- We found insufficient evidence to support the claim.
- Mr Buckley’s claim relies on analysis produced by the Centre for Migration Control. Our research identified a number of serious limitations in the underlying CMC data, making nationality-based crime comparisons highly uncertain.
- The University of Oxford’s Migration Observatory argues that there are no reliable official statistics on crime by nationality, and comparisons between nationalities are often hampered by poor population data.
- The UK Government acknowledges significant gaps in this area of data collection, meaning there is currently no reliable evidence base to support claims of this nature.
On 9 June 2026 (listen from 31:15) DUP MLA Jonathan Buckley appeared on BBC Radio Ulster’s Talkback show, commenting on a knife attack in Belfast the previous evening. The suspect, who has since been charged with attempted murder, is Sudanese. However at the time of broadcast, the PSNI had reported that he was Somalian.
During the programme, Mr Buckley stated:
When we’ve looked at some of the statistics of those who have come to this country, via asylum seekers or whether it’s small boats or otherwise to the United Kingdom, as a whole it shows that the statistics read that those of Somalian descent are more likely to commit some of these crimes in comparison with other nationalities who have come and settled in the United Kingdom. There is a problem, there is a significant problem and I think that most people in the general street understand that. I think there has been an inability of political representatives to be truthful and honest with the public.
The programme’s host, William Crawley, enquired from where he got the statistics. The MLA responded:
“They were actually Home Office statistics […] There was a question in Parliament that outlined the criminal convictions of those from an asylum seeker perspective that came to the United Kingdom.”
- Court proceedings
Before we evaluate the claim, it is worth making clear that the attack in North Belfast is now subject to criminal proceedings and therefore subject to UK law around reporting of ongoing court cases. The details discussed in this fact check, however, are unrelated to the details of this case.
We should also underline that the alleged assailant is Sudanese. Mr Buckley was working from the PSNI information available at the time of speaking.
Nonetheless, this illustrates how inaccurate information can spread during heated moments of political crisis.
- Source of the Data
We contacted Mr Buckley to establish the source of the data. His office replied to say that he was ”referring to the arrest rate for serious crime in general for foreign criminals in the UK” as evidenced in a report by the Centre for Migration Control (CMC).
The reply went on to say that “these statistics presented by the Home Office outline that the arrest rate for people from Somalia in England and Wales is 64.6 per 1000 people. Comparatively, the arrest rate for people from England and Wales was 12 per 1000.”
It added, “Jonathan is also mindful of international statistics from the Journal of Scandinavian Studies in Criminology and Crime Prevention (Vol.15, Iss.2) which demonstrates the marked differences between immigrant groups with regard to violent offending. In the table relating to Norway, immigrants from Somalia had three fold or larger ratios when compared to the majority population.”
- What do these sources say?
On 6 January 2025, the Centre for Migration Control (CMC, more on it later) published an analysis of arrest statistics in 2024 that it received from police forces in England and Wales through a Freedom of Information (FOI) process, and compared them to population estimates by the Office of National Statistics (ONS) in 2021.
The CMC asked all 43 police forces in England and Wales for “a breakdown of arrested non-UK nationals by their nationality” in the first ten months of 2024. On the basis of the replies, the CMC created a “migrant crime league table”, which ranked Somalia as the national grouping with the sixth highest arrest rate in the UK.
In April 2025, the Centre for Migration Control (CMC) published further analysis, this time of conviction statistics between 2021-23 that it received from the Department of Justice in an FOI. The underlying data comes from police forces. It compares convictions to the same ONS population study in 2021. This report lists Somalia as the national grouping with the fifth highest conviction rate in the UK, which it says is 8.05 times greater that of the ‘British population’.
It is possible that this latter report is the source of some of the figures the MLA was referring to. So the analysis that follows is based on both reports, noting when a point may apply to one or both.
