Why is data on grooming gangs so bad?
28:48 Share

This BBC podcast is supported by ads outside the UK.

Toyota is the best resale value brand for 2025, according to kellybluebookskbb.com. And with a wide range of dependable vehicles for any lifestyle, you can get everything you need in a vehicle today while investing in tomorrow. So choose Toyota and choose value. Shop via toyota.com for great deals and more. Vehicles projected resale value is specific to the 2025 model year. For more information, visit kellybluebookskbb.com. Kelly Blue Book is a registered trademark of Kelly Blue Book Co. Inc.

Toyota, let's go places.

Deploy your career in IT today. Learn more at mycomputercareer.edu. Skillbridge and other VA benefits are available to those who qualify. BBC Sounds. Music, radio, podcasts. Hello and welcome to More or Less, where your weekly guide to the numbers all around us in the news and in life. And I'm Tim Harford.

This week, in a Curtis versus Curtis cage fight, we try to understand to what extent the 1979 election set the tone for today's political arguments about immigration. We explore the numbers behind nuclear enrichment and we ask an economist to explain why being pillaged by a Viking might be more lucrative than you would imagine.

But first, last week, Baroness Louise Casey published an audit on group-based child sexual exploitation. These are cases of child sexual exploitation with two or more perpetrators, what has been termed grooming gangs in much of the press coverage. The report said that the ethnicity of perpetrators had been shied away from by authorities.

When official data on ethnicity has previously been published, it's been reported that 83% of perpetrators are white, roughly in line with the population as a whole, and that only a small proportion, around 7%, are Asian. However, there is a lot of missing data. Around two-thirds of the cases don't have a record of the ethnicity of the perpetrator. Now you might say, how can that be? And you might also say...

Hang on, what do you mean by perpetrator? A suspect? Someone convicted in court? Well, we will get to that because the answers to those questions are connected. Baroness Casey has cautioned against presenting this data and this idea that 83% of perpetrators are white without including the large number of missing cases. That, she writes in her recent audit, would be misleading.

When Baroness Casey gave evidence to the Home Affairs Select Committee last week, she said this about the lack of data on ethnicity. There's an irony to sometimes we decide to find this data writ large and other times we decide not to. So, I mean, I'd add religion into that as well. You could easily run a deep dive into a set of data to establish all sorts of drivers, but nobody's bothered. Sorry, that's too harsh. Nobody has...

seen as a priority. Our reporter Josephine Casserly has been looking into this. Hello, Jo. Hello. So first, let's just go through what the data looks like with and without that missing data. Yeah, so this data is for England and Wales and for 2023. And as you said, when it was published, it was reported that 83% of perpetrators of group-based child sexual exploitation and abuse are white and around 7% are Asian and 5% black.

That's all roughly in line with the population as a whole. So it doesn't suggest that there's any particular ethnic group that's disproportionately involved in grooming. However, there's no data in 66% of cases. So if you include that in the total, you get 28% of perpetrators are white, 2% are Asian, 2% are black, and for 66%, we don't know.

And Baroness Casey is saying that we shouldn't be declaring that 83% of perpetrators are white men when we don't know the ethnicity in so many cases. That's right. So this is what Baroness Casey said to the Home Affairs Select Committee about that missing data. My view is collect something or don't collect something. For God's sakes, don't half collect it. That's a bloody disaster. So Jo, how is it that for so many of these cases we just don't know?

So it's all about the question that you raised earlier about what we mean by perpetrator. So we're not talking here about people who are convicted of rape or other offences. This data is recorded right at the beginning of an investigation. So when the police first become aware that an offence might have been committed, often when the victim first comes to the police. But at this point, the police might not have arrested anyone. They might not even have a suspect.

So they could only have a victim's description and the victim might not know the perpetrator's ethnicity and even if they do, they could get it wrong. And in some cases, they'll never have a suspect. What about when an arrest is made? The police should then ask the person that they've arrested what their ethnicity is. But in some cases, this just doesn't happen. And in other cases, the suspect might refuse to answer.

And in any case, if it's more than a few months down the line from when the crime was first reported, it's going to be too late to be included in that data because the data is just a snapshot of the cases reported in each quarter. No one's going back to fill in the blanks once more is known about a suspect. So I guess the next obvious question is that if two thirds of the data is missing, is there any reason to believe that any particular ethnicity is more likely to

to be in that missing data or whether the missing data is basically going to be the same as the data we do have? So that's the big question and we do not know the answer. There are different theories which suggest that the missing cases will contain more Asian men or fewer Asian men, but ultimately we just don't know.

