Ten decisions across half a century, but this installment does not line them up by date: it groups them by terrain - health, local government, and regulation and public employment. The Institut Thomas More report examines the medical numerus clausus (student admissions quota, 1979), the “lois Defferre” (Defferre decentralization laws, 1982), regulatory inflation (1989), the national health-insurance spending target or ONDAM (1996), the loi SRU (solidarity and urban-renewal law, 2000), tarification à l’activité (activity-based hospital funding, 2003), the creation of the regional health agencies, or ARS (2009), the surge in civil-service headcount (2012), the loi NOTRe (new territorial organization law, 2015), and the abolition of the taxe d’habitation (residence tax, 2017). That is, in fact, what makes it the densest installment in the series.
Health: four steering mechanisms that never aligned
The medical numerus clausus, created by the law of 12 July 1971 to limit the number of doctors trained, received a first correction with the law of 2 January 1979: the number of students admitted was now to take into account « des besoins de la population » (“the needs of the population”). The report sees a genuine shift here, but judges, in hindsight, that the text amounts to a « occasion manquée majeure » (“major missed opportunity”), since it « identifie le problème sans modifier les mécanismes qui structurent la politique démographique médicale française » (“identifies the problem without changing the mechanisms that shape French medical-workforce policy”). The numerus clausus hit a historic low in the late 1990s, around 3,500 admissions a year, according to the Direction générale de l’offre de soins (the health ministry’s care-provision directorate). A consequence quantified by the report: the number of practicing general practitioners fell from 91,524 in 2012 to 81,870 in 2025, a drop of 10.55% in thirteen years, according to RecoMédicales. The numerus clausus, the report concludes, acts as a « goulot d’étranglement démographique » (“demographic bottleneck”).
The 1996 Juppé ordinance of 24 April created the national health-insurance spending target (objectif national de dépenses d’assurance maladie, ONDAM), voted each autumn as part of the social-security financing law. The report judges that it « impose un pilotage centralisé, rigide et court-termiste, qui empêche les réformes structurelles nécessaires au système de santé » (“imposes centralized, rigid, short-termist steering that blocks the structural reforms the health system needs”). The record is quantified: from 1997 to 2009, ONDAM was respected only once - in 1997, the year it was created - , with some years running at twice the voted rate, according to FIPECO (an independent public-finance research body); it has been systematically exceeded every year since 2020, according to a Senate report. In 2025 it reached €265 billion, the health-insurance branch posted a €13.8 billion deficit in 2024, and the Cades (the agency that manages social-security debt) has absorbed more than €258 billion in cumulative social debt since 1996 - three figures the report attributes to the Cour des comptes (the national audit office).
Tarification à l’activité (T2A, activity-based hospital funding), introduced by the social-security financing law of 18 December 2003 and applied from 2004, replaced the block-grant system with funding proportional to the number and type of hospital stays. The report documents what it calls a « course à l’activité » (“race for activity”): by rewarding volume, T2A mechanically pushes hospitals to produce more of it. A caveat belongs here: the body of the text attributes the 15% rise in medicine, surgery and obstetrics stays between 2005 and 2018 to the Agence technique de l’information sur l’hospitalisation (ATIH, the hospital-information technical agency), while its own footnote points instead to a February 2026 article by AMAF - an internal discrepancy this installment does not resolve. The 2019 “rapport Aubert,” quoted by the report, concludes that « la tarification actuelle valorise insuffisamment la prévention, notamment secondaire et tertiaire dans le cadre de la prise en charge des maladies chroniques » (“current pricing under-values prevention, particularly secondary and tertiary prevention, in the management of chronic disease”) - conditions that, according to the report, affect 35% of the population and account for more than 60% of health-insurance spending.
The HPST law of 21 July 2009 finally created the regional health agencies (agences régionales de santé, ARS), meant to unify at regional level powers previously scattered across several institutions. The report finds the opposite of the stated goal: « la création des ARS, censée rendre le système plus agile, conduit à un alourdissement de son pilotage et à la complexification considérable de la gouvernance des hôpitaux » (“the creation of the ARS, meant to make the system more agile, has instead produced heavier steering and considerably more complex hospital governance”). Drawing on the Cour des comptes and the IGAS (the health and social-affairs inspectorate), it points to a mismatch between the scope of responsibilities given to the ARS - planning care provision, organizing patient pathways, correcting territorial imbalances - and their actual capacity to act, since the law layered onto existing mechanisms rather than replacing them.
Local government: half a century of recentralization through complexity
The first “loi Defferre” of 2 March 1982 abolished prefects’ administrative oversight (tutelle) of local authorities and transferred executive powers to the elected presidents of the general and regional councils - despite François Mitterrand’s own wariness; he reportedly warned Pierre Mauroy, according to the report: « Vous n’allez pas me ressusciter les comtes de Flandre et les ducs de Normandie ! » (“You’re not going to resurrect the counts of Flanders and the dukes of Normandy on me!”). Between 1982 and 1986, nearly twenty laws organized this “acte I de la décentralisation” (first act of decentralization), during which local-government spending rose by about 30%, according to a 1993 Insee study. The report attributes to the granting of a general competence clause (clause générale de compétence) to every tier of government the « péché originel » (“original sin”) of a poorly designed decentralization, the root of the « mille-feuille territorial » (“territorial layer cake”) - a complexity that Cannes mayor David Lisnard sums up, quoted by the report, as the « République des Cerfa » (“Republic of Paperwork Forms”).
