Comparison · checked 22 October 2026

refnation vs the polls

Recent refnation results beside YouGov, Find Out Now and other national polls on the same issues.

The finding

On 24 of 27 recent issues that national pollsters also asked about, refnation's result came down on the same side as the poll. The typical gap was 7.5 points, and 17 were within 10.

A plain headcount of the same votes was on the same side 21 times, with a typical gap of 15.7 points. The difference is the balancing: refnation works out each member's political leaning from how they vote, then gives all seven leanings, from Hard Left to Hard Right, an equal say in every result. It brought the result nearer the poll on 18 of 27.

Where it misses

The widest gaps were on real-name social media (refnation 55%, YouGov 87%), rent caps (refnation 59%, YouGov 89%) and the two-child cap (refnation 52%, YouGov 72%). refnation came down on the same side each time, but less one-sided than the poll.

3 came down on the other side: Andy Burnham's job as Prime Minister (refnation 46%, YouGov 53%), scrapping inheritance tax (refnation 47%, YouGov 66%) and the triple lock change (refnation 43%, YouGov 63%).

Polls differ among themselves too. YouGov asked about limiting rent rises three ways this year, and support with don't knows set aside ran from 72% to 89% depending on the wording.

Every comparison

On each track, running from 0% to 100% Yes with a line at 50: refnation · the poll, don't knows set aside · plain headcount.

Should the Budget scrap the windfall tax on North Sea oil and gas to boost UK production?

46%

refnation

29%

Survation

points apart

refnation, on the ballot 21 October 2026: 45.6% Yes from 1,940 votes (plain headcount 60.6%).

Survation for the End Fuel Poverty Coalition, 2–7 Apr 2026, 2,047 adults: “…Do you support or oppose, or neither support nor oppose, the current windfall tax on oil and gas company profits?” Oppose the tax 17%, Support the tax 41%, Neither or don't know 42%.

The poll asked about support for the tax as it stands, so opposing it is counted as the scrap side.

Should migrants have to wait 10 years, not 5, before they can settle permanently in the UK?

58%

refnation

65%

YouGov

points apart

refnation, on the ballot 21 October 2026: 57.5% Yes from 1,949 votes (plain headcount 71.7%).

YouGov, 20 Mar 2026, 3,881 adults: “Would you support or oppose increasing the amount of time that migrants have to live in the UK before they are legally entitled to stay on a permanent basis?” Support 53%, Oppose 28%, Don't know 19%.

The poll does not name the 10-year wait.

Should MPs face a mandatory by-election if they defect to another political party?

84%

refnation

87%

YouGov

points apart

refnation, on the ballot 21 October 2026: 83.5% Yes from 1,960 votes (plain headcount 83.5%).

YouGov, 27 Jan–1 Feb 2026, 4,212 adults: “If an MP defects from a political party… do you think there should or should not be a by-election in the constituency they represent?” Should 68%, Should not 10%, Don't know 22%.

Should social media users have to verify their real identity to post?

55%

refnation

87%

YouGov

points apart

refnation, on the ballot 19 October 2026: 55.2% Yes from 1,794 votes (plain headcount 51.5%).

YouGov, 19 Jun 2025, 5,538 adults: “When it comes to people's identity on social media, which of the below comes closest to what you think should happen?” Display it, or disclose it to the network 73%, Use social networks without either 11%, Don't know 16%.

Should trans women be allowed to use women's toilets and changing rooms?

30%

refnation

36%

Ipsos

points apart

refnation, on the ballot 18 October 2026: 29.7% Yes from 1,768 votes (plain headcount 17.8%).

Ipsos, 24 Apr–8 May 2026, about 1,000 adults: “Transgender people should be allowed to use single-sex facilities (e.g., public restrooms) that correspond to the gender they identify with.” Agree 30%, Disagree 53%, Neither or don't know 17%.

About trans people in general, not only trans women, and adults aged 16–74. YouGov's latest question offered four options rather than a yes or no.

Should police always reveal a suspect's nationality and ethnicity when they announce a charge?

54%

refnation

61%

YouGov

points apart

refnation, on the ballot 17 October 2026: 53.5% Yes from 1,773 votes (plain headcount 67.5%).

YouGov for The Times, 28–29 May 2025, 2,055 adults: “Generally speaking, when the police arrest a suspect in a high profile crime, do you think they should or should not release details of the suspect's ethnicity?” Should 50%, Should not 32%, Don't know 18%.

The poll asked about ethnicity in high-profile arrests.

Is it acceptable for the government to raise income tax, National Insurance or VAT in the Budget, despite promising not to?

36%

refnation

39%

Set aside

A poll was found for these, but it asked something the referendum didn't, so they are not in the figures:

Is rejoining the EU worth an estimated £40 billion a year? No poll put a cost on rejoining.
Should the UK adopt the euro? No poll has asked it since 2020; recent ones ask about the euro only as a condition of rejoining.
Should the next Budget tax wealth more rather than cut pensions and welfare? Polls asked about a specific 2% tax on fortunes over £10 million, without the trade-off.
Should 18 to 21-year-olds not in education, work or training have to do national service? Polls asked about national service for all young people.
Has Donald Trump's second term been good for the UK? Polls ask whether people approve of Trump, not about his effect on the UK.
Should the Royal Family's finances, including the Sovereign Grant, face an independent review? Polls ask whether the Royal Family is good value for money.
Should personal drug use be decriminalised and handled by the NHS instead of police and courts? Polls ask about cannabis alone, or split soft and hard drugs.

No poll since January 2025 was found on scrapping photo ID to vote, Swiss-style citizen-initiated referendums, banning fireworks sales, a national DNA database, teaching LGBTQ+ relationships in schools, suspending the asylum system, capping water and energy bosses' pay, scrapping energy standing charges, cutting VAT for hospitality and freezing council tax.

How the comparison is made

180 referendums ran on the daily ballot from 29 September 2026 to 21 October 2026. The 44 most likely to have been polled were searched against YouGov's published questions and Find Out Now's tables since January 2025, then other British Polling Council members'.
Every match is listed, close or not. Where more than one poll matched, the closest wording was used, then the most recent.
Poll figures set aside don't knows: Yes ÷ (Yes + No). A refnation vote is Yes or No, so that is the like-for-like figure.
Wording and dates are not identical to the referendum's, so each poll's own question is quoted, with a note where the difference matters.
refnation figures are the published results. The method is on How results are calculated.

Common questions

Is refnation accurate?+

On 24 of 27 recent issues that national pollsters also asked about, refnation's result came down on the same side as the poll, with a typical gap of 7.5 points. A plain headcount of the same votes was typically 15.7 points away, so the balancing roughly halves the gap. Every comparison is on this page, misses included.

Is refnation representative of the UK?+

Members join by choice, so refnation doesn't simply count heads. It works out each member's political leaning from how they actually vote, then gives all seven leanings, from Hard Left to Hard Right, an equal say in every result, so no one side of politics can decide it by turning up. Results are reported as refnation members' votes. The balancing covers political leaning, not age or region, which is one reason a result can differ from a poll.

How is that different from how polls weight?+

Polls weight a sample to the population by age, region and who people say they voted for last time. refnation reads leaning from a voting record instead: each referendum is scored for the political direction a Yes and a No point in, so a member's leaning is built from their own votes, referendum after referendum.

Why compare without the don't knows?+

A refnation vote is Yes or No; a skip is not counted. A poll's Yes and No figures, with don't knows set aside, are the like-for-like comparison. Each poll's published figures, don't knows included, are shown beside it.

Also: refnation vs YouGov · How results are calculated · Press