Three Decades Of Covering Up Sexual Offending By Foreigners

Last month, The Economist published a calculation no British government department has been willing to put its name to. Using Ministry of Justice figures prised out by Oxford's Migration Observatory through Freedom of Information requests, it estimated Afghan nationals in England and Wales are convicted of sex crimes at seven times the British rate, even after adjusting for age and sex.

On the Telegraph's reading of the accompanying chart, nationals of the Middle East and north Africa came in at around three times the British rate, and sub-Saharan Africans at around twice.

The government knows all of this and publishes "guidance" for migrants about not raping the locals or their children. The cat is firmly out of the bag about the worst rape statistics in British history.

It is news to absolutely no-one who has paid any attention. Or anyone who has any knowledge of the repulsive "culture" in that part of the world, like the rampant rape of children. Africa doesn't need explaining with its relentless genocides and mass rape.

Then came the spin and the apologism.

Afghan convictions (not incidents or arrests) had moved from 92 in 2010 to 2012 to 102 in 2022 to 2024, while the Afghan population roughly tripled. Ten more convictions, the magazine concluded, "is not an epidemic". Four-fifths of sex crimes, it added, "were committed by British people".

A small, sad anomaly, safely filed. Not every migrant; just a few bad apples among the pitiful refugees seeking a better life. Just as British as you or I.

Both reassurances collapse once the figures beneath them are examined.

The police disclosures, prosecution records, and official audits now in the public domain describe something worse than one bad ratio: a state which knows far less than it claims about who commits sexual violence in Britain, which has repeatedly promised to find out, and which is currently paying lawyers to stop the public reading tables drawn from its own records.

Rare Convictions In A Tripled Population

In late 2021, the median wait in rape cases from offence to completion stood at 839 days. By September 2025, adult rape cases were taking an average of 417 days from charge alone. On those delays, a rape conviction recorded between 2022 and 2024 will typically concern an offence committed between late 2019 and the summer of 2022.

By the magazine's own count, some 38,000 Afghans came to Britain in the year and a half after the Taliban took Kabul in August 2021. Around half of the 29,000 people resettled under the two Afghan schemes landed in the third quarter of 2021. Another 8,633 Afghans crossed the Channel by small boat in 2022, and thousands more followed in 2023.

The Economist manipulatively set the old crimes of a small community people don't accuse or prosecute because of imagined "racism" against the headcount of a much larger one, and conveniently for the Establishment, pronounced the result as nothing to see here.

In Bristol in November 2025, three men and a 17-year-old boy, all Afghans, gang-raped a 17-year-old girl over several hours, then fled to France in a lorry. They were sentenced on October 6th.

None of it appears in the 102, because the data stops in 2024.

Police records from last year hint at what is still working its way through. Greater Manchester Police logged 23 Afghan-national offender records for sexual offences in 2025. The Metropolitan Police logged 16 records and Merseyside six, according to disclosures gathered by The Reasonable Adjustment.

Police records are not convictions; one man can appear several times, and Greater Manchester's table includes undecided cases.

Even so, three forces out of 43 generated 45 entries in a single year, against a national court total averaging 34 a year.

Operation Stovewood, the National Crime Agency's inquiry into Rotherham, had identified more than 1,100 victims and secured 42 convictions by the time of the 2025 national audit.

Count Rotherham the way The Economist counted Afghans and Britain's most notorious child abuse scandal comes to 42.

Nor are the 92 and the 102 really "crimes". The Ministry's extract counts cautions as well as convictions, by sentencing occasion, and its broad headings say nothing about seriousness.

The sex offences in the dataset range, in the magazine's own words, "from rape to consensual sex in public".

Each scores one.

Nothing to see here. Please refer questions to RICU at the Home Office.

How 20x Magically Became 7x

The seven was offered as a correction to Nick Timothy, the shadow justice secretary, who has claimed Afghans are more than 20 times likelier than Britons to commit sex crimes. His figure rested on an estimate of about 13,000 Afghans, extrapolated from a survey sample of fewer than 60 people, a series the Office for National Statistics stopped publishing in 2021.

Set beside earlier conviction data, which put the British rate at 2.66 per 10,000 between 2021 and 2023, it produced a rather startling multiple.

The Economist made two changes.

It threw out the 13,000 for the 2021 census count of roughly 21,400, then went further, estimating the population had doubled to 45,000 by the end of 2022 and averaging its figure across 2020 to 2022. Every extra head in the denominator lowers the rate before anyone has touched age. The census was taken five months before the Taliban took Kabul, and a higher number was always going to be needed.

The difficulty lies in the timing.

