France's Controlled Experiment: What LFI, NPA-R, and NPA-A Reveal When the Far Right Goes After French Women
Postscript: why LFI leads despite everything
TL;DR Three organisations, three strategic weaknesses, one question the French municipal elections have sharpened without resolving. LFI’s republican universalism repeatedly dissolves women’s specificity into a class subject that doesn’t resemble most women’s lives. NPA-R’s workerist cadre model produces correct analysis of the pension reform and sub-3% results wherever it stands. NPA-A’s united front orientation is vindicated in Alençon and Toulouse but absent in Hénin-Beaumont and Douai, the places the RN is actually winning women. The RN, meanwhile, has moved from conquest to consolidation. Bardella is offering to step aside for the traditional right where it leads: hegemony-building, not just vote-winning. The second round is 22 March. The strategic question it will not resolve is which formation, if any, is building care-economy infrastructure in the places where precarious women actually live — and that question will still be open on 23 March. A postscript explains why LFI leads despite everything.
France has become a country of laboratories. The first is well documented: between 2002 and 2024, the Rassemblement national transformed itself from a party women avoided into one they now vote for at rates matching or exceeding men. The second laboratory has received less attention. In 2022–23, the Nouveau Parti Anticapitaliste split into two formations with opposed strategic orientations. La France Insoumise had already staked out a third path. Three organisations, three models, one question: which of them is actually contesting the terrain where the RN recruits women?
The answer was uncomfortable in advance of 15 March. It remains uncomfortable after it. None of them, not fully. But the first round of the 2026 municipal elections has now made the reasons why legible in results, not just in theory. Perpignan re-elected Louis Aliot in the first round, without needing a second. Hénin-Beaumont, Hayange, Fréjus: the same story. In dozens of intermediate towns, RN slates lead the count. Jordan Bardella is proposing that the RN withdraw where the traditional right leads — a confident move, the move of a party that has finished the territorial conquest phase and is now managing the consolidation.
The differences between the three formations are not incidental. They go to the strategic DNA of each one.
What the RN is actually doing
Start with what the RN is actually doing. The closure of the gender gap, from a gap of six to ten points in 2002, depending on the survey, to effective parity by the 2024 European elections, did not happen through accident or cultural drift.¹ It happened because Marine Le Pen ran a decade-long project to make the party legible to women in precarious economic situations: women in service-sector employment, rural women without access to proximity public services, single mothers in peri-urban zones where the left’s local presence has been hollowing out since the 1980s. The RN’s rhetorical offer to these women combines purchasing power, neighbourhood security, and a particular deployment of secularism as a proxy for women’s protection. It does not stand up to scrutiny on any of these: the party’s voting record on equal pay directives, the Istanbul Convention, and care worker pay is one of consistent abstention or opposition. But the rhetorical offer is addressed to real anxieties. The 2026 municipalities are the laboratories in which the RN is proving that its governance model can survive contact with actual power — a 74% satisfaction rate in previously won cities is not nothing. The left’s response, by contrast, has addressed a different question.
¹ Post-election surveys vary on 2002: Ipsos/Vizzavi recorded 21% of men vs 11–12% of women in the first round; other analyses give 20% men vs 14% women — a six to ten-point gap depending on methodology. By June 2024, Ipsos/Talan for France Télévisions and IFOP both recorded effective parity, with some surveys showing women marginally exceeding men. See Nonna Mayer, Ces Français qui votent Le Pen (2002); Mayer’s subsequent work confirms the gender effect had largely vanished by 2012.
The moral barrage and its exhaustion
The dominant left response to far-right advance has been what the research literature calls the barrage républicain: a strategy of moral condemnation, anti-racist mobilisation, and appeals to republican values designed to construct a cordon sanitaire around the RN. This strategy was never without merit. But it has accumulated a long history of failure in the constituencies that matter most here, and that history demands some precision.
The classical barrage républicain was a cross-party mechanism: the left stood down for centrists and centrists stood down for the left, the whole operation underwritten by an appeal to republican values that in practice meant subordinating the left’s programme to Macronist management of the crisis. That version reached its exhaustion point in 2022. The 2024 legislative elections produced something different in form: the NFP coordinated withdrawals internally, the left standing down for the left rather than for Borne or Darmanin. Whether that represents a qualitative break with the barrage républicain logic or a narrower version of the same strategy is precisely what the 2026 municipal elections are now testing. The NPA-A’s post-first-round statement names the obstacle clearly: PS and Place Publique sectarianism is blocking the left-only version of unity that would distinguish it from the cross-party capitulation of 2022. If that distinction cannot be maintained in practice, the two strategies collapse into each other.
