Key points

The right to levy taxes remains. What is weakening is the capital of trust that turns a legal obligation into a politically accepted contribution. This report distinguishes law from perception and analyzes the moment when the levy on the fruit of one’s work can be morally experienced as spoliation, once the link between levy, results and consent breaks down.

NOTE ON RIGOR: A POLITICAL INDICTMENT ONLY CARRIES WEIGHT IF IT IS FACTUALLY UNASSAILABLE

This article defends a severe thesis: in France, consent to taxation can no longer be taken for granted. The claim is not that every French person now refuses any common contribution. That would be false. The most robust surveys show, on the contrary, that a large majority still regards paying taxes and social contributions as a civic act. But they show at the same time that the concrete justification for taxation, trust in the State that spends it and the perceived quality of public services have crossed thresholds that now rule out speaking of a serene, stable and widely shared tax consent. This report therefore examines the tipping point at which a legally owed contribution can cease to be experienced as participation in a common endeavor and be felt, politically and morally, as a levy on the fruit of one’s work.

To say “consent to taxation has completely disappeared” would therefore be excessive. To say “it is eroding a little further” would, conversely, be too weak. The wording most faithful to the data is more serious: tax consent can no longer be taken for granted; it is no longer a national consensus; it must be won back.

This demand for precision also applies to spectacular examples. The French tax administration does not officially use satellites or drones to hunt down a vegetable greenhouse: the “Foncier innovant” system uses public aerial images from the IGN (the French national mapping institute) with artificial intelligence, then has anomalies checked by an official. Ukraine is not “the most corrupt country in the world”: Transparency International’s CPI 2025 gives it 36 points out of 100 and 104th place out of 182, which remains bad but does not justify that phrase. Finally, saying in 2026 that France “has no money to buy Canadair water bombers” is no longer accurate: two additional aircraft were ordered in June 2026. On the other hand, the Senate has documented the aging of the fleet and the risk of a capacity shortfall. Correcting these three shortcuts does not disarm the criticism; it makes it much harder to dismiss.

The thesis of this report: France still retains tax civic-mindedness, but it is exhausting the capital of trust that makes that civic-mindedness politically sustainable.

LET’S STOP TALKING ABOUT “EROSION”: TAX CONSENT CAN NO LONGER BE TAKEN FOR GRANTED

The most revealing signal is not 43.6%, nor 57.3%, nor €3,536 billion. It is the 51% result, provided it is read correctly. In November 2025, Elabe found that 51% of French people considered paying taxes justified because it funds public services, ten points fewer than in March 2024. The survey covered 1,000 people and, around 50%, Elabe states a margin of error of about 3.1 points at a 95% confidence level. It therefore cannot be asserted that a statistically certain majority has turned against the justification for the levy. It can, however, be asserted that the country is practically split in two on this justification, after a ten-point drop. That is enough to say that tax consent is no longer a stable political given. [1]

This result must be read alongside another, apparently contradictory figure: 79% of French people still regard paying taxes and contributions as a “civic act” in the 2025 barometer of the Conseil des prélèvements obligatoires (Council on Compulsory Levies). But in the same survey, only 22% say they trust the State to use public funds well, compared with 33% two years earlier. [2] The paradox is extraordinarily powerful: citizens still recognize the principle of the common contribution, but they no longer sufficiently believe in the quality of its use.

This is precisely the boundary this report analyzes, between tax obedience and tax consent. Obedience can be produced by law, withholding at source, audits, surcharges, seizures and penalties. Consent, in the socio-political sense, requires something else: a willingness to accept the levy because it is judged necessary, fair and linked to a legitimate collective use. The Organisation for Economic Co-operation and Development (OECD) defines tax morale as the intrinsic motivation to pay taxes and highlights in particular the role of trust in government, perceived fairness and the quality of public services in voluntary tax compliance. [29] A State can collect taxes. It cannot manufacture trust by decree.

51 %

still consider paying taxes justified because it funds public services

Elabe, Nov. 2025

87 %

believe public money is poorly used

Elabe, Dec. 2025

22 %

trust the State to use public funds well

CPO, 2025

76 %

perceive elected officials and political leaders as rather corrupt

CEVIPOF, 2026

These four indicators do not describe a passing mood. They outline a tax pact under extreme strain. And when the common contribution remains compulsory while its subjective legitimacy collapses, the risk is not only budgetary. It is the relationship between the citizen and the State that changes in nature.

1. THE TAX PACT IS NOW BROKEN IN TWO

The French debate is often badly framed. People ask whether the French “like” taxes, an almost absurd question in a country where no citizen spontaneously rejoices at handing over part of their income. The real question is whether the levy is still perceived as the legitimate counterpart of a common good. Recent polls show that this link has become extraordinarily fragile.

Elabe finds that 86% of French people believe that the amount of taxes is not well used by the public authorities. In a separate survey published a month later, 87% more broadly consider that public money is poorly used and 87% that the State is not sufficiently transparent about that use. [1][3] These are no longer majorities of discontent; they are near-consensuses of distrust.

The problem hardens further when one looks at the service delivered. Again according to Elabe, 62% of French people say they are dissatisfied with the quality of public services and 66% feel it has deteriorated in recent years. Justice, roads, hospitals, administrative services and schools concentrate high levels of dissatisfaction. [3] In other words, the taxpayer is not only saying “I pay a lot.” They are increasingly saying: “I pay a lot and I don’t see enough of what I get in return.” That is what turns tax pressure into a legitimacy problem.