And I should say that overall, this data really doesn't tell us very much, partly because so much of it's missing, but also because the data doesn't just cover what's commonly referred to as grooming gangs, but any sexual exploitation of children that involves two or more people working together.

So that could be parents or step parents colluding to abuse their children or two school kids abusing other children at school or two priests in a religious institution. And I'm sorry to list all these horrendous possibilities, but as far as we can work out, abuse by what might be considered grooming gangs could only account for around 20 percent of all child sexual exploitation by groups of people.

And all this ethnicity data that we've been discussing is for all those categories put together. There isn't a specific subset that's for grooming gangs. Right. So there are various layers of problems with the data. And together, they make it impossible to draw any conclusions, really, about the ethnicity of perpetrators.

Yes, but there are three police forces that have more complete data. So that's Greater Manchester Police, West Yorkshire and Rotherham. And their data has fewer gaps? It does. So their data is recorded in a different way. So once they have a suspect in a case, they're going back and filling in the blanks about the characteristics of that person, rather than just taking this snapshot of when the crime is reported. So in Greater Manchester, for example, that means that they have ethnicity data for all suspects. And how does the data look?

Yeah, so sticking with Greater Manchester Police, it's a very different picture. It's a much higher proportion of suspects who are Asian, around 54%, and that's compared to 21% of the population of Greater Manchester.

White and black suspects are both underrepresented. Is the data from West Yorkshire and Rotherham similar? It is. So both of those other data sets also show a disproportionately high number of Asian suspects. The Rotherham data is a bit different. It's linked to a specific police operation called Stovewood. And in this data, 64% of the suspects are Pakistani, compared to around 4% of Rotherham's population. Wow. And that is very different.

It is, but remember that Greater Manchester, Rotherham and West Yorkshire are all areas where there have been big investigations into child sexual exploitation committed by Asian men. So these investigations will have unearthed a lot of cases related to these communities. But the situation may be very different in other parts of the country. There's no suggestion from Baroness Casey's team that the cases of Greater Manchester, Rotherham and West Yorkshire are representative of the country as a whole.

And there's another question that I'd add, which is that now that police and criminologists are focused on Asian grooming gangs, is it possible that similar crimes committed by white men are now being overlooked because they don't fit this pattern? And that's just speculation, but one child abuse expert I spoke to worried that this could be a possibility.

It's also important to note that group-based child sexual exploitation and abuse only accounts for about 5% of total cases of child sexual abuse. So it's a very small proportion. And generally, it's thought that only a minority of child abuse is ever reported. So all the data that we're talking about here is about cases that are reported to the police. There'll be many others which are not. Josephine Cassley, thank you very much.

At the weekend, the US military bombed three sites in Iran linked with its nuclear weapons programme. The UK says it wasn't involved in the strikes, but the Foreign Secretary David Lammy, speaking on the Today programme, did say that Iran's nuclear programme was not acceptable. Yes, they can have a civil nuclear capability that is properly monitored.

that involves outsiders, but they cannot continue to enrich at 60%. We here in the UK, in our own sites at URENCO and Sellafield, we're enriching to about 6% to 8%. Why do they need 60% enrichment if it is not to get a nuclear weapon? We weren't exactly sure what these numbers really meant, so we enlisted the help of Geoff Parks. He's a professor of nuclear engineering at Cambridge University.

Let us start with the basics. The fuel for the fission reaction that makes nuclear power and nuclear bombs possible is uranium. Uranium is mined from deposits of ore. Uranium can be mined from places all over the world. The largest supplier of uranium at the moment is Kazakhstan, but it's also available from Canada. I think they're second in the league table. Namibia are third. Australia are fourth.

The peculiar thing is that the uranium you get out of the ground is actually two different forms of uranium mixed together like in a cocktail. They're called isotopes and they're identified by their numbers uranium-235 and uranium-238. The uranium as it comes out of the ground

is 0.7% uranium-235 and 99.3% uranium-238. So those two

It is all frankly a bit confusing, so let's continue with the cocktail metaphor. Uranium-235 is like the spirit in the cocktail, the isotope you need to get the fission reaction going for power stations and bombs. And uranium-238 is sort of like the mixer, it doesn't really do very much.