The loi SRU of 13 December 2000 imposed a minimum 20% social-housing quota on towns of more than 3,500 residents - raised to 25% in 2013. The report finds that none of its three stated goals - social mixing, lower prices, curbing urban sprawl - has been met a quarter-century later, on any of the indicators it cites: demand for social housing rose from around one million people in 2000 to 2.2 million in 2020 and 2.8 million in 2025, according to the Cour des comptes and the Union sociale pour l’habitat (the social-landlords’ federation). The HLM (subsidized housing) turnover rate fell to 8% in 2025, from 10.5% in 2010, and average occupancy length doubled, from 8 to 17 years, according to Apur (the Paris urban-planning agency). The report also documents a geographic concentration running counter to the mixing goal: Seine-Saint-Denis, home to just 12% of the Paris-region population, holds about 17% of the region’s social housing, with rates reaching 60% in Dugny and 51% in L’Île-Saint-Denis, according to the ministère de la Ville et de la Rénovation urbaine (the Ministry for Cities and Urban Renewal).
The loi NOTRe of 7 August 2015 abolished the general competence clause for departments and regions and raised to 15,000 residents the population threshold for the intercommunal cooperation bodies (EPCI) that towns are required to join. The report calls it « l’un des pires textes en cinquante ans de décentralisation » (“one of the worst pieces of legislation in fifty years of decentralization”): it creates, in its view, an additional tier with no constitutional status as a local authority yet equipped with its own staff, budget and taxing power, driven by a technocratic « big is beautiful » logic rather than a competence-sharing plan thought through beforehand. A 2019 National Assembly information report, quoted by the report, puts the savings actually achieved at €9.5 billion - far short of the €50 billion announced.
The 2018 finance law of 30 December 2017 enacted the phased abolition of the taxe d’habitation (residence tax) between 2018 and 2021, a campaign pledge of Emmanuel Macron. The report puts the cost to the state - which compensates towns « à l’euro près » (“euro for euro”) - at around €18 billion a year, according to economist François Ecalle (FIPECO). It focuses above all on a political effect: abolishing the only tax paid by every resident of a town, whose rate was set by the mayor, severed the financial link between tenants - 40% of households, according to INSEE - and their town, which has become dependent on state grants. The report quotes Senate president Gérard Larcher, for whom the town is a « petite République dans la grande » (“little Republic inside the greater one”).
Regulation and public employment: a double inflation
The decree of 12 September 1989 tasked the Commission supérieure de codification (the codification oversight body) with « œuvrer à la simplification et à la clarification du droit » (“working toward the simplification and clarification of the law”). The report draws the opposite balance sheet: French law comprised around 366,000 legislative and regulatory articles in early 2026, up 54% in twenty years - and 84% by word count - , according to the Secrétariat général du gouvernement (the government’s general secretariat), which also tracks the code du travail (labor code) growing from 5,131 articles in 2004 to 11,301 in 2024 (+120%), the code de la consommation (consumer code) from 633 to 2,172 (+243%), and the code de l’action sociale et des familles (social and family welfare code) from 653 to 3,584 (+448%). The report cites the OECD, for which France has « une passion » (“a passion”) for codification, and which puts the administrative burden of regulation at 3% to 4% of GDP, or €90 to €120 billion a year. More than twenty simplification plans have followed one another in forty years; « tous les Premiers ministres successifs, à l’exception d’Edith Cresson, ont produit leur texte » (“every prime minister in succession, with the sole exception of Édith Cresson, has produced their own text”), the report notes.
On 10 July 2012, at the close of a “conférence sociale” (tripartite social conference), the Ayrault government ended the Révision générale des politiques publiques (RGPP, the general review of public policies), launched in 2007 by François Fillon, and replaced it with the Modernisation de l’action publique (MAP). Between 2007 and 2012, the RGPP had cut around 93,000 posts in national education, the prefectures, the finance administration, agriculture and defense - while over the same five years, local and hospital civil-service headcounts grew by 200,000 posts. The report sees this, as with public debt at the same moment, as a missed opportunity. Over 1975-2025, civil-service headcount grew 57% while the population grew 30%, according to figures combining INSEE, the DGAFP (the civil-service statistics directorate), DREES and the relevant ministries - a second aside: the report’s body text puts this at “plus 55 %” where its own table gives +57%, the only figure consistent with the headcounts it lists (3.75 million, then 5.88 million agents): +125% for local-government staff, +110% for the justice system, +90% for hospital staff, -55% for the armed forces. In international comparison, France in 2025 shows public spending at around 57% of GDP and 86 public employees per 1,000 residents, against 49% and 58 employees for Germany, according to the OECD and Eurostat. The report adds to this the proliferation of agencies: in 2012, the Conseil d’État counted 103 of them while the Inspection générale des finances counted 1,244; the budget of the state’s operators alone reached €81 billion in 2023, according to the Journal du Dimanche.
What this installment finds
Across these ten decisions, the same tension runs through all three terrains: a stated goal - clarifying the law, controlling health spending, bringing decisions closer to citizens, restoring purchasing power - against a documented outcome that, according to the report, runs the other way every time. Two internal inconsistencies are worth repeating, both flagged along the way: the body of the text attributes a figure to the ATIH while its own footnote points to the AMAF, and it puts at 55% a rise in public headcount that its own table establishes at 57%. The figures - Cour des comptes, INSEE, DREES, OECD, Eurostat, FIPECO, Apur, Ifrap - can be checked independently of the framing phrases around them, “missed opportunity,” “demographic bottleneck,” “original sin,” “big is beautiful.” Those remain the report’s own; the overall assessment belongs to the final installment.
This eight-part series offers an outside reading of the report - it is not a publication of the Institut Thomas More. Read the full report (PDF, 112 pages, in French) - 1975-2025 : les 50 décisions qui ont coulé la France, coordinated by Jean de Belot, Tarick Dali and Jean-Thomas Lesueur, Institut Thomas More, June 2026.
