Madagascar’s Gen Z uprising has achieved the objective that appeared almost impossible only weeks ago: President Andry Rajoelina is no longer governing the country. But the manner in which he fell should complicate any simple celebration of revolutionary victory. After weeks of demonstrations over electricity and water shortages developed into a nationwide revolt against corruption, poverty and political privilege, the decisive moment came when the elite CAPSAT military unit broke with Rajoelina and sided with protesters. Other security institutions subsequently defected, Rajoelina fled the country after saying his life was in danger, and parliament voted to impeach him even after he attempted to dissolve the National Assembly. On 14 October, CAPSAT commander Colonel Michael Randrianirina announced that the armed forces had taken power, suspended several major institutions and proposed a military-led transition lasting as long as two years before elections. The protesters who had spent weeks demanding Rajoelina’s departure celebrated alongside soldiers in Antananarivo, and their relief is understandable. But Madagascar has been here before. CAPSAT was the same military unit whose intervention helped Rajoelina take power during the 2009 political crisis. Sixteen years later, it has helped remove him. The circularity should make the central problem obvious: if political leaders ultimately enter and leave power according to which side the military chooses during a crisis, changing the president does not necessarily change the political system.
The distinction between popular revolution and military coup therefore matters enormously. The protests were real and broad. They emerged from genuine failures in electricity, water, employment and governance, and the United Nations reported at least 22 deaths during the early unrest, although the government disputed that figure. Rajoelina had lost substantial political legitimacy before the military changed sides. But protesters themselves did not possess the coercive power that finally removed him. CAPSAT did. Once soldiers refused to suppress demonstrations and then assumed control of the armed forces, the balance of power shifted decisively. This creates a dangerous precedent even when the military intervenes on behalf of a popular cause. Future commanders learn that sufficient public mobilisation can provide political legitimacy for military intervention, while future protest movements learn that obtaining military support may be more effective than constructing electoral majorities. The result can be a system where the street supplies legitimacy and the barracks supplies authority. That is not the same thing as democratic government, because the decisive political institution becomes the organisation controlling weapons rather than the institution counting votes.
The historical parallel with 2009 makes this especially important. Rajoelina himself rose during mass demonstrations against President Marc Ravalomanana, after which CAPSAT broke with the government and helped force Ravalomanana from power. Rajoelina initially governed as the leader of a transitional authority before eventually returning through elections. The political justification then was also that the existing government had lost legitimacy and that intervention was necessary to respond to popular demands. Today’s protesters understandably believe their cause is different because Rajoelina himself became the entrenched political elite they oppose. Institutionally, however, the mechanism is remarkably similar. A precedent does not care whether today’s beneficiary deserves sympathy. Once military intervention becomes an accepted method for resolving political legitimacy crises, future officers inherit the same option against governments the current protesters may actually support.
This is why Randrianirina’s announcement that the military could govern for up to two years should receive much more scrutiny than the immediate popularity of Rajoelina’s removal. Two years is an extremely long political transition if the principal objective is organising credible elections. Temporary governments almost always possess reasons to extend temporary authority. Electoral rolls need updating. Constitutional reform needs consultation. Corruption needs investigating. Security needs restoring. Institutions need restructuring. Every justification can be individually plausible while collectively producing an indefinite transition. The relevant question is therefore not whether Randrianirina promises elections eventually. It is whether Madagascar establishes binding institutional milestones that make transferring power more likely even when the military later discovers reasons it would prefer not to.
The suspension of institutions makes that concern more serious. Military leaders announced the suspension of the Senate, High Constitutional Court, Independent National Electoral Commission, High Court of Justice and High Council for the Defence of Human Rights and the Rule of Law, while preserving the National Assembly. Some of these institutions may indeed have suffered from politicisation or weak public confidence under Rajoelina. Suspending them nevertheless creates a paradox: institutions accused of failing to constrain executive power are being replaced temporarily by a military leadership subject to even fewer institutional constraints. This is the revolutionary clean-slate fallacy. When existing institutions are imperfect, destroying them can appear easier than reforming them. But institutional emptiness does not produce neutrality. It transfers discretion to whoever controls the transition.
