Official: Hate Crime prosecutions down, convictions down, since Brexit vote

An irresponsible, largely media-induced panic.


Today (17 Oct 2017), the Crown Prosecution Service released its official annual Hate Crime report for 2016-2017 (along with associated data tables).  And it is breathtaking - because it almost single-handedly refutes the idea that the Brexit vote caused a substantially large increase in actual hate crimes being committed.


I've already written two major posts on Hate Crime, please see these for the evolution of my thinking:

  1. Was there an explosion of "hate crime" after the Brexit referendum? (April 2017). Discussion of all of the 'perception" based definitions of Hate Crime, why these are problematic and 6 reasons to be skeptical of the media claims. But, it concludes it was likely/probable there was a genuine increase in Hate Crime, just smaller than claimed.
  2. Hate Crime and Brexit: Exclusive Update (Oct 2017). This used a rare analysis method to show that police recorded racial/religiously aggravated Hate Crime offences rose faster two years prior to the referendum, and barely more than equivalent non-aggravated offences after it.


Racial/Religious hate crime by the five specific racial or religiously aggravated offence types recorded by the police.  From Office of National Statistics Hate Crime data (Source)

Update 17 Oct 13:15: Added wording on "Offences" to make clear the distinctions between total hate crime and specific offences, plus another version that shows that more clearly.

All Racial/Religious hate crime - specific aggravated offence types plus any other crime "flagged" as as a racial and/or religious Hate Crime - recorded by the police.  From Home Office Hate Crime data (Source). 

The trend in falling prosecutions/convictions is the same, if anything referrals to the Police fall slightly faster.


Brexit isn't even mentioned in the CPS's 15 page annual report, or the 28 pages of supporting data. Wow. This is in contrast to the focus on Brexit immediately afterwards - the Director of the CPS (Alison Saunders) had only 6 meetings with outside groups after the referendum - with 50% of these dedicated to Hate Crime. Hate Crime is around 1.3% of all crimes and 2.5% of the CPS's caseload. This indicates reacting to media hysteria (see end of post), rather than rational allocation of workload and resources to spend half the directors external meeting time on this.

Race/religious Police Referrals to the CPS are down from 25.8%/19.9% to 21.9%/15.6% (depending on definition used; as a percentage of Police recorded Hate Crime). This is when the Police pass the case onto the CPS for a decision whether to charge a suspect. This is the fourth straight year "Police recorded Hate Crime" has gone up, while referrals have plummeted. The most plausible explanation is that "recorded Hate Crime" is less and less reflective of actual, genuine Hate Crimes that could end up in court (as opposed to those that exist in the "perception" of a victim, bystander or even a Police officer). The only other alternative explanation is the Police are recording genuine Hate Crimes but not referring them to the CPS - i.e. not trying to take the crime to court. That would be extraordinary and require extraordinary evidence, especially considering prosecutions/convictions are very stable, and divergence had been increasing for four years.

Race/Religious Prosecutions are down 7.89%. Imagine the pressure the CPS would have been under post referendum to show they were "cracking down on Hate Crime", yet prosecutions are substantially down. The most plausible explanation is they did not have sufficient evidence to proceed, even with the cases the police did refer.

Race/Religious Convictions are down 7.87% (as a percentage of Police Recorded Crime). The conviction rate itself was unchanged at 83.8%.

Sentence uplifts were increased from 34.8% to 55.8% - uplifts are applied when "hostility" is proven in court. This indicates cases that actually reached court were more likely to be genuine Hate Crimes.

Here is how the CPS present this data in their data tables. Spot the non-existent Brexit "surge". Note that there were zero racial/religious murders, less robbery, less criminal damage and half as many public order offences. The only main increase was "offences against the person" (and the vast majority of these are non-injury, i.e. verbal, still nasty but "low-end' as the Police refer to them).

Total Hate Crime Prosecutions and Convictions. Source: CPS, Oct 2017.

Total Hate Crime (Racial/Religious) Prosecutions and Convictions. Source: CPS, Oct 2017.

Hate Crime by "strand" / offence category. Source: CPS, Oct 2017.

*Racial/Religious hate crime is the type that would target nationality and ethnicity, rather than age or sexuality, so would be the most sensitive to intolerance resulting from or caused by the Brexit vote. This reasoning is also used by the Police and Home Office statisticians. It is also quite complex to compare Police Recorded Crime with the CPS data as they operate slightly differently when it comes to victims, and there is a lag between being recorded by the Police and being prosecuted through the courts. Unless stated otherwