Military officers seized power in Gabon on Wednesday shortly after electoral authorities announced that President Ali Bongo Ondimba had won another term. The officers cancelled the election result, dissolved several state institutions and placed Bongo under house arrest, later naming General Brice Oligui Nguema as transitional leader. For many Gabonese, however, the immediate reaction was not simply fear of another African military takeover. Crowds celebrated Bongo’s removal, reflecting the extraordinary depth of frustration with a political system dominated by one family for more than half a century.
The history explains much of that reaction. Omar Bongo governed Gabon from 1967 until his death in 2009, making him one of the world’s longest-serving presidents. His son Ali Bongo then won the presidency and remained in office for another fourteen years. The family therefore dominated Gabonese politics for approximately 56 years, an extraordinary concentration of political power even before considering the networks of patronage and economic influence surrounding the presidency. The existence of elections during much of that period did not necessarily mean citizens believed political alternation was realistically available.
This is what makes Gabon’s coup different from Niger’s last month. Niger’s soldiers justified their takeover partly through deteriorating security and jihadist violence. Gabon does not face an equivalent insurgency. The military instead presented its intervention as a response to an election and political system it considered uncredible. The coup therefore raises an uncomfortable possibility: what if the constitutional order soldiers overthrow is itself only weakly democratic?
The easy democratic answer is that coups are illegitimate because soldiers do not possess electoral authority. That remains correct, but it is incomplete. If elections are systematically manipulated, opposition parties cannot compete fairly and incumbents become effectively impossible to remove, simply defending “constitutional order” can amount to defending institutions that have lost much of their democratic substance. Constitutions can protect democracy, but authoritarian governments can also operate through constitutions.
This is why elections should be understood through contestability rather than ritual. A country does not become meaningfully democratic merely because citizens periodically enter polling stations. The governing party needs to face a genuine possibility of losing. Opposition parties need sufficient freedom to organise and campaign. Electoral institutions need enough independence that citizens trust the count. Media and civil society need enough space to scrutinise government. Most importantly, an electoral defeat needs to result in an actual transfer of power.
If those conditions weaken sufficiently, elections can become mechanisms for legitimising incumbents rather than mechanisms for replacing them. The ballot remains, but the uncertainty that gives the ballot political meaning disappears. Gabon’s long period of Bongo family rule inevitably created questions about whether elections still provided a credible route towards alternation, particularly after disputed contests in previous years.
Yet acknowledging that problem does not make Wednesday’s military intervention democratic. The soldiers did not correct the electoral process and then allow citizens to determine the result. They cancelled the election and assumed political authority themselves. The mechanism through which Bongo was removed therefore gives Gabonese citizens no more direct control over who governs next than the disputed election did.
This distinction matters because coups often acquire legitimacy by comparing themselves with the government they remove rather than with the institutions they create. If Bongo’s system was corrupt or electorally manipulated, military rule appears better by comparison. But removing a bad institution is only the first half of political change. The relevant question is what replaces it.
A military junta can be preferable to a particular dictatorship and still create another dictatorship.
This is why the phrase “democratic coup” is so dangerous. It contains an understandable intuition: if soldiers remove an authoritarian ruler whom citizens could not realistically remove electorally, perhaps the intervention creates an opportunity for democracy. Historical cases exist where military interventions eventually produced civilian transitions. The danger is converting a possible outcome into an assumption about military intentions.
The people possessing weapons after the coup also possess incentives to retain power.
A general can promise elections in two years. When two years arrive, the general decides whether conditions are ready. Political opponents can then be described as threats to stability. Elections can be delayed because reforms remain incomplete. Temporary restrictions become necessary for national unity. The transition develops reasons to continue.
This is the time-consistency problem of military rule from Niger again. Soldiers may genuinely believe intervention is temporary at the moment they seize power. Their incentives change once they control appointments, budgets and state institutions. A promise to surrender power later becomes less attractive precisely when the moment to surrender arrives.
Gabon therefore needs the military’s democratic intentions to be demonstrated through constraints rather than speeches. Is there a clear transition timetable? Are civilian political organisations free to operate? Will electoral institutions become independent? Can opposition figures compete? Are media restrictions removed rather than expanded? Most importantly, is the military itself excluded from manipulating the eventual result?
These questions matter more than whether crowds currently celebrate.
Popular support for a coup is not equivalent to democratic legitimacy because public opinion immediately after regime collapse measures relief as much as consent. A citizen celebrating Bongo’s removal may not support permanent military government. They may simply prefer almost any immediate alternative to another Bongo term.
Political preference is comparative.
This is the same mechanism that explains why voters sometimes support opposition parties they do not particularly like: removing the incumbent can become the dominant objective.
The danger is that negative legitimacy expires quickly. A new government can initially derive support from not being the old government. Eventually citizens begin judging what it actually does. The revolutionary coalition that agreed Bongo should leave can fragment once questions shift towards economic policy, constitutional design and who should govern permanently.
Gabon possesses substantial oil wealth, which makes the governance problem particularly important. The country is one of sub-Saharan Africa’s significant oil producers and has a relatively high GDP per capita compared with many countries in the region. Yet aggregate national income conceals substantial inequality, and many citizens believe resource wealth has benefited political and economic elites disproportionately.
This is the familiar resource-curse problem. Oil can provide governments with large revenues without requiring the broad tax base through which states normally develop fiscal relationships with citizens. Governments financed heavily through natural resources may become less dependent on ordinary taxpayers, while political competition increasingly revolves around control over resource rents.
