{
  "id": "eu-2014-council-dual-majority-two-counts",
  "provision": "Two counts never merged: 55 per cent of member states and 65 per cent of population, both required",
  "source": {
    "constitution": "Treaty on European Union and Treaty on the Functioning of the European Union (consolidated versions, OJ C 202 of 7 June 2016), with Protocol (No 36) on transitional provisions and Council Decision 2009/857/EC of 13 December 2007",
    "jurisdiction": "European Union (Council of the European Union and European Council)",
    "year": 2014,
    "article": "Article 16(3), 16(4) and 16(5) TEU; Article 238(1), 238(2), 238(3)(a), 238(3)(b) and 238(4) TFEU; Protocol (No 36) Article 3(1), 3(2), 3(3), 3(4); Council Decision 2009/857/EC Articles 1, 2, 4, 5"
  },
  "text_summary": "Article 16(3) TEU: 'The Council shall act by a qualified majority except where the Treaties provide otherwise.' Article 16(4) TEU, first sentence: 'As from 1 November 2014, a qualified majority shall be defined as at least 55 % of the members of the Council, comprising at least fifteen of them and representing Member States comprising at least 65 % of the population of the Union.' Second sentence: 'A blocking minority must include at least four Council members, failing which the qualified majority shall be deemed attained.' The two counts are separate and both must be met; neither is converted into the other and no single composite number is published. Article 238(2) TFEU raises the state count to 'at least 72 % of the members of the Council' while keeping population at 65 per cent where the Council does not act on a proposal from the Commission or the High Representative. Article 238(3)(a) applies the same pair of figures to the participating states where not all members vote, and adds that a blocking minority there must include 'at least the minimum number of Council members representing more than 35 % of the population of the participating Member States, plus one member, failing which the qualified majority shall be deemed attained.' Article 238(4) keeps abstentions from preventing adoption of acts requiring unanimity.",
  "period": "In force since 1 November 2014 (Protocol No 36, Article 3(1)); the right of any Council member to demand the old Nice calculation instead ran from 1 November 2014 to 31 March 2017 (Protocol No 36, Article 3(2)) and has expired; the Ioannina style delay mechanism of Council Decision 2009/857/EC continues from 1 April 2017 at lowered thresholds (Articles 4 and 5). Still in force; consolidated text of 7 June 2016 checked.",
  "history": "Before 1 November 2014 the Council counted weighted votes, not states. Protocol (No 36), Article 3(3) still prints the table: Germany, France, Italy and the United Kingdom 29 votes each, Spain and Poland 27, Romania 14, Netherlands 13, Belgium, Czech Republic, Greece, Hungary and Portugal 12, Bulgaria, Austria and Sweden 10, Denmark, Ireland, Croatia, Lithuania, Slovakia and Finland 7, Estonia, Cyprus, Latvia, Luxembourg and Slovenia 4, Malta 3, a total of 352 after Croatia joined on 1 July 2013. An act needed 'at least 260 votes in favour representing a majority of the members' on a Commission proposal, and 260 votes representing two thirds of the members otherwise, which is 73.9 per cent of the votes. Population entered only on demand: a member could 'request that ... a check is made to ensure that the Member States comprising the qualified majority represent at least 62 % of the total population of the Union.' A delay mechanism for near misses had existed since the informal foreign ministers' meeting at Ioannina on 27 March 1994: if members holding between 23 and 26 votes signalled opposition, the Council would do all in its power to reach a solution adoptable by at least 68 of 87 votes. Nice superseded it in 2001. The move to two counts was fought over twice. At the Brussels European Council of 12 December 2003 the draft Constitutional Treaty failed, with Spain and Poland the main opponents of the double majority; the Polish line was to keep the Nice weights, captured in Jan Rokita's slogan 'Nicea albo smierc' (Nice or death). The square root alternative is a separate and later proposal: Wojciech Slomczynski and Karol Zyczkowski of the Jagiellonian University published the Jagiellonian Compromise in 2004, giving each state a weight proportional to the square root of its population with an optimal quota of roughly one half plus one over the square root of pi times the number of states, following Penrose's square root rule. Poland pressed it at the European Council of 21 and 22 