Is the private schools VAT change causing more parents to lie?
Welcome to the 145th edition of The Week in Polls (TWIP) which dives into the polling details behind a definitive looking headline from The Times. Will we find that the poll results back up the headline, or not?
Then it is a summary of the latest national voting intention polls and a round-up of party leader ratings, followed by, for paid-for subscribers, 10 insights from the last week’s polling and analysis.
This week, those ten include wisdom from Paula Surridge on why FindOutNow may be getting higher figures for Reform than other pollsters.
But first, a trio of follow-ups to last time’s edition about polling regulation.
First, I worded poorly the part where I said pollsters should have to be more open about what other questions were asked in the same poll. I wrote, “We simply have to take on trust that the pollster has not put other questions in first which end up biasing the answers to subsequent questions due to the framing effect.”
What I should have pointed out is that the British Polling Council’s rules have a relevant section, stating, “The client and survey organisation may keep other findings (that have not been published) confidential except where such findings are relevant to the topics covered in questions that have been published or where the question order is relevant to the published results.”
My point was (and is) that this is a fairly limited rule, as ‘relevant to the topics’ can be interpreted narrowly, and on question order, although a strict implementation of this wording would be fine, I am gently sceptical about how strictly it and the relevance rule really are being implemented as it is exceptionally rare to find any additional questions being published as a result.1
Second, thank you to one of those behind BPIX getting in touch to confirm that yes, it was run on an administrative shoestring by academics. Chris Prosser also kindly pointed out this previous write-up of BPIX’s work.
Third, I have been in touch with YouGov and the British Polling Council about the YouGov website terms and conditions which I mentioned. Thank you both for their constructive responses, and hopefully a fuller update can follow in a later edition.
Thank you also for all the other feedback, and with that let us see what new feedback I can provoke this week.
Want to know more about political polling? Get my book Polling UnPacked: the history, uses and abuses of political opinion polling.
VAT up, parental lying up?
Last week, when talking about how polling regulation could be improved, I said:
Outside of the years of the big polling errors - poor reporting of polls is a consistently bigger problem than those with the polls themselves.
In other words, blame the journalists, not the pollsters.
By spooky (although given my point, also predictable) coincidence, the last few days have given us another example to test out whether the problem with a newspaper piece is with the polling or the journalism.
Let us dive in to see all the gory detail.2
A clear headline, but is it backed up by a clear poll?
Earlier in the week The Times ran this headline:
The story goes on:
The research by Zoopla, the property website, found that 28 per cent were “lying, bending rules or playing the system” to get their children into their preferred state schools, a rise from 24 per cent in late 2022 when it previously carried out the research.
Already the story is watering down with “lying” in the headline becoming “lying, bending the rules or playing the system”. Although headlines are necessarily stripped of some nuance, this gives a clue that we need to look at the story’s details with care.
The benefits of polling transparency
Unlike political polling stories, the story does not include basics such as name of pollster and fieldwork dates which we are used to seeing for stories about voting intentions and leader approval ratings. The Times is by no means alone in omitting such information from these sorts of broader polling stories, though it is hard to see from a good journalism viewpoint why you would regularly provide basic polling information for some polling stories and yet not for others.
Luckily in this case the press release from Zoopla is available online, so the curious reader can go hunt out more information about the poll. Their press team also responded quickly to my questions. That gets us to knowing that the pollster was Mortar Research and that the fieldwork was 3-6 January.
Mortar is a member of the British Polling Council, and replied impressive quickly and outside of normal working hours when I asked them for the data tables with the cross tabs. (The cross tabs also, hooray!, come with formatting and colour - one of the more niche things I called for more of last week.)