So, to make the uranium more useful for power stations and bombs, you have to increase the proportion of the 235, making the cocktails stronger. That is what enrichment is. And the process works like this: You can turn uranium into a gas by combining uranium and fluorine.

and then you can use that gas in a centrifuge-based process. So you feed your gas into the centrifuge, spinning at high speed. You may have heard of these centrifuges. When you spin the gaseous uranium compound in a container at supersonic speeds, the heavier uranium-238, the mixer, goes to the outside and the lighter uranium-235, the spirit, stays towards the middle.

So you then sort of suck out stuff from the middle of your centrifuge and that then is a gas with proportionately more uranium-235. One centrifuge only makes a small difference to the proportion of the uranium you want, so you have to repeat the process of spinning and siphoning multiple times. If you're using that process...

to produce fuel for a civil nuclear reactor, where you're looking for about 4% enrichment, then you need about 25 centrifuge stages to get from 0.7% uranium-235 as it comes out of the ground to about 4%. So that's how you get the fuel for nuclear reactors. To get the fuel for bombs, you just keep on going, spinning and siphoning. You can in principle get a nuclear weapon from...

enrichments of above 20%, but they're very, very large. And the higher the enrichment, the smaller the weapon becomes, so more easily transported and delivered, so to speak. For a practical nuclear weapon, you need to enrich until you reach 90% uranium-235. It's believed Iran currently has uranium at 60% enrichment, so 60% uranium-235.

Unless all that uranium has been destroyed, how far is it from 60% to 90%? It's harder to get from uranium as it comes out of the ground to 60% enriched than to get from 60% to 90%. It's just a question of more stages.

of the centrifuge separation process I've described. We asked Geoff for a technical explanation of how big a centrifuge is and he said somewhere between the size of a fridge and the size of a small car. But you need a lot of these centrifuges all connected together to enrich uranium all the way up to the level needed for a nuclear weapon. So basically you need a space the size of a small factory.

In the case of Iran, one of these small factories was hidden under a mountain, which was then hit by very large bombs. It is not yet known exactly what damage was done. Thank you to Professor Geoff Parks. You're listening to More or Less.

We get it. There are too many car insurance companies trying to convince you that they have the best car insurance rates. We don't think we need to convince you. We're rude, and we do car insurance differently. We don't think it makes sense to only base your car insurance rate on things that have nothing to do with your driving, like your occupation or education.

We'll be right back.

Good drivers could save up to $900 a year when they switch to Root. We're taking an old industry and making it fair. Root Insurance, because better drivers deserve better rates. Download the app today and see how much you could save. Terms and conditions apply, subject to underwriting review. See Root.com for details.

Deploy your career in IT today.

Learn more at mycomputercareer.edu. Skillbridge and other VA benefits are available to those who qualify. The BBC recently released Shifty, a series of documentaries about British politics and culture during the 1980s made by the filmmaker Adam Curtis.

In the films, he argues that Margaret Thatcher and Britain fell into a land of make-believe, where the concept of a shared reality on which we can all depend has dissolved, and with it, any hope of a functioning democracy.

But is Adam Curtis dissolving reality a little himself? Right at the start of the first episode, the text on the screen makes the following claims. Mrs Thatcher was behind in the 1979 election campaign. Then she made a speech attacking immigration.

She knew it would appeal to swing voters in the old industrial towns. I believe we shall only succeed in maintaining and securing our traditional tolerance and fairness in this country if we cut the number of immigrants coming in now. Immediately, she took the lead in the polls.

Maths professor and blogger Oliver Johnson wrote a post suggesting that this was itself make-believe. The speech wasn't during the election campaign, and Thatcher was never behind in that campaign anyway. So who is right? Who better to arbitrate than another famous Curtis, this one a knight of the realm, the king of election night, Sir John Curtis, Professor of Politics at the University of Strathclyde.

Let's go through this line by line. First, is it true that Mrs Thatcher was behind in the 1979 election campaign? That isn't true. No, the first, Mrs Thatcher was ahead, both on average and...

In all the polls by one, there was one exceptional poll that had Labour very, very narrowly head towards the end of the campaign. But Mrs Thatcher went into an election, or rather, Labour were precipitated into an election...

with the Conservatives well ahead in the polls in most instances by double digits. OK. What's more, she didn't make a speech attacking immigration during the campaign. That's my understanding. I've certainly checked, you know, the standard text on elections. There was a reference in a radio phone-in to immigration, but there wasn't any significant speech by Margaret Thatcher during the election campaign. Right.

