zone after zone calculated that its safety lay in its own regional exit rather than in a shared house: Sidama voted itself a region in 2019, a South West cluster followed in 2021, and further splits were legislated in 2023. And, in the purest illustration the theory could ask for, the federal move in 2023 to dissolve the regional special forces into the national army — a demand, in effect, that everyone disarm at once. In a system built on trust it would have been the obvious good. In a prisoner’s dilemma it was close to unthinkable, because to lay down your arms first, while others might not, is exactly the move the structure punishes worst. In Amhara the order was widely read as disarmament before rivals, the Fano mobilisation followed, and the equilibrium reasserted itself in blood. Tigray before it and Oromia around it tell the same story in different keys.

Here we must hold the discipline this series owes itself — the discipline Curtis himself so often abandons. The prisoner’s dilemma illuminates; it must not be permitted to exonerate. An equilibrium does not demolish a village or shell a town. People do, and they choose to, and some choose it for gain rather than survival. The frame explains why trust was so structurally fragile; it does not dissolve the responsibility of those who exploited the fragility. To let the game theory become a theodicy — a proof that the blood was no one’s fault, merely the mathematics working itself out — would be to fall into the very trap we are trying to describe: the seductive total system that explains everything and therefore excuses everything. The model is a light to think by, not an alibi.

What it does show, unmistakably, is that our national suspicion is not a cultural failing or an ancient hatred waiting in the blood. It is a designed outcome — engineered into a founding text, suppressed for a generation by a party, and released by a liberation. We were handed our freedom in 2018, and the freedom was real. But the self the constitution had modelled for us — the wary, calculating actor — was waiting inside the gift. We opened the door, and it walked in with us. That is the first jaw of the trap. The second is quieter, and wears the face of competence. It counts.

Sources

The architecture of suspicion. John von Neumann & Oskar Morgenstern, Theory of Games and Economic Behavior (1944); John Nash, “Equilibrium Points in n-Person Games” (1950) and “Non-Cooperative Games” (1951); the prisoner’s dilemma formulated at RAND by Merrill Flood and Melvin Dresher (1950), named by Albert Tucker. The link between Nash’s illness and the model is Curtis’s interpretation, presented here as such.

The ethnic security dilemma. Barry R. Posen, “The Security Dilemma and Ethnic Conflict,” Survival 35:1 (1993).

The founding text. Constitution of the Federal Democratic Republic of Ethiopia (1995), Preamble and Article 39, quoted from the official English text.

The Ethiopian record. Sidama regional referendum (2019); formation of the South West Ethiopia Peoples’ region (2021); further southern restructuring (2023); the 2023 federal move to integrate regional special forces and the ensuing Amhara/Fano mobilisation — to be cited to the Tribune’s own reporting and named external sources at production.

THE TRAP · PART II OF V · THE ETHIOPIAN TRIBUNE