A stronger transition would therefore restore or replace independent institutions rapidly rather than governing through prolonged exceptional authority. An electoral commission can be reconstituted with representatives acceptable across political factions. Courts can review transitional decisions. Civil society can monitor procurement and appointments. A clear election date can be established well in advance, with changes requiring transparent legal procedures rather than military discretion. The objective should be creating credible commitment by the transitional government to its own disappearance. This is one of the hardest things any temporary authority can do because successful reform eventually makes the reformers less powerful. That is precisely why constitutional constraints matter.
The protesters themselves appear much less unified about what comes next than they were about what needed to end. Reuters interviewed demonstrators who were unequivocal that Rajoelina needed to leave but frequently could not identify who should replace him or what precise political structure should follow. This is not evidence that the movement was unserious. It illustrates the fundamental asymmetry between opposition and governance. Thousands of citizens can agree that electricity shortages, corruption and unemployment are unacceptable without agreeing on tax policy, privatisation, constitutional reform, foreign investment or the appropriate role of the military. Protest aggregates dissatisfaction. Government allocates scarce resources among competing demands. The second task necessarily creates disagreements the first can temporarily conceal.
This is the negative-coalition problem. A political movement organised around removing one leader can become extraordinarily broad because participants postpone disagreements until after the common opponent disappears. Students, workers, opposition politicians, civil-society groups, business owners and soldiers can all demand Rajoelina’s resignation for completely different reasons. Once he leaves, the coalition’s unifying variable disappears. The movement then needs to transform “Rajoelina must go” into answers about electricity pricing, mining contracts, anti-corruption institutions, employment, public spending and elections. That transformation is where revolutions frequently fragment.
Madagascar’s economic constraints will make those disagreements immediate. Roughly three-quarters of the population lives in poverty, while the country’s long-run economic performance has been exceptionally weak; the World Bank estimates cited by Reuters indicate that GDP per capita fell by about 45% between independence in 1960 and 2020. Madagascar simultaneously exports valuable products including vanilla, nickel, seafood, textiles and minerals. The coexistence of natural wealth with widespread poverty understandably produces anger, but it can also create unrealistic expectations that removing corrupt politicians will release enough hidden wealth to solve development problems quickly. Corruption matters substantially. It is not the only constraint. Weak infrastructure, limited industrial productivity, low human capital, geographic isolation and repeated political instability all reduce growth. There is no confiscated pot of elite wealth large enough to substitute for decades of productivity growth.
The electricity and water problems that began the uprising therefore will become the new government’s first credibility test. Protesters will reasonably expect visible improvement because these were the grievances around which mobilisation initially formed. But repairing electricity systems requires capital, fuel, maintenance, technical expertise and politically difficult decisions about tariffs and subsidies. If the national utility sells electricity below sustainable cost, somebody eventually finances the difference through taxes, debt or deteriorating infrastructure. Raising prices can improve utility finances while angering precisely the households who protested over living costs. Maintaining low prices without financing investment perpetuates outages. The transition therefore immediately confronts the difference between political demands and budget constraints.
This is why blaming every failure on corruption can become dangerous after a revolution. Corruption allows politicians to explain poor outcomes morally: remove corrupt people and services improve. Some improvements genuinely can occur through better procurement and reduced theft. But infrastructure also contains real resource constraints. A power station costs money even under perfect governance. Fuel needs importing. Engineers need salaries. Distribution networks need maintenance. If the new authorities promise that integrity alone will make electricity simultaneously cheaper and more reliable, disappointment will arrive quickly. Good governance improves the efficiency of difficult choices; it does not eliminate the choices.
The same applies to youth employment. Madagascar’s median age is around 19, making it one of the youngest societies in the world. A youthful population can generate extraordinary growth when education, investment and employment expand together. It can generate extraordinary frustration when large cohorts enter adulthood without sufficient opportunity. The new government cannot manufacture hundreds of thousands of productive jobs through public employment without eventually creating a fiscal crisis. Sustainable employment requires private investment, entrepreneurship, infrastructure and access to markets. That means the transition needs precisely the thing political instability tends to destroy: investor confidence in the durability of rules.