The CMC findings were widely reported in the media, on TikTok and repeated by some politicians, particularly those from Reform. It is possible they may also have been spoken about in the NI Assembly and UK Parliament, however our searches of Hansard did not turn up anything specifically relating to Somali crime rates. As always, we are happy to update our analysis and amend this article if readers can point us to relevant contributions.
However, the wider question is, are the figures in the CMC reports accurate, and did Mr Buckley communicate them accurately?
FCNI have identified a number of problems with CMC’s use of data, including significant criticisms about the organisation’s lack of rigour in its analysis. Many of these shortcomings are laid bare in internal documents from the Office for National Statistics (ONS) and Oxford University analysts The Migration Observatory.
- Issues with the CMC analysis
- Arrests ≠ crimes.
The January report uses arrest data. This has then been loosely translated by Mr Buckley on Talkback as “crimes committed” and “criminal convictions.” But these are all different things. An ONS internal review of the CMC’s figures (released under FOI) points out that arrest figures represent people charged, not convicted, and underline that “time and time again audits have found that the police in the UK falsely charge foreign nationals at a higher rate.”
This would inflate Somali arrest rates relative to their actual offending rates. This report takes no account of arrests that did not lead to charges, never mind convictions. This point does not apply to the April report on convictions.
- Underestimation of foreign nationals
Both CMC reports use ONS survey data from 2021 to estimate the size of each nationality group – the Annual Population Survey for the arrests report, and the Labour Force Survey for the convictions report. Both surveys share a fundamental weakness: as the ONS internal review of the CMC’s figures puts it, “population surveys don’t pick up foreign nationals well.”
Both are household surveys conducted in English, which systematically undercount hard-to-reach migrant communities. For small nationality groups like Somalis, this matters enormously: any uncertainty in the baseline population figure gets multiplied when calculating a per-1,000 rate. The ONS advised Sky News to use the more reliable 2021 Census instead, and even that predates a significant migration spike, meaning the true 2023/24 population would be higher still.
The practical effect of using an underestimated population baseline is that crime rates for small foreign national groups appear much higher than they actually are. Sky News, working with the ONS, the Home Office and Oxford’s Migration Observatory, found that the CMC’s methodology produced an Afghan sexual offence conviction rate seven times higher than a corrected calculation, 22 times the UK-born rate versus a more accurate estimate of around three times. If a similar degree of distortion applies to the Somali figures, the CMC’s headline numbers are not a reliable guide.
On top of this, the CMC states it chose the upper end of the ONS population estimate range for each nationality – a decision that is not explained by CMC. Regarding this, ONS stated:
“It’s instructive that the CMC took the upper end of the estimate; the aim may have been to make the figure look as large as possible.”
- Incomplete data from police forces.
There is no adjustment or weighting for the five missing forces in the report on arrests (Cleveland, Hampshire, Thames Valley, Warwickshire and Sussex), which could be significant for a small population like Somalis, who may be geographically concentrated. Contrary to other methodological decisions, this might have deflated Somali arrest rate figures.
- The police forces that did reply provided data on offences at “varying degrees of granularity“. Nationality is often mixed up with ethnicity, which can lead to inaccuracy. An internal ONS review highlights that “even some of those FOIs that forces put out, in the notes they explain how it’s a bit of a bodge.” Somalian population too small for meaningful data.
The Migration Observatory at Oxford has flagged serious problems with nationality-level crime statistics involving Somalis. The 2021 census counted only 2,130 non-naturalised Somali passport holders in England and Wales, but Somali-born migrants were more likely than average to hold no passport, meaning a further 6,200 people of probable Somali nationality were uncounted.
The Migration Observatory advised Sky News that with such small populations, an inaccuracy of just a few thousand people can have a very large impact on estimated rates. This point also applies to the report on conviction rates.
- The ‘British citizens’ category includes foreign nationals.
The reports compare foreign national arrest/conviction rates against a “British” arrest rate derived from the Home Office’s national figure. But this national figure includes foreign nationals (as total arrests include everyone), it does not attempt to isolate British nationals only. So the comparison is not like-for-like.
