Hello, and welcome to More or Less with me, Tim Harford. This week's programme line-up is something to warm the heart of Tony Blair himself. Education, education, hang on, immigration. We resolve a friendly wager between two policy wonks about what VAT on private schools would do to demand for private schooling. We revisit Welsh literacy, putting your questions to an expert. A recent A-level maths paper caused controversy for being unfair and confusing. We'll figure out what went on. But first, certain claims floating about the internet got the attention of not one, not two, but five loyal listeners recently. Chloe, John, Steve, Sarah and Emma all got in touch to ask about headlines such as this from the Daily Mail. 27 young migrants are hired for every British youngster as youth worklessness fuelled by soaring non-EU immigration analysis reveals. The news articles all cite research from the right-leaning think tank the Centre for Social Justice or CSJ. The headline for their article is this. 27 young non-EU migrants hired for every young Brit since 2020 analysis reveals. Our five listeners wanted to know if this ratio was reputable. Our famous five correspondent Lizzie McNeill has been taking a look. Hello, Lizzie. Hi, Timmy. No. Noted. So this would be remarkable if it's true, although I'm not sure how it could be. There must be two or three million young people being hired over this time period. And I don't think there are tens of millions of young migrant workers are there. You think the headline means that 27 young non-EU migrants are hired for every young Brit since 2020? That is literally the headline of the report. Yeah, that's what it says, but it's not what it's saying, you know? Right. So what is it saying? Well to see what they're saying, you have to completely disregard the headline and look at what the analysis actually does. Which is? Right. So the CSJ used payroll data from HMRC and compared the employment numbers for under 25s at two points in time, January 2020 and December 2025. With you so far? Excellent. They then looked at the proportion of those workers who were UK nationals or non-EU nationals. In 2020, the number for UK nationals was around 3.84 million and this rose to 3.85 million in 2025. So about a 10,000 increase. Yeah, in fact closer to 11,000. OK, but how many young UK nationals were hired in that period? According to the Centre for Social Justice, 11,000. That's the maths. No, that is not the maths. I'm going to guess that almost all of these 3.85 million young people with jobs in 2025 were hired at least once in the last five years and many of them will have been hired several times. I mean, I think I was hired four or five times before I turned 25 and in general the UK labour market has quite a bit of turnover. Leaving and entering jobs all the time. Quite. Emma Monk, who is an oil listener and social media debunker, made this point on Twitter. But both you and Emma are missing the point, which is that when the Centre for Social Justice used the word hired, they don't mean it. They mean aggregate increase in the number of jobs over five years. And that's nonsense. I'm presumably you challenged them on this point. Yeah, they said that we were splitting hairs. That's not splitting hairs. That's using the English language. And this really matters because in a few years demographic change is going to mean that there's no increase at all in the number of native born young people with jobs because the number of native born young people is going to be falling. Will they say then that literally no young person has been hired? Shall I go through the rest of the maths for you? OK, go on. So the number of young UK nationals with a job increased by 11,000 between 2020 and 2025 and the number of non-EU nationals in work rose from 80,000 to 370,000, an increase of 290,000. Keep going. The CSJ then compared these increases and divided one by the other. That's how they get their 27 to one figure because 290,000 divided by 11,000 equals about 27. That's number one. But actually several million young UK nationals have been hired, probably more than once. Yes. And by the same logic, most of the 370,000 young non-EU nationals too. So to fix the headline around nine or 10 British young workers hired for everyone non-EU migrant? Yep. So the original headlines, no good. It is true that the number of employed young UK nationals has been pretty flat over the past five years and the number of non-EU working people has gone up a lot. But it's not clear whether that's a good or a bad thing. Yes. And it's missing one part of the story. Oh. For some reason, the CSJ completely ignored young workers with EU nationality. The number of these has collapsed over the same period from nearly 240,000 in 2020 to 63,000 in 2025, a fall of more than 170,000. Which is the post-Brexit change in migration. Yeah. And even if you do include EU nationals into the equation, migrant employment has still increased, whereas UK national employment is pretty flat or has declined. And we also know, highlighted in a recent government report, that we do have an issue with needs and they're young people, not in employment, education or training. Over the last five years, the proportion of young people who are needs has gone up from just under 12% to just over 13%. Although, need data also includes non-EU nationals. And are migrant workers driving this increase in needs? It's hard to say. The government report found that migration wasn't a primary driver of the increase in the need population, but