Are hate crimes against Jews being deliberately exaggerated?
JVL Introduction
When we see media headlines about hate crime being the ” highest ever” it’s natural to believe them.
They usually refer to police recorded offences – but the Home Office itself tells us these figures used alone are not reliable!
This is shown graphically by the fact that while police data indicate that hate crime had increased by over 300% since 2011, the Crime Survey for England and Wales (CSEW), the other key source of hate crime statistics, shows a decrease of around 30% over the same period.
The Home Office decided the increases in police reports were mainly related to increased victim reporting and improved police recording practices, causing an artificial inflation.
Based on these, a headline figure in 2023 suggesting that Jews are 5 times more likely than Muslims to be victims of a hate crime, increasing recently to 9 times, based on police religious hate crime figures, has been used as part of a coordinated campaign to show Jews as the most victimised minority in England and Wales (see image above).
Yet Crime Survey stats suggests this is the opposite of the truth, with Muslims experiencing twice the rate of religious hate crime and five times the rate of racial hate crime than Jews.
How do we explain the divergences?
Police recorded hate crime data has recognized limitations. They are exposed to a significant variation in victim reporting rates and police recording procedures, and they do not publish victim religion under the much larger strand of Race.
CSEW data relates to a direct victim survey and is therefore neither affected by variable victim reporting rates, nor by differences in police recording. It also gives victim religion under the strand of Race, when strand distinction is not always obvious, and where many more Muslims than Jews are found.
The Home Office’s advice always to present CSEW data alongside police statistics for a more reliable assessment has largely been ignored by the media, politicians and, worryingly, even governments. Is this sometimes deliberate? As a result, public perceptions and the allocation of government resources shaped by police statistics alone are likely to be misplaced.
In his analysis for JVL, Dr Alan Maddison teases out the reality which careful analysis of the available stats reveals, and mentions the political motivation which may lead to the distorted use of highly partial statistics.
PS: Updated 31 Aug
Are hate crimes against Jews being deliberately exaggerated?
Dr Alan Maddison, 29 August 2026
In his July 2026 review of ‘antisemitism and other forms of racism in the NHS and healthcare regulatory system’, Lord John Mann asserted:
Jewish people proportionately face the highest level of reported hate crime of any religious group.
In an open letter to the incoming prime minister Andy Burnham that same month, Gideon Falter of the Campaign Against Antisemitism (CAA) likewise alleged:
Jewish people are over nine times more likely to be victims of hate crime than any other religious group in the UK.
National media similarly reported:
Jewish people are almost ten times more likely than Muslims to be victims of religious hate crimes. (The Times, 30 April 2026)
Per head of population Jews experienced more than eight times as many hate crime incidents as Muslims. (BBC Panorama, 25 April 2026)
The suggestion that British Jews are not just threatened but uniquely so is not new. The Campaign Against Antisemitism advanced it back in 2019 and, in 2023, used it as the basis for a a billboard campaign across major British cities.
But is it true?
Why does the answer matter?
There at least two reasons.
First, to ensure that government resource allocation and action plans are correctly and equitably distributed across victim groups.
Second, the claim that British Jews are uniquely targeted has been deployed to delegitimise Palestinian rights advocacy and “significantly constrain” (Human Rights Watch) British citizens’ right to protest.
To evaluate the claim, we need to examine two key sources of hate crime statistics: police records and the victim-based Crime Survey of England and Wales (CSEW). We use the latest installments of each of these in the analysis below.[1]
Each of these sources has known limitations. The Office for Statistics Regulation (OSR) has accordingly advised that police and CSEW statistics be “used alongside each other” in order to provide a more “complete picture of hate crime”.
The police use five “motivation strands” for recording and classifying hate crimes: race or ethnicity, religion or belief, sexual orientation, disability, and transgender identity. It is the first two that concern us here.
