What restoring means - The European rule of law after the Hungarian alternation

Introduction: a deadline and its paradox

By 31 August 2026, the Hungarian government must have demonstrated what Brussels calls, in its manager’s vocabulary, “super-milestones”: the restoration of judicial independence, effective anti-corruption arrangements, the reform of public procurement. Hungary will then have one month to submit the evidence; the Commission will analyse it until 20 November; and if the examination concludes favourably, 16.4 billion euros of frozen funds will begin to flow before the end of the year1. This timetable proceeds from the agreement concluded on 29 May between the Commission and Péter Magyar, who had become Prime Minister twenty days earlier2. The National Assembly passed a first anti-corruption package on 23 June3.

There is something dizzying about the speed of this sequence, and it is what detains us. On 12 April, Hungarian voters ended sixteen years of government by Viktor Orbán, in proportions no poll had fully anticipated4: 53.56% of the vote for the Tisza party, a turnout of 79.5%, and above all 138 seats out of 199 — the two-thirds majority, the very one that had allowed Fidesz to rewrite the Constitution in 2011. Orbán conceded on election night. Seven weeks later, a defrosting plan had been signed.

We should like to take this sequence seriously — that is, not to content ourselves with welcoming it. For it puts to European constitutional law a question it has, in truth, never had to decide in a live case: what does it mean to restore the rule of law when its degradation was itself a work of law — accomplished through statutes regularly passed, appointments regularly made, constitutional revisions regularly adopted? And by a happy accident of the calendar, Europe has for its reflection not one case but two: at the moment Hungary begins its restoration armed with a constitutional majority, Poland is pursuing its own, begun in December 2023, without ever having possessed that instrument — and finds itself, thirty months on, largely immobilised. Two countries, one enterprise, opposite means: political reality rarely composes so clean an experiment.

Capture as a work of law

To measure what restoring demands, one must first describe exactly what was undone — and the manner of its undoing. That manner is the heart of the problem.

The transformation of the Hungarian state between 2010 and 2026 did not proceed by force. It proceeded through law, and this is precisely what legal scholarship came to call, with Kim Lane Scheppele, autocratic legalism: the methodical use of legal forms to dismantle the guarantees those forms were meant to protect5. A two-thirds majority, obtained in 2010 under the electoral rules then in force, adopted a new Fundamental Law in 2011; that Fundamental Law multiplied “cardinal laws”, amendable only by qualified majority, in fields — taxation, pensions, the status of judges, the media — that no constitutional tradition ordinarily reserves to that rank; and a series of long-mandate appointments placed loyalists at the head of the oversight institutions for terms exceeding, by design, several parliamentary cycles. The electoral terrain itself was reshaped — constituency boundaries, a majority premium, a media landscape absorbed by allies — to the point that international observers described elections that were free but deprived of fair conditions of competition. Each step, taken in isolation, had regularity on its side. The whole produced what jurists came to call constitutional entrenchment: a state of the law such that any future majority, however absolute, would collide with sitting mandates, qualified thresholds and pre-empted institutions.

That entrenchment was no side effect; it was the very object of the operation. A legal capture of the state differs from a dictatorship precisely in this: it organises its own electoral survival — it makes alternation costly to obtain, then sterile to exercise. The literature that documented it — Scheppele for Hungary, Wojciech Sadurski for Poland, Laurent Pech for the European dimension of the phenomenon — converges on this point: the decisive test of such a regime is not its longevity but what becomes of its locks on the day it loses6.

That day has come, in both countries, and the locks have held very differently.

A decade of instruments, built to point the other way

Before examining the two restorations, we should recall where the instruments that now claim to measure them came from — for their origin explains their present embarrassment.

The European rule-of-law toolkit was built by sedimentation, each layer answering the failure of the last. The “rule of law framework” of 2014, a mere dialogue mechanism, was the Commission’s response to the discovery that the treaties provided nothing between the isolated infringement action and the nuclear weapon of Article 7. Article 7 itself was activated in its preventive limb against Poland in December 2017, on the Commission’s initiative, then against Hungary in September 2018, on Parliament’s — and both procedures promptly silted up in a Council where each of the two defendants could count on the other’s vote, a full-scale demonstration that a unanimity-minus-one mechanism does not work when the defendants are two7. Then came the annual Rule of Law Reports, from 2020, which document without compelling; and, the same year, the true innovation: Regulation 2020/2092, which for the first time ties the disbursement of European funds to respect for the rule of law where its breach threatens the Union’s financial interests. Hungary and Poland challenged its validity; the Court of Justice, sitting in full court — a rare solemnity — dismissed their actions on 16 February 2022, and the mechanism was activated against Hungary before that year was out, leading to the suspension of cohesion commitments to which the blockages under the Recovery Facility were added8.

