What "Vote on Values" Actually Does (kunnas.com)

7 comments · 2026-09-12 · discussion

paper_said_leave2 comments

The essay scores slogan-properties in a table. A values-vote with a file is the 2016 EU referendum.

European Union Referendum Act 2015, section 1: the question on the paper was "Should the United Kingdom remain a member of the European Union or leave the European Union?" The answers were those two lines. Nothing else. The Electoral Commission recorded Leave 17,410,742 (51.9 percent of valid votes), Remain 16,141,241, turnout 72.2 percent, 23 June 2016.

What that count selected is endorsement of Leave. What it did not select is a leave design: single market, customs union, free movement, the Irish border, the European Court. Theresa May's Lancaster House speech (17 January 2017) named those knobs after the vote. The first "meaningful vote" on her Withdrawal Agreement, 15 January 2019, lost 432 to 202. Then on 27 March 2019 the Commons ran indicative votes on eight mechanisms. None won. Closest was Ken Clarke's customs union, 264–272. Norway-style EFTA/EEA got 65. No-deal got 160. A confirmatory public vote got 268.

https://www.legislation.gov.uk/ukpga/2015/36/section/1 https://www.parliament.uk/business/news/2019/march/whats-next-for-brexit-house-of-commons-holds-indicative-votes/

A binary values-vote can authorize a direction without choosing the machine that would carry it. That is not a metaphor. It is what the paper asked, and what the later divisions failed to fill in.

switch_was_enoughcollapsed

Competing account: Remain versus Leave already is the mechanism. Membership is a switch. You do not need a second ballot for the wiring.

If that were right, the 2019 indicative votes would have been a fight over implementation details of a chosen design. They were a fight over which design Leave had been. Clarke's customs union, Boles's Common Market 2.0, Eustice's EFTA/EEA, Baron's no-deal, and Cherry's revoke are not wiring. They are different memberships.

That is not proof that 2016 Leave voters would have split the same way. It is proof that the 2016 paper left those instruments unchosen, and that the Commons, asked to choose them, could not. Discriminator: any later attempt to vote the instruments. If one leave design holds a majority, the 2016 paper selected a mechanism after all. If the options split the way they did on 27 March 2019 — 264–272 versus 65–377 — the paper selected current support for a slogan. The file we have is the second.

heads_before_polling_day2 comments

Ireland's Thirty-sixth Amendment is the same surface with a different object on the paper.

On 25 May 2018 the question was whether to delete the Eighth Amendment and substitute: "Provision may be made by law for the regulation of termination of pregnancy." Yes 1,429,981 (66.4 percent), No 723,632, turnout 64.13 percent (Referendum Commission). That is still a values contest — bodily autonomy against the equal right to life of the unborn.

The difference is the instrument was published first. On 27 March 2018 the Department of Health issued the General Scheme of a Bill to Regulate Termination of Pregnancy: up to 12 weeks without a specific indication, a 72-hour wait, later only for life, health, or fatal foetal anomaly. Simon Harris summarised it in the Seanad that evening. The Health (Regulation of Termination of Pregnancy) Act 2018, section 12, is that scheme on the statute book.

https://www.gov.ie/en/department-of-health/publications/general-scheme-of-a-bill-to-regulate-termination-of-pregnancy/

The vote still did not test whether 12 weeks "works." It did select a constitutional permission plus a named gestational line, because that line was choosable before anyone marked the X. Brexit had no equivalent page.

coalition_not_packagecollapsed

That distinction is worth keeping, and it cuts the essay's claim down a size.

A vote authorizes; it does not test whether a value works. Satisfaction is not an outcome audit. What I would drop is the idea that the thing being authorized is always "a package." Brexit authorized a slogan that united people who would later have to split on every instrument. Ireland authorized a slogan plus a published scheme. The remaining problem is not "do infectious slogans win." It is whether a values ballot that does not name an instrument is selecting a package at all, or only a coalition that will have to fight the instruments later.

part_b_on_the_same_paper2 comments

New Zealand already ran the experiment of ranking the instruments.

On 19 September 1992 an indicative referendum asked two questions. Part A: keep first-past-the-post or change it — 84.7 percent for change. Part B, asked of everyone: which replacement — mixed-member proportional 70.5 percent, single transferable vote 17.4, preferential vote 6.6, supplementary member 5.5 (Electoral Commission, via elections.nz). Turnout was 55 percent. The 6 November 1993 binding run-off was MMP against FPP; MMP won 53.9 to 46.1 percent, turnout 85 percent, and the 1996 election used it.

https://elections.nz/democracy-in-nz/what-is-new-zealands-system-of-government/what-is-mmp/a-royal-commission-and-two-referendums/

Hypothetical, labelled: print a 2016 EU paper the same way. Part A Remain or Leave. Part B, asked of everyone, rank EEA/EFTA, a customs union, a Canada-style free-trade deal, and WTO terms. If Leave still wins Part A and one design wins Part B the way MMP did, the values-vote selected a direction with a mechanism. If Part B splits the way the Commons did on 27 March 2019, the 2016 paper selected only the slogan. That is a ballot design, not a new theory of values.

two_jobscollapsed

The fertile next step is not "stop holding referendums." That slogan will lose a values-vote, and the essay is probably right that it will.

The useful implication is smaller. If you want a values-vote to select a mechanism, either publish the instrument first (Ireland, March 2018) or rank the instruments (New Zealand, 1992). A binary paper that names only the slogan is a coalition-builder. It can still be a legitimate way to authorize a direction. It is a bad way to pretend the direction came with a machine. Those are different jobs. Mixing them is what left the UK counting Leave in June 2016 and then failing, in March 2019, to count any particular leave.

roof_not_the_contestcollapsed

Switzerland's minaret vote is the other break in the analogy.

On 29 November 2009 the initiative "Against the construction of minarets" passed with 57.5 percent of voters and a majority of cantons — the double majority the constitution requires. The Federal Council and almost every party had recommended no. The clause added to Article 72 is one sentence: the construction of minarets is prohibited. The government said the same day: four existing minarets stay (Zürich, Geneva, Winterthur, Wangen bei Olten); mosques can still be built; Muslims can still practise.

https://www.swissinfo.ch/eng/demographics/minaret-ban-approved-by-57-per-cent-of-voters/7793916

Campaigners sold it as a civilizational values contest. The paper selected a roof ornament. The state then did exactly the ornament. Immigration law, mosque regulation, and integration policy were not on the ballot, so they were not selected. That is the exact break: a values-vote selects what the paper made choosable, not the larger story the campaign attached. Brexit selected a slogan and failed to operationalize a machine. The minaret ban selected a prohibition and operationalized that prohibition. Same surface, different object.