The assertions of Lord Mann and others, cited at the head of this article, are based on a very limited subset of the available statistics. They are based solely on police-recorded hate crimes, ignoring data from the CSEW; and solely on crimes classified in police records under the strand of “religion”, ignoring those classified under the strand of “race”.
Police figures (see Table 1) show that nearly half of religious hate crimes involved Muslims, who were in absolute terms the most targeted group. Adjusting for population size, however, Jews experienced a much higher rate of religious hate crime (nearly 9x the rate of Muslims) than any other minority group—just as Lord Mann and others said.
However, this only applies to offences recorded by police under the motivation strand of “religion”. The number of offences recorded under the strand of “race” was around ten times larger, and these also involved Muslims, Jews, and the other above-mentioned victim groups.
Police statistics unfortunately do not disaggregate how many Muslims or Jews were victims of “racial” hate crimes. This is where the CSEW data provides a more complete picture, as it specifies victim religion and ethnicity in both “race” and “religion” strands for the three-year period ending in March 2025. (See Figure 1)
Under the strand of “religion,” Muslims reported the highest rate of hate crime victimization of all groups, both ethnic and religious—twice that for Jewish people. This relationship is completely at odds with that indicated by police data shown in Table 1, and clearly inverts the assertions by Mann and others.
Under the strand of “race”, Muslims also reported higher rates than any other religious or ethnic victim group—five times that of Jewish people.
When combining the strands of Religion and Race the hate crime rates for Muslims also reported the highest rates—three times that of Jewish people.
Lord Mann and others totally ignore the OSR’s authoritative guidance and repeatedly cite police-recorded figures in isolation, as if the confounding CSEW data did not exist. This is particularly irresponsible given that, unlike the CSEW data, police-recorded hate crime figures are not officially accredited by the OSR.[2]
Alarmist claims by Mann and others, based on cherry-picked data, should not influence either Government policy on hate crime or public opinion.
But why do police-recorded and CSEW numbers differ so radically?
CSEW v. police-recorded statistics
Reliable statistics for crime are notoriously difficult to compile and anyone in the business knows to treat them with caution.
CSEW data has the distinct advantage of covering hate crimes not reported to police, and so is not affected by variations in group reporting rates and police recording practices. However, the CSEW does not include victims who are under 16 or living in communal residences or hotels; general “public order hate crimes” that are not directed against an individual victim; or hate crimes against religious and other organisations. In addition, CSEW reports are only published every three years, in order to acquire three annual data sets for more reliable figures.
The annual police reports cover a broader victim population and all “public order hate crimes”, including those without a specific victim. Unfortunately, the police only record victim religion for those crimes classified under the “religion” strand, and currently only provide what they call indicative statistics on victim ethnicity over the combined strands of “race” and “religion” (indicative because the police only record self-described ethnicity in around 40 percent of cases).
The net effect of all this in the latest figures was that overall police figures for hate crime incidents under the combined motivation strands of “race” and “religion” in the three-year period April 2022–March 2025 were 18 percent lower than those from the CSEW estimates.[3]
How did CSEW and police hate-crime rates differ?
The CSEW data showed that Muslims experienced religious hate crimes at a rate of 80 per 10,000 (see Figure 1). As noted, police-recorded hate crimes under the strands of “race” and “religion” combined were 18 percent lower than CSEW results. We might therefore expect police records to show anti-Muslim religious hate crimes occurring at a rate of 66 per 10,000 (66 being an 18 percent reduction from 80). But in fact, the police recorded anti-Muslim religious hate crimes at a rate of only 10 per 10,000—less than one-sixth the rate expected based on CSEW data.
A similar discrepancy is visible with respect to religious hate crimes against Jews—except this time, it runs the other way. Based on CSEW data, one would expect police records to show anti-Jewish religious hate crimes at a rate of 33 per 10,000 people. But the actual police rate was 94 per 10,000—some three times the rate expected based on CSEW data. (See Table 2 below for a summary)
It should be noted that due to the small sample size for Jewish victims, the CSEW rates (shown in Figure 1 and Table 2) carry higher margins of error and should be treated with caution.
