Niger orders France's ambassador out. Paris says he is staying.

France's ambassador in Niamey became the subject of an unusually literal diplomatic dispute on 25 August. Niger's military authorities told Sylvain Itté to leave within 48 hours. Paris refused, saying the soldiers who had overthrown President Mohamed Bazoum had no authority to make that demand.

Three days later, Emmanuel Macron used his annual address to French ambassadors to back Itté's continued presence. By 31 August, the junta had escalated again, claiming to remove his diplomatic privileges and ordering his expulsion. France said he remained at his post.

The confrontation distils the wider crisis. France and the EU recognise Bazoum as Niger's legitimate president. The officers who detained him on 26 July control the state machinery through which embassies, military partnerships and aid programmes have to operate. Denying their legitimacy does not make that practical control disappear.

In Toledo, where EU ministers met at the end of August, Josep Borrell welcomed Bazoum's foreign minister and the president of the ECOWAS Commission. The guest list made Europe's political position clear. It did not provide an answer to the question now confronting Paris: how long can a diplomatic presence remain effective when the authorities controlling its surroundings want it gone?

An embassy becomes the front line

The junta's expulsion demand came a month after members of the presidential guard detained Bazoum. General Abdourahamane Tiani emerged as the military government's leader. The takeover interrupted a presidency that European governments had treated as a central security and development partner in a region where earlier coups had already narrowed their options.

An ambassador normally represents a state to the authorities of another state. A coup complicates that relationship when the sending government refuses to recognise the new rulers. France's position is that Bazoum remains the competent authority. The junta seeks to demonstrate the opposite by giving an order and demanding visible compliance.

Neither side can treat Itté's location as a purely symbolic matter. The ability to work depends on security, communications, access and the movement of staff and supplies. Maintaining a flag at a compound is different from maintaining an embassy capable of performing its functions. Those operational conditions are central to the decision Paris must make.

Macron's 28 August speech defended the existing policy and praised the diplomats remaining in place. He also rejected the junta's depiction of France as Niger's enemy. This was an attempt to separate support for an elected president from an argument about colonial power, even as the military authorities made France's role a focus of mobilisation.

France's history in Niger gives that mobilisation a receptive audience. It does not establish that every demonstrator supports military rule or that opposition to a foreign troop presence amounts to consent for Bazoum's detention. Public protest, attitudes towards France and approval of a coup are separate questions, even where the junta tries to combine them.

The same caution applies to European descriptions of Niger as a lost partner. A government's relations with Paris cannot stand in for the political preferences of an entire country. A policy framed only around retaining access and influence risks reinforcing the claim that Europe is concerned principally with its own strategic position.

The Toledo meeting also underlined the difference between representation abroad and control at home. Bazoum's foreign minister could address European counterparts and explain the elected government's position, but his presence could not ensure that instructions reached Niger's ministries or security forces. Europe must preserve the political channel without overestimating what it can implement. One practical priority is access to the detained president and reliable information about his condition. Another is keeping regional diplomacy connected to people inside Niger who can carry out an agreement. A settlement negotiated only among outside capitals would still require local authority to release prisoners, change orders and permit a return to civilian government.

For Paris, recalling the ambassador could be presented by the junta as a victory. Keeping him in place at any cost could make one compound the centre of a confrontation without restoring Bazoum. The relevant diplomatic calculation is therefore what each choice preserves: personal safety, communication, a legal position and the possibility of influencing the crisis. Those objectives may no longer point in the same direction.

ECOWAS threatens force and keeps talking

West African leaders, rather than European institutions, have taken the lead in demanding the restoration of constitutional government. The Economic Community of West African States imposed sanctions after the coup and set an initial deadline for the officers to reverse it. That deadline passed without Bazoum returning to power.

At a second extraordinary summit on 10 August, ECOWAS directed the activation of its standby force while maintaining the pursuit of a diplomatic solution. The two positions were intended to reinforce each other: military preparation could lend pressure to negotiations. They also created a burden of credibility if the junta judged that intervention would not follow.

An instruction concerning a standby force is not evidence that an operation has begun. Troop contributions, readiness, command, logistics and political approval all matter. Public discussion that treats a summit decision as a deployed force compresses a difficult chain of national choices into a single headline.

Nigeria is particularly important because of its size, position and President Bola Tinubu's role as ECOWAS chair. A regional operation would have consequences for Nigerian security and communities connected across the border. Tinubu's regional responsibilities therefore cannot be separated from the domestic political case he would have to make for military action.

Mali and Burkina Faso, themselves governed by military authorities, opposed intervention and warned that it could draw them into the confrontation. Their position makes the threat of regional escalation concrete. It does not establish the number of forces they could deploy or how they would operate, questions that require evidence beyond declarations of solidarity.

European support for ECOWAS raises another distinction. Backing the organisation's demand for constitutional government does not automatically define what European countries would contribute to a military operation. Transport, intelligence, financing and combat participation carry different obligations. A general offer of support leaves those choices unresolved.