The perception of the contributory pact is just as stark: 59% of French people feel they contribute more to the system than they benefit from it; only 7% think they benefit more than they contribute, and 23% think they are at break-even. [3] Here again, this measure is not an individual actuarial audit. It does not say what each household actually receives in schooling, health care, pensions, security, infrastructure or transfers. It measures something even more important politically: how the social contract is felt.

When 59% think they give more than they receive, 87% consider public money poorly used and only 51% still find payment justified by public services, the word “consent” can no longer be used as if it went without saying.

France therefore retains a strong civic core: 79% still associate taxation with a civic act. That is good news. But it is also a warning: this civic-mindedness now acts as a reserve that the State is drawing down. The more it demands, the more it must demonstrate. The more it audits, the more it must account for itself. The more it levies, the more visible results it must produce. Otherwise the citizen still obeys, but no longer truly consents.

2. FOUR BUDGET FIGURES ARE ENOUGH TO UNDERSTAND THE ANGER

The feeling of tax pressure is not a purely psychological construct. France objectively belongs to the developed economies where the weight of compulsory levies and public spending is highest. That alone does not prove that this spending is useless: a welfare state funds more pensions, health care, benefits and collective services. But it does mean that it carries an obligation to deliver results proportionate to the effort demanded.

43,6 %

compulsory levies / GDP in 2025

Insee

57,3 %

public spending / GDP in 2025

Insee

€152.5 bn

public deficit in 2025

Insee

€3,536.1 bn

public debt in Q1 2026, i.e. 117.5% of GDP

Insee

In 2025, compulsory levies amount to 43.6% of gross domestic product (GDP) according to Insee. [6] For 2024, the latest comparable year available in the revenue statistics of the Organisation for Economic Co-operation and Development (OECD), France ranked second in the OECD with 43.5%, behind Denmark at 45.2%, against an average of 34.1%. [7] France is therefore not simply a “somewhat more taxed” country: it sits at the top of the ranking of developed economies.

On the spending side, French general government accounts for 57.3% of GDP in 2025 according to the national accounts. [8] In the European comparison published by Insee based on Eurostat, France reaches 57.2%, against 49.5% for the 27-member European Union. [9] Here again, this level can be defended politically if the country embraces a more collective social model. But it becomes much harder to defend when 62% of citizens say they are dissatisfied with public services and 87% consider public money poorly used.

Above all, this level of levies is not enough to balance the books. The public deficit still reaches €152.5 billion in 2025, or 5.1% of GDP. [8] In the first quarter of 2026, public debt reaches €3,536.1 billion, or 117.5% of GDP. [10] The political equation then becomes explosive: a country levies among the highest rates in the developed world, spends far more than the European average, remains heavily in deficit and accumulates a historic debt. For the citizen, the question “where does the money go?” ceases to be a populist slogan; it becomes an elementary accounting question.

The serious answer cannot be “everything is wasted.” A large share of public spending funds pensions, health care, teachers, police officers, infrastructure, social benefits and debt interest. But the opposite answer, “there is no problem since everything serves a purpose,” is just as untenable. In an economy of this size, a few points of inefficiency represent tens of billions. The level of levies therefore obliges the State to hunt down waste with at least as much energy as it hunts down taxpayers’ reporting errors.

3. THE TAX AUTHORITY ENTERS THE AGE OF PROFILING: THE POLITICAL SYMBOL IS DEVASTATING

This is probably one of the most underestimated angles of contemporary tax consent. French taxpayers no longer face only an administration that receives returns and checks documents. They face an administration that cross-references masses of data, uses data mining, relies on predictive models and deploys artificial intelligence tools to schedule a growing share of audits.

The CFVR system, “ciblage de la fraude et valorisation des requêtes” (fraud targeting and query enhancement), is described by the DGFiP (the French public finance directorate) itself as a system that profiles the persons concerned through predictive analysis in order to identify potential tax breaches. It can draw on tax, banking and asset data, data from other administrations, social security bodies and private databases, and certain indicators from platforms or social networks; any case it proposes must nevertheless be examined by an official before any action targets the taxpayer. [12]

In February 2026, the government told the National Assembly that data analysis was the source of 50% of audits of businesses and, since 2025, of 50% of audits of individuals, reaching two years early the target initially set for 2027. It also confirmed the intensification of data exchanges between the DGFiP and social security bodies, notably the Urssaf, with quarterly exchanges since 2025. [13]

The Cour des comptes (the French Court of Audit) had already quantified the organizational break: for a large part of audit scheduling, a team of 32 people assisted by data mining replaced in 2023 a setup that mobilized about 500 people in 2018, i.e. 427 fewer full-time equivalents, with no reduction in the volume of scheduling. [14] Technically, it is a productivity success. Politically, it is also proof that the State knows how to modernize its capacity to target the taxpayer extraordinarily fast.

The Foncier innovant system produces an even more visible symbol. Contrary to the often repeated caricature, it does not officially involve satellites or drones flying over gardens. The DGFiP uses public aerial images from the IGN and algorithms to extract the outlines of buildings and swimming pools, compare them with tax and planning declarations, and then submit any anomaly to human verification. The experiment made it possible to tax more than 20,000 swimming pools that had previously gone untaxed. [11]

The caricature of the “greenhouse spotted by satellite” is false. The official reality remains politically striking: public aerial images + artificial intelligence + data cross-referencing + human review + possible taxation.