Why was police-recorded hate crime against Muslims so low?
In the latest report to the end of March 2025, the total size of the strand of “religion” in the police data was only a third of that found in the three-year CSEW report. In an analysis for the previous period to March 2020, we suggested that the most plausible explanation was that police staff, misguided by victim skin colour, had wrongly classified around three out of every four non-White victims of a religious hate crime under the motivation strand of “race” rather than “religion”, as illustrated in Figure 2 below.
Though there is a margin of error inherent in the CSEW estimations, our calculations for the latest three-year data set (to March 2025) on this “strand transfer” phenomenon suggest that police statistics under the strand of “religion” captured a very similar 1 out of every 4.4 religious hate crimes targeting Muslims and other non-White victims, including Hindus, Sikhs, and Buddhists. This proposed erroneous “strand transfer” explanation is supported by the indicative police ethnicity data.[4]
Another significant factor contributing to the low rates of recorded anti-Muslim hate crime could well be under-reporting by Muslim victims. A January 2026 survey by The Women and Equalities Committee, appointed by the House of Commons, indicated reporting rates as low as 12.5 percent among Muslim females. Together with the plausible strand-error phenomenon posited above, this under-reporting could largely explain why police rates for Muslims were less than a sixth of those expected based on adjusted CSEW rates.
Why was police-recorded hate crime against Jews so high?
The anomalously high rates of police-recorded religious hate crime against Jews may reflect the contribution to police statistics made by the Community Security Trust (CST). The CST is a charity that tracks and publishes antisemitism incidents annually, and has an agreement with police forces for the automatic mutual sharing of antisemitism hate crimes, holding a third-party reporting status with them.
The final annual Home Office published police rates for Jews including and excluding the CST contribution, are illustrated in Figure 3 below.
The CST contribution for the last three years appears to have increased the initial police-recorded rates by a factor of three. This alone could largely explain the divergence with CSEW rates. In part it could be related to CST’s increased victim-reporting capacity and also significant proactive efforts by its own staff to detect and report incidents.
About half of CST-recorded incidents were not reported by victims but instead by CST staff/volunteers and other witnesses. As well as being inflationary many such hate crimes, including those targeting synagogues and other Jewish communal venues, which are often guarded by CST volunteers, could therefore lack an individual victim and thus not appear in the CSEW data.
The CST also advises the Home Office and police on how to apply the contested IHRA working definition of antisemitism, and offers training on it. This definition has been criticised for excessively conflating legitimate criticism of or opposition to Israel with antisemitism (see here and here). The CST reports that around 80 percent of its recorded incidents over the last three years involved anti-Israel and/or anti-Zionist elements. Our published analyses (here and here) suggest that only a small minority of these are likely to have been antisemitic; unwarranted conflations by CST staff could have had a significant impact on the police rates, too.
Other third parties such as Tell MAMA (which monitors anti-Muslim incidents) do not appear to have had the same level of inflationary impact; the relevant police rates for those groups were never higher than those expected based on CSEW data.
Summary and implications
Unlike Lord Mann and others, we have followed the OSR’s advice and examined CSEW statistics, as well as data from other sources, alongside police records.
The conflict between rates from the two key sources can be explained by the recognised limitations of police-recorded statistics, in particular variations in victim reporting rates, perhaps inflated in this case by the CST, and errors in police staff strand allocation and recording. The CSEW data is not affected by such limitations.
From this analysis, the OSR-accredited CSEW data, despite its limitations and error margins, appears to provide a far more reliable picture of hate-crime rates. It shows that it is Muslim not Jewish people who face the highest per capita rates of hate crimes of any religious, or indeed ethnic, group in England and Wales.
The earlier anti-Jewish hate crime rate claims used alone are grossly misleading and this need to cease. Any related prioritisation of resource allocation and action plans influenced by them need urgent review.
But are hate crime rates really the best measure for any prioritisation assessment?