The EU's 12 August statement welcomed continued diplomacy and noted the standby-force decision. It also expressed readiness to support ECOWAS through measures including a new framework for individual sanctions. At that point, readiness to establish a sanctions regime should not be reported as a completed EU listing of coup leaders.

Negotiation presents its own risk. Talks can secure access to Bazoum or clarify a possible way out; they can also allow the junta to consolidate control. The distinction depends on whether discussions produce verifiable steps. Access, release of detainees and a credible return to civilian authority would mean more than another meeting with no change in conditions.

Europe's security partner is now its operational problem

Before the coup, Niger occupied a growing place in European security planning for the Sahel. Cooperation with military authorities elsewhere had deteriorated, making the relationship with Bazoum more valuable. That concentration increased Europe's exposure when his own presidential guard turned against him.

The EU announced the suspension of budget support and security cooperation after the takeover. These were direct policy changes under European control. They expressed non-recognition and prevented business as usual, but they could not by themselves reverse the balance of force inside Niger.

A foreign military presence depends on more than the physical ability to remain at a base. It requires political consent, access to local partners and arrangements for airspace, supplies and movement. When the recognised president is detained and the officers controlling those arrangements reject cooperation, the mission's practical basis becomes contested.

The need to protect personnel can then consume resources originally intended for counter-terrorism operations. Commanders must distinguish between maintaining a presence, defending that presence and conducting the mission that justified it. These are different tasks, with different requirements and political implications.

The argument is particularly difficult because insurgent groups do not pause while governments debate recognition. FEWS NET's August assessment described insecurity disrupting livelihoods and trade, including in regions far from the embassy dispute. A struggle over power in Niamey can divert attention from the areas where people face armed attacks.

It would still be wrong to infer that any arrangement with the junta necessarily improves security. Resuming cooperation could provide resources and legitimacy without changing how the military governs. Conversely, ending cooperation can remove channels and capabilities that were useful. European governments need to identify the specific effects of each choice rather than assuming either contact or isolation supplies a complete policy.

The possibility of alternative external partners is part of the junta's bargaining position. Yet a meeting, a flag at a rally or an expression of interest does not prove that another provider can replace existing training, logistics and surveillance. Assessments of changing influence should follow the personnel, contracts and equipment, not only the political symbols.

For Europe, the deeper vulnerability is reliance on executive relationships that can disappear in a single domestic confrontation. Working with elected leaders remains important, but durable partnerships also need parliamentary scrutiny, civilian administration and public explanation. Otherwise a security programme can become identified so closely with one presidency that its wider constituency remains uncertain.

There is also a distinction between information-sharing and operational partnership. Diplomats may need contact with the officers controlling checkpoints to secure a journey or arrange a delivery. That does not necessarily amount to political recognition. Making this distinction explicit could preserve necessary practical channels while keeping the demand for constitutional government intact. Refusing every form of contact would give the junta more control over access without necessarily weakening its authority. The challenge is to define the purpose and limits of each engagement so that emergency arrangements do not quietly become a normalised security relationship.

Sanctions reach the market before they reach the barracks

The crisis began during a period when food insecurity was already severe. The World Food Programme reported continuing food distributions and treatment for malnutrition in August. Its work shows why humanitarian access cannot be treated as a secondary issue to resolve once the constitutional argument has ended.

Border restrictions, interrupted financial flows and uncertainty over transport can raise the cost of essentials even where the intention is to pressure military rulers. Households pay those costs through prices and availability. Officials with access to state resources may be better able to protect themselves from the same measures.

FEWS NET described disruption to trade and aid in its August outlook, alongside the continuing effects of insecurity. The analysis does not support attributing every hardship to one cause. The coup, the sanctions, restrictions on movement and pre-existing conflict interact, and the mix differs between regions.

A humanitarian exemption needs to work operationally. Traders, banks, transport companies and border officials must understand what is permitted. A rule allowing food imports has limited value if a payment cannot be processed or a driver cannot cross the border. This is where sanctions policy meets the administrative details that determine whether assistance actually arrives.

Exemptions also need monitoring. Authorities may seek to divert supplies or use access as a bargaining instrument. The answer is to improve verification and delivery arrangements, not to assume that every shipment benefits the junta. Civilian need remains a distinct consideration even when those controlling territory behave coercively.

Migration adds another dimension. Niger is a country of transit, destination and return, with movements extending well beyond Europe-bound routes. Closing borders can trap people who are trying to go home as well as those travelling north. A European policy that counts only departures towards the Mediterranean will miss much of the immediate human consequence.

Sanctions should therefore have stated political conditions and a way to adjust measures in response to verified action. If the junta cannot identify which step would produce which change, pressure becomes easier to portray as punishment of the population. Clarity does not guarantee compliance, but it makes the policy more intelligible to those affected by it.

The embassy confrontation remains unresolved at August's close, as does the larger question of Bazoum's return. Europe has a consistent recognition policy and a regional organisation it supports. Its influence will depend on whether those positions can secure access, protect civilians and produce a negotiation with consequences. The fact that France's ambassador remains in Niamey demonstrates resolve. It cannot, by itself, demonstrate that the strategy is working.