more competition for jobs will obviously make it harder to get jobs, particularly in low-paid work. Also, those 12 to 13% of young people are not necessarily all looking for jobs. More than half of that group are described as not actively looking for work or unable to work. And this number has risen sharply since 2020, but there's been a slight decline this year. Now, the employment rate for all young people has fallen over this time period, from 55% in 2020 to 51% in 2025. But that's all nationalities lumped together, though. We'd love to have the employment rate for young people broken down by nationality, but no one collects that data. And what did the Centre for Social Justice say when you asked them about all this? They told us... It is indisputably the case that for every 27 net additional young non-EU migrants on payrolls since 2020, there was just one net additional young UK national. I mean, that is indeed what those two numbers say, but as for how the change in those two numbers should be interpreted, whether the increase in non-EU nationals in the workforce is causing an increase in needs, replacing EU workers, helping to staff public services, or is bad for Britain, is certainly highly disputable. And the claim we've been looking at doesn't really help us to figure that out. Thank you, Lizzie. And if you're interested in that big change in the labour force since we left the EU, then we have got a Radio 4 series for you. Ten years after Brexit is a special, special series marking the 10-year anniversary of the Brexit referendum. Here's a friend of the programme, Madeline Sumption, from the Migration Observatory, explaining everything. Well, after the post-Brexit immigration system was introduced, there was a big increase in overall net migration to the UK. And I think that this happened partly by design and partly by accident. So you had the war in Ukraine and people coming from Hong Kong. There was then more people coming in as international students. And then there was an increase in people coming on work visas, initially in some of these middle skilled jobs, like butchers and chefs that hadn't been eligible for work visas under the previous system when there was free movement, but then became eligible. And then perhaps most dramatically in the care sector starting in 2022. And my suspicion is that if the government had known how many people would take up these policies, it would have introduced a much more restrictive offer. It's just that the enthusiasm for these new policies among prospective migrants was much greater than they had anticipated. Madeline Sumption, and if you want a whole lot more expert analysis on the impact of Brexit, search out Ten Years After Brexit on BBC Sounds. We're not much of a betting program here on more or less. Spoiler alert, the house always wins. But it is sometimes useful to get people to put their money where their mouth is. Or mouths are, whatever. Back in September 2024, there was a lot of speculation about the effect that the introduction of VAT on private school fees would have on pupil numbers. Some thought a drop of 25% was a realistic possibility. So we decided to host a bet that had been cooked up on the internet. In one corner, substacker Sam Friedman, who bet that private school pupil numbers would fall by 5% following the introduction of VAT. In the other, Maxwell Marlowe from the Adam Smith Institute, who bet they'd fall by 10%. Tellingly, neither of them seemed to take the more dramatic predictions seriously. Anyway, the numbers are in from the government school census. And the actual change in pupil numbers since before the VAT introduction was announced. Drumroll please. Well, let's let Sam tell you. The fall according to DFE data was 5.6%. That's pretty close to 5%. So you won, I think. I'm going to call it. You won. Yes, yes. I actually did think it might be a little bit lower than that, but I'll take the win. Yeah, well, I think we can discuss the details. But Maxwell, do you agree that Sam has won? Yes, I can see it. And I'll be emailing a receipt of my donations to the Trustal Trust, I believe, as £100. Sam, I'll send that over to you afterwards to confirm. Great, thank you very much. So the bet is done. But joviality aside, the stat in question represents thousands of children. The number of children in private schools was down by 33,000 over the last two years. Their lives have been changed by a change in tax policy. Yes, obviously it has affected some families, some young people who've had to move school. But most of the fall has come from in the entry years. So it's people not starting school who otherwise would have. Most parents have kept their children in school rather than move them around. But you have seen a drop off in the numbers applying to private school as a result of VAT. That's where most of the fall is happening. And those young people will mostly be going to state schools instead. And that's what 93% of the population do already. So we have very good state schools in this country. So yes, it will have affected some people, but it's mostly children not going to private schools from the start who otherwise would have done. There is another part of this that's worth noting, the demographic background to these figures. So the state sector has seen a drop of just under 2% over the same period. If you take that into account and assume that levels are dropping equally on both sides, that suggests that about 3.7% of the drop in private school pupils is to do with VAT. And the rest is