In Figure 4 we have used the recent CSEW data to estimate the average daily number of hate crime offences experienced by each victim group.
To derive these daily estimates we simply multiplied CSEW rates by total victim population sizes and divided by 365.
We saw earlier that the latest CSEW “race” and “religion” hate crimes against Muslims at three times the per capita rate experienced by Jews. If we revert from rates back to absolute offence numbers, the divergence is even starker: the number of racial and religious hate crimes against Muslims was forty-eight times higher than the number targeting Jews.
Given this data, any prioritisation of antisemitism in Government anti-racism strategies over and above racism directed against other victim groups would indicate that those strategies have little to do with fighting hate crime but are instead motivated by ulterior political agendas, for instance the effort to silence critics of Israel in the NHS, universities, on pro-Palestinian marches and elsewhere.
Acknowledgement
I would like to express my immense gratitude to Dr Jamie Stern-Weiner for his significant contribution to this article.
Footnotes
[1] Hate crime, England and Wales, years ending March 2023, 2024, 2025, and CSEW data, Section 3 and appendix tables 16 and 17, See e.g. Hate crime, England and Wales, year ending March 2025, published 9 October 2025
[2] Accredited means the CSEW statistics gathering and reporting were assessed by the OSR and officially recognised as meeting its high standards and follow their Code of Practice for statistics, previously known as National Statistics. Police-recorded crime statistics in England and Wales lost their National Statistics accreditation in January 2014. The UK Statistics Authority (now operating via the Office for Statistics Regulation) removed the status because evidence showed the underlying data was unreliable. The police forces/Home Office have made improvements but not yet regained ORS accreditation. Both sources are regularly assessed. The CSEW briefly lost its accreditation around the Covid shut-down because of the inability to perform face-to-face interviews. It was rapidly restored for the last two data sets.
[3] CSEW data showed 131,000 hate crimes under the combined strands of “race” and “religion”. Police-recorded statistics tallied 108,000 over the same period—18 percent lower.
[4] The police indicative ethnicity data , covering 40% of recorded hate crimes, suggested that Asian victims accounted for 31.5 percent of offences in the combined strands of “race” and “religion” over these three years. Based on the assumption that this 31.5% can be applied to the total of 108,000 police-recorded hate crime offences, it provides 34,000 offences against Asians. This would equate to 14,000 offences under the strand of “religion”. We estimate around 3,500 Asian victims appear to be included under the police strand of “religion” (Table 1), which amounts to one in four.
Great article and a very useful resource. The key point is that hate crimes against Jews are classified by the police as religious hate crimes, while those against Muslims are split between two categories, religious- and race-based hate crimes, with the great majority of cases in the latter.
Its doubtful if those attacking (presumed) Muslims verbally or physically separate out their motivations. But what is clear is that Jews per capita experience an order of magnitude less attacks.
The spin machine claiming otherwise is of course part of the campaign that also tries to undercut criticism of Israel by claiming it is motivated by antisemitism. Jews are to be understood as victims.
Well done JVL. Brilliant article.
Benedict Anderson (Imagined Communities) made the point regarding print capitalism. …media controls and shapes people not through direct force but by constructing their ‘shared’ sense of reality and political ‘identity’.
The power of print!
“Facts get their importance from what is made of them in interpretation… for interpretations depend very much on who the interpreter is, who he or she is addressing, what his or her purpose is.” Edward Said
Thank you for this informative review. There currently an effort across the media to present the idea of Jews under siege from antisemitic violence. The lack of concern for groups which suffer a higher per capita level of hate crimes reveals the political motivations behind the exaggerated reporting. This article will help us all to respond to that on the basis of the facts.
Hi Alan,
I was wondering whether you could provide an explanation of how you came to the 3/4 figure?