What both versions share, and what neither has resolved, is the experience of the strategy from the receiving end. In peri-urban and deindustrialised constituencies, being asked to vote tactically against the RN rather than for something has felt, for two decades, like a lecture from a distant state. Not a lecture precarious women necessarily rejected on its terms; one they had stopped listening to because it consistently failed to speak to the cost of their commute, the closure of their local maternity ward, or the question of who looks after the children when the shift finishes late.
Ethnographic work, particularly the Collectif Rosa Bonheur’s research on centralité populaire in peripheral cities, captures this with precision. For women in these communities, politics is not an abstract field of rights and values but is deeply embedded in the reproduction of daily life. When left discourse prioritises the rights of minorities without addressing the material infrastructure those women depend on, it does not produce converts. It produces the specific resentment the RN has learned to address, framing itself as the only party that talks about what actually matters.
The SOS Racisme model exemplifies the failure. Built on symbolic cultural events and high media access, it was a product of the 1980s PS ecosystem and had already exhausted its capacity to reach the working class by the early 2000s. The contrast with the Maison des Femmes network could not be more stark. Where SOS Racisme mobilised identity, the Maisons des Femmes, anchored in Saint-Denis and now present in Clermont-Ferrand and beyond, mobilise through service delivery: healthcare, legal aid, psychological support, economic empowerment, concrete solidarity rather than abstract solidarity. In high-RN areas, SOS Racisme has retreated to symbolic presence. The Maisons des Femmes have stayed. That is why they work.
LFI and the republican cage
La France Insoumise represents the largest attempt at a left alternative in contemporary France and the most instructive failure. Its programme on women’s material interests is not empty: universal public childcare, mandatory pay transparency, a billion euros annually for domestic violence services, pension reform resistance explicitly framed around women’s broken careers. On paper, this is a “for” politics rather than an “against” politics.
The practice is different. Three moments define the pattern.
First, the hijab. LFI’s positioning on secularism has consistently collapsed the gendered and racialised dimensions of the debate into a question of republican state neutrality. By doing so, it has systematically failed to support the specific struggles of Muslim women in the workforce and public space, and has reinforced precisely the laïcité-as-women’s-protection framing that the RN uses to recruit them. This is not a tactical error; it is a structural consequence of subordinating feminist analysis to a republican universalism that treats the state as a neutral arbiter of competing claims rather than an active participant in exclusion.
Second, the Quatennens case. In September 2022, Adrien Quatennens acknowledged that he had struck his wife. Mélenchon’s response was to publicly renew his “confidence and affection” for his deputy. Forty-five LFI deputies subsequently voted to reinstate Quatennens after a four-month suspension; fifteen voted against. Clémentine Autain documented what followed with precision: the generational split within the party, where it was consistently the younger members who could not understand the decision, the consequences “already devastating in feminist movements and among young people,” and the departure of LFI members in significant numbers. The denouement was baroque. In June 2024, Quatennens was reinstated while Autain, Alexis Corbière, and Raquel Garrido were purged for insufficient loyalty. Autain’s comment was direct: in LFI, it is now better to have been convicted of domestic violence than to have disagreed with the leadership on questions of democracy or anti-Semitism. She has since left the movement entirely.
Third, the 2023 pension mobilisation. The gendered dimensions of the reform were not invented by the far left: the government’s own impact assessment acknowledged that women would have to work seven months longer on average than men, because of broken careers; that women’s pensions were already forty percent lower than men’s; and that the minimum pension uplift would largely exclude part-time workers, of whom twenty-eight percent are women. Révolution Permanente documented this with force. LFI’s campaign framing, by contrast, subordinated the specifically feminist analysis of the reform to a generalised defence of “the social state” and the “theft of life.” Women’s pensions appeared as a secondary illustration of an argument about the working class, not as the primary terrain of a feminist reading of austerity.
The structural pattern in all three cases is the same. LFI’s universalism repeatedly dissolves women’s specificity into a class subject that resembles, in practice, the traditional male industrial worker. Its “for” politics is for the people, with the people defined in ways that systematically flatten what is specific about women’s oppression.
Now we have a fourth test, live. In Paris on 15 March, Sophia Chikirou ran an autonomous LFI list at 11.7%, explicitly refusing the Socialist-led bloc. She qualified for the second round. The question now posed to her, by 22 March, is whether she merges with Emmanuel Grégoire’s 38% or maintains her independence. The republican cage thesis, laid out above, makes a confident prediction: LFI’s republican universalism will struggle to articulate a reason for unity that doesn’t simply reproduce the barrage républicain logic it claims to have superseded. Bardella is watching.
