Glasgow sex crimes went up 28%. Was it migrants?

A post on X says Glasgow had a “30% increase in sexual assaults, driven by the migrants that linger around at night”. We pulled the police records, the census and the Home Office files. Here is what they show. Every number links to a public document.

The number is real

Police Scotland reported it to Glasgow's Safe Glasgow Partnership on 13 March 2025. It covers April to December 2024, against the same months in 2023.

Receipt: Police Scotland, Glasgow City Q3 2024/25 report, appendix page 4.

So the post has the size about right. It missed something bigger. Rape went up more than sexual assault. And it was not only Glasgow. Across Scotland, rape and attempted rape went up 15% in the full 2024/25 financial year. In England and Wales, all recorded sexual offences went up 11% over that financial year.

Glasgow still stood out. Sexual assaults fell 4% across the rest of Scotland while Glasgow's went up 28%. That needs explaining.

Two things changed in April 2024

The rise is visible in the April to June 2024 quarter. Two changes need to be considered when comparing these records.

1. The rule for what counts as evidence changed

A recent account of an attack, together with distress observed by another person, can help corroborate the account. Police Scotland told Glasgow councillors the guidance “came into effect in April 2024” and applies to “not only domestics but all incidents”. It gave this as the reason domestic abuse cases rose.

Receipt: Police Scotland, Glasgow City Q2 2025/26 appendices, page 3.

2. The police got a new computer system for recording crime

It is called UNIFI. Police Scotland say it was “fully rolled out across all Police Scotland divisions as of 5 December 2023”. It replaced the old systems of eight former forces. From 1 April 2024 all crime figures come from it.

Receipt: Police Scotland FOI 24-0634; SPA performance report Q4 2024/25, appendix.

Police Scotland has explained the domestic abuse rise by the new evidence rule, and says the rape rise reflects more reporting. The reports checked here give no explanation for the sexual assault rise and do not attribute it to migration. They do not measure how much of that rise came from the evidence rule or the recording-system change.

What kind of crimes went up

A “detected” crime has an identified accused and enough evidence to justify consideration of proceedings. It is not a conviction.

Receipt: Police Scotland Q3 2025/26 management information workbook (Glasgow City rows).

Rape detections doubled. Recorded crimes are counted when the report is raised; detections are counted when the crime is detected. A detection in 2024 can therefore concern a crime recorded earlier. These totals do not show how many of the newly recorded crimes were detected, or whether the accused were partners, strangers or recent arrivals.

The smaller increase in sexual-assault detections does not tell us how many of the extra recorded assaults had a known suspect. An undetected crime can also have a suspect without enough evidence to meet the detection threshold.

Receipt: Police Scotland, recorded and detected date definitions.

What the accused records show

UNIFI provides a searchable nationality field for recent crime reports. The box is not compulsory. In the published tables, nationality is unknown or unrecorded for about half the accused entries.

Greater Glasgow, 2025: 117 accused entries linked to the detected rape-related crimes listed in the disclosure.

Receipt: Police Scotland FOI 26-0040. The response does not establish a count of unique people.

Nationality and place of birth are different things, and half the forms are blank. If every blank form were a foreign national, the non-British share would be 63%. If every blank form were British, it would be 10%. So this table cannot settle whether migrants are over- or under-represented. Most entries with a known nationality were British; that does not tell us the nationality of the missing entries.

A second table gives 153 accused–crime entries for Greater Glasgow rapes recorded in 2024, excluding restricted records. Ethnicity is recorded for 125 entries; 92 were White Scottish or White British. These are not unique people.

Scotland-wide, for sexual crimes recorded from January 2024 to May 2025, 19% of accused entries with a known nationality were non-British. The census says 10% of residents were born outside the UK. Those measure different things and cannot establish an offending-rate comparison. Nationality was unknown or unrecorded for 54% of entries, and no age or sex adjustment has been made.

Receipt: FOI 25-0272; FOI 25-1795; Census 2022 table UV204.

These tables do not provide a comparable nationality breakdown for 2023 and 2024. Police Scotland refused requests for 2023 because producing the figures would require a manual review of legacy records beyond the FOI cost limit. That does not mean the information was never recorded. Asylum status is not a readily searchable field either.

Receipt: FOI 24-1614; FOI 26-0373; FOI 25-3132.

Could new arrivals have done it?

First, the dates. The 28% is for April to December 2024. Anyone who arrived in Glasgow after that cannot have caused it. So the question for the 2024 rise is who was here in 2024.

Glasgow housed about 4,000 asylum seekers then, and it still does. The Home Office count was 3,939 in December 2023, 4,193 in December 2024 and 3,938 in June 2026. What did jump was refugees who had just been given the right to stay, leaving Home Office housing and asking the council for a home. Most of them already lived in Glasgow. The council says 1,048 households in 2024/25 came to Glasgow from elsewhere after being granted status.

Receipt: Home Office asylum support by local authority, to June 2026; Glasgow HSCP annual report 2024/25, section 2.6.4.

Now the sums. Glasgow recorded 266 more rapes, attempted rapes and sexual assaults in 2024/25 than the year before. Glasgow has about 271,000 men aged 16 and over. The city's rate is 3.8 of these crimes per 1,000 men a year.

Receipt: Scottish Government, Recorded Crime in Scotland 2025-26, Table 1; National Records of Scotland, mid-2025 population estimates.

These are scenarios, not measurements. The group sizes are assumptions, the city rate counts records rather than offenders, and one person can be behind several records. Even so, at five times the city rate the men in the new households would account for about 7% of the rise, and every man among the asylum seekers about 18%.

