French mobilisation reveals a delivery gap

On 17 October, protesters forced open the doors of Marseille’s Hôtel de Ville on the Vieux-Port, using barriers and kicking the double wooden doors while shouting “C’est la Révolution, on prend la Bastille”, before overturning a bust of Marianne, the embodiment of the French Republic. In Limoges, the town hall was set alight; its roof and clock tower collapsed.

The revolutions that swept across Europe in 1848 overthrew Louis-Philippe’s monarchy in France and challenged established rulers elsewhere. They posed two practical tests: whether those in power had addressed the grievances that mattered most, and whether enough people would defend existing institutions under pressure.

Britain entered the 1848 upheavals with a record of reform, including an expanded franchise under the Reform Act of 1832 and the repeal of the Corn Laws in 1846. Before the Chartist gathering at Kennington Common on 10 April, more than 80,000 special constables had been recruited to defend London, while troops were deployed in the capital. In France, Louis-Philippe had resisted electoral reform, and members of the National Guard refused to act against the demonstrators when the crisis came.

The value of the comparison is not that history repeats itself. It supplies two tests of institutional resilience: whether those in power have addressed the grievances that matter, and whether enough people regard the institutions as theirs to defend under pressure. What a crowd chooses to attack reveals another dimension of that relationship.

From expenditure to provision

France’s latest mobilisation was preceded by weeks of lycée protests. By 28 September, protests and school blockades in Île-de-France had been under way for a week, with some involving property damage and fireworks used as projectiles, as disruption spread to other regions.

By 7 October, Education Minister Édouard Geffray had told a parliamentary committee that 24 lycées had been burned or ransacked so badly that they could not reopen for several weeks, leaving approximately 30,000 pupils without in-person classes. On 9 October, the justice ministry reported that 6,877 people had been placed in police custody during the lycée mobilisation, 86 per cent of them minors.

Schools are meant to fulfil a civic role in forming citizens of the French Republic. At the Lycée Le Corbusier in Aubervilliers, pupils were already missing lessons because of teacher absences before the school blockade began. Reporting from the school captured the contradiction: they were missing lessons to protest against the teaching they had already lost.

Their complaints were concrete: teacher shortages, overcrowded classes, deteriorating buildings and overloaded timetables. At Le Corbusier, pupils also sought changes to Parcoursup, the national higher-education admissions platform.

The clearest illustration of the delivery gap came in the government’s staffing response. Following the opening of a national consultation on 5 October, Geffray announced on 8 October that approximately 3,000 substitute teachers would be deployed. These were existing replacement teachers without assignments, not new recruits.

According to Le Monde, Geffray credited new software with identifying 3,024 replacement teachers without assignments. The ministry had classified 114 lycées as “red” and another 400 as “orange” for staffing needs, without defining those categories. On 14 October, he told the Senate’s Cultural Affairs Committee that two-thirds of the available replacements had been assigned.

The administrative test is straightforward: why had replacement teachers remained unassigned while lessons went uncovered, and how many missed lessons did the intervention actually prevent? Institutions resolve grievances when provision reaches the people experiencing the failure, not when deployments are announced.

Fiscal constraints

These failures occurred despite high public expenditure. INSEE figures show that public expenditure stood at 57.3 per cent of GDP in 2025. OECD data for 2023 showed spending per upper-secondary pupil roughly 24 per cent above the OECD average, a comparison also highlighted in AFP’s reporting on the education divide. Teachers with fifteen years’ experience nevertheless earned 15–20 per cent less than the OECD average.

Fiscal pressure constrains the response. France has not produced a balanced budget since 1974, while pensions accounted for 24.3 per cent of public expenditure in 2025, according to the Conseil d’orientation des retraites. The IMF’s July assessment put public debt at 115.7 per cent of GDP at the end of 2025 and the deficit at 5.1 per cent, calling for spending to be reprioritised while protecting priority needs.

On 19 October, French ten-year government bond yields rose to 4.9 per cent, above Italy’s 4.58 per cent. The question is not simply whether France can spend more, but whether it can direct existing resources towards unmet needs.

Measuring discontent

The mobilisation did not begin with a conventional national campaign. Reporting tracing its origins describes a TikTok account called “V2” posting a call on 7 September for protests against fuel prices, choosing a Saturday at the start of the school holidays to encourage participation. The call subsequently spread into wider cost-of-living and education protests. Videos mentioning the date had reportedly accumulated 132 million views, according to Visibrain figures cited in that reporting.