It would also be false to say that the DGFiP legally uses “counter-terrorism means.” Tax auditing and counter-terrorism share neither the same law, nor the same purposes, nor the same powers. But the comparison reflects an understandable feeling: technologies long associated, in the public imagination, with mass surveillance and intelligence — data analysis, profiling, relationship networks, aerial imagery, automation — are becoming ordinary tools of the tax administration.

The problem is therefore not to refuse all technology. Fighting fraud is a requirement of fairness, and 82% of French people even think that the public authorities do not fight tax fraud enough. [1] The problem is the symbolic asymmetry: when citizens see the State becoming extraordinarily effective at detecting an undeclared swimming pool or a tax anomaly, while at the same time reading reports on hundreds of millions of euros swallowed up by failed public projects, consent deteriorates at high speed.

4. WHEN THE TAXPAYER NO LONGER SEES A REFEREE, BUT AN OPPONENT

A tax administration can only function if it is both feared by fraudsters and recognized as legitimate by taxpayers acting in good faith. It is a delicate balance. Too much weakness feeds fraud and injustice. Too much perceived power without matching transparency feeds hostility toward the institution itself.

Here, law must be distinguished from perception. The DGFiP is not the taxpayer’s “enemy” in the institutional sense; it applies the law as voted, collects the revenue needed for the functioning of public authority and offers procedural safeguards and avenues of appeal. But an administration can be legally legitimate and politically experienced as hostile. And it is precisely this perception that a responsible State must watch.

Everyday vocabulary says a great deal: “levy,” “bludgeoning,” “racket,” “spoliation,” “confiscation.” It would be too easy to dismiss these words as mere verbal excess. They reflect a break in perception: that of a State which always knows how to ask for more, which has ever more powerful means of control, but which seems far less able to demonstrate, euro by euro, that the same demand for performance, economy and accountability applies to itself. The word “spoliation” is not here an overall legal characterization of the tax system; it describes the political feeling that arises when taxpayers no longer recognize, in the share taken from their work, a sufficiently legitimate counterpart in a common good.

The situation becomes particularly dangerous when the tax administration is technically modernized faster than public services are perceived to improve. The State knows how to profile, cross-reference, automate and send reminders. But the citizen sees court delays, hospital emergency rooms, closures of services, deteriorating roads, overcrowded classrooms or the difficulties of certain administrations. They do not compare budget nomenclatures; they compare their personal experience with the power they see deployed against breaches.

The word “confiscatory” is, moreover, not foreign to French law. The Conseil constitutionnel (Constitutional Council) recalls that the requirement of equality before public charges would not be respected if a tax were confiscatory in nature or imposed on a category of taxpayers a burden excessive in relation to their ability to pay. [32] This case law obviously does not allow French taxation as a whole to be described as “confiscatory” in the legal sense: that assessment depends on the characteristics of each tax and on the taxpayer’s situation. But it establishes an essential point: even positive law recognizes that there is a boundary beyond which the levy can become excessive.

Taxation is legally neither theft nor spoliation: it funds the common contribution provided for by our constitutional order. But this legal characterization does not settle the political question. When the State takes a considerable share of the fruit of our labor, when 87% of French people consider public money poorly used and only 22% say they trust the State to use these funds well, the common contribution can cease to be experienced as such. For a growing share of taxpayers, it becomes a levy imposed on the product of their work, to the point of taking on, in their perception, the justified and justifiable features of a genuine spoliation of the fruit of their labor. It is precisely at that moment that tax obedience persists while consent disappears.

FROM BASTIAT TO TODAY: WHEN EXCESSIVE TAXATION BECOMES SPOLIATION OR CONFISCATION

This idea was not born with contemporary tax debates. It runs through a significant part of French economic thought. These authors obviously do not all speak of the France of 2026 and their doctrines differ; quoting them therefore does not turn their texts into statistical proof of the current situation. They do show, however, that the boundary between legitimate contribution, excessive taxation, infringement of property and a feeling of confiscation is an old debate in political economy.

Frédéric Bastiat, The Law (1850): “When a portion of wealth is transferred from the person who owns it — without his consent and without compensation [...] — then spoliation is perpetrated.” [33]

Jean-Baptiste Say, Treatise on Political Economy: “All excessive taxes [...] no longer bring in new revenue, without thereby ceasing to cause new misfortunes.” [34]

Jean-Marc Daniel, Les impôts. Histoire d’une folie française (2017): “Bad taxation [...] stifles the economy [...] and sustains the feeling of confiscation and abuse.” [35]

Patrick Artus, 2024: “The rate of compulsory levies [is] markedly higher in France than in the other euro-area countries.” [36]

Comparing them with current data is therefore not a matter of making Bastiat or Say say that they had diagnosed the France of 2026. It is a matter of noting that France today combines a level of levies among the highest in the OECD [7], massive distrust about the use of funds [1][3] and very low trust in the State to use them well [2]. In this context, the feeling of confiscation is no longer a marginal word: it becomes an intelligible political consequence of the break between the effort demanded and the perceived legitimacy of its use.

5. THE TAX CRISIS IS SUPERIMPOSED ON A CRISIS OF POLITICAL TRUST

No tax system floats above the political system that decides it. Consent to taxation depends on the trust placed in those who vote spending, set rates, create tax loopholes, close schemes and explain sacrifices. Yet French political trust stands at an extraordinarily low level.