Margaret Thatcher did make a big speech on immigration, but that was early in 1978, long before the election campaign. More famous still were her comments in an interview about how some people feel swamped by immigration. But that was also from early 1978. People are really rather afraid that this country might be rather swamped by people with a different culture.

And you know, the British character has done so much for democracy, for law, and done so much throughout the world, that if there's any fear that it might be swamped, people are going to react and be rather hostile to those coming in. How about the third sentence? Did she know that this speech, which didn't happen in 1979, would appeal to swing voters in the old industrial towns?

Now, actually, the community of people to whom it was widely thought that Margaret Thatcher focused, which became, as it were, the illustrious story. What was then being talked about was Essex man and Essex woman. And this was essentially was a focus on people who are skilled working class voters who

who, rather than being left behind, were the people who were advancing economically, who were perhaps some of them interested in the Conservatives' support for being willing to sell council housing. But these were people who, therefore, because of their economic advancement, were regarded as now potentially winnable over by the Conservatives because they were now looking forward to, in a sense, enjoying a middle-class lifestyle. So again...

Yes, it was argued that there was a challenge to Labour's grip of the working class, but it's not the challenge of the left behind over immigration. It was the challenge of a society becoming more affluent and what therefore would happen to the skilled working class voters who are at the more affluent end of Labour's traditional working class coalition. I think you know where this is going, but the final statement is that...

Immediately, she took the lead in the polls. No, I mean, the speech to which references made, I mean, both the speech, the swamping comment that you made and the speech which is made reference to in the documentary is not followed by any significant change in the Conservative position in the opinion polls. There was something that led to a big shift in the opinion polls later on, the winter of discontent. The polls shifted...

At the turn of the year of 1978 to 1979, when the winter of discontent broke out, there were significant public sector strikes. The graves weren't being dug. The bins weren't being collected. It was a pretty harsh winter and support for the government fell away quite heavily.

When we asked Adam Curtis about all of this, he acknowledged a mistake and took full responsibility for it. The original caption, implying the speech was during the campaign, had been oversimplified in error, he told us, and he would be changing it as soon as possible. It should have referred to the polls leading up to the 1979 election campaign, although, as you heard, Sir John Curtis isn't convinced by that claim either.

Adam Curtis also completely agrees that the winter of discontent was central to the Conservative election victory. At the same time, Adam Curtis told us that election analysis isn't really the point of the film. The series is not in any way a traditional political documentary. Its main aim is to focus on what was going on in the minds of millions of British people, on their thoughts, feelings and fears that rose up over the past 40 years.

In this section, I wanted to show how Mrs Thatcher touched on something that was going to spread much further over the coming decades and have a profound effect on our society. That was a statement by Adam Curtis and our thanks to Sir John Curtis. No relation. Duncan Weldon, an economics writer and friend of the programme, has a new book out. Blood and Treasure is about what economics can teach us about war and conflict.

Timely, you might think. But we weren't going to let Duncan come on here and share his opinions willy-nilly about the economics of Viking raids. Because as anyone knows who's been watching the former Fox journalist Tucker Carlson, there is an essential rite of passage for anyone who wants to share an opinion. Here's Carlson interviewing US Senator Ted Cruz. How many people live in Iran, by the way? I don't know the population. At all? No, I don't know the population.

You don't know the population of the country you seek to topple? How many people live in Iran? 92 million. Okay. So time for a game. Population estimates mean prizes. And in this case, the prize is a chance to plug your book. Please do play along at home.

Duncan has been given about 20 seconds notice of this. Tim, I promise not to Google. Yes, well, you better not. If you do tolerably well at this quiz, I will put in a word for you with the Trump administration. There is probably a cabinet post available for you. So, population of Turkey? Turkey, I'm going to say about 80 million. 85 million, not bad. South Africa? Ooh, South Africa. I'm going to go...

90 million? It's 63 million. Vietnam? Vietnam, quite high. 110 million? 101 million. Colombia? Colombia, lower. 30 million? It's 52 million. Not sure that's very close. Saudi Arabia? Quite low. 22 million? It's actually 35 million. I think that's been a rapidly growing population. You may be just slightly out of date with that. Really tough one to finish. The UK? The UK.

The UK, ooh, 70 million? Yeah, 68 million. I would have thought you'd get closer, but it's not bad, Duncan. I'm going to give you Turkey, Vietnam and the UK. I think you got South Africa and Colombia and Saudi Arabia wrong, although none of them grotesquely wrong. I think you were within the ballpark.

With that highly meaningful test of competency sorted, let's get to the book and one fascinating example of the way that an economic analysis of conflict can produce counter-intuitive results. Let us talk about the Vikings and no, not the rugged, sexy Vikings of my dreams, but the unspeakably violent raiders of yore. I would have thought that being pillaged by Vikings is very bad news for your economy.

Yes, I mean, you know, being raided by the Vikings is quite a zero-sum game. You know, some people turn up in a boat, they kill a lot of people, they carry off movable wealth. The economics of that aren't particularly interesting. I'm going to suggest negative sum and I'm going to suggest that the Vikings are usually on the winning side. Yes, I will completely agree with that. But what is interesting as the Viking Age moves on is the Vikings begin to change tactics.

Rather than a handful of Vikings turning up in a boat, grabbing some loot and making their getaway, you start to see larger and larger Viking forces appearing in the countries that become England, France, even in the Iberian Peninsula. In economic terms, the Vikings start to move from being roving bandits to stationary bandits.

And the incentive structure for a roving bandit and a stationary bandit are rather different. A roving bandit is incentivised to grab as much as they can carry and do a lot of damage if they have to to grab it. Whereas a stationary bandit, someone who is going to be in the same area for a long time, it starts to look more like a mafia protection racket.

They're extorting money from people. And if you're running a protection racket, if you're a stationary bandit,

you're incentivised not to kill the goose that laid the golden egg. This theory of the stationary bandit originally comes from a political economist called Mansa Olsen, and it makes a lot of sense. A stationary bandit might build infrastructure to protect that shiny egg and prevent roving bandits from making off with it. You know, another thing we can look at is, you know, payments of tribute to make the Vikings go away, you know, the Dane gold.

The Vikings would turn up in what became England or France, demand a lot of silver in order to go away. Now paying the Dane gold, paying the Dane geld, has received a bad reputation in history. So bad that Kipling wrote a poem about why it's bad to give in to these menaces. You never get rid of the Dane, right? You never get rid of the Dane.

But it's worth asking what the second part of this is. Once you've got a transfer of financial resources, which is ultimately what handing over a load of silver is, you then tend to see a transfer of real resources. The Vikings were not pirates in a children's cartoon who buried this treasure and put an X somewhere on a map.

they generally spent it. And it would appear that a lot of what they spent was spent buying goods and services from the people that had paid it. What's really interesting is, and as much as we can put the evidence together, if we look at Europe in the 9th, 10th, 11th centuries, the so-called Viking Age, when Vikings were prowling the coasts,

we actually see a pickup in trade, exchange and production, perhaps stimulated by these sort of transfers of tribute payment. Thanks to Duncan Weldon, and since he did pretty well at the quiz, a reminder that his book is called Blood and Treasure. That is all we have time for this week, but if you're at a loss while you wait for us, a reminder that we do have a shorter podcast on Saturdays,

And you can also find my other podcast, Cautionary Tales with Tim Harford, on BBC Sounds. Or perhaps you'd like to meet some of the pioneering people who've helped shape modern statistical research, such as Thomas Bayes or Florence Nightingale. You can do that. Head to bbc.co.uk, search for More or Less, and follow the links to the Open University.

If you've seen a statistical claim that makes you sceptical or curious or both, then tell us about it. We're at moreorless at bbc.co.uk. We'll be back next week and until then, goodbye. More or Less was presented by me, Tim Harford, and produced by Tom Coles, with Nicholas Barrett, Josephine Cassidy, Lizzie McNeill and David Verry.

The production coordinator was Brenda Brown. The programme was recorded and mixed by Gareth Jones. And our editor is Richard Varden. We pulled into what felt like an old compound. You think, wow, this is a very old property. I'm Danny Robbins and Uncanny is back. We have three brand new summer special episodes and things are about to get scary. I could feel something moving up the side of the bed and I can't quite believe what I'm seeing.

A trip to a tiny medieval town in central Spain turns into a holiday from hell. I could make out its long, matted fur, and I am absolutely petrified. Absolutely petrified. This was just pure terror. And we'll investigate more spine-chilling cases in an episode recorded live at the Hay Festival. Uncanny Hay audience, who is feeling team believer? And who is team sceptic?

Listen now on BBC Sounds.