From annulled elections in Romania to re-run referendums in Denmark and Ireland, a single architecture emerges to ensure that popular sovereignty never disturbs the integration project
Editorial Analysis | November 2026
The historical record of European referendums reveals that this logic is not a recent innovation but a constitutive feature of the integration project itself. Danish voters rejected the Maastricht Treaty in June 1992, and the treaty was accepted in a second referendum in May 1993 after the negotiation of four opt-outs (Danish Parliament, 2019). Irish voters rejected the Nice Treaty in June 2001 by a margin of 54 per cent to 46 per cent, and a second referendum in October 2002 approved it by 62.89 per cent to 37.11 per cent (BBC News, 2002). Irish voters rejected the Lisbon Treaty in June 2008 by 53.4 per cent to 46.6 per cent, and a second referendum in October 2009 approved it by 67.1 per cent to 32.9 per cent (House of Commons Library, 2026). French and Dutch voters rejected the European Constitution in May and June 2005, and most of its substantive provisions were subsequently enacted through the Lisbon Treaty without any further popular vote (Cato Institute, 2005). Greek voters rejected the creditors’ bailout terms by 61.3 per cent to 38.7 per cent in July 2015, and the Greek government signed a memorandum of understanding containing further austerity measures within days of the result (LSE, 2015). The pattern across four decades is unambiguous: when voters reject a treaty, the treaty is renegotiated at the margins and resubmitted; when they reject a constitution, its contents are repackaged as an ordinary treaty and ratified by parliaments; when they reject austerity, the government accepts austerity anyway.
The Romanian case in 2024 represents the most direct application of this logic to a candidate rather than a treaty. Calin Georgescu, an independent candidate with no party apparatus, won the first round of the presidential election on 24 November 2024 with 22.94 per cent of the vote, and the Constitutional Court annulled the entire electoral process on 6 December, two days before the second round was due to be held (Xinhua, 2024). The court’s decision rested on declassified intelligence alleging Russian interference and irregular campaign financing, and the Directorate for Investigating Organised Crime and Terrorism opened a criminal investigation into Georgescu’s campaign (Xinhua, 2024). The annulment was justified by reference to the integrity of the electoral process, and its practical effect was to remove from contention a candidate whose programme threatened the consensus on continued alignment with Brussels and the war effort in Ukraine. Georgescu was subsequently barred from standing in the re-run election, which completed the removal that the annulment had begun. The sequence is instructive because it demonstrates that the legal machinery of electoral integrity can be deployed to produce an outcome that the electorate declined to produce.
The Hungarian case operates through a different mechanism but achieves a comparable result. Viktor Orbán was defeated in the April 2026 parliamentary election by Péter Magyar’s Tisza party, and the new parliament used its two-thirds majority in June 2026 to amend the constitution so that no person who has already held the premiership for eight years may hold it again (MIA, 2026). The amendment applies retrospectively from 1990, which means that it disqualifies Orbán specifically while imposing a general rule that also binds Magyar after two terms (Telex, 2026). The measure was presented as a term-limits reform of the kind common in presidential systems, and its design ensures that a politician who dominated Hungarian politics for two decades cannot return through the ballot box. The constitutional amendment is not a ban on a party or a criminal conviction but a structural alteration of the eligibility criteria, and it demonstrates that the containment of populist challenges can be achieved through constitutional engineering as effectively as through judicial action.
The German case illustrates the use of the intelligence apparatus and the party-ban procedure as instruments of political exclusion. The Federal Office for the Protection of the Constitution classified the Alternative for Germany as a confirmed right-wing extremist organisation in May 2025, a designation that carries significant consequences for state funding, surveillance and public legitimacy (Der Rechte Rand, 2025). A Cologne administrative court temporarily prohibited the agency from maintaining that classification in February 2026 pending the outcome of the main proceedings, and the AfD celebrated the decision as a victory for the rule of law (European Conservative, 2026). The Social Democrats, the Greens and Die Linke nevertheless continue to pursue a formal ban before the Federal Constitutional Court, and the political pressure to proscribe the party has not abated despite the court’s intervention (European Conservative, 2026). The German case is significant because it shows that even when the judiciary imposes a procedural check on the executive’s classification power, the political system continues to seek the party’s prohibition through other channels. The commitment to excluding the AfD from the political process is therefore not contingent on any single legal instrument but operates across the institutions of the state.
The French case demonstrates the use of criminal law to shape the field of presidential candidates. Marine Le Pen was convicted in 2025 of embezzling European Parliament funds and sentenced to a term of imprisonment and a period of ineligibility that would have prevented her from standing in the 2027 presidential election. The Paris Court of Appeal upheld the conviction in July 2026 but reduced the sentence to one year of house arrest under electronic monitoring and fifteen months of ineligibility that had already been served, which left her free to stand (Radio France, 2026). The court’s decision was described by legal commentators as relatively lenient in comparison with the first-instance ruling, and Le Pen immediately announced her candidacy while indicating that she would appeal to the Court of Cassation (Politico, 2026). The appellate outcome therefore achieved a compromise: the conviction was affirmed, the symbolic disqualification was maintained, and the practical ability to stand was restored. The episode demonstrates that the judicial process can function as a mechanism for calibrating the degree of electoral disruption that the political system will tolerate, and that the calibration is subject to contestation at every stage.