just because numbers are falling across the system because we've had lower birth rates over the last decade or so. At the same time, the headline numbers also hide an important distinction. Because as we spoke about on a previous programme, there are two different things going on when you compare mainstream private schools with private special schools. Yes, absolutely. So these special schools, of which 146 new special schools have opened over the past 18 months, that's an increase of around 9,900 students for the sector. What this means is these schools are really quite specialist. There's more staffing, there's more care and attention made for students who have a learning disability or other disabilities that would make their education much harder, whether that is in a state school or in a mainstream private school. So a lot of this has been driven by private equity, so there's a lot of money flowing in. But it does mean fundamentally that if they have this educational health and care plan, an EHCP, which is given out by the Council and the NHS in conjunction, it means that there's essentially VAT exempts, there's some taxpayer funding there even, which actually affects the overall VAT take from this policy. If you remove the growth in the pupil count in private special schools, then the decline in the numbers in mainstream private schools is bigger. I would say that looking at the arithmetic, kind of the mainstream independent decline, for England at least, is between June 24 and June 26 would be about 43,100 odd students. So 9,900 is a transfer from those mainstream schools to those special schools. And if we were to redo the BET, which we're not doing, because I've conceded already, it would be about 7.7 to 7.9% if you include the 2,600 lost students in Scotland. So that's really where a lot of the transfer is going and I still concede and I will be making my donation, but I think this is very nuanced and I expect this to carry on into the long term as this debate and the sector continues to evolve. All of this has an impact on whether the predictions of financial benefit from the policy come to fruition. The fall was a bit larger than the government forecast, which means they'll have raised a bit less VAT revenue. And it also means they'll have to pay for children who go instead to state schools at taxpayer expense. I asked Sam if we can figure out the costs and benefits to the Treasury. So it's quite difficult to figure that out for two reasons. Firstly, because the numbers in the state system have also fallen because people numbers are going down generally, the transfer of pupils from the independent sector to the state sector is pretty marginal in the scheme of things. And the marginal cost of each of those pupils is very low. If you imagine primary school in London, which has a few additional students from a local independent school but whose numbers were dropping anyway, they don't need to hire any new teachers to accommodate those pupils. They don't need to pay for any more space. So the actual cost of accommodating those pupils is very, very low. The government will collect a little less in VAT than they thought they would because numbers are down a bit more than they were predicting. But again, it's fairly marginal in the scheme of things. And of course, the people who are not spending that money on private schools anymore will be buying other things with that money, most of which has VAT on it as well. Maxwell, there were some fairly alarming predictions in some corners of the press. Some suggestions that pupil numbers might fall by 25% or more. Do you think that this is going to be a crisis or a minor problem for the private sector? I mean, as I said, the last time I was on, we made the bet. I think the 25% plus is total exaggeration and it's a shame that filtered through. But I do think that if we look at the way the government have phrased or at least forecast the migration of students, they've already made up quite a lot of that budget already. The government predicted that by this point, 14,000 pupils would have left the private school population due to the introduction of VAT. If you take off the demographic change from the headline number of 33,000, the actual figure comes in at around 22,000. Thanks to Sam Friedman and Maxwell Marlow. In last week's programme, we asked whether one in five Welsh children really are leaving school functionally illiterate. We heard from Professor Cathy Rastall, who told us that Welsh school children perform particularly poorly on the international PISA tests. Professor Rastall suggested that this is because Wales has not required its schools to teach phonics, as is the case in England. But several of our loyal listeners wrote in to suggest another reason that Welsh pupils might be performing poorly on these tests, the fact that they speak Welsh. You asked us, are children taking the test in their first language? Does being bilingual mean you spend less time on each language? To find out more, we spoke to John Jarrim, Professor of Education and Social Statistics at University College London. Welcome to more or less. The first thing to clear up, do school children in Wales take the PISA tests in English or in Welsh? They can actually pick the language of the test that they take it in. So if in Wales about 17% of the pupils took the test in Welsh. And the child chooses or the school chooses? Mostly the school chooses in terms of Welsh medium schools will often instruct the school to take it in the Welsh language. If