The last 12 months: a second test

A lot of people believe Glasgow has had a big wave of migrants in the past year. The records say the wave is one particular group: refugees who have been granted the right to stay and then turn up homeless. The council expected about 3,500 refugee households to apply as homeless in 2025/26, up from 2,750 the year before. About 900 of them travelled to Glasgow after being granted status somewhere else. By March 2026, refugee households were 44% of new homeless cases and half of the live caseload.

These measures cover different dates and populations. Support counts and homelessness applications are not direct counts of new arrivals; the small-boat figures cover the whole UK.

Receipt: Glasgow HSCP Strategic Plan Monitoring Report, April 2026, page 27; HSCP Chief Officer update, March 2026; Home Office asylum support datasets, June 2026; NRS international migration; DWP National Insurance registrations to March 2026; Home Office irregular migration, June 2026.

The later figures provide another comparison, but they cannot identify who caused the earlier rise. More homelessness applications do not necessarily mean more people newly arriving in Glasgow. Here is what happened to recorded crime in the year to June 2026.

Receipt: Scottish Government, Recorded Crime in Scotland, year ending June 2026, Table 1. Police figures for April to June 2026 say the same: rape 108 against 117 a year earlier, sexual assault 192 against 179.

Rape did not go up. Sexual assault moved by the same 3% as the rest of Scotland. The 9% rise in all sexual crimes is online offences, mainly “causing to view sexual images” (+93) and “communicating indecently” (+67). Glasgow's latest year looks like Scotland's.

Single cases in the news are real, and each one matters to the person it happened to. But a case is not a rate. The records above are the rate.

The England test

Here is a test with far more places in it. England and Wales count crime under one set of rules. Some councils have no asylum seekers on Home Office support. Some have 8 per 1,000 people, more than Glasgow's 6 per 1,000. If asylum seekers drive sex crime, the councils with more of them should show bigger rises.

The result is mixed. A straight-line comparison across the councils finds a small link: councils with more asylum seekers at the end of 2023 had slightly bigger rises on average. That comes mostly from a few high-asylum councils with big rises. But councils whose asylum numbers went up during 2024 did not see bigger rises than councils whose numbers went down. And the test that follows the 276 councils quarter by quarter does not establish a clear association. The sums are in the next section.

Receipt: Home Office, police recorded crime Community Safety Partnership open data; our workings are in the downloads below.

How we did the sums

Three bits of maths sit behind this article. Here they are without the jargon. The full workings are in the downloads.

1. Was Glasgow's jump just luck?

Imagine every sexual assault recorded in Scotland in those nine months of 2023 and 2024 is a marble. Glasgow had 495 then 634. The rest of Scotland had 3,402 then 3,280. If Glasgow were really following the same trend as everywhere else, how often would you get a split that lopsided by chance? The test is called Fisher's exact test. Answer: about one time in 84,000 under that simple model. Glasgow's recorded counts did move differently, but this is not the real-world odds that luck or a particular cause explains the change. The test assumes independent records and does not account for local effects of recording changes.

We also asked how unusual it was for Glasgow. We had seven earlier years where the same April-to-December comparison could be made. In none of them did Glasgow pull away from the rest of Scotland by anything like this much. Under the old pattern, a gap this size would happen about once in 190 tries under the model's assumptions. With only seven earlier comparisons, archive gaps and recording changes, that is an illustrative check rather than reliable odds about underlying offending or migrant responsibility.

2. Do more asylum seekers mean more sex crime?

A correlation is a number from −1 to +1. Zero means no straight-line relationship; other patterns can still exist. Across the 276 English and Welsh councils, the correlation between how many asylum seekers a council houses and how much its sexual offences rose was 0.15. Between the change in asylum seekers and the change in offences it was 0.03. Both are small, but the first is not zero. A straight line through the councils slopes slightly upwards, and that slope passes the usual test for chance (p = 0.025). It stays when London is left out and when councils are compared only with others in the same police force. It is a comparison between places, and places differ in many ways.

Then the within-council version. We used the available observations from 276 councilsacross 14 quarters. We let every council have its own normal level, so a big city is not compared with a village. We let every quarter have its own shared effect. That accounts for common changes, but a law or recording change can still affect places differently. Then we asked: when a council's asylum numbers go up, do its sexual offences go up more than everyone else's that quarter?

The answer was +0.7% in offences for each extra asylum seeker per 1,000 residents, and the honest range around that answer runs from -0.8% to +2.3%. That range includes zero. In plain terms: no clear evidence of an association within councils over time, not proof of no effect. The same test on Scotland's 32 councils gives -3.8% with Glasgow included and -1.0% without. Both ranges include zero too.

3. Could the new arrivals have done it?

This one is division. Take the 266 extra rapes, attempted rapes and sexual assaults. Divide by the number of men in the group being blamed. Compare with the city's own rate of 3.8 per 1,000 men. The group would need to offend at 28 to 70 times the city rate. The group sizes are assumptions, so treat this as a scale check, not a measurement.

What we still do not know

We do not know why Glasgow's rise was bigger than the rest of Scotland's. Police Scotland has not explained it. Here is what they should be asked.

• •Why do officers leave the nationality box blank on half of crime reports? The Scottish Government says an arrested person “must tell the police your name, address, date and place of birth and nationality”.
• •How much of Glasgow's 2024 rise came from the April 2024 evidence rule and the new computer system?
• •How many of the extra rapes were by partners or ex-partners?

Until police answer, anyone who says they know who did it, in either direction, is guessing. Those are the facts. What you make of them is up to you.

Check every number

Every figure on this page can be checked. Download the data, the workings and the script results, or open the original documents through the source list below.

The request we are sending to Police Scotland

See also our figures on why poverty is worse in Glasgow and the drug-deaths report, built the from official records with linked sources.