For all those positives about Zoopla and Mortar, there is though a difficulty even for any Times reader who has dug beyond the minimalism of the newspaper’s story. Because Mortar is a relatively low profile polling firm, and because it does not do public election polls that can be tested against reality, we have no real way of knowing how good, or not, their polls are. (Again hark back to last week and the advantage of my suggestion of polls including test questions so that we all can see the quality of the sample.) Moreover, as this poll is of “people who applied for a school place for their child(ren) in the last five years” there is not any comparison against reality of basic demographics etc. that could be carried out as we don’t know what those should be.
It could be a high quality poll. It might be mediocre. Or it could be a stinker. The public information gives us no real clue as to which, even though the speed of response from the pollster and the layout of their crosstabs makes me warm to them.
It is though definitely just the single new poll from Mortar, compared by Zoopla and The Times to a previous one in late 2022.3
That makes it potentially useful as a piece of information but also risky to rest a story and newspaper headline just on that data point. It is a single sourced story, with the source of unknown robustness. (Again, note the contrast with what journalists and their editors would consider appropriate for a non-polling story, where single-sourcing is usually frowned on and when used has to be justified by the story’s importance and/or the known quality of the source).
What the polling numbers say
What about the specific claims made in the story? The Times was clear, at least at first, in its core claim: a rise in the number of parents “lying”, to use the terminology of the headline and first paragraph and “using dishonesty” to use the terminology of the subhead.
That was watered down to “lying, bending rules or playing the system” by the second paragraph.
Checking the details of the poll, moreover, requires us to do more watering down.
The 2025 poll has 27% “of UK parents admit to flouting the rules to get their children into schools”. That is up from 24% in 2022.
A 3 percentage point increase is not large enough for us to be confident that it it a genuine increase rather than statistical noise. If there was a series of polls all showing a 3 point or similar increase, we could draw such a conclusion with decent certainty. But for just the one poll, it adds to the picture of the fragility of the story.4
What is more, the slipping around of different terms used by The Times indicates that the 27% comes from a fairly broad category of behaviour. Diving into the poll details we see that the 27% includes those who said they had “made voluntary donations to the school in advance of my child’s place being confirmed”, “attempted to involve myself with the school prior to applying - for example by volunteering” and “tried to become friends with senior figures at the school”.
Calling that trio of activities ‘playing the system’ is fair enough. But is it “lying” or “dishonesty” to try to become friends with someone? Yet it is only by saying that it is that you can get to the 27% figure.5
Is VAT to blame?
Next, what about the claim that - if the poll is right, the increase is real and we’re happy to call it dishonesty/lying - this change in behaviour is down to the Labour government’s VAT policy?
Remember the certainty of that headline, subhead and opening paragraph.
Here things get even harder to sustain. The polling is of people who applied for a school place in the last five years. For most of that time Labour was not in government. For much of that time, Labour was not looking that likely to be in government in the near future either. It is quite a stretch to say that someone who lied in, say, 2020 when Boris Johnson was popular was prompted to lie solely by the thought that Keir Starmer might be a success after all, win a general election in a few years and put VAT on school fees after that.
If there has been an increase in lying, it is quite possible it is due to other factors, such as perhaps the extra strains of lockdown and COVID making some parents even more anxious to get the school they want.
The Zoopla press release, to be fair, has a go at fleshing this out:
Several factors have likely contributed to the increase, including the addition of VAT on private school fees … Amongst those who bent the rules, 12 per cent said their children previously went to a private, fee-paying school, but the fees were no longer affordable, or that they intended to send them to one. Of those who could no longer afford the fees, nearly half (45 per cent) said the VAT increase was the reason.
Reading that carefully, we have the original 27% flouting the rules, then 12% of the 27% being parents who switched their kids from private schools due to the costs (so that’s 3.2% of parents overall) - and then it is 45% of those who say the VAT increase was the issue (which brings it down to 1.4%.)
In other words, the survey has 1.4% of parents saying they flout the rules and that the VAT increase has been the reason for switching from private to state schools.
Or, digging into the crosstabs to check that calculation, we see that 45% is 45% of 33 (!) people, i.e. 15 people out of the original overall sample of 1,019 which is 1.5%. (0.1% higher than the 1.4% above due to rounding.)