NPA-R and the workerist transposition
NPA-Révolutionnaires emerged from the 2022–23 NPA split committed to building a cadre party on the Lutte Ouvrière model: workplace implantation, ideological coherence, strict independence from reformist forces. Its analysis of women’s material interests is, on the specific question of the pension reform, genuinely sharp. Its January 2023 tract named the mechanisms with precision: broken careers, time-partial work, the 40% pension gap, the government’s claimed “feminist” measures exposed as applicable only to complete careers that most women will not have. NPA-R understands, at the level of analysis, why the pension reform is a feminist issue.
The problem is not the analysis. It is the organisational logic that follows from it. NPA-R approaches a supermarket, a hospital, or a care home as if it were a car plant. The LCR made this critique of Lutte Ouvrière in 1971, citing Lenin’s argument in What Is To Be Done? that class consciousness cannot be built through economic struggle alone but requires engagement with the full range of social contradictions. It remains accurate. For the modern female proletariat, the decisive terrain is not the factory floor but the reproduction of daily life: the school run, the care rota, the question of who covers the emergency. NPA-R’s model, however correct its analysis of women’s pension position, has no organisational answer to this question. It builds cadres; it does not build community.
The 2026 municipal results confirm this in numbers. NPA-R fielded independent “Workers’ and Revolutionary” lists in over twenty cities, including Paris (Blandine Chauvel), Lyon, and Marseille. In Lyon, “Lyon Ouvrière et Révolutionnaire” captured approximately 1.5% of the vote. Nationally, no standout figures emerged. Under 2–3% wherever they stood, no runoff qualifications, no territorial breakthrough. The workerist transposition produces correct analysis and marginal politics. Every time.
The NFP leadership has called this “playing into the hands of Le Pen.” That charge is too simple: the NPA-R’s analysis of the PS as a decades-long administrator of austerity is not wrong. But the strategic conclusion, total independence, means contesting terrain where the RN is weakest (the urban left-wing university ward, the transport hub, the hospital entrance) while ceding the terrain where the RN is strongest: the peri-urban housing estate, the deindustrialised northern town, the places where women in precarious work have been waiting for a left that speaks to them since Jospin.
NPA-A and the partial exception
NPA-Anticapitalistes is the closest to an answer, and the most instructive in its limitations. Through its participation in the NFP and its organic relationship with autonomous feminist movements including Nous Toutes and feminist strike collectives, NPA-A has maintained what the LCR tradition always insisted on: that feminist struggle is not a secondary contradiction to be resolved downstream of the economic question, but a constitutive dimension of anti-capitalist politics. Philippe Poutou’s articulation of the NFP strategy is not merely electoral. It is about occupying the space where reformist forces have retreated, maintaining an independent anti-capitalist pole while influencing the wider movement to go on the offensive.
The 15 March results vindicate this orientation at the local level. In Alençon, a united Gauche Unie list that included NPA-A secured 15.2%. In Toulouse, the list led by François Piquemal won over 25%. In Gérardmer, NPA-A comrades secured 19.5%. These are not transformational numbers on a national scale. They are, however, exactly what a united front strategy looks like when it works: anti-capitalist forces retaining their identity while building the broadest possible alliance against the right.
The NPA-A statement issued on the night of 15 March was, on this score, sharper than any of the formation’s recent national communications. It did not celebrate its own results. It named the threat: “the entrenchment and dangerous rise” of the RN, dozens of towns and cities where RN slates lead, Aliot re-elected in Perpignan without needing a second round. Then it named the responsibility, clearly: “where the right and far right can win or retain municipalities, it is the responsibility of left-wing lists that have come out on top to create the conditions for unity.” And then it named the obstacle: “the sectarianism of the PS and Place Publique is driving the entire left into a dead end.” That is not barrage républicain language. That is united front logic, applied in real time, under electoral pressure.
What NPA-A does not have, or not yet, is the territory. Its 2026 performances are understandably concentrated in cities where it already had organisational presence, where it had movement relationships to build on. Good results in Alençon and Gérardmer, not in Douai or Hénin-Beaumont. The Maison des Femmes model is not NPA-A’s model: it is a different organisational form, rooted in service delivery rather than political mobilisation, building what the Collectif Rosa Bonheur calls centralité populaire through years of concrete presence. NPA-A has the feminism. It has the movement relationships. It has, now, a demonstrated capacity to win within united front frameworks. What it does not have is organisational geography in the places where the RN is winning women.