The Polish case represents the most overt deployment of physical coercion in the service of judicial control. The centrist government of Donald Tusk deployed riot police at the Constitutional Tribunal on 30 October 2026 after a group of government-backed judges attempted to take control of the court from its chief justice, who was aligned with the previous Law and Justice government (New York Times, 2026). The confrontation followed the suspension of the tribunal’s president and his deputy for a month, and the police were deployed to prevent demonstrators from entering the building (BBC, 2026). The Tusk government framed the intervention as a restoration of the rule of law, while the opposition characterised it as a coup d’état against a co-equal branch of government (Times of India, 2026). The Polish case is analytically distinct from the others because the dispute concerns control of the court rather than the disqualification of a candidate or the annulment of a vote, and it nevertheless belongs to the same family of phenomena: the use of state power to determine which institutions and which personnel will preside over the resolution of constitutional disputes. The outcome of that determination will shape the legal framework within which future elections and referendums are conducted.
The historical referendums demonstrate that the containment of the ballot has a longer pedigree than the current wave of judicial and constitutional interventions. The Danish rejection of Maastricht in 1992 was reversed in 1993 after the negotiation of opt-outs that left the treaty’s core provisions intact, and the second referendum was approved with 56.7 per cent of the vote on a turnout of 86.5 per cent (Danish Parliament, 2019). The Irish rejection of Nice in 2001 was reversed in 2002 after a series of declarations and assurances that did not alter the treaty’s legal content, and the second referendum produced a majority of 62.89 per cent on a turnout that had risen substantially (BBC News, 2002). The Irish rejection of Lisbon in 2008 was reversed in 2009 after the negotiation of legal guarantees that were designed to address the specific concerns that had motivated the No vote, and the second referendum produced a majority of 67.1 per cent (House of Commons Library, 2026). The French and Dutch rejection of the Constitutional Treaty in 2005 was not reversed by a second referendum, but the Lisbon Treaty retained the substantive innovations of the rejected constitution and was ratified by parliamentary processes that avoided popular votes in most member states (Cato Institute, 2005). The Greek rejection of the bailout terms in 2015 was followed within days by the government’s acceptance of a memorandum that contained more stringent conditions than those that had been rejected (LSE, 2015). The common structure across these cases is that the electorate is permitted to vote, but the outcome of the vote is treated as provisional until it aligns with the preferences of the governing institutions.
The question that these cases collectively pose is not whether any individual decision was legally correct but whether the cumulative architecture is compatible with the democratic legitimacy that the European Union claims for itself. Each intervention can be defended on its own terms: the Romanian court acted on intelligence assessments, the Hungarian amendment pursued a legitimate term-limits objective, the German intelligence agency followed its statutory mandate, the French courts applied the criminal law, the Polish government invoked the rule of law, and the Danish and Irish governments returned to their electorates with improved terms. The difficulty arises from the pattern rather than from any single instance, because the pattern reveals that the institutional response to electoral rejection is consistently to find a mechanism for reversing or circumventing the rejection. A political system that responds to every adverse vote by re-running the vote or disqualifying the voter’s preferred candidate is not a system that respects popular sovereignty, and the accumulation of such responses across four decades and a dozen jurisdictions cannot be dismissed as coincidence.
The material interests that underpin this architecture are not difficult to identify. The European Union’s integration project has been driven by the interests of export-oriented capital, financial institutions and the professional classes that staff the Brussels institutions, and each of these constituencies benefits from the legal and regulatory harmonisation that treaties provide. The enlargement of the union, the single market, the common currency and the sanctions regimes that accompany them create economies of scale and regulatory rents that are distributed to organised interests within the member states. When voters reject a treaty or support a candidate who opposes further integration, they threaten the continuation of those arrangements, and the institutional response is calibrated to preserve them. The referendums on Maastricht, Nice, Lisbon and the Constitutional Treaty were not about abstract constitutional questions but about the distribution of sovereignty and the direction of economic policy, and the reversal of their outcomes preserved the arrangements that the treaties established. The Romanian annulment, the Hungarian amendment, the German classification and the French conviction each operate within the same field of forces, because each removes or neutralises an actor whose programme would alter the distribution of authority within the union.