This genealogy teaches two things. The first is that every instrument was conceived, calibrated and tested to measure a degradation: to identify breaches, freeze disbursements, document regressions. None contains a symmetrical procedure for certifying a recovery — no exit criteria, no methodology for verifying a restoration, no case law of the thaw. The second is that this asymmetry was no oversight: nobody, in 2020, was working on the hypothesis that a captured regime might lose an election. The European instruments tacitly shared their adversaries’ prognosis — that of irreversibility. The Hungarian alternation wrong-foots them as much as it wrong-foots Fidesz: the Union must improvise the grammar of recovery with tools that have only ever conjugated in the passive.

The Polish precedent, or restoration without instruments

Poland voted first. In October 2023, a coalition led by Donald Tusk prevailed over the Law and Justice party after eight years in power; the new government took office in December, carrying an explicit commitment to restore judicial independence and to mend relations with the European institutions9.

Thirty months later, the legislative record of that restoration can be stated in a few lines, and the fewness is instructive. The centrepiece — the reform of the National Council of the Judiciary, the judicial appointments body whose composition had been handed to the parliamentary majority in 2017, in defiance of what both the Court of Justice of the Union and the European Court of Human Rights have since held — was passed by the Sejm, then vetoed by the President of the Republic10 on 19 February 2026. Karol Nawrocki, elected in 2025 with the support of the party now in opposition, has also declined to sign forty-six judicial appointments, on the ground that the candidates supported the government’s “unconstitutional actions”11. The government, which possesses neither the three-fifths majority needed to override a veto nor a Constitutional Tribunal that is not itself a product of the contested period, can neither push through nor go around. The Polish restoration thus hangs on the presidential election of 2030 — or on sub-legislative expedients whose legal fragility feeds, in turn, the accusation of acting by the very means one claims to correct.

For that is the second Polish lesson, more uncomfortable than the first. Unable to legislate, the Tusk government has acted through administrative decisions, resolutions, the handling of appointments and budgets — in the public media from December 2023, in the prosecution service, in the management of institutions whose composition it contested. The takeover of the public media in December 2023 was effected through the detour of ordinary company law, up to and including the liquidation of the entities concerned, while the rulings of a Constitutional Tribunal regarded as irregularly composed were treated as nullities. Each of these acts could claim a defensible reading of the law in force; several rested on foundations that jurists sympathetic to the restoration themselves judged acrobatic. The Venice Commission, consulted in various respects, has held a constant line in its opinions: correcting past illegalities does not dispense with forms, and the fate of judges appointed under the contested rules — the “neo-judges” — cannot be settled wholesale without regard to individual situations12. The dilemma is real and seems to us to require stating without pretence: treating each tainted appointment one by one will take a decade; annulling them collectively would inflict on legal certainty exactly what the capture had inflicted on it. There is no clean solution to this dilemma, only more or less avowable trade-offs — and the quality of a restoration is perhaps judged first by what it avows.

The Union, for its part, did not wait for the dilemma to be resolved. As early as February 2024, the Commission released the essentials of the Polish funds — up to some 137 billion euros across all envelopes —; in May 2024, it closed the Article 7 procedure opened in 2017, in view of a “new situation” and a programme of reforms then barely begun — and whose essentials remain, as we have just seen, unaccomplished13. We shall return to what that precocity reveals.

The Hungarian temptation, or instruments without precedent

The Hungary of 2026 presents itself as the exact inverse. Where Tusk lacks everything, Magyar has everything: a two-thirds majority permits revision of the Fundamental Law itself, the rewriting of the cardinal laws, the overhaul of the entrenched institutions. The constitutional steamroller Fidesz built is intact; it has merely changed drivers.

That is a historic opportunity, and it is a logical trap. For if the new government uses the two-thirds instrument without restraint — revising the Constitution at forced march, replacing the holders of long mandates with its own people, redrawing the oversight institutions to its convenience — in what respect will its work be formally distinguishable from the work it claims to undo? The objection is not a rhetorical device of the old regime, though the old regime will not fail to make it; it goes to the heart of what the rule of law means. If formal legality sufficed, the work of Fidesz was legal; if it does not suffice, then one must say what, beyond legality, separates a restoration from a revenge — failing which the concept of the rule of law reduces to the approval of whoever wins.

