Can street protests build lasting institutions, or only tear down governments?
Street protests can remove a ruler but cannot build what replaces him. That takes years of deliberate sequencing after the crowd goes home.
Arendt distinguishes power from authority: Sudan's 2019 coalition and Bangladesh's 2024 movement both toppled leaders without founding anything to replace them. Hirschman's test is irreversibility; Sri Lanka's 2022 protests were undoable in a way that installing Wickremesinghe through parliament was not. Sirleaf shows the alternative actually working: Liberia rebuilt revenue collection, banking, police and courts in sequence over years after 2006, not overnight. Nyerere and Fanon both warn that whoever fills the vacuum, a technocrat like Yunus or an administrative elite, tends to reach for shortcuts or capture the state the crowd vacated.
Rousseau holds that the Dhaka students' will carries its own legitimacy and needs no unelected government to broker it. Arendt and Sirleaf counter that without deliberate institution-building, that will has nowhere to live once the square empties.
Confidence summary: The council treats the power/authority distinction as settled analytic ground; it disputes sharply, and without resolution, who has standing to build what follows.
1. The core argument
The uncomfortable finding is not that protests fail. It is that they succeed, repeatedly and completely, at the one task that turns out not to matter most. Bashir fell. Rajapaksa fled. Hasina fled. In every case the crowd achieved its stated goal within months. And in every case the thing that followed, a power-sharing council, an establishment prime minister, an unelected interim government, was improvised by whoever happened to be standing in the vacated room. The council's sharpest move is separating two achievements usually spoken of as one: generating enough collective power to remove a government, and founding durable authority to replace it. These require different timescales, different skills, and arguably different people entirely. Liberia's post-2006 reconstruction took years of sequenced institution-building that no crowd, however large, can compress into a season of occupation. The protest is the easy part. The council's live disagreement is not whether this gap exists but who is entitled to close it.
2. How each member frames it
Hannah Arendt draws the line that structures the entire debate: power (people acting in concert) and authority (institutions that outlast the acting) are not degrees of the same thing but different categories of political achievement. Her sharper, less comfortable point is that America's constitutional convention succeeded partly because it deliberately excluded the crowd from the drafting room, a move she would apply to Sudan's unresolved civilian council and find wanting, not for lacking street input but for lacking deliberation at all.
Jean-Jacques Rousseau rejects Arendt's framework at its root: he does not concede that deliberation by a small elite constitutes legitimate founding, however orderly. His harder claim, easy to soften on the card, is that Yunus's technocratic competence is irrelevant to his legitimacy; competence was never Rousseau's test. An unelected government that governs well is still, on his reading, the "elites reorganising power while calling it reform" that he condemned in 1755.
Albert O. Hirschman supplies the mechanism that makes the Arendt/Rousseau quarrel practical rather than purely theoretical: irreversibility. His uncomfortable admission is that his own framework offers no answer to Rousseau's objection, it tells you which decisions are hard to undo, not who deserves to make them. Sri Lanka's parliament, not its protesters, made the irreversible choice, and Hirschman treats that as a fact about power, not a verdict on its fairness.
Ellen Johnson Sirleaf brings the one case in the set that actually worked, and her candid limit is important: her sequencing (revenue, banking, security, courts) depended on negotiated legitimacy and international backing that Sudan's 2019 movement simply did not have. She does not claim her model is portable without those preconditions, which makes her implicit judgment on Sudan less a reproach than a diagnosis of missing inputs.
Julius Nyerere turns his own record into the argument, which gives his position unusual weight: he had more legitimacy than almost any post-independence leader and still reached for forced relocation when his institutions failed. His point against Sirleaf is that her sequencing worked partly by luck of circumstance; most leaders holding conviction without a tested plan, himself included, improvise coercively rather than patiently.
Frantz Fanon supplies the darkest reading and refuses Nyerere's self-criticism as sufficient explanation: the pattern he names is structural, not personal failing. His concession, offered almost reluctantly, is that Yunus does not fit his bourgeois-capture model neatly, a technocrat is not a comprador class, which he flags as a genuine anomaly worth watching rather than resolving.
3. Where the council agrees
The non-trivial agreement is that speed is the enemy of durability. Every member, across four centuries of political experience, treats the interval right after a government falls as the most dangerous moment, not the protest itself. This is surprising because the public narrative around Sudan, Sri Lanka and Bangladesh treats the toppling as the climax; the council treats it as the opening move. They also agree, against the instinct to romanticise crowds, that street power generates no automatic claim on what replaces the fallen government: someone fills that vacuum, and the crowd rarely gets to choose who. Finally, Sirleaf's sequencing (revenue before politics, security before elections) is accepted by every member as evidence that institution-building is a multi-year craft, not a founding-document moment, even by those who reject her conclusions about who should perform it.
4. Where the council splits
The real fault line is whether popular will, expressed only through the act of removal, carries legitimacy that survives into the post-protest period, or whether legitimacy must be freshly constructed through deliberation, negotiation, or elections afterward. Rousseau stands alone in insisting the Dhaka students' will is self-legitimating and needs no further authorisation; an unelected government governing competently is still, for him, illegitimate by definition. Arendt, Hirschman and Sirleaf stand against him: they treat legitimacy as something built, sequentially, by people who may never have been in the square. Neither side is wrong because they are answering different questions, one about the source of legitimate authority, the other about the conditions for institutions that actually function.
5. For a policymaker to decide on
An interim government, Bangladesh's being the live case, must choose between calling elections quickly to restore a mandate, honoring Rousseau's demand that the street's will not be indefinitely deferred, or delaying elections to sequence revenue, security and judicial reform first, following Sirleaf's Liberia model. Quick elections risk handing power back to the old apparatus before it is reformed, as Fanon warns. Slow sequencing risks the technocrat becoming permanent, the fraud Rousseau names. The council cannot adjudicate this because it is a judgment about which illegitimacy a given country can better survive.