- No demographic standardisation.
The internal ONS review contains the critique that foreign nationals in the UK are not demographically representative of the British population. Young males are most likely to commit crime, and migrants are more likely to fall into this group. The internal ONS review says that all samples like this should be standardised for age and sex in order to accurately compare it with the entire British population, which includes the children, the elderly and women. This point applies to both reports.
- Issues with communication
Ultimately the CMC analysis is beset with a swathe of problems, outlined above.
Mr Buckley’s presentation also has some problems.
- Data attribution is misleading.
On Talkback – and in his reply to us – the MLA claims he was referring to Home Office figures. While there is some Home Office data in the CMC Report (such as the 12 per 1000 arrest rate for people from England and Wales figure cited above), the arrest by nationality data – ie. data about Somalis – comes from CMC’s FOI requests to individual police forces rather than the Home Office.
- His claim of “Somalian descent” goes further than the CMC analysis
The specific claim was about people of “Somalian descent”. CMC itself acknowledges that the arrest figures exclude migrants who have subsequently been awarded British citizenship, and second-generation migrants. “Somalian descent” would include these groups, and is a much wider claim.
- Other data
The reply from Mr Buckley’s office also referenced an article in the Journal of Scandinavian Studies in Criminology and Crime Prevention (Vol.15, Iss.2).
This report does not support Mr Buckley’s claim that “those of Somalian descent are more likely to commit some of these crimes in comparison with other nationalities who have come and settled in the United Kingdom”.
This study was published in 2014 and relies on analysis of crime figures related to immigrant populations to Norway (from crimes occurring in 2008 and 2009) and Finland (crime data from 2010 and 2011). Being over a decade old and not in relation to the British Isles – which are somewhat comparable, as they are other European nations – does place limitations on the usefulness of this report.
However, the key issue is that, while the paper does examine crime rates among a select number of immigrant population subgroups based on nationality and finds that there were differences in crime rates among these subgroups in the dataset – with Somalis at the upper end) – it does not conclude that people certain nationalities are fundamentally more likely to commit crimes than others.
The analysis put forward by the academics, when adjusted for demographics such as age and gender, states that nation of origin is “one important detail” linked to varying levels of violent crime but that “[explaining] such variations might require studying each group in greater depth.”
The analysis compared populations from 18 different nations, the former Yugoslavia (as one group), as well as various subcontinental groupings. It found that, adjusting for certain demographic factors, out of the 25 national/supranational groups assessed, Somalis were one of the groups most likely to commit violent crimes.
The researchers did not include any weighting for socioeconomic circumstances in their final analysis, but stated that this is obviously a factor worthy of consideration. They also said that they had not been able to properly address any possibility of racial bias in the justice system and its effect on the underlying figures and said this may be a factor in variable crime rates between individuals of different nationality – although they said this was unlikely to be a defining factor on its own.
The academics behind the study also cautioned against drawing firm or simplistic conclusions from their work, stating:
“[The] groups with the highest crime risks are those where the share of humanitarian immigration, comprising individuals who often have experienced wars or other traumatic crises, is the largest: Africa, Iraq, Afghanistan and former Yugoslavia share the same pattern…
“While searching for a theoretical explanation for immigrants’ higher level of crime is warranted, we suggest that one should not be tempted to propose one such theory as a comprehensive answer, and global emigration patterns should be taken into account in such explanations. Established criminological theories should be expected to apply to immigrants as well as to the majority populations, but there might be some factors that only apply to the immigrant population—or perhaps to only some of the immigrant groups.”
Overall, this study is an interesting – and, in contrast to the work of CMC, valid – piece of research but does not represent proof that Somalis are more likely to commit violent crimes than people from other countries, nor does it claim to represent such proof (and, of course, it’s worth repeating that immigrant populations from fewer than 20 individual nations were even subject to direct analysis in this work).