We believe that dividing line exists, and that it can be stated with the help of three criteria which doctrine and the practice of the Venice Commission allow us to ground14.

The first concerns procedure. The capture was carried out by surprise, without consultation, by amendments passed at night and statutes voted in urgency; a restoration worthy of the name imposes upon itself the delays, the opinions, the publicity and the adversarial debate it is restoring. Time forgone is here the proof given. In this respect, the Hungarian calendar of the coming months will provide an indicator of rare legibility: a judicial package passed after referral to the Venice Commission does not resemble a package passed in three days.

The second concerns destination. The decisive question, for each revision, is whether it returns power to independent institutions or transfers it to the new incumbents. Re-establishing a pluralist judicial council, returning to the rank of ordinary statute what should never have been cardinal, shortening mandates one could have filled oneself: these are acts that relinquish power. Their opposite — new twelve-year mandates, new proprietary foundations, new qualified thresholds cut to one’s own measure — would reproduce the entrenchment while changing its beneficiaries.

The third concerns reversibility, and it is the most counter-intuitive. A successful restoration must be capable of being undone by a subsequent alternation playing by the ordinary rules; that is, indeed, how one recognises it. The success of the Magyar government will be measured not by the solidity of the positions it will have acquired, but by the fact that in 2030 a simple majority — whichever it may be — will find a state governable without a qualified majority. The most telling indicator will fit in a single number: the count of cardinal laws at the close of the parliamentary term, compared with their count at its opening. If it falls, the two-thirds majority will have served to make itself less necessary; if it holds or grows, it will have served what two-thirds majorities in Hungary have always served.

There remains, in Hungary as in Poland, the problem the criteria do not dissolve: that of persons. Hungarian entrenchment is not made only of texts; it is made of incumbents. A prosecutor general elected for a long, renewable mandate by qualified majority alone; constitutional judges appointed for twelve years by successive Fidesz majorities; a media council cast from the same mould; and that late innovation, the public asset management foundations, to which universities and considerable patrimonies were transferred, their boards filled, for life or nearly so, before the alternation — so many positions expressly designed to survive an electoral defeat, and surviving one at this very moment15. Shortening those mandates by statute invites the same objections from legal certainty as were raised against Warsaw; respecting them means cohabiting for years with the old regime’s garrisons at the control points. The two-thirds majority gives Budapest a resource Warsaw lacks — constitutional revision permits what ordinary statute forbids — but it also sharpens the temptation, since everything is possible. We shall refrain from settling what those concerned will themselves have to arbitrate; we observe only that the way this precise point is handled will tell, better than any speech, which of the two logics has prevailed.

What the Union rewards

There remains the third actor, who is also the judge: the European Union. The instruments it forged during the decade of regression — the conditionality mechanism of Regulation 2020/2092, the milestones of the Recovery Facility, Article 7, the annual Rule of Law Reports — face for the first time their inverse test: measuring not a degradation but a recovery. Their behaviour in this new configuration calls for a critical observation, which the juxtaposition of the two files makes hard to dismiss.

The Hungarian file itself already offered a troubling precedent, prior to the alternation. In December 2023, on the eve of a European Council at which the opening of accession negotiations with Ukraine required Budapest’s assent, the Commission had released 10.2 billion euros to the Orbán government, on the ground that the agreed judicial reform had met its milestones. The European Parliament saw in this a bargain and took the matter to the Court of Justice16. Whatever the outcome of that action, the episode had established one thing: the calendar of thaws follows the Union’s political needs at least as much as the member state’s reforms. What follows should be read in that light.

In the Polish case, the essentials of the funds were released and the Article 7 procedure closed while the legislative restoration had not begun to succeed — nor, as we have seen, has it since. In the Hungarian case, a 16.4-billion defrosting plan was agreed twenty days after the government took office, before a single reform had been voted. Verification exists, and we do not minimise it: the milestones of 31 August, the examination until 20 November, the staggering of disbursements. But the direction of movement precedes the proof. What the European instruments reward, in both cases, is not an observed state of the law; it is a declared trajectory — to put it more bluntly, a political alignment.

Two readings of this finding are possible, and honesty requires giving each its force. The first sees in it a disqualifying weakness: if the funds follow the colour of governments rather than the state of institutions, conditionality is not a legal instrument but a premium on friendship, and its case law will one day be turned against the Union — the first hostile government to come along will observe that the money flowed for reforms promised, not accomplished. The second reading answers that an incentive is worth nothing unless timely: a restoration government faces adversaries entrenched in the state apparatus, an impatient public and narrow budgetary margins; haggling over European support on the ground that not everything is yet accomplished would weaken the only force capable of accomplishing anything at all. The Commission has manifestly embraced the second reading — in Poland yesterday, in Hungary today.