June 2007 and was left almost isolated; Sweden had floated a square root rule at the 2000 intergovernmental conference and it was turned down then too. What Poland got instead in 2007 was time and a talking shop: Lisbon was signed on 13 December 2007 and entered into force on 1 December 2009, but Article 16(4) was held back to 1 November 2014, and from then until 31 March 2017 any single Council member could demand that an act be put through the Nice calculation instead. Alongside it, Council Decision 2009/857/EC of 13 December 2007 revived Ioannina in percentage form: from 1 November 2014 to 31 March 2017, members representing at least three quarters of the population or three quarters of the number of states needed for a blocking minority could force the Council to 'discuss the issue' and to 'do all in its power to reach, within a reasonable time ... a satisfactory solution'; from 1 April 2017 the same applies at 55 per cent instead of three quarters, so the mechanism did not lapse, it got easier to trigger. Two things in the practical record could not be verified for this entry. First, no source was found showing whether any member state ever actually invoked the Protocol No 36, Article 3(2) right to the Nice calculation between 2014 and 2017; that is unverified here. Second, and more important for us: no source was found that reports how often the population count rather than the state count was the binding one. The Council's own votes-on-acts dataset records the voting rule for each act but not which of the two criteria was decisive, and the Council's voting results pages on consilium.europa.eu returned HTTP 403 to every attempt here, so they were not opened. VoteWatch Europe, which ran from 2008 to June 2022 and whose Council and Parliament data was published in September 2022, was likewise not opened. What the recorded votes do show is that the question rarely arises, because the Council almost never lets it come to a contested vote. The Stiftung Wissenschaft und Politik collected more than 1,300 public Council votes from the entry into force of Lisbon to September 2023 (1,397 votes as of 12 April 2023 in the dataset by Nicolai von Ondarza and Paul Bochtler, DOI 10.7802/2560) and found a consensus rate, meaning no votes against, of almost 83 per cent, stable over time: 82 per cent before Brexit and 85 per cent since. Where abstentions are also excluded, 64 per cent of votes were fully unanimous. The United Kingdom was outvoted 167 times, 16 per cent of public votes while it was a member; France was outvoted five times in the whole period, well under 1 per cent. No vote was found in which all four Visegrad states were outvoted together. The best known contested case under the new rule is Council Decision (EU) 2015/1601 of 22 September 2015 on relocating 120,000 asylum applicants from Italy and Greece under Article 78(3) TFEU, carried by qualified majority with the Czech Republic, Hungary, Romania and Slovakia against; Finland's abstention is reported in secondary accounts seen only as a search snippet. Four states against is exactly the floor the second sentence of Article 16(4) sets, and their combined population was nowhere near 35 per cent, so the state count and not the population count was the live constraint there. Brexit changed the arithmetic without changing the text: 55 per cent of 27 is 15 states, which is also the floor written into Article 16(4), and the reinforced majority of Article 238(2) now needs 20 of 27.",
  "assessment": {
    "verdict": "mixed",
    "reasoning": "As a design this is the strongest in force example of two counts that must both be met and are never merged. Both numbers are in the treaty text, both are checked separately, and the population figures are published and updated rather than folded into a weight. The rule has survived an enlargement and a departure without amendment. But three things qualify it. The blocking minority floor of four members is a hard override, not a safeguard: when only three states object the population count is switched off entirely and the majority is 'deemed attained', even though Germany at 18.7 per cent and France at 15.1 per cent of the EU-27 population are already at 33.8 per cent between them. So the population count protects a coalition of four or more and nobody else. Second, we could not establish from any public source how often each count actually bound. The Council publishes voting results and the voting rule, not which criterion decided, so the central empirical question about a two count rule is unanswerable from the record. Third, the rule mostly operates as a threat. On the SWP figures almost 83 per cent of public votes have no vote against at all, and 64 per cent are fully unanimous. A decision rule that is applied in anger in roughly one vote in six is shaping negotiations, not deciding them.",