The CEVIPOF Political Trust Barometer published in February 2026 measures only 22% trust in politics. Parties fall to 15%, the President of the Republic to 18% and the National Assembly to 20%. CEVIPOF stresses that trust in national political institutions has reached some of the lowest levels in the series. [4]

The research note published a few weeks later shows something even more corrosive: 76% of respondents believe that French elected officials and political leaders are “rather corrupt,” compared with 65% in 2022 and 2023. [5] It is essential to state this figure correctly. It does not prove that 76% of political leaders are corrupt. It measures citizens’ perception. But from the standpoint of tax legitimacy, perception alone is already enough to cause considerable damage.

Why willingly accept a high levy when three quarters of the population broadly suspect those who decide how it is used? Why believe in the need for a new tax when 87% already consider public money poorly used? Why consent to additional sacrifices when 88% believe that elected officials do not take responsibility when they fail and 80% say that the government is “flying blind”? [4]

The problem is not that leaders’ corruption has been 76% proven. It has not. The problem is that a fiscal democracy cannot function durably as if the perception of corruption by three citizens out of four had no consequences.

Tax consent rests on a moral chain: I contribute because the rule is legitimate; the rule is legitimate because it is decided by institutions in which I place a minimum of trust; I accept the effort because I believe the money will be used in the general interest. When every link weakens at the same time, the chain can continue to hold legally while being politically broken.

6. 1789 HAD ALREADY WRITTEN IT ALL: CONTRIBUTE, CONSENT, FOLLOW THE USE, DEMAND AN ACCOUNT

Criticism of compulsory levies is not a denial of taxation. The 1789 Declaration of the Rights of Man and of the Citizen explicitly states that a common contribution is indispensable for the maintenance of the public force and for administrative expenses, and that it must be shared among citizens according to their means. [15] Taxation is therefore not legally theft: it is one of the foundations of the modern State and of national solidarity.

But the Declaration does not stop at the obligation. Its Article 14 gives citizens the right to ascertain the necessity of the contribution, “to consent to it freely, to follow its use” and to determine its main terms, by themselves or through their representatives. Article 15 adds that society has the right to demand an account from any public agent of its administration. [15]

A legal clarification is indispensable. In French constitutional law, the principle of consent to taxation in Article 14 does not mean that each taxpayer individually consents to each tax. The Conseil constitutionnel recalls that the provisions of Article 14 are implemented by Article 34 of the Constitution and that, in a representative democracy, tax consent is exercised mainly through Parliament, clearly informed and free in its vote. [30] This report therefore also uses the word “consent” in a socio-political sense, close to the tax morale studied by the OECD: the adherence and trust that make the levy durably acceptable. This distinction takes nothing away from the political force of Articles 14 and 15: the necessity of the contribution, following its use and rendering accounts remain at the heart of the tax pact.

This is where the current crisis takes on a historic character. The State’s right to levy taxes remains intact. But the political legitimacy of the levy is never definitively acquired. It must be reproduced in each generation through proof of usefulness, fairness and accountability. An administration can automate collection; a democracy cannot automate consent.

7. TAXPAYERS SEE PUBLIC MISTAKES, AND THEY PAY FOR THEM TOO

No large State can run thousands of projects without failures. Innovation involves risk, and an administration paralyzed by fear of error would also be costly. But some fiascos become politically toxic because they give taxpayers the feeling that the State’s mistakes do not carry the same moral price as their own.

The SIRHEN program of the French Ministry of Education became a symbol. In 2017 the Cour des comptes had warned about a project whose initial estimate of €60 million had risen to €323 million. [16] In its 2020 Annual Public Report, the Court stated that at the end of 2019 total spending exceeded €400 million and that the tool managed only 2% of Education ministry staff. [17] The exact scope of the figures varies according to dates and assessments, but the order of magnitude of the overrun is directly documented by the financial court.

In August 2026, a parliamentary question brought the police program Scribe, renamed XPN, back into the spotlight. It cites a settlement order of the Cour des comptes of October 16, 2025 assessing the overall cost of the program at €257.4 million, including development expenses and the loss linked to investigators’ wasted time; the initial budget was €11.28 million and, nearly ten years after the launch, no operational deployment had taken place according to this parliamentary document. [18] This overall figure is therefore reported here with its exact nature: it comes from a written question that attributes these findings to an order of the Court. Independently of this amount, the Cour des comptes had already documented in 2024 the structural weaknesses of the Scribe project, its freeze in March 2021, its abandonment in October 2021 and €13.3 million of spending on the program between 2016 and 2022. [31]

Here again precision is needed: according to the parliamentary question, the €257.4 million does not correspond solely to IT invoices paid to a contractor. The amount includes a broader cost, notably the loss resulting from lost operational time. It should therefore not be presented as €257.4 million of IT spending actually disbursed. For the citizen, this nuance does not remove the problem: a public failure can destroy value in the form of spending, hours of work, delays and lost opportunities.

A taxpayer who forgets a return may receive a reminder, a surcharge or an audit. When a public project overruns by tens or hundreds of millions, they ask: who personally pays for the mistake? Who answers to them?