The revelation: what the controlled experiment shows
Here is what twenty years of strategic argument on the French far left has not yet produced: an organisation that combines NPA-A’s feminist politics and united front instincts with the Maisons des Femmes’ territorial presence. The service delivery model is doing more to build durable ties with working-class women in high-RN zones than any of the three formations. Not because it avoids politics, but because it starts from material life, from the healthcare appointment, the legal advice, the childcare barrier to work and political participation, and builds from there toward something that might, in time, be called political loyalty. It is Lenin’s ‘tribune of the people’, translated into the terrain of deindustrialising technofeudalism.
The care economy — childcare, eldercare, domestic violence services, care worker pay — is the primary terrain. NPA-R names this in its analysis of the pension reform and then returns to the factory. LFI names it in its programme and then subordinates it to republican universalism. NPA-A has the right political instincts and the wrong organisational geography. The RN names it in its campaigns and votes against it in every chamber where it has a vote.
March 15 has sharpened the question without resolving it. The RN is now in a consolidation phase. Bardella’s offer to withdraw in favour of the traditional right where the latter leads is the move of a movement that has internalised Gramsci without acknowledging him: it is building hegemony, not just winning votes. The left’s response, in the five days between 15 and 22 March, will be conducted in the familiar register of “will they unite.” The answer matters electorally. It does not answer the strategic question: which of these formations is building the kind of organisational presence, in the places where women in precarious situations actually live, that outlasts an election?
No, that is not the whole answer. Anti-racist politics remains essential. The feminist strike, which mobilised around 8 March in direct anticipation of the first round, remains essential. The united front is not a mistake. But these are necessary conditions, not sufficient ones. Until the revolutionary left takes the care economy as seriously as it takes the industrial workplace, and until it builds organisational geography to match its programmatic commitments, the Maison des Femmes in Saint-Denis will be doing more to contest the terrain the RN has occupied than three parties combined.
The controlled experiment is not over. Round two is 22 March. But the first results are in.
Postscript: why LFI leads despite everything
The analysis above stands. The Quatennens case, the hijab, the pension mobilisation, the Bonapartist internal culture: none of that has been answered by 15 March. But the results demand an honest addendum. LFI has emerged from the first round in a structurally stronger position than the PS, despite every justified criticism levelled at it, and that fact requires explanation rather than avoidance.
The numbers first. In 2020, LFI fielded 75 municipal lists. In 2026, it fielded 504, deploying more than a quarter of its parliamentary deputies as list candidates. The result: qualification for the second round in 96 communes, doubled, tripled or in some cases quadrupled scores compared to 2020, and outright victory in Saint-Denis in the first round, with Bally Bagayoko winning 50.77% against the Socialist incumbent. Roubaix, a city with 100,000 inhabitants and 62% abstention, has LFI’s David Guiraud at 46.6%, far ahead of anyone else. In the three largest French cities: Paris 11.7%, Marseille 11.94%, Lyon 10.41%. Qualification in eight of the ten largest cities in the country.
More significant than the headline figures is the geography. Over 60% of LFI’s qualifying lists are in banlieues populaires. This is not the movement-based urban feminism of NPA-A’s comfort zone. This is the peri-urban and suburban working-class territory the article identifies as the RN’s primary recruiting ground among women. LFI is, slowly and unevenly, building the municipal presence the article says is missing from the revolutionary left. It is doing so with all its political pathologies intact. That is the uncomfortable finding.
The PS, meanwhile, is in the kind of crisis that makes Faure’s pre-election posturing look like a man shouting at a tide. His national line — no accord between the PS and LFI — lasted approximately twelve hours before the local reality imposed itself. Toulouse: LFI’s Piquemal finished ahead of the PS candidate, and a joint list was announced within twenty-four hours, with LFI taking the mayoral candidate and the PS taking the métropole presidency. Limoges: the PS stands down behind the LFI candidate. Nantes: Johanna Rolland, among the PS’s most prominent anti-LFI voices, concluded a fusion deal by Monday morning because the arithmetic left her no choice. Brest, Clermont-Ferrand, Besançon: the same story repeated. Marine Tondelier of Les Écologistes put it with characteristic directness: those in the PS refusing alliance are “aspiring to be kings of the cemetery.”
The chaos runs in every direction. In Strasbourg, the PS incumbent Catherine Trautmann arrived first in the first round, watched the outgoing Écologiste mayor fuse with LFI, and responded by allying with the centrist Horizons list, for which Faure expelled her from the PS. Hollande, publicly demanding no alliance with LFI, was quietly contradicted by one of his own former advisers making precisely such a deal in Corrèze. The PS national line and PS local practice have never been further apart.