    "dissent": {
      "verdict": "critical",
      "reasoning": "On a stricter reading the second count is close to decorative and the double majority quietly handed power to the largest states. Axel Moberg of the Swedish foreign ministry argued in 2007 that 'in practice, only criterion (3) remains', that is population, because the majority of states requirement bites only in the unrealistic case of all the large states against all the small ones. On his simulations Germany's blocking potential doubles and France, the United Kingdom and Italy gain roughly 40 to 50 per cent, while medium sized states fall below 60 per cent of their Nice blocking potential and the smallest lose more. Spain and Poland, who fought hardest in 2003, lost ground against the big four. A second and compatible criticism is that the whole construction is theatre: the Council decides by consensus, the formal rule is rarely reached, and the real filter sits in Coreper and the working groups, where there are no published counts at all. Helen Wallace put it to the House of Lords that explicit voting 'occurs on only a minority of issues, and does not tell us much about how QMV currently operates in practice'. On this reading the two counts are not a protection for small members; they are a scoreboard nobody looks at while the bargaining happens elsewhere.",
      "evidence": [
        "https://www.realinstitutoelcano.org/en/work-document/is-the-double-majority-really-double-the-second-round-in-the-debate-of-the-voting-rules-in-the-eu-constitutional-treaty/ (opened; Axel Moberg, 30 May 2007, 'in practice, only criterion (3) remains', Germany's blocking potential doubles, medium states below 60 per cent of Nice)",
        "https://www.files.ethz.ch/isn/167306/SN06646.pdf (opened; House of Commons Library SN/IA/6646, 23 May 2013, Helen Wallace quote on explicit voting, 65 per cent of 2009 to 2012 QMV decisions still unanimous)",
        "https://www.swp-berlin.org/publications/products/comments/2024C16_QualifiedMajorityDecisions.pdf (opened; consensus rate almost 83 per cent, France outvoted five times, no Visegrad four outvoting found)"
      ]
    }
  },
  "lesson": "Copy the shape: write both counts into the founding text with their own figures, check them separately, publish each number beside the other, and never let anyone convert one into the other or quote a single blended figure. Copy also the slow mover, a duty to keep discussing when a near minority objects, because it keeps a losing side talking instead of leaving. Do not copy the floor clause that switches the second count off when fewer than four objectors appear; a population or machine count that silently stops applying below a headcount is not a second count at all. Do not copy a regime whose own records show which rule was used but not which count bound, because then nobody can audit whether the second count ever did any work. And expect what the Council got: once both counts are known, almost everything is settled before the count is taken, so publish the near misses and the objections too, or the two numbers will describe an event that almost never happens.",
  "evidence": [
    "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A12016M016 (opened; primary text, Article 16(3), 16(4) both sentences verbatim, 16(5) on the transitional provisions, consolidated TEU)",
    "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A12016E238 (opened; primary text, Article 238(1) to (4) TFEU verbatim, the 72 per cent variant and the 35 per cent plus one blocking minority)",
    "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A12016E%2FPRO%2F36 (opened; Protocol No 36 Article 3(2) verbatim, the full Nice weighted vote table, 260 votes, the 62 per cent population check on request)",
    "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32009D0857 (opened; Council Decision 2009/857/EC, three quarters thresholds to 31 March 2017 in Articles 1 and 2, 55 per cent from 1 April 2017 in Articles 4 and 5, 'do all in its power')",
    "https://www.swp-berlin.org/publications/products/comments/2024C16_QualifiedMajorityDecisions.pdf (opened; SWP Comment 16, April 2024, von Ondarza and Sturzer: more than 1,300 public votes to September 2023, consensus almost 83 per cent, 82 before and 85 after Brexit, 64 per cent fully unanimous, UK outvoted 167 times or 16 per cent, France five times, Germany 18.7 and France 15.1 per cent of EU-27 population)",