Gilets jaunes, lycée protesters and trade unions converged around a shared date without a single national organising structure, although their demands and local arrangements differed. Timur Kuran’s work on preference falsification and social cascades offers one explanation for how participation can expand rapidly: people become more willing to express discontent when they see others doing so.

Interior Minister Laurent Nuñez reported 550,000 participants nationwide in the 17 October mobilisation. Speaking on France Inter on 19 October, he said officials had anticipated a little over 200,000 participants. The ministry’s own estimate was more than twice its forecast.

French President Emmanuel Macron called for lessons on intelligence and anticipation. But identifying organisers is not the same as measuring the conditions in which participation can cascade. The forecasting error is consistent with officials underestimating how visible grievances could draw separate groups into a mobilisation without central direction; it does not, by itself, establish why the forecast failed.

A test of public order

The attacks on town halls in Marseille, Lille, Angers, Grenoble, Calais and Limoges, as well as Nantes, targeted municipal institutions and symbols of the Republic. Limoges Mayor Guillaume Guérin spoke of hatred of institutions and everything representing state authority. The significance lay not only in the destruction, but in the targets selected.

An initial Interior Ministry tally recorded 2,397 arrests and 997 injured police officers and gendarmes, seven seriously, on 17 October. Nuñez acknowledged a loss of control in major cities. By 22 October, Justice Minister Gérald Darmanin had reported 2,700 people placed in police custody following the weekend’s violence and looting, with 217 incarcerated.

The Bouches-du-Rhône prefecture announced a department-wide ban on demonstrations for 24 October, including those unrelated to the unrest, citing security concerns and the availability of security forces. It also extended the curfew to adults in designated areas, including parts of central Marseille, from 23:00 to 05:00. On 23 October, however, the prefecture authorised a union demonstration in Marseille as an exception to the ban.

In Lyon, an initially announced blanket ban was followed by an order covering a broad central perimeter on 24 October. Authorities said they feared multiple confrontations could overwhelm the public-order operation.

In Montpellier, a declared march was prohibited and gatherings were restricted within a central perimeter. Arrangements continued to change elsewhere: Marseille withdrew its proposed blanket public-transport shutdown, while in Paris an initially agreed Bastille–Nation march became the subject of a dispute over a proposed alternative route on 23 October.

These measures did not suspend the right of assembly nationwide, but they showed the limits of the authorities’ capacity to protect it while containing further violence. In several cities, the response included prohibiting demonstrations rather than safeguarding them. Further mobilisation is planned for Saturday, 24 October. If the authorities hold the streets, that will demonstrate control of the immediate public-order threat, not that the underlying grievances have been resolved.

Unequal provision and institutional failure

Seine-Saint-Denis, which includes Aubervilliers and the Lycée Le Corbusier, exemplifies the delivery gap. Reporting from schools in and around the department describes recruitment and retention difficulties alongside inadequate facilities. An OECD education analyst interviewed by AFP described disadvantaged schools as having younger teachers, more temporary staff and higher absenteeism.

Responsibility is divided: regions are responsible for lycée buildings, while the state manages teaching personnel. Administrative responsibility is divided, but the experience of failure is not.

Seine-Saint-Denis’s poverty rate was 29.5 per cent in 2023, compared with 15.9 per cent across metropolitan France in the same statistical series. Separately, INSEE’s national figures for 2024 put the poverty rate in metropolitan France at 15.4 per cent, the joint-highest rate since the comparable series began in 1996.

Immigrants accounted for 31.2 per cent of the department’s population in 2021, the highest share of any metropolitan department. The integration question is whether immigrants and their children encounter institutions as routes to opportunity and belonging, or as repeated experiences of exclusion. Where schools, housing and access to work fail to offer a credible route into shared citizenship, delivery failure can also undermine integration.

Political beneficiaries

With France’s 2027 presidential election approaching, Marine Le Pen, of the far-right Rassemblement national, was already leading presidential voting-intention polls before the mobilisation. An Ifop-Fiducial survey for LCI, Le Figaro and Sud Radio, conducted from the evening of 17 October to 19 October, put her first-round support at 34–38 per cent across the scenarios tested, up one to two percentage points from the preceding poll.

Jean-Luc Mélenchon, of the hard-left La France insoumise, stood at 15–20 per cent across the scenarios in the official polling notice, while candidates from the presidential centre recorded the clearest declines in the reported comparison with the preceding poll. Macron’s position was already weak: Ifop’s October barometer, conducted before the national mobilisation, put his satisfaction rating at 17 per cent.

The Ifop-Fiducial polling notice also put Le Pen ahead in all five tested run-offs, from 56–44 against Édouard Philippe to 68–32 against Mélenchon. These are hypothetical voting intentions, not election predictions.