The British case illustrates the same dynamic in a different institutional setting, because the United Kingdom’s departure from the union has been followed by efforts to reverse or dilute the decision through administrative and legal means. The government initially approved delays to local elections in thirty councils until 2027, and it abandoned those delays in February 2026 after legal advice indicated that they would be unlawful and after Reform UK launched a legal challenge (BBC, 2026). The postponements had been justified by reference to the cost of running elections in councils that were due to be abolished, and the reversal demonstrated that the legal framework can constrain the executive’s capacity to manage the electoral calendar. The episode is a reminder that the containment of the ballot does not always succeed, and that legal challenge and political mobilisation can restore the electoral process even when the executive would prefer to delay it. The lesson is not that the British system is immune to the pattern but that the pattern is contestable, and that the outcome depends on the availability of legal remedies and the willingness of actors to use them.
The final question is the one that the pasted material poses directly: who protects the voters when the politicians who claim to defend democracy decide that the voters are the problem. The answer that emerges from the record is that no institution within the European architecture is charged with that protection, because the institutions that might perform the function are the same institutions that are engaged in the containment. The courts are staffed through political processes, the intelligence agencies answer to executives, the constitutional courts are composed through partisan appointments, and the referendums are called and framed by governments that have a stake in the outcome. The only protection available to voters is the one they create for themselves through organisation, litigation and political mobilisation, and the historical record suggests that such protection is possible but precarious. The Danish and Irish reversals were not inevitable, and the Greek capitulation was not the only possible response to the creditors’ demands. The Romanian annulment, the Hungarian amendment, the German classification, the French conviction and the Polish police deployment are not the end of the story but moments in a contest whose outcome remains open.
Authored By: Global GeoPolitics
Thank you for visiting. This is a reader-supported publication. If you believe journalism should serve the public, not the powerful, and you’re in a position to help, becoming a PAID SUBSCRIBER truly makes a difference. Alternatively you can support by way of a cup of coffee:
https://buymeacoffee.com/ggtv |
https://ko-fi.com/globalgeopolitics |
Bitcoin: 3NiK8BoRZnkwJSHZSekuXKFizGPopkE7ns
Reference List
BBC News (2002) ‘Q&A: Ireland’s referendum’, BBC News, 21 October. Available at:
(Accessed: 31 October 2026).
BBC News (2026) ‘Police deployed during stand-off over control of Poland’s top court’, BBC News, 30 October. Available at:
(Accessed: 31 October 2026).
BBC News (2026) ‘Councils face “uphill struggle” after elections delay U-turn’, BBC News, 17 February. Available at:
(Accessed: 31 October 2026).
Cato Institute (2005) ‘Defeat of a proposed constitution for the European Union by voters in France and the Netherlands’, Cato Institute. Available at:
(Accessed: 31 October 2026).
Danish Parliament (2019) The Danish opt-outs from EU cooperation. Copenhagen: The Danish Parliament. Available at:
https://www.thedanishparliament.dk/
(Accessed: 31 October 2026).
Der Rechte Rand (2025) ‘Neue Entwicklungen’, Der Rechte Rand, 2 June. Available at:
https://www.der-rechte-rand.de/
(Accessed: 31 October 2026).
European Conservative (2026) ‘After AfD secures key court victory, pressure to ban party continues’, European Conservative, 27 February. Available at:
https://europeanconservative.com/
(Accessed: 31 October 2026).
House of Commons Library (2026) The Treaty of Lisbon after the second Irish referendum. London: House of Commons. Available at:
https://commonslibrary.parliament.uk/
(Accessed: 31 October 2026).
LSE European Politics and Policy Blog (2015) ‘Experts react: Greek referendum’, LSE, 6 July. Available at:
(Accessed: 31 October 2026).
MIA (2026) ‘Orbán barred from office for life under new Hungarian constitution’, MIA, 15 June. Available at:
(Accessed: 31 October 2026).
New York Times (2026) ‘Poland deploys riot police in standoff over country’s top court’, New York Times, 31 October. Available at:
(Accessed: 31 October 2026).
Politico (2026) ‘Le Pen verdict: Surprise ruling complicates French far right’s presidential plans’, Politico, 7 July. Available at:
(Accessed: 31 October 2026).
Radio France (2026) ‘Marine Le Pen, condamnée et candidate: quelles conséquences pour la présidentielle?’, France Culture, 8 July. Available at:
(Accessed: 31 October 2026).
Telex (2026) ‘Hungarian Parliament adopts amendment to Fundamental Law which prevents Viktor Orbán from ever being PM again’, Telex, 15 June. Available at:
(Accessed: 31 October 2026).
Times of India (2026) ‘“Coup d’etat” or rule of law? Why Poland’s PM and president are locked in a bitter standoff’, Times of India, 31 October. Available at:
https://timesofindia.indiatimes.com/
(Accessed: 31 October 2026).
Xinhua (2024) ‘Roundup: Romania to restart presidential election, probe launched against leading candidate’, Xinhua, 7 December. Available at:
(Accessed: 31 October 2026).