Taxation does more than raise money. It creates bargaining. Citizens paying substantial taxes have strong incentives to demand information about how government spends them, while governments dependent on those taxpayers have stronger incentives to remain responsive. Resource revenue can weaken that relationship if the state receives a large share of income from extracting commodities rather than taxing a broad productive economy.
This does not make authoritarianism inevitable. Norway demonstrates that oil wealth can coexist with strong democratic institutions. Botswana has historically managed diamond revenues more effectively than many resource-rich states. Institutions determine how rents are distributed and constrained.
Gabon demonstrates the opposite danger: natural-resource wealth can make political control unusually valuable because controlling the state means influencing access to enormous economic rents.
That helps explain why dynastic politics can become durable. Political power creates economic networks, and economic networks reinforce political power. Businesses benefiting from government relationships have incentives to support continuity. Officials whose careers depend on the ruling network fear political turnover. Patronage distributes enough resources to maintain coalitions.
The state becomes an economic ecosystem organised around incumbency.
This is why simply removing the president may change less than supporters expect. If the military leadership inherits the same networks of oil revenue, procurement and patronage, Gabon could replace the people controlling the system without changing the system itself.
The critical question is therefore whether the coup produces elite rotation or institutional transformation.
Elite rotation changes who receives access to state resources.
Institutional transformation changes the rules determining how those resources are allocated.
The first can happen overnight.
The second requires transparency, independent courts, competitive elections and administrative reform.
Military takeovers are unusually good at producing the first and have no automatic advantage in producing the second.
General Oligui himself illustrates the complexity because he emerged from within Gabon’s existing security establishment rather than from outside the political system. Coups are frequently presented as revolts against elites, but military leadership is usually itself part of the state elite. The institution capable of seizing the presidency is almost by definition one that already possesses substantial resources and proximity to power.
This is why anti-establishment rhetoric should be examined carefully regardless of who uses it. An actor can oppose the current ruler without opposing the institutional arrangements that concentrated power around that ruler.
The military may object to Bongo.
That does not necessarily mean it objects to concentrated executive authority.
For ordinary Gabonese, the rational response may nevertheless be cautious optimism rather than immediate rejection. If the previous electoral system genuinely offered little possibility of alternation, regime disruption creates uncertainty, and uncertainty can contain opportunity.
The mistake would be treating opportunity as achievement.
A transition becomes democratic only when the people who seized power construct institutions capable of removing themselves from power.
That is an extraordinarily demanding test because it asks rulers voluntarily to build constraints against their own interests.
Yet that is precisely what constitutional democracy requires.
The international response creates another dilemma. The African Union suspended Gabon, while regional and foreign governments condemned the coup. That is important because normalising every military takeover creates incentives for officers elsewhere to imitate successful examples. West and Central Africa have now experienced a striking sequence of coups or military transitions, and each successful takeover changes expectations about what is politically possible.
Coups can therefore produce a demonstration effect. Soldiers in one country observe officers elsewhere seize power, survive international condemnation and eventually become accepted as political realities. The expected cost of attempting a coup falls.
Regional organisations consequently need credible anti-coup norms.
But automatically demanding restoration of the previous government creates its own problem when that government possessed weak democratic legitimacy. Restoring constitutional order and restoring one particular ruler are not necessarily identical objectives.
The better principle should be restoring civilian political choice.
That could mean returning Bongo if credible constitutional institutions determine he legitimately won. It could mean organising a genuinely competitive new election if the previous process cannot command confidence. What should not become normal is allowing the military to determine the answer indefinitely.
This distinction would make regional democratic norms more credible because they would defend institutions rather than individual incumbents.
Citizens understandably become sceptical when international organisations condemn coups immediately but tolerate years of electoral manipulation preceding them. If democracy matters only when soldiers overthrow presidents, but not when presidents weaken democratic competition gradually, anti-coup principles can appear like incumbent-protection mechanisms.
The African Union and regional organisations therefore need to defend constitutional democracy rather than constitutional incumbency.
That means scrutinising elections before coups happen.
It means responding to term-limit manipulation, repression and electoral fraud with greater consistency.
Otherwise the international community intervenes politically only after democratic institutions have already lost enough legitimacy for soldiers to receive applause when they destroy them.
This is where Gabon’s coup provides a more difficult lesson than Niger’s. Niger demonstrated the danger of using security failure to justify military government. Gabon demonstrates the danger created when civilian institutions become so distrusted that military government initially looks like political liberation.
Both routes end at the same institutional question.
Who controls the people with guns?
And how can citizens remove them?
Democracy cannot be defined simply as opposition to military coups because civilian governments themselves can hollow out democratic institutions. Nor can democracy be rescued simply by overthrowing those governments because military rulers possess no inherent democratic legitimacy.
The standard has to remain contestable civilian power.
Can citizens choose leaders?
Can those leaders lose?
Can opposition organise?
Can courts constrain government?
Can the military remain subordinate regardless of who wins?
Those conditions matter more than whether the person occupying the presidential palace wears a suit or a uniform.
Ali Bongo’s removal may end one of Africa’s longest political dynasties.
That could become an opening for democratic reform.
It could also become another transfer of concentrated power between elites.
The difference will not be determined by how many people celebrated on Wednesday.
It will be determined by whether the soldiers who cancelled an election eventually build a system in which they themselves can lose one.