Best,
Joshua Alston
This a very welcome article but does seem to have to this topic a bit late. After the 2014 bombing of Gaza – 551 children killed along with 2000+ adults – I did a media representation case study of the story of a Jewish lad badly hurt in a gang in Manchester. Originally reported as ‘drug related’ by The JC, it was taken down and re-spun as an anti-Semitic attack across the media for many months despite the Police, DPP, and Judge saying otherwise. There were similar listed cases around this time. See Arena ‘Anti-Semitism Moral Panics’. https://arena.org.au/anti-semitism-moral-panics-by-gavin-lewis/
Frequently too by comparison Black victim experiences are frequently labelled merely ‘alleged racism’ – if at all. Whereas anti-Semitism allegations seems to require little corroboration. There is also the issue of like-for-like standards. BBC online recently reported “Jewish man ‘kicked like animal'”. Leaving aside the tabloid imagery of kicking animals as a British norm, actually he was an Israeli. Would anyone have reported the religion of a white South African or Rhodesian hurt in a revenge attack – and those colonial projects had armed revolutionary resistance movements.
Thank you for a very useful article, that put things in perspective.
Still there is antisemitism and probably increasing thanks to Israel’s best efforts.
Although, it is good to know that the scale is being exaggerated, for the individual Jew at the receiving end, the hurt and harm is very real.
In the same way that Islamophobia, or racism in every form, is very real to the individuals and groups that experience it.
Statistics are useful up to a point.
Btw it is also worth mentioning the comparative ease with which so-called anti-Semitism dossiers are presented as establishment credible and authentically newsworthy. From memory, it might have been Saul Staniforth that critiqued one of these as ‘padded’. But the real issue is that if you’re from an oppressed group – particularly of people of colour – you’re not likely to be going to the authorities, Police who shoot-to-kill, beat, disproportionately stop and imprison you, the Education authorities who strip-search your children, or the Health Managers who marginalise your complaints and treatment. Parts of the media are even suggesting Housing authorities should be privileging indigenous Britons? By comparison look at the massive advantage in access, networking, lobbying and uncritical acceptance of claimed anti-Semitism grievances of Israel’s lobby. And significantly these advantages are hardly ever equivalently enjoyed even by non-Zionist or anti-Zionist Jews.
This is important – the exaggeration of victimhood is deliberate. From my own experience I have rarely witnessed any antisemitism but frequently here condemnations of ‘Muslims’.
I think the headline is unfortunate because it makes this careful analysis seem biased and denialist. It’s therefore hard for me to use the findings, even though I believe them entirely and can see the facts on the ground every day where I live.
The following query was posted on a thread elsewhere on the web:
“Do you know where they got the figure that 3/4 non-white victims of religious hate crimes are miscategorised under race? There isn’t really a basis for this explained or a linked article. ”
Alan Maddison replies:
Under Fig 4 of an earlier article based on the previous CSEW data set, I show Muslim hate crimes under the strand of Religion as a share of the combined religious and racial hate crimes as 3.5% in police reports and while the CSEW estimate was 14%. This suggested an undercount of 1 in 4 (3.5/14) for Muslims under this strand. The most recent CSEW data set period to March 2025 gives a similar ratio between police and CSEW shares.
But I then took another additional approach.
The Police strand of Religious hate crime totaled 10,065 offences, and the CSEW reported 38,000 offences which adjusted (-18%) to an expected 31,000 offences for the police strand, just over three times higher than the actual figure of 10,065. So who were these missing victims?
We analysed published CST antisemitism incidents over this period. These cover both strands, and include non-criminal incidents. The CST numbers were very close to published police religious hate crime figures for Jews. This suggested that almost all the hate crime offences involving Jews had been registered by police under the strand of Religion. This made sense as 98% of British Jews are White, and police error based on skin colour would be minimal, so no under-reporting here.
So increasing the police strand of Religion three-fold to align with CSEW estimates, would be largely down to a higher than three-fold increases from non-White victims. Based on our estimates of their number this worked out to represent a 4.4 fold increase.
The indicative police ethnicity data tend to support this hypothesis (see footnote 4).
I hope this answers your query.