children are taking the test in their second language, does that affect their results? So yeah, so we've done some analysis of this issue previously and we found that it does matter which language the children take it in, with those taking it in Welsh, tenured performed lower than those taking it in English. Any idea why? I mean, there's a lot of potential reasons why this could be the case, including several methodological issues. So one of the things that happens in Wales when young people take their GCSEs is they often get a choice whether to answer in English or in Welsh, or rather they get to see questions in both languages. That's not the case in the PISA study where essentially they have to choose one or the other. So, you know, that's something that could affect the results and it is something that's probably a bit unusual to the children. There's also been potentially longstanding issues with the translation obviously in PISA into other languages as well, and Welsh is one of those. And because it's a relatively minor language, with not that many people obviously internationally, speaking Welsh and taking it, it doesn't have quite the same level as verification as some of the more major languages out there. So they are potential kind of methodological reasons why there's a difference. And I suppose another reason is if the child is learning two languages, there are advantages to that, of course, but maybe it's just harder to learn two languages and it slows them down a bit. Do we know about that? Yeah, sure. That is also a possible reason as well. I think the interesting thing that we see in PISA is that it's not only reading where Welsh children tend to perform lower than the rest of the UK. You also see a difference in other subjects like mathematics and science as well. So although it is pronounced in reading and slightly more pronounced than in other subjects, you know, you do see it kind of across the board which suggests it's not only a language issue while Wales is underperforming relative to the rest of the UK. So there seems to be some disadvantage for reasons that we cannot fully account for in taking the test in Welsh. If we kind of just none of the numbers back up again and say, well, let's adjust for that, does that then fully compensate and put the Welsh school system back on par with the school system in England? I mean, it seems not. Well, at least in the PISA study from my previous analysis, I put the difference at around 10 PISA test points. So if you took into account the Welsh language issue, it might boost Wales's performance and literacy up to around the OECD average. That's still well behind the rest of the United Kingdom, England in particular. So I think it accounts for part of the difference, but not all of it. Do we know then why Welsh children seem to be doing worse than children in other parts of the UK? Not exactly. And it's likely to be a multitude of reasons, right? So one of the things that you often see with these international studies is people go to their pet explanation as, aha, that's the thing that's driving it. That's the reason why. And in Wales, you can look at things like there being less emphasis on league tables, not measures like progress 8 that there are in England or whatever, and all these things could potentially be explanations. And it's very hard to tease out the impact of one from another. Yeah. I mean, one thing that Professor Astel mentioned to us was the fact that Wales doesn't mandate the use of phonics as a possible explanation for the poorer performance of Welsh children. I think they can still use phonics if they want to. So I was curious as to whether that sounded plausible as a partial explanation to you. I mean, there's definitely some credibility to putting that argument forward. How much of a difference it makes, we don't know. And it would be surprising if it can explain the entire difference between Welsh children and English children all the way up to age 15. So again, it might be a contributing factor, but no one really knows how much of the difference that can explain between, say, England and Wales. It's very hard to tease it apart. Our thanks to John Gerrim, Professor of Education and Social Statistics at UCL. Have you ever had an anxiety dream? Maybe the one where, well, your teeth fall out. Or you realise you're in Cafe Nero and you haven't got any pants on. That's just you, Tim. Or maybe it's the old classic. You're sitting down to do a maths exam, you open the paper, and you can't make sense of anything. Well, for thousands of A-level students up and down the country, that nightmare was their waking reality last week when they sat the EdXL Pure Maths Paper One. It was hard, much harder than they expected. So hard that it became a national news story, reported in The Times, Telegraph, Daily Mail, even the BBC. So why was the paper so hard? And has it meant that thousands of students have lost out on the grade they desperately wanted? To find out, I spoke to Sebastian Bison, a former school maths teacher, now maths YouTuber, who specialises in maths and further maths A-level. There was a bit of fuss recently about an A-level maths paper. What exactly happened? So the maths paper was sat last week. It's the first A-level maths paper. And usually after an exam, you can kind of get a sense of how students have found it. And I would say when the students came out of the exam hall, there was just an overwhelming sense of, wow, that was tough, that was really challenging. And since that happened, there