The convictions in the sum date from 2020 to 2022, and on the delays above many concern offences committed before August 15th, 2021, when a large share of the people in the denominator were still in Afghanistan. The Migration Observatory's own warning applies too: no reliable estimates exist of how many people of each nationality live in Britain.

The second change was age and sex standardisation.

Men commit 98 per cent of sex crimes, and Afghans here are about a decade younger than the general population. The adjustment trimmed the figure further.

• It tells us how a young Afghan man compares with a young British man.
• It tells us nothing about what the country has taken on.
• Britain admitted actual men at actual ages.
• The standardised population exists only on paper, and no woman has ever been assaulted by it.

The magazine called its estimate still uncertain, perhaps skewed by a few serial offenders or by quirks of detection.

Bullshit.

All the doubt attaches itself to the seven. None attaches itself to the ten.

A Third Of Sex Crime Arrests Are Foreigners

The comforting 80 per cent spans 2010 to 2024, a period beginning long before the arrivals now under discussion. The magazine concedes the police recording of nationality is "messy", then states the result as fact, although cautions and convictions can only show who was caught.

It also depends on what "British" means in the data, which is very little.

Nationality in these records is self-reported. A dual national is logged under whichever nationality he states as primary.

A naturalised citizen counts as British, and government crime statistics contain nothing on country of birth. It's highly unlikely anyone English would consider these people "British."

By March 2021, most people born in Afghanistan and living in England and Wales had already become UK citizens, according to the same Oxford briefing. Any of them convicted of an offence lands in the British column. The man in the dock helps to fill in the form, and the magazine then presents the form as demography.

The broad heading of "sexual offences" flatters the picture further. It sweeps together rape, exposure, and consensual sex in public.

Separate the serious contact crimes, and the foreign share climbs sharply.

Those figures come from Ministry of Justice data obtained by the Centre for Migration Control. Once the unknowns are included, foreign nationals could account for up to 34 per cent of convictions for sexual assault on a female.

At the 2021 census, 9.9 per cent of residents held a non-UK passport. Oxford puts non-citizens at roughly 13.5 per cent of adults by 2025.

One in five rape convictions and one in four convictions for sexually assaulting a woman: this is what lies beneath the sarcastic and insensitive words "hardly an epidemic".

London is starker.

1. Of 6,061 people the Metropolitan Police proceeded against for sexual offences between 2019 and 2022, 1,671, or 27.6 per cent, were foreign nationals.
2. A separate 2025 dataset of 1,599 custody-based records puts the share at 34.9 per cent, with another 8.7 per cent unrecorded.
3. Londoners with a non-UK passport made up 23.3 per cent of residents in 2021.
4. In stranger rape cases from 2015 to 2023, only 774 of 1,296 accused were recorded as UK nationals, with the remainder a mixture of foreign nationals, suppressed small counts, and gaps.

Force by force, the 2025 disclosures run as follows.

None of these is a rate, and none can be ranked against the others. Every percentage is a share of whatever one institution happened to tally.

In a country with a functioning statistical service, no-one would have to guess. But then again, we'd find out, and the government would have to concede.

A Third Of Nationality Data Is Missing

The Economist found convictions by ethnicity "more reassuring". Its age-standardised run for 2020 to 2022 suggested "Pakistani, other Asian and other non-black minority men" were no more likely to have been convicted of sex offences than white British men.

Black men are thereby left out of the reassurance, and the article never says what the magazine found for them. Nor does it say how much of the ethnicity column is empty, and none of it answers the grooming gang question, which concerns one form of offending in particular towns.

According to Ministry of Justice figures obtained by the former minister Neil O'Brien, the share of people convicted of sexual offences whose ethnicity went unrecorded rose from 15 per cent in 2010 to 29 per cent in 2024. For child sex offences it rose from 11.6 per cent to 28.7 per cent.

The system is getting willfully more blind as the question gets louder.

The Crown Prosecution Service's 2024/25 figures for rape-flagged cases show the same gap: 638 of 2,623 convictions, or 24.3 per cent, carry no stated ethnicity.

Set the remaining 1,985 against census shares and the raw picture looks like this.

Both black groups missing from the magazine's reassurance stand at about twice their census share or more.

These are raw shares, unadjusted for age, and younger groups would be expected to fall once adjusted.

A conviction in a rape-flagged case is also not always a conviction for rape.

They also assume the 638 blanks are spread evenly across every group. Nobody can test this assumption, because the state did not write the information down.

A comparison requiring precision to one decimal place cannot be run on a dataset missing a quarter of its entries, and anyone offering it as a verdict is overselling.