True, it is possible that the statistical noise means the 1.5% may be a bit of an underestimate.
But even so, 1.5%.
And remember, this 1.5% is the top of the range. All those issues about five years, what counts as lying and so on would bring that number down.
This is a long way from The Times’s headline and story.
To be fair to Zoopla, their own press release - and their press team in answer to my questions - have provided rather more transparency than the newspaper. The crosstabs from the pollster are also better than most, not only for the formatting but also for the clarity of showing percentages both ways when a question was asked of a subset - both as a percentage of the subset and of the full sample.
The press release is not perfect, as it gives a breakdown of one polling answer by region but without sample sizes give for each region. As their press team then told me, those sample sizes vary from 27 to 157. The regional numbers are very tasty looking for regional and local media coverage, but with those sort of sample sizes should come with health warnings.
Yet given the cynicism often attached to polling done for PR purposes,6 it is ironic that the press release does better than the journalists.
The Times also gets one number wrong in their write-up. It said, “Zoopla said that 12 per cent of the 1,019 parents questioned in the survey, which was carried out this month, had previously educated their children in private schools but could no longer afford the fees.”
That 12 per cent figure as described by The Times should ring alarm bells given that only around 6 per cent of all children go to private schools.7 If we do then go and check it, we find that the Zoopla press release states, “Amongst those who bent the rules [my emphasis], 12 per cent said their children previously went to a private, fee-paying school, but the fees were no longer affordable, or that they intended to send them to one.”
That 12 per cent as described by The Times should have been 3 per cent.8
It’s about journalists, not pollsters
To summarise, read carefully the poll does suggest that “lying, bending the rules or playing the system” is likely to have risen a little in the last five years. It is only one poll, and from a pollster without a clear public track record, so caution should apply, but if you forced me to bet, I would bet on the truth being that such behaviour has risen a little.
That is as far as the data takes us.
Which leaves us stranded a long way short of The Times having as one of its top stories a headline and story leading the reader to conclude as solid fact that Labour’s VAT rise has prompted an increase in lying.
So while there are big questions about this poll based story, the questions are predominantly ones for the journalists, not the pollsters.
I was prompted to write this piece by seeing the critiques of the poll from Anthony Wells and Jim Waterson. As the latter pointed out, the story was the third top on The Times’s homepage for a while. Other media outlets have also written up the poll, generally with better, though not perfect, stories. For example, The Independent - often a source of my raised eyebrows for using dramatic language to describe small movements in voting intention polls - this time did better, caveating the VAT claim thrice over, attributing it to Zoopla, saying it was a claim about something that may have happened.
Voting intentions and leadership ratings
Here are the latest national general election voting intention polls, sorted by fieldwork dates:
Next, a summary of the the leadership ratings, sorted by name of pollster:
For more details, and updates during the week as each new poll comes out, see my regularly updated tables here and follow The Week in Polls on Bluesky.
For the historic figures, including Parliamentary by-election polls, see PollBase.
Catch-up: the previous two editions
My privacy policy and related legal information is available here. Links to purchase books online are usually affiliate links which pay a commission for each sale. For content from YouGov the copyright information is: “YouGov Plc, 2018, © All rights reserved”.
Quotes from people’s social media messages sometimes include small edits for punctuation and other clarity.
Please note that if you are subscribed to other email lists of mine, unsubscribing from this list will not automatically remove you from the other lists. If you wish to be removed from all lists, simply hit reply and let me know.
The surprising similarity between Tony Blair and Nigel Farage, and other polling news
The following 10 findings from the most recent polls and analysis are for paying subscribers only, but you can sign up for a free trial to read them straight away.
- joined me for the latest episode of Never Mind The Bar Charts as
Keep reading with a 7-day free trial
Subscribe to The Week in Polls to keep reading this post and get 7 days of free access to the full post archives.