The central conclusion from this research is that nationality is a relevant factor that should be considered when analysing crime rates in a given population, not that individuals from specific countries are more prone to violence per se – and, specifically, the report does not conclude that 12 years after its publication, in another country, the Somali sub-population will be more prone to violence than other national groups. The report states:
“One might worry that a highly disaggregated analysis of crime amongst immigrants might lead to the groups with highest levels of crime being singled out and, thus, lead to increased labelling—even if the overwhelming majority of any group analysed here did not commit crimes. However, the alternative is to lump all immigrants into one group, disregarding variation within that group. In our opinion, if social problems accumulate amongst particular groups of immigrants, this issue needs to be addressed at that level. This holds also if a part of the phenomenon is racial bias in the justice system or discrimination in society in general. More nuanced descriptions promote more precise and nuanced explanations, which require disaggregation.”
Broadly speaking, these findings are similar to the comments from ONS and the Migration Observatory in that more details are needed before conclusions can be made.
What about some more recent figures that apply to the UK directly?
- Other research
By far the most comprehensive data on foreign nationals in the criminal justice system in England and Wales is the Ministry of Justice’s (MoJ) prison population statistics. They do not measure crime rates, but they do record the nationality, age, sex and offence category of people held in prison on a given date.
Using the MoJ prison figures, alongside their own FOI requests to MoJ, the Migration Observatory found that foreign nationals account for roughly the same share of convictions and imprisonment as their share of the adult population, and “if we control for age and sex, non-citizens are underrepresented in the prison population.”
However, the Migration Observatory researchers also stress that there are no reliable official statistics on crime by nationality and that comparisons between nationalities are often hampered by poor population data.
It is worth noting that the UK Government has itself acknowledged the data gap. In April 2025, Home Secretary Yvette Cooper instructed officials to publish data on criminal convictions of foreign nationals, and in August 2025 described existing data on crimes by ethnicity and immigration status as “patchy,” with changes underway.
Statistics on the nationalities and offences of deported foreign national offenders were eventually published by the Home Office in May 2026. But data on conviction rates across the broader foreign national population – which would be needed to evaluate claims like Buckley’s – has not been published.
- Conclusion
Available evidence on migration and crime in the UK is very limited.
Based on the evidence, we cannot conclude that people of Somalian descent are more likely to commit crimes than other migrants to the UK.
That said, the MLA’s claim drew on a report from the CMC, an organisation committed to controlling and reducing migration to Britain, which had been widely discussed in the media. During a live broadcast interview, he made some errors in communicating the data.
However, an internal ONS review sums up the situation well when it says that the CMC report material is “clearly of public interest ” but “where data is available, it’s partial or poor quality” and “could potentially mislead the public debate.”
Given that the attacker was Sudanese and not Somali at all, we think this is a good casestudy of how misunderstandings and poor information can spread during times of political crisis when emotions run high. A few hours later, more accurate information about the background of the suspected attacker would have negated the opportunity to share this inaccurate claim.
- A word on the CMC
The various analyses outlined above make clear how questionable the work of the Centre for Migration Control has been in its evaluation of any link between nationalities and crimes. But what is the CMC?
The CMC describes itself on its Substack as a “think tank committed to controlling and reducing migration to Britain.” Its founder and research director is Robert Bates, a former Reform UK activist, whose own comments are included in the report, “[m]ass migration has made us all less safe […] the Home Office has been too concerned with appeasing open-border lobbyists.”
Think tanks are research organisations that rely on expertise to analyse issues, generally across politics, economics or society. However, given the frankly substandard statistical analysis contained in reports that rely entirely on statistical assessment of data, it would be remiss to say that the CMC’s efforts display any actual expertise.
The Centre for Migration Control is better understood not as a think tank but as a blog. None of this necessarily invalidates the CMC’s findings – they are invalidated by all the material points above – but it is useful context.