The same interrogation holds for Article 7, whose Hungarian limb, opened in September 2018, now awaits its dénouement. The Polish precedent argues for swift closure: it was in May 2024, a few months after the Tusk government took office and on the strength of a trajectory more than a record, that the Council closed the procedure opened in 2017. Transposed to Budapest, that precedent would close the file before winter. We should like, however, to point out the paradox such a closure would carry: the Article 7 procedure addresses a “clear risk of a serious breach” of the values of Article 2, and that risk does not vanish with the government that embodied it — it persists in the mandates, the foundations and the cardinal laws so long as they remain. To close the procedure because the new incumbents are well disposed would confirm that the instrument was aimed at a government rather than at a state of the law; to keep it open against a friendly government, until the locks are actually lifted, would conversely be the best proof that the Union can tell the two apart. We do not hide from ourselves that the second branch has little chance of prevailing; merely formulating it allows one to measure what the first concedes.

We shall not arbitrate this dispute in the abstract, for it will not be settled in the abstract: it will be settled on 20 November. If the examination of the Hungarian evidence is conducted with the rigour of an audit — if milestones not met entail disbursements deferred, even at the price of a crisis with a friendly government — then the instrument will have shown that it measures something other than alignment, and the precedent will hold for all future governments, whatever their colour. If the disbursements follow the calendar whatever becomes of the evidence, the instrument will have revealed its nature, and the doctrinal consequences will have to be drawn: European conditionality sanctions enemies and credits friends — a conceivable policy, but one that should then no longer be stated in the vocabulary of law.

Conclusion: the test of truth

The Hungarian alternation has been welcomed, rightly, as good news for the European rule of law. We should like, at the end of this journey, to shift the accent slightly: it is above all its test of truth — and doubly so.

A test for the new governors, first. Poland has been demonstrating for thirty months that a restoration without instruments silts up in vetoes and expedients; Hungary is about to demonstrate what becomes of a restoration that possesses every instrument. Between the impotence that corrupts forms by dint of circumvention and the omnipotence that corrupts them by dint of ease, the path is narrow, and no one in Europe has yet walked it: the Hungarian case has no precedent, which also means it will set one. The three criteria we have proposed — the procedure one imposes on oneself, the destination of the power one takes back, the reversibility of what one builds — are nothing like an algorithm; they form, we believe, the minimal grid for distinguishing, in the months ahead, the work of restoration from its imitation.

A test for the Union, second, and perhaps more so. The Orbán years established that its instruments could — slowly, imperfectly — sanction a degradation. The coming months will establish whether they can verify a recovery, or merely register a rallying. The distinction may seem byzantine; it will nonetheless decide what the signature of Article 2 of the Treaty will be worth for decades to come. For the state captures of the past decade prospered on a simple discovery: the European guarantees were declaratory. Restoration offers the Union the occasion to demonstrate the contrary — that its conditions are verified, even among friends, above all among friends. The appointments are fixed, and they fit in a few dates: 31 August and 20 November for the Hungarian verification; the second attempt at reforming the Council of the Judiciary, or its abandonment, for Poland; and, for whoever wants a simple measure of the Hungarian trajectory, the count of cardinal laws at the opening and at the close of the parliamentary term. On 20 November, in examining Budapest’s evidence, the Commission will not be judging Hungary alone. It will be judging itself, and we shall then know what restoring means.

Bibliography

Institutional and normative sources

Treaty on European Union, Articles 2 and 7.
Regulation (EU, Euratom) 2020/2092 of the European Parliament and of the Council of 16 December 2020 on a general regime of conditionality for the protection of the Union budget.
Venice Commission, Rule of Law Checklist, CDL-AD(2016)007, March 2016.
Council of the European Union, closure of the Article 7(1) TEU procedure in respect of Poland, May 2024.
European Parliament, resolutions on the situation in Hungary and the freezing of Union funds, 2023-2024.