This is not a call to punish civil servants for every project that fails. It is a demand for symmetry of accountability. The more the State perfects its instruments for detecting individual error, the more it must make visible the mechanisms that detect, stop and sanction public inefficiency.

8. “WHERE DOES THE MONEY GO?”: THE QUESTION HAS SPILLED OVER INTO EVERY POLITICAL CAMP

The unease no longer belongs to a single political family. For a long time there have been associations, think tanks, elected officials and essayists seeking to quantify the spending they consider useless. Their figures must be read as political theses or activist work when they do not come from an audit institution, but their proliferation shows that the subject has become central.

Nicolas Dupont-Aignan published in 2021 “Où va le pognon ?” (“Where does the dough go?”), subtitled “how to recover €100 billion a year of fraud and waste to rebuild France.” [24] The €100 billion figure is his political thesis, not a single official estimate certified by the Cour des comptes. But the very success of this kind of title reveals the power of a question that has become popular: why increase levies before demonstrating that every major existing expense is useful?

Sarah Knafo, in her Orange speech in August 2025, defended a counter-budget claiming €63 billion of immediate savings. [25] Here again, this is a partisan proposal each line of which must be debated, not an amount objectively available without trade-offs. But it is part of a broader phenomenon: the challenge to spending no longer concerns only the tax rate; it concerns the very architecture of the State.

Contribuables Associés published in 2025 “Le Livre noir de l’argent public” (“The Black Book of Public Money”) by Jean-Baptiste Leon, presented as a collection of about a hundred cases of waste. [26] GenerationLibre published in March 2026 “Le grand gaspillage” (“The Great Waste”), devoted to municipal subsidies in Paris, Lyon and Marseille, with €473 million of spending analyzed and a rationalization scenario costed by the think tank at €185 million. [27] This work is not of the same nature as a judicial audit. It must therefore never be presented as official truth. But ignoring it would be just as absurd: it shapes public perception of waste.

The central point is not whether all these estimates are accurate to the euro. The point is that a State already facing massive tax distrust can no longer afford to answer these criticisms with contempt. It must open its data, publish results, compare costs, explain failures and demonstrate why each euro levied could not have been saved before asking for the next one.

9. UKRAINE: ONE CAN DEFEND SUPPORT AND UNDERSTAND THE TAXPAYER’S ANGER

The debate on aid to Ukraine perfectly illustrates the difference between a legitimate strategic decision and how it is received fiscally. Since the Russian invasion of February 2022, France has chosen to support Ukraine for reasons of European security, respect for the sovereignty of an attacked State and stability of the continent. This policy can be coherently defended. But French taxpayers cannot be forbidden from comparing the billions committed abroad with the needs they see unmet at home.

The Ministry of the Armed Forces stated that as of May 1, 2024 France had delivered €3.035 billion of military equipment to Ukraine and paid €2.1 billion to the European Peace Facility, i.e. more than €5.135 billion of military support within that scope. [19] In March 2025, the President of the Republic announced €2 billion of additional military support. [20] As of July 1, 2026, the Ministry for Europe and Foreign Affairs also counted €471.6 million of French humanitarian aid since the start of the full-scale war. [21] These figures must not be added up mechanically as if they all belonged to the same budget scope and the same period, but they show the scale of the effort.

The question of Ukrainian corruption must also be dealt with without propaganda. Calling Ukraine “the most corrupt country in the world” would be false. Transparency International’s 2025 Corruption Perceptions Index gives it 36 points out of 100 and ranks it 104th out of 182 countries. [22] That is a poor position, far from Western European standards, and Transparency International Ukraine itself highlights the persistent difficulties of reforms. But several dozen countries are ranked lower.

This reminder does not cancel the taxpayer’s question: when a heavily indebted country, among the most taxed in the developed world, devotes billions to a foreign policy, the government must explain with extreme rigor why this effort is a priority, how it is controlled and what mechanisms prevent misappropriation. The weaker domestic trust is, the stronger the demand for external traceability must be.

The contrast with national civil protection resources is often used simplistically. Saying today that “France has no money to buy Canadair water bombers” is inaccurate: on June 4, 2026, the State ordered two new aircraft for nearly €200 million, after other investments in the aerial fleet. [23] But the Senate had documented that as of July 1, 2025 only 9 of the 12 Canadair CL-415s were operational, with an average age of 27.5 years, and that simultaneous major fires had revealed a risk of capacity shortfall. [28]

The serious criticism is therefore not “the State gives everything to Ukraine and nothing to the French.” It is harsher because it is true: a State under budgetary pressure must rank competing emergencies, and it loses consent when it no longer convinces the population that this ranking is coherent, controlled and compatible with the fundamental needs of the territory.

10. TAX OBEDIENCE AND TAX CONSENT: THE DIFFERENCE THAT CAN TIP A COUNTRY OVER

A tax system can remain extremely effective at collection even as its political consent collapses. Withholding at source, pre-filled returns, automated data exchanges and collection procedures make taxes hard to avoid. But this efficiency can mask a more dangerous phenomenon: the transformation of a contribution deemed legitimate into a levy merely endured.

The CPO also notes a certain growing understanding toward avoidance behaviors. In its 2025 barometer, only 25% of respondents believe it is “never justified” to move abroad to reduce one’s taxes, while 9% consider such expatriation always justified. [2] These answers must not be extrapolated into a prophecy of generalized fraud. They nonetheless signal a normative shift: certain behaviors once perceived as a breach of civic duty are increasingly excused.