Le Pen’s strength among working-class voters predates the unrest: an Odoxa–Backbone survey published on 15 October, conducted from 24 September to 7 October, put her first-round support among manual workers at 58 per cent. The latest polling suggests she is well placed to benefit electorally from the mobilisation, although it does not establish that the unrest caused her gains. Those taking to the streets and those rewarding a political response at the ballot box need not be the same people.

Nevertheless, uncertainty over Le Pen’s eligibility complicates the electoral picture. In July, the Cour de cassation said it could rule on her anticipated appeal by early April 2027, although the timetable remained subject to change. The security stakes extend beyond the election: Le Pen has said she would maintain assistance to Ukraine for six months following her election, then end financial support.

For Britain, continuity also matters in nuclear cooperation with France. The Northwood Declaration established a UK–France Nuclear Steering Group, co-chaired by senior officials from the Élysée and the Cabinet Office, to provide political direction and coordinate work across nuclear policy, capability and operations. 

UK resilience

Britain’s debate over the French unrest has focused on its causes. Kemi Badenoch has linked the risk of similar unrest in Britain to immigration policy, while Nigel Farage has emphasised borders and migration, calling for “law, order, and borders”. Angela Rayner, speaking to Sky News, rejected Trump’s immigration explanation and described the violence as thuggery.

These competing explanations focus on who is involved and how they behave. They do not establish whether institutions are responding effectively to grievances, or whether enough people are available to sustain essential services under pressure. They argue about the spark; the security question also requires measuring the tinder.

The Strategic Defence Review and National Security Strategy call for a whole-of-society approach to defence and security, including wider civilian participation in national resilience. This envisages a collective endeavour involving the military, government, industry, civil society and the population at large.

Meanwhile, Socialist Students is promoting coordinated walkouts and youth strikes on 20 November. Such mobilisation must be assessed on its own terms. Lawful protest is not violence, nor does opposition to government policy establish unwillingness to help a community in an emergency.

A resilience plan dependent on civilian participation must establish not only how people will be mobilised, but whether they are willing and able to respond. The special constables of 1848 illustrate the historical importance of civilian participation; their modern successors provide one measurable indicator of it. England and Wales had 5,196 special constables at 31 March 2026, down 5.9 per cent in a year and well below the 2012 peak of 20,343.

Attitudinal polling raises a related but distinct question. An August 2026 Demos/YouGov poll found that net agreement with the statement “I am proud to be a British citizen” had fallen from 74 per cent in 2011 to 47 per cent in 2026. Yet pride in local communities rose from 29 to 40 per cent, suggesting that declining national pride need not imply weakening attachment to local communities.

The John Smith Centre at the University of Glasgow’s UK Youth Poll found that 51 per cent of 16- to 29-year-olds were unwilling to take up arms for Britain. An Ipsos survey, conducted from 29 September to 1 October 2026, found that 48 per cent of adults aged 18–75 said there were no circumstances in which they would take up arms and go to war for Britain. Neither finding establishes how many people would undertake civilian emergency roles.

The security concern is the gap between plans that depend on civilian participation and evidence of how many people would enrol, remain available and respond. Britain cannot assume that the allegiance and capacity those plans require will be there when called upon.

Measuring the tinder

The practical response is not to predict the next mobilisation, but to identify where institutional vulnerability is accumulating. Publish the delivery gap by French department and British local authority, using comparable measures where possible: frontline vacancies in schools, policing and social care; service waiting times; poverty and housing pressures; the time taken to acknowledge and remedy failures; and volunteer recruitment, retention and availability.

Official explanations deserve scrutiny: how quickly responsibility is attributed, what evidence supports the explanation, and whether it is revised when that evidence changes. Measurement must lead to accountability, linking each service failure to a responsible institution, remedy and deadline. Participation figures should distinguish willingness to serve from enrolment, and enrolment from readiness to respond.

The argument should remain open to challenge. Evidence that a central organising structure directed the mobilisation’s growth from the outset would weaken the spontaneous-cascade explanation, while a reversal of Le Pen’s polling gains would weaken the case that she is securing a lasting electoral benefit. In Britain, a sustained rise in national attachment or willingness to serve, measured through comparable polling, would also weaken the assessment.

Restoring order cannot, by itself, repair the relationship between institutions and citizens. France’s warning to Britain is not that similar unrest is inevitable, but that institutional delivery and public allegiance belong within the security assessment. A deployment can hold a street. Resilience depends on institutions that deliver and people prepared to stand behind them.