A foreign or domestic company considering a twenty-year investment in Madagascar now needs to ask whether contracts signed by the next government will survive another political upheaval. Every coup increases the political-risk premium attached to investment. Investors demand higher expected returns to compensate, avoid long-lived projects or leave altogether. Political instability therefore can become economically self-reinforcing. Weak growth produces protests. Protests produce unconstitutional transitions. Unconstitutional transitions reduce investment. Lower investment weakens growth further. Breaking this instability-development trap requires making the next transfer of power boring.
That may be the most useful objective for Madagascar’s transition. Not another charismatic saviour, revolutionary council or national refoundation, but a political system in which an unpopular government loses an election and leaves office without soldiers deciding the outcome. Successful democracy eventually becomes administratively mundane. Votes are counted. Governments change. Contracts survive. Bureaucrats continue working. The absence of drama is evidence of institutional strength.
CAPSAT’s popularity with protesters makes civilian control particularly difficult because the soldiers currently possess revolutionary legitimacy. Randrianirina reportedly asked crowds whether they were prepared to accept military rule and received enthusiastic support. This is understandable after security forces helped remove a government protesters despised. It is also precisely when democratic safeguards matter most. Institutions are rarely surrendered permanently to actors citizens distrust from the beginning. They are surrendered to actors citizens trust during emergencies. The relevant question is not whether Randrianirina personally intends dictatorship. It is whether Madagascar should construct a political system whose democratic future depends on one colonel’s intentions.
The answer should be no regardless of his intentions. Good institutions are designed around the assumption that future leaders may be worse than current ones. If the military receives broad political authority because today’s commander appears aligned with popular demands, tomorrow’s commander inherits the authority without necessarily inheriting the same alignment. This is the institutional version of the veil-of-ignorance test: would protesters want the same military powers available if CAPSAT later supported a government they opposed?
The international response therefore should focus on the timetable and structure of civilian transition rather than attempting to restore Rajoelina personally. His political legitimacy appears severely damaged, parliament impeached him and significant sections of the security establishment abandoned him. Trying to reverse that reality externally could deepen instability. But recognising the military takeover unconditionally would create another harmful precedent. The African Union has long maintained formal opposition to unconstitutional changes of government because coups historically destabilised large parts of the continent. The principle matters even when the deposed government was unpopular. Otherwise every military intervention can claim to represent authentic popular dissatisfaction.
External actors should therefore use recognition, aid and economic cooperation as conditional incentives for a credible civilian transition. A clear electoral calendar, restoration of judicial institutions, protection for opposition parties, independent media and civilian oversight of the security forces could unlock deeper international support. Failure to meet benchmarks could delay recognition or financial assistance. This is conditionality with a defined off-ramp, much more useful than permanent isolation because it tells the transitional government exactly what behaviour produces normalisation.
France faces a particularly delicate position because Rajoelina reportedly left Madagascar aboard a French military aircraft, according to sources cited by Reuters, while many protesters already viewed him as excessively close to Madagascar’s former colonial power. Rajoelina’s French citizenship had become politically controversial, and French involvement in his evacuation can reinforce narratives that Paris protects friendly elites when domestic legitimacy collapses. France therefore needs to avoid appearing to determine Madagascar’s political transition. Former colonial powers possess influence precisely because of historical relationships, but visible use of that influence can reduce legitimacy rapidly.
This is the post-colonial influence paradox. France can possess substantial diplomatic, economic and security relationships in a former colony while discovering that overt intervention makes those relationships politically toxic. The more visibly Paris attempts to preserve influence, the more attractive alternative partners such as China, Russia or others can become to nationalist political movements. Sustainable influence therefore increasingly depends on supporting institutions and economic relationships rather than individual leaders.
India should draw a similar lesson because Madagascar sits within the wider Indian Ocean strategic environment. New Delhi has strong reasons to maintain maritime-security, development and economic ties with Antananarivo, particularly as competition with China expands across the Indian Ocean. But political instability means relationships tied too closely to one administration become vulnerable when that administration falls. The most durable Indian strategy is therefore institutional rather than personal: coast-guard cooperation, infrastructure, disaster assistance, education, trade and capacity-building that remain valuable regardless of which party governs.