has kind of gathered quite a bit of pace online with the petition of students saying this exam was too hard, this exam was too tough. Do you have any idea why they struggled? So there's a few things. I think one of the things was that many of the questions had quite horrible answers. And what I mean by that is in maths, if you do a question and you expect to get an answer, there's an integer, a whole number, or something that looks quite clean, it usually gives you a little sort of wink and a nudge of, yeah, you've done this question correct. But instead, they were ending up with quite sort of nasty looking numbers. I would never say an ugly number. I would never say that as a mathematician, but things like 231 over 47 or long decimals. And what that tends to do is create a sense of sort of like maths paranoia of, am I doing this question correct? And for young people, naturally, where it's quite a high stakes exam, this can lead to a bit of a domino effect of starting to panic. And that panic sort of seeps into other areas of the exam paper as well. Interesting. So if you put down your, you did your working and you came up with the answer is 12x, you would feel one way about it. And if instead you got the answer is 12.833x, or the root square root of something, or you would just start wondering whether you've done it right. And then you're taking time to check and definitely. So does it matter? I mean, the reason I ask this is because I understand that the grade boundaries are difficulty adjusted. So in the end, who cares? So I think when we look at the bigger picture, I would agree with that. The same proportion of students will end up with the A stars, the same proportion of students will end up with the A's. But I think when we zoom into an individual level and we have a student who for whatever reason, maybe don't have a stronger sense of resilience in an exam, and they wanted to get a particular grade, which they've been on track for, they might have panicked and it might have set them on a different kind of trajectory, which means they're not gonna necessarily end up with that grade. So we need to think about that difference between like the bigger picture, nothing's really changed, the grade boundaries will be set accordingly, with maybe those individual cases of students who have panicked and not done as well as they could have done. Has it got easier or maybe even more difficult to do A level maths over the years? This is where it gets contentious. I personally think that the current specification that we have, there was a reform in 2018, I think it's much tougher than the specification that I actually went through A levels with in the previous iteration. The types of things we're expecting students to be able to do, the volume of things that we want them to have understood and learned is much higher than when I took my A levels. And that has been reflected in the exams themselves, the questions are tougher, but each year we do have a roughly similar proportion of students that are getting A stars. So we don't actually get to reflect the fact that the students are doing better, they're able to learn these concepts better, teachers are getting better at teaching these concepts. So there's a lot of things going on, but I would not be of the mind of saying that things are getting easier, that's not my view. Well, from what you're saying, a student who gets a top grade now is actually performing at a higher level and doing more difficult, more challenging things than a student who got a top grade 10 years ago. I do believe that, and there is nuance to that because if you go back sort of 50 years, 60 years, the A levels were objectively more demanding with some very advanced kind of maths in it. I think 34 years, I think you'll find. 1992 was in fact when A level maths was absolutely at its peak. I wonder why. What advice would you give to candidates who have got through that first paper and are just about to take on two more A level maths papers? So I would say if you come across a question that's a show that's where you can't quite work out what the numbers might be, have a guess at what those numbers could be in the first part of the question, because then in later parts of the question, you can still gather up lots of those method marks and show the examiner that you really know what you're doing. Our thanks to YouTube maths teacher, Sebastian Bison. That's Mr. Bison to his pupils. Best of luck to any young, loyal listeners undergoing exam season, thoughts and prayers for the parents helping them through it. And that's all we have time for this week, but please keep your questions and comments coming in to more or less at bbc.co.uk. We will be back next time, and until then, goodbye. More or less was presented by me, Tim Harford, the producer was Tom Coles, with Nathan Gower, Josh McMinn and Lizzie McNeil. The production coordinator was Brenda Brown. The program was recorded and mixed by James Beard, and our editor is Richard Varden. Why do some brilliant business ideas come a cropper? I'm Sean Farrington, and in a new series of Toast, I'll be looking at five more brands, businesses and wonder products, which offered a lot, but didn't stick around, including a budget cinema selling tickets for 20 pence and the toilet paper we knew from school. I remember worrying about getting paper cuts in an area that you wouldn't wish to be getting a paper cut. Finding out what we can learn from their disappearance. Toast from BBC Radio 4. Listen first on BBC Sounds.