This is where the State takes a historic risk. The more it replaces consent with the automation of levies, the longer it can delay the visible manifestation of the crisis. Revenue keeps coming in. Audits work. Databases grow richer. Algorithms target better. But beneath this apparent efficiency, legitimacy can drain away.

A fiscal democracy is not just a system capable of collecting. It is a system in which a sufficiently broad majority accepts the principle and the destination of the contribution. When that majority falls to around half on the very justification for payment, when trust in the State drops to 22% and the perception of misuse exceeds 85%, the problem is no longer a matter of communication. It is a matter of the architecture of the tax pact.

11. WHY THE FRENCH NO LONGER WANT TO “PAY MORE”

It would be tempting to reduce tax anger to selfishness: everyone wants public services but no one wants to fund them. Surveys show a more complex reality. A majority of French people still call for more spending on health, security, education, justice or defense. [3] They are therefore not mechanically demanding a minimal State.

What they contest is the combination of a high levy, an impression of waste and a public return judged insufficient. They want cuts to consulting firms, the State’s operating expenses, institutional communication, certain public agencies and waste, while a majority refuses any reduction of everyday public services. [3] The message is therefore not “abolish the State.” It is “stop confusing the useful State with the apparatus that feeds itself.”

This nuance is fundamental. It means that winning back tax consent does not only require a uniform cut in all spending. It requires a readable hierarchy. Citizens more easily accept a high level of contribution if they see the hospital working, justice ruling within reasonable time frames, schools educating, security protecting, infrastructure being maintained and those responsible answering personally for major overruns.

Conversely, every misunderstood expense becomes fiscally toxic. A subsidy judged absurd, abandoned software, an empty building, an agency with vague missions, a disproportionate consulting contract, a scheme created and then abolished: taken separately, each may represent an infinitesimal fraction of the budget. But politically, they add up in the taxpayer’s memory and build a conviction: “I am being asked to make efforts that the State does not impose on itself.”

12. WHAT THE STATE SHOULD DO IF IT WANTS TO AVOID A LASTING BREAKDOWN OF THE TAX PACT

The answer cannot be a communication campaign explaining to the French that they do not understand the complexity of public finances. When 87% say public money is poorly used, the problem is no longer changing the slogan. The proof must change.

First requirement: make every major expense traceable and understandable.

Citizens should have access to a genuine public “tax receipt,” not a marketing infographic, but a standardized presentation showing where €100 of levies goes, which results are associated with the main policies and how those results evolve. Each major mission should publish its objectives, its full cost, its performance indicators and the gaps with what was promised.

Second requirement: impose a “stop-loss” doctrine on the State for major projects.

Any digital, real estate or industrial program exceeding a significant threshold should be subject to mandatory stop points. When cost or schedule drifts beyond a predefined threshold, the project must be publicly revalidated, restructured or stopped. The cost of abandoning it must be compared with the cost of continuing. Major public failures must no longer be able to survive for years through institutional inertia.

Third requirement: apply the same culture of detection to waste as to fraud.

The analytical power that the DGFiP applies to tax anomalies shows that the State knows how to use data. Why not impose the same ambition on detecting administrative duplication, overpaid public services, off-contract purchases, subsidies without results, under-used buildings or drifting digital projects? Taxpayers would more readily accept tax AI if they simultaneously saw AI become a hunter of public waste.

Fourth requirement: democratically audit tax algorithms.

Predictive profiling is not trivial. The models that steer half of the audits of certain categories should be subject to independent review on bias, false positives, taxpayers’ rights, data security, proportionality and the possibility of challenging a human decision based on an algorithmic recommendation. Performance cannot become the sole criterion of legitimacy.

Fifth requirement: no increase in levies without a prior review of the corresponding spending.

Whenever a government proposes a new tax or a rate increase, it should simultaneously publish the savings studied, the spending cut, the schemes evaluated and the reasons why the need cannot be financed otherwise. This is not a universal economic rule; it is a rule of trust. Taxpayers must see that the new tax comes as a last resort, not as a reflex.

Sixth requirement: make accountability visible.

Accountability is not necessarily criminal. It can be administrative, budgetary, managerial and political. But a public project that drifts massively must have an identified person in charge, a public timeline, documented decisions and professional consequences when management faults are established. Perceived impunity is the direct enemy of tax consent.

13. THE DANGER IS NOT A “TAX REVOLT”: IT IS THE NORMALIZATION OF A COUNTRY THAT PAYS WITHOUT BELIEVING IN IT ANYMORE

The most dangerous scenario is not necessarily a spectacular tax insurrection. France has known outbursts of anger linked to taxation in its history, but the contemporary State has collection mechanisms so powerful that a breakdown of consent can remain invisible in revenues for a long time.

The real danger is quieter: a country where people pay because they cannot do otherwise; where taxation ceases to be associated with the common good and psychologically becomes an imposed loss; where aggressive tax optimization becomes morally acceptable; where tax expatriation ceases to be stigmatized; where every public expense is presumed suspect; where the administration is perceived as an opponent; where political trust collapses and the State responds with more control.

This spiral feeds itself. The less citizens trust, the more the State controls. The more the State controls without restoring trust, the more citizens feel suspected. The more they feel suspected, the less they perceive taxation as a voluntary contribution to the common good. And the more consent decreases, the more the administration must rely on technical constraint rather than adherence.