Press and monitoring (April-August 2026)

France 24, “Législatives en Hongrie : le parti de Magyar obtient une supermajorité des deux tiers au Parlement”, 12 April 2026; Euronews, same date, on Viktor Orbán’s concession.
La Libre Belgique, “Le Premier ministre Magyar et la Commission s’accordent sur un plan pour dégeler 16,4 milliards d’euros destinés à la Hongrie”, 29 May 2026; Euronews, 29 May 2026.
Franceinfo, “Les députés hongrois approuvent des mesures anticorruption, avec le dégel des fonds européens en vue”, 23 June 2026.
Notes from Poland, “President vetoes bill reforming judicial body at heart of Poland’s rule-of-law crisis”, 19 February 2026; RTBF and Euractiv on the refusal of forty-six judicial appointments, 2026.
Fondation Jean-Jaurès, “Après le scrutin du 12 avril 2026, année zéro pour la Hongrie ?”, 2026; German Marshall Fund, “In Uncertain Waters: The Restoration of the Rule of Law in Poland”, 2026.

Scholarship

SCHEPPELE K. L., “Autocratic Legalism”, University of Chicago Law Review, vol. 85, 2018.
SADURSKI W., Poland’s Constitutional Breakdown, Oxford, Oxford University Press, 2019.
PECH L. and SCHEPPELE K. L., “Illiberalism Within: Rule of Law Backsliding in the EU”, Cambridge Yearbook of European Legal Studies, vol. 19, 2017.
CLOSA C. and KOCHENOV D. (eds), Reinforcing Rule of Law Oversight in the European Union, Cambridge, Cambridge University Press, 2016.
MÜLLER J.-W., “Should the EU Protect Democracy and the Rule of Law inside Member States?”, European Law Journal, vol. 21, 2015.

Notes

On the 31 August deadline, the one-month period for submitting evidence, the Commission’s examination until 20 November and the prospect of disbursements before year-end: press coverage of the May 2026 agreement, cited in the bibliography. The detail of the milestones belongs to implementing documents we have not been able to consult directly.
La Libre Belgique and Euronews, 29 May 2026. The figure of 16.4 billion euros aggregates funds frozen under the conditionality regulation, cohesion policy and the Recovery and Resilience Facility; before the alternation, some 6.3 billion in cohesion commitments and 10.4 billion under the Facility remained blocked.
Franceinfo, 23 June 2026.
Official results of the 12 April 2026 election: Tisza, 53.56% of the vote and 138 seats out of 199; Fidesz, 37.86% and 55 seats; turnout 79.50%. The Magyar government took office on 9 May 2026.
SCHEPPELE K. L., “Autocratic Legalism”, art. cited.
Besides the works cited, see PECH L. and SCHEPPELE K. L., art. cited, on rule-of-law backsliding as an incremental and legally clothed process.
The Article 7(1) TEU procedure was initiated in respect of Poland by the Commission on 20 December 2017 and in respect of Hungary by a European Parliament resolution of 12 September 2018. The former was closed in May 2024; the fate of the latter, after the alternation of April 2026, had not been formally settled at the time of writing.
CJEU, full court, 16 February 2022, Hungary v Parliament and Council, C-156/21, and Poland v Parliament and Council, C-157/21. The Council implementing decision suspending cohesion commitments in respect of Hungary under Regulation 2020/2092 was adopted in December 2022.
The Tusk government took office in December 2023, following the elections of October.
Notes from Poland, 19 February 2026. On the judgments concerning the National Council of the Judiciary as recomposed in 2017, see the case law of the Court of Justice and of the European Court of Human Rights referred to by the same source; their enumeration would exceed the scope of this note.
RTBF and Euractiv, 2026. The head of government publicly denounced a veto cast “blindly”; we report the phrase as a statement of position.
The Venice Commission’s opinions on the treatment of contested judicial appointments in Poland advocate an individualised approach respectful of acquired situations; we render their general orientation, the detail varying from text to text.
On the closure of the Article 7 procedure in respect of Poland in May 2024 and the release of funds in the first half of 2024. The earlier release of 10.2 billion euros to Hungary, decided in December 2023 by the Commission and challenged by the European Parliament before the Court of Justice, belongs to the same file; the outcome of that action was not known to us at the time of writing. The Polish release of February 2024 was announced by the Commission as amounting to up to some 137 billion euros across all envelopes.
Venice Commission, Rule of Law Checklist, doc. cited. The three criteria formulated here — procedure, destination, reversibility — constitute our synthetic reading of that grid applied to the case of a restoration; they do not appear there in this form.
We describe here categories of arrangements abundantly documented by the literature and the institutional reports cited; the precise count of current mandates and their expiry dates calls for case-by-case verification which this contribution does not claim to provide.
Commission decision of December 2023 concerning some 10.2 billion euros; action for annulment brought by the European Parliament in March 2024. The outcome of that action was not known to us at the time of writing.