Cooking The Books To Get To 83% White

Baroness Casey's national audit published on June 16th, 2025 showed how this cleber accounting trick is performed. The national police dataset for group-based child sexual abuse in 2023 covered 6,670 suspects. Ethnicity was not declared for 4,404 of them.

Remove those people and the remainder come out 83 per cent white, conveniently close to the census.

Put them back and 28 per cent are known to be white, 2 per cent known to be Asian, and 66 per cent are a mystery.

Casey's audit found the same conclusion repeated for years, by senior officials reading flawed data and in a 2020 Home Office paper whose finding of a majority white offender population she found hard to understand and could not see evidenced.

Her verdict deserves framing:

Instead of examination, we have seen obfuscation.

Where police actually wrote the data down, the pattern changed.

West Yorkshire's table still has 24 per cent of entries blank.

In Rotherham, nearly two-thirds of the 323 designated suspects in Operation Stovewood were recorded as being of Pakistani background, in a town where Pakistanis make up about 4 per cent of residents.

Casey also found Greater Manchester's wider child sexual abuse caseload sits much closer to the local population, with 32 per cent of suspects unknown.

Offences differ, offender groups differ, and the single heading of "sex crimes" averages them into oblivion.

She further noted a significant proportion of the dozen or so live operations she saw appeared to involve non-UK nationals or asylum claimants. No national series can tell the public whether those operations were typical.

Nobody Knows Anyone's Nationality

On February 7th, 2025, Greater Manchester Police refused a request for the nationalities of foreign nationals arrested for sexual offences. Nationality, it explained, was "not a searchable field" on its custody system, and checking 2,961 reports by hand would take around 246 hours.

Sixteen months later, the same force, querying its crime records, produced a table of 2,691 offender records for 2025, listing nationalities from Afghanistan to Zimbabwe, with 267 unknown.

The information existed. Somebody had to be asked in the right way.

Elsewhere the field is simply blank.

The Reasonable Adjustment asked 25 forces for their 2025 figures. Sixteen supplied something usable, six refused, and three gave no substantive answer.

• Avon and Somerset, Cheshire, and the Police Service of Northern Ireland pleaded cost.
• Humberside reported an IT fault and sent ethnicity data instead.
• South Yorkshire asked for clarification of offence codes already in the request, then went quiet.
• Dyfed-Powys said it does not record offender nationality, which will surprise readers of the table of 1,065 arrests by nationality on its own disclosure log.

No two of the sixteen counted the same thing.

Their arrests, offences, custody bookings, and offender links cannot be added together, and four forces warned in writing against comparing their figures with anyone else's.

Every force was asked to mark counts under five as "<5".

• Merseyside and West Midlands printed exact numbers regardless; Merseyside stating small counts are not personal data.
• Greater Manchester and North Yorkshire withheld the same kind of figure as exempt personal data.
• Derbyshire reproduced the instruction, then listed several nationalities at exactly one.
• Devon and Cornwall hid nearly every nationality behind "<5", leaving UK at 1,279 and Unknown at 72.

North of the border, Police Scotland admits nationality has been an optional field since January 2024, filled in around half the time, and its crime system has no field for immigration status at all.

At national level the shelf is bare.

Alex Norris, then a Home Office minister, told the Commons in a written answer on April 28th the police recorded crime series

does not include data on the nationality and immigration status of suspects.

On October 21st, four days before The Economist published, the Office for National Statistics confirmed it holds no criminal justice outcome data, has never linked such data to Home Office immigration records, has no plans to try, and regards the work as outside its remit.

The UK Statistics Authority made no recommendations on the matter in 2025; the one internal briefing its regulator could produce concerned press coverage.

When Casey asked the courts for figures on group-based exploitation, HM Courts and Tribunals Service asked its staff to recall ten years of cases from memory and the result was unusable.

Prison populations are helpfully broken down by nationality.

The Home Office proposed more detailed tables by the end of 2025 and delivered them in May 2026, covering foreign offenders leaving detention or removed from the country, with entries into detention added in August.

In other words, we may now read about the men it has already caught and is showing off to the door. If it can get them through it, which it can't.

Data on foreign offenders living in the community, the group the public might reasonably worry about, still needs "further development".

25 Years Refusing To Record Characteristics

The Children's Commissioner called for police to agree a method of recording perpetrator ethnicity in 2012.

14 years ago. 10 years after the grooming gang scandal became widely known.

The following year, the government told the Home Affairs Committee local boards could decide how to collect their own data. In 2022, the Independent Inquiry into Child Sexual Abuse recorded a widespread failure to do it.