That is why the phrase “consent to taxation is eroding” no longer matches the data. It conveys the image of slow erosion at the edges of a pact that is still solid. Yet the pact is now fractured at its center: only 51% still find payment justified by public services; 87% think public money is poorly used; 22% trust the State to use the funds well; 76% perceive political leaders as rather corrupt.

This is no longer a crack. It is an alert at the level of the tax regime.

Taxation remains legal. It remains indispensable. Fraud remains reprehensible. But none of these three truths answers the central political question: how long can a democracy durably levy at such a high level when surveys simultaneously show a very large majority distrusting the use of public money, strong dissatisfaction with the services delivered and very low trust in the institutions that decide on spending?

The 1789 Declaration did not reduce the citizen to the act of paying: it linked the contribution to its necessity, to consent exercised by citizens themselves or through their representatives, and to the right to follow its use. In contemporary constitutional law, this consent is expressed mainly through the vote of Parliament. [15][30] Two centuries later, France is rediscovering the political force of this link between contribution, representation and accountability.

The State can automate tax returns. It can generalize withholding at source. It can cross-reference data. It can profile tax risk. It can analyze aerial images. It can increase penalties. It can improve collection.

But there is one thing that no algorithm at Bercy (the French finance ministry) will ever be able to collect by force: trust.

And when tax trust is lost, taxpayers do not necessarily stop paying. They stop believing that they are paying for a common endeavor. The amount levied then becomes again, in their concrete experience, a share of the fruit of their labor of which they are deprived before even being convinced of its use. It is at this precise point that the contribution, still legal and compulsory, can be morally experienced as spoliation: not because taxation could legally be equated with theft, but because the link between levy, results and consent has broken. The State then ceases to be perceived as the organizer of the common good and becomes, for a growing share of the population, the one that levies before having convinced.

At this stage, asking for still more without a radical reform of spending, transparency and accountability is no longer merely imprudent. It means risking the destruction of what no budget law can recreate in an emergency: the very consent that makes taxation bearable in a democracy.

WHAT THE ARTICLE ASSERTS — AND WHAT IT DOES NOT

Statement Assessment
Tax consent can no longer be taken for granted. SUPPORTED ANALYTICAL THESIS. The 51% result places opinion at near-equality, given the poll’s margin of error. The ten-point drop, the massive distrust about the use of funds and the decline in trust support the phrase “can no longer be taken for granted,” without allowing the claim that it has completely disappeared.
All French people now refuse to pay taxes. No. 79% still regard payment as a civic act.
Taxation is legally theft. No. The common contribution is constitutionally grounded and indispensable. The report distinguishes this legal characterization from the political judgment that, when consent and trust collapse, the levy on the fruit of one’s work can be experienced as a genuine spoliation.
The report asserts that French taxation as a whole is legally confiscatory.No. It argues that, given the level of levies and the measured distrust of their use, the feeling of spoliation can be politically and morally justified. In law, the confiscatory character is assessed according to the characteristics of the tax and the ability to pay; the Conseil constitutionnel explicitly recognizes this limit. [32]
The DGFiP uses satellites or drones to find garden sheds. No, as regards the official system cited. Foncier innovant uses public aerial images from the IGN and human verification.
Ukraine is the most corrupt country in the world. No. The 2025 Corruption Perceptions Index gives it 36 points out of 100 and ranks it 104th out of 182. The level remains worrying, but this phrase would be false.
France is not buying Canadairs. False in 2026: two additional aircraft have been ordered. The documented difficulties concern age, availability and capacity risk.
All political leaders are corrupt. No. The 76% figure measures respondents’ perception, not the criminal reality of elected officials.

SOURCES AND REFERENCES

Polls measure opinions and perceptions, not individual facts. Work by associations, political figures and think tanks is cited as a contribution to public debate and not as official audits. Budgetary and institutional data are preferred where they exist. The term “tax consent” is distinguished, in the body of the text, from its strict constitutional meaning; likewise, the term “spoliation” is used in a political and moral register when it does not refer to a precise legal characterization. Historical quotations are used as landmarks of economic thought, not as statistical proof about contemporary France. Main links and figures consulted and checked: August 21, 2026.

[1] Elabe, “Les Français, le budget et la fiscalité” (The French, the budget and taxation), November 5, 2025 — source — 86% consider taxes poorly used; 51% consider payment justified by public services. Sample of 1,000 people; around 50%, stated margin of error of about 3.1 points at a 95% confidence level.

[2] Conseil des prélèvements obligatoires / Cour des comptes, 2025 Barometer — source — 79% “civic act”; trust in the State to use funds: 22%.

[3] Elabe, “Les Français et l’utilisation de l’argent public” (The French and the use of public money), December 3, 2025 — source — 87% public money poorly used; 62% dissatisfied with services; 59% think they contribute more than they benefit.

[4] CEVIPOF, 2026 Political Trust Barometer — source — 22% trust in politics; very low levels for national institutions.

[5] CEVIPOF, research note, March 2026 — source — 76% perceive elected officials and political leaders as corrupt.

[6] Insee, compulsory levy rate, May 2026 — source — 43.6% of GDP in 2025, provisional figure.

[7] OECD, Revenue Statistics 2025 — source — France 43.5% in 2024, 2nd in the OECD; OECD average 34.1%.

[8] Insee, general government accounts in 2025 — source — Spending 57.3% of GDP; public deficit €152.5 bn.