    "https://www.swp-berlin.org/publikation/public-voting-data-of-the-council-of-the-eu (opened; the dataset behind it, von Ondarza and Bochtler, DOI 10.7802/2560, January 2010 to March 2023, 1,397 votes as of 12 April 2023)",
    "https://github.com/EUCouncil/CouncilVotesOnActsDatasetSample (opened; the Council's own votes-on-acts data cube records country, vote, act type, policy area, voting procedure and voting rule, but no field for which criterion was decisive)",
    "https://www.files.ethz.ch/isn/167306/SN06646.pdf (opened; House of Commons Library note, Nice arithmetic 255 of 345 and the 62 per cent check, the 1 November 2014 change and the request right to March 2017, VoteWatch finding that 65 per cent of QMV decisions from July 2009 to June 2012 were adopted unanimously)",
    "https://epthinktank.eu/2014/12/09/changed-rules-for-qualified-majority-voting-in-the-council-of-the-eu/ (opened; European Parliament Research Service, 9 December 2014: new thresholds, blocking coalition of at least four states and 35 per cent of population, 74 per cent of weighted votes under Nice, request right until 31 March 2017)",
    "https://en.wikipedia.org/w/index.php?title=Voting_in_the_Council_of_the_European_Union&action=raw (opened; 255 votes from January 2007, 260 from July 2013, total 345 then 352, 62 per cent population, Penrose quota of 61.4 per cent, transition to 31 March 2017)",
    "https://en.wikipedia.org/w/index.php?title=Ioannina_compromise&action=raw (opened; meeting of 27 March 1994, the 23 to 26 vote band, 'do all within its power' to reach a solution adoptable by at least 68 of 87 votes, superseded by Nice 2001, revived for 2014 to 2017)",
    "https://en.wikipedia.org/w/index.php?title=Jagiellonian_compromise&action=raw (opened; Jagiellonian University 2004, square root weights, optimal quota formula one half plus one over the square root of pi times the number of states)",
    "https://www.realinstitutoelcano.org/en/work-document/is-the-double-majority-really-double-the-second-round-in-the-debate-of-the-voting-rules-in-the-eu-constitutional-treaty/ (opened; Axel Moberg, Swedish Ministry of Foreign Affairs, 30 May 2007, the case that the state count is practically irrelevant)",
    "https://arxiv.org/pdf/2312.16878 (opened; Petroczy and Csato, sensitivity analysis of Council voting power, large states gain under 55 per cent plus 65 per cent compared with Nice, blocking minority of four)",
    "https://arxiv.org/abs/0712.2699 (opened; Martin Kurth, 17 December 2007, revised 19 May 2008, on the Jagiellonian Compromise quota and rounding effects)",
    "https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32015D1601 (opened; Council Decision (EU) 2015/1601 of 22 September 2015, Article 78(3) TFEU, 120,000 applicants; the text itself records no vote breakdown)",
    "https://ecre.org/eu-council-agrees-on-relocation-of-thousands-of-asylum-seekers/ (opened; 'Despite Hungary, the Czech Republic, Slovakia and Romania voting against the proposals', 'the use of a qualified majority vote saw the proposals accepted')",
    "https://eumigrationlawblog.eu/relocation-and-its-numbers-which-role-for-the-courts/ (opened; Decision 2015/1601 'was not adopted unanimously but with a qualified majority in the Council, with vehement opposition from the Czech Republic, Romania, Slovakia, and Hungary')",
    "https://www.consilium.europa.eu/en/council-eu/voting-system/qualified-majority/ (not opened; HTTP 403 on every attempt, including the european-council.europa.eu redirect and the open data and voting results pages, so no Council figure is cited from it)",
    "https://journals.sagepub.com/doi/10.1177/14651165211004754 (snippet; Pircher and Farjam, European Union Politics 22(3), 2021, report 36 per cent contestation for 2010 to 2019 against about 16 per cent pre Lisbon; paywalled, abstract seen only in search results)",
    "https://en.wikipedia.org/wiki/Jan_Rokita (snippet; the slogan 'Nicea albo smierc', Nice or death, as the Polish delegation's line from late 2003)",
    "https://www.eurofound.europa.eu/en/publications/all/european-council-fails-agree-constitutional-treaty (snippet; Brussels European Council of 12 December 2003 failed with Spain and Poland the main opponents of the new voting rights)"
  ],
  "adopted_in": [
    "Article 16.6",
    "Article 2.3",
    "Article 10.7"
  ],
  "status": "proposed",
  "people": [
    "Jakob Hirn"
  ],
  "agents": [
    "claude-for-jakob"
  ]
}