What explains LFI’s advance? Three things, none of which require revising the political critique.
First, organisational investment. Mélenchon sent his deputies door to door in municipalities where the left had no recent presence. This is substitutionism of a recognisable kind: the party apparatus substituting for the missing social infrastructure. It works electorally in the short term. It does not build the Maison des Femmes in Saint-Denis. But it builds council seats, and council seats build resources, visibility and local presence that the revolutionary left will spend years trying to match.
Second, banlieue sociology. LFI’s base in Seine-Saint-Denis and comparable territories is rooted in communities, working-class, racialised, often Muslim, that the PS managed for decades without transforming and that the revolutionary left has addressed primarily in the register of solidarity rather than political representation. Bagayoko’s 50.77% in Saint-Denis is not just an electoral number: it is the capture of an institutional resource by a constituency that was previously administered by a party that no longer speaks to it. The political limitations of LFI’s programme for that constituency, the hijab question above all, have not yet been paid for at the ballot box.
Third, the PS’s accelerating decomposition. A weakening rival concentrates its vote losses on the most competitive terrain, and that terrain increasingly belongs to LFI. The PS is not losing to the right uniformly: it is losing to LFI in its urban working-class bastions and to the centre-right in its bourgeois liberal constituencies. Both losses are structural, not conjunctural.
None of this resolves the strategic questions the article poses. LFI’s banlieue presence does not address the peri-urban and deindustrialised zones where the RN recruits women: it addresses a different working-class geography. The Quatennens legacy has not been settled; the feminist movements that experienced it as devastating are still watching. And the Bonapartist internal culture means that the municipal gains of 2026 will be managed by an apparatus without democratic accountability to the communities that produced them.
But the revolutionary left has to reckon with a formation that, for all its failures, is building the kind of local institutional presence that the article identifies as decisive. Dismissing LFI as simply the republican cage with better results misses what has actually happened. The question is not whether LFI’s political limitations are real. They are. The question is what the revolutionary left offers to a constituency that is currently voting for Bagayoko in Saint-Denis and Guiraud in Roubaix, and what it would take to build something better in those same streets.
Afterword: what the days before the second round confirmed
This article was published before the second round. Two findings from the between-rounds period are worth recording as a coda.
First, Sébastien Delogu’s withdrawal in Marseille. After three weeks of LFI treating Benoît Payan as the primary enemy, Delogu withdrew his 12% to leave a clear left vote against the RN. Payan’s response was to deposit his second-round list at the prefecture before opening negotiations with LFI. The article argues that PS sectarianism is the primary obstacle to the left-only version of the barrage républicain. Marseille is the case study in miniature: LFI held the united front line; the PS refused it until electoral arithmetic made refusal impossible.
Second, the data on where LFI fuses. Mediapart’s breakdown shows a sharp gradient: in cities over 100,000 inhabitants, LFI maintained independently in ten contests and fused in eleven. In cities under 25,000, it maintained in fourteen and fused in two. The fusion strategy is overwhelmingly urban and metropolitan. In small towns — precisely the intermediate-city terrain where the RN recruits women, which is this article’s central subject — LFI either has no presence or no bargaining power. The political scientist Rémi Lefebvre added the sharpest observation of the between-rounds period: the alliances are forming where the threat is the moderate right (Toulouse, Limoges), but where the far right is the primary threat (Marseille) or the hard right leads (Paris), there is no alliance. LFI’s leverage is greatest where PS incumbents fear the moderate right, not where the RN is the real danger.
Both findings confirm the article’s argument rather than revising it. LFI is building successfully in its own geography: urban, banlieue, places where its demographic coalition is strongest. The peri-urban and deindustrialised terrain where the RN recruits working-class women remains, after the second round, largely outside its reach. The controlled experiment continues.
Notes
¹ Post-election surveys vary on 2002: Ipsos/Vizzavi recorded 21% of men vs 11–12% of women in the first round; other analyses give 20% men vs 14% women — a six to ten-point gap depending on methodology. By June 2024, Ipsos/Talan for France Télévisions and IFOP both recorded effective parity, with some surveys showing women marginally exceeding men. See Nonna Mayer, Ces Français qui votent Le Pen (2002); Mayer’s subsequent work confirms the gender effect had largely vanished by 2012.
Collectif Rosa Bonheur, centralité populaire: developed in Sociétés Contemporaines and elaborated in Prendre la ville (La Découverte, 2022). The Autain documentation is from Franceinfo and Le JDD (public statements, 2022–2024). The Révolution Permanente pension analysis was published January–April 2023.