[9] Insee / Eurostat, public spending in the European Union, July 2026 — source — France 57.2% of GDP in 2025; EU-27 49.5%.

[10] Insee, public debt in the first quarter of 2026 — source — €3,536.1 bn; 117.5% of GDP.

[11] DGFiP, rollout of Foncier innovant — source — Public IGN aerial images, AI, detection of buildings/pools and human verification.

[12] DGFiP, presentation document of the CFVR system — source — Profiling through predictive analysis and categories of data processed.

[13] National Assembly, answer to written question No. 10400, February 10, 2026 — source — Data analysis at the origin of 50% of business and individual audits; exchanges with social security bodies.

[14] Cour des comptes, AI at the Ministry of the Economy and Finance, 2024 — source — CFVR: productivity gains and tax audit scheduling.

[15] Declaration of the Rights of Man and of the Citizen, Articles 13 to 15 — source — Common contribution, consent and the right to demand an account.

[16] Cour des comptes, SIRHEN program, March 6, 2017 — source — Estimated cost risen from €60 M to €323 M at that stage of the project.

[17] Cour des comptes, 2020 Annual Public Report, “Le système d’information des ressources humaines de l’Éducation nationale : une modernisation dans l’impasse” (The Education ministry’s HR information system: a modernization at a dead end) — source — More than €400 M of spending at the end of 2019; SIRHEN then managed only 2% of Education ministry staff.

[18] National Assembly, written question No. 17697, August 11, 2026 — source — Parliamentary question citing a settlement order of the Cour des comptes: €257.4 M of announced overall cost, including development and lost operational time; project presented as not operationally deployed.

[19] Ministry of the Armed Forces, review of military support to Ukraine as of May 1, 2024 — source — €3.035 bn of equipment + €2.1 bn EPF = more than €5.135 bn within that scope.

[20] Élysée, meeting with Volodymyr Zelensky, March 26, 2025 — source — Announcement of €2 bn of additional military support.

[21] France Diplomatie, humanitarian aid to Ukraine — source — €471.6 M as of July 1, 2026.

[22] Transparency International Ukraine, CPI 2025 — source — Ukraine: 36 points out of 100, 104th out of 182.

[23] Ministry of the Interior, forest fires, June 5, 2026 — source — Order of two additional Canadairs, nearly €200 M.

[24] BNFA, bibliographic record of Nicolas Dupont-Aignan, “Où va le pognon ?” — source — Political essay published in 2021; thesis of €100 bn of fraud/waste.

[25] Sarah Knafo, Orange speech, August 2025 — source — Counter-budget claiming €63 bn of immediate savings.

[26] Contribuables Associés, “Le Livre noir de l’argent public” — source — Activist presentation of about a hundred cases.

[27] GenerationLibre, “Le grand gaspillage,” March 2026 — source — Analysis of municipal subsidies; the think tank’s own scenarios.

[28] Senate, 2026 Budget Bill, Civil Protection — source — Canadair fleet: average age 27.5 years, 9 of 12 operational as of July 1, 2025; capacity risk.

[29] OECD, “Tax Morale: What Drives People and Businesses to Pay Tax?” — source — Tax morale is the intrinsic motivation to pay taxes; trust in government, perceived fairness and the quality of public services are among the factors of voluntary tax compliance.

[30] Conseil constitutionnel, decision No. 2010-5 QPC of June 18, 2010 and commentary — source — The principle of consent to taxation in Article 14 is implemented by Article 34 of the Constitution; case law links it to the consent of Parliament, not to the individual agreement of each taxpayer.

[31] Cour des comptes, “Le pilotage de la transformation numérique de l’État par la direction interministérielle du numérique” (Steering of the State’s digital transformation by the interministerial digital directorate), 2024 — source — Scribe project: structural weaknesses identified; frozen in March 2021, abandoned in October 2021; €13.3 M of spending between 2016 and 2022.

[32] Conseil constitutionnel, decision No. 2026-1189 QPC of March 27, 2026 — source — Article 13 of the 1789 Declaration would be disregarded if a tax were confiscatory in nature or imposed a burden excessive in relation to the ability to pay.

[33] Frédéric Bastiat, The Law, 1850 — source — Definition of “legal plunder” and analysis of the infringement of property when the law organizes a transfer without consent or compensation.

[34] Jean-Baptiste Say, Treatise on Political Economy, Book III, Chapter X — source — Criticism of excessive taxes and their economic and social effects.

[35] Jean-Marc Daniel, Les impôts. Histoire d’une folie française, Tallandier, 2017; passage quoted in a 2018 review — source — Bad taxation can stifle the economy and sustain a feeling of confiscation and abuse.

[36] Patrick Artus, “Quelle réforme fiscale en France ?” (What tax reform for France?), Cercle des économistes, October 29, 2024 — source — Notes a compulsory levy rate markedly higher in France than in the other euro-area countries.

EDITORIAL NOTE

This article is a critical analysis of public policy. It explicitly distinguishes established facts, perceptions measured by polls, legal characterizations and political or moral judgments. It makes no individual accusation of corruption, misappropriation or offense in the absence of a ruling or a competent source. When it uses the notions of “spoliation” or “confiscation” outside their strict legal sense, it presents them as a political characterization of the relationship between the level of levies, perceived results and consent. The economic data cited may be revised by statistical bodies.