China will likely pursue the same opportunity through investment and resource relationships. Madagascar’s mineral deposits make it potentially important to global supply chains for materials such as nickel, cobalt and graphite. Competition among external partners can benefit Madagascar if contracts are transparent and governments bargain effectively. It becomes harmful when elites exchange opaque concessions for political or personal benefits. The transitional government therefore has an opportunity to publish major resource contracts and establish competitive procurement rules. Transparency can convert geopolitical competition from a source of corruption into a source of bargaining power.
The youth movement should demand exactly these structural reforms rather than assuming its role ends because Rajoelina has fallen. Protest movements often become demobilised after the visible target disappears, allowing organised institutions such as the military, old political parties and business networks to dominate the transition because they possess permanent organisational capacity. Gen Z Madagascar has enormous mobilisation capacity but appears deliberately decentralised. That protected it from easy repression and co-option. It now creates a disadvantage because decentralised movements can remove leaders more easily than they can negotiate constitutions.
The movement therefore needs some institutional form without necessarily becoming one political party. Civic organisations can monitor the transition. Independent groups can publish military appointments, procurement contracts and election preparations. Young candidates can contest future elections across multiple parties. Digital platforms can track whether electricity, water and anti-corruption promises actually improve. The objective should be maintaining pressure without claiming that protesters themselves represent every Malagasy citizen automatically.
That distinction matters because street numbers are not a substitute for elections. Tens of thousands of protesters can demonstrate extraordinary dissatisfaction while millions of citizens remain outside the streets for many reasons. Some may support the protests. Some may fear violence. Some may support Rajoelina. Others may simply prioritise work and family. Democratic legitimacy cannot be inferred perfectly from crowd size, just as it cannot be inferred from military support. Elections remain the mechanism through which these hidden preferences become countable.
The eventual electoral system therefore needs to be credible enough that the losing side accepts the result. That requires an independent commission, transparent counting, observer access and rules established before candidates know whether they benefit from them. Suspending the electoral commission may have been politically convenient during the takeover; replacing it with a trusted institution should become one of the first transitional priorities.
The parliament’s impeachment of Rajoelina adds an unusual constitutional layer because lawmakers acted while the president attempted to dissolve the legislature, producing competing claims about which institution retained authority. This ambiguity gives every side legal arguments supporting its preferred narrative. Rajoelina can call the takeover unconstitutional. Parliament can claim it removed him lawfully. The military can say it intervened after constitutional institutions deadlocked. The existence of multiple plausible narratives makes restoring an independent constitutional arbiter particularly important. Otherwise the winner’s interpretation becomes law simply because the winner controls coercive power.
This is precisely what courts are supposed to prevent. A constitutional court does not need to be politically popular. Its value comes from creating a rule-bound mechanism for resolving disputes that would otherwise be settled by force. Suspending the High Constitutional Court therefore may provide short-term flexibility while removing one of the institutions Madagascar needs most during a constitutional crisis. If the court itself lacked legitimacy, reform its appointment process. Eliminating adjudication leaves politics without a referee.
The deeper economic problem remains poverty. Three-quarters of the population living in poverty means almost every institutional reform will operate under severe material constraints. Citizens need immediate improvements, while government needs to avoid unsustainable promises. This makes international financing useful but also dangerous. Emergency assistance can stabilise public services. Excessive donor dependence can recreate a state whose programmes disappear when external support changes.
Madagascar therefore needs a development model capable of converting its agriculture, minerals, tourism, textiles and strategic location into broader productivity growth. That requires electricity first, which brings the entire revolution back to where it began. Reliable power allows manufacturing. Manufacturing creates employment and exports. Export revenue finances imports and investment. Higher incomes expand the tax base. A stronger tax base finances infrastructure. Infrastructure attracts more investment. Development can become self-reinforcing in the same way institutional failure can.
This is the positive development loop Madagascar needs to build. The protests identified the first broken link very clearly: electricity and water. The military can remove a president. It cannot generate megawatts through political authority. That requires engineers, investment, regulation and competent administration. The legitimacy of the transition will therefore eventually depend on bureaucratic outcomes rather than revolutionary symbolism.
There is something almost perfectly circular about Madagascar’s current moment. In 2009, CAPSAT helped remove a president amid protests and enabled the