In March 2023, Chris Philp, then a Home Office minister, assured Parliament it had been mandatory since April 2022 for forces to record the ethnicity of those arrested in grooming gang cases.

Sir Iain Duncan Smith told MPs this year only one force in 43 had complied with the 2022 request.

On June 16th, 2025, Yvette Cooper told the Commons the ethnicity and nationality of suspects in child sexual abuse cases would be recorded on a mandatory basis for the first time. It had thus been mandatory for three years, and was now about to become mandatory. In July 2025, she wrote to chief constables setting out an expectation.

In January 2026, ministers in the Lords resisted an amendment to make it law, offering legislation when parliamentary time allows.

On June 1st, 2026, safeguarding minister Natalie Fleet promised a power to mandate the collection of ethnicity data.

By September 15th, Shabana Mahmood could report the drafting of a police reform bill was under way.

Two things went missing on the journey.

1. Nationality has quietly dropped out of the promised power, at least in the minister's description of it.
2. And ministers keep describing the duty in terms of child sexual abuse and exploitation.

On this description, a grown woman raped in a park remains outside it.

Astute readers will recognise this timeline of refusing to record "protected characteristics" is roughly the same to the Stephen Lawrence enquiry and its fallout.

Offenders' Rights Outweigh Disclosure

On June 22nd, 2025, the Centre for Migration Control asked the Ministry of Justice for convictions from 2018 to 2024, by nationality and detailed offence, offering to accept "<5" wherever numbers were small. The Ministry refused within eight days, citing section 40(2) of the Freedom of Information Act 2000, which protects personal data. During the regulator's investigation it added section 12, the £600 cost limit.

The Information Commissioner had to serve a formal Information Notice on May 5th, 2026.

The decision came on June 22nd, 2026, a year to the day after the request. The Ministry had failed to show the cost limit applied, and privacy protected only numbers below five. In one of the drier rebukes in recent Whitehall history, the Commissioner noted complexity "alone does not obviate the requirement to provide a properly evidenced estimate".

The full notice ordered disclosure.

The Ministry appealed on July 17th, three days before Andy Burnham replaced Keir Starmer and David Lammy lost his job.

His successor is the same Alex Norris who told Parliament in April the data was not in the crime series. He has left the appeal in place, and the Centre says the department has since sought a 28-day extension.

Court papers quoted by the requester show the Ministry arguing the rights of offenders "outweigh the legitimate interests in disclosure", and citing a risk of identification from the combined weight of statistics, court reports, and press coverage.

These are people convicted in open court, and West Midlands Police has released nationality counts as low as one under the same Act. The labour in dispute, on the Ministry's own estimate as recorded by the requester, is around.... 88 hours.

At the last public report, in late September, the tribunal had not ruled.

The Ministry says it does not recognise claims of a cover-up, and Lammy told Timothy publication options were under continuous review. Lammy might have consulted the absurd Lammy Review of 2017, which argued criminal justice agencies should publish their ethnicity data by default, because institutions earn trust by opening themselves to outside scrutiny.

Nobody outside the Ministry knows what the full tables show, and nobody can prove why it is fighting. We already know. We all know.

Governments do not generally hire counsel to suppress good news.

The Economist could compare Britain with Austria, Denmark, Finland, Germany, and Sweden because those countries keep usable records. Statistics Denmark routinely classifies convicted persons by national origin, country of origin, age, sex, and offence, and calculates a standardised index for comparing groups.

Record And Publish The Damned Data

Parliament should place a statutory duty on the Home Office and the Ministry of Justice to publish arrests, charges, convictions, and sentences every quarter, by detailed offence, with age and sex.

Each table should cover nationality, country of birth, ethnicity, and immigration route, with cells below five suppressed.

Rape should stand apart from exposure, stranger attacks from domestic ones, and group offending from lone offending.

Every published percentage should include the blanks, as Casey did, so no official can ever again turn 28 per cent into 83 per cent by deleting two-thirds of a dataset.

Recording nationality and ethnicity at charge should be compulsory for all sexual offences, for adult victims as well as children, with each chief constable answerable to the Home Secretary for the completion rate.

Police and court records should be linked to Home Office immigration records, a job the Office for National Statistics says it has never attempted and considers somebody else's.

The appeal should be withdrawn this week. It would cost the Ministry nothing but its dignity, an account overdrawn for some time.

The Economist asked whether immigrants from some countries commit more sex crimes. Its own numbers answered yes, and its article answered here's why that's a good thing.

The second answer was never available to it.

Nobody can rule out an epidemic from a ledger with a third of its pages torn out, least of all a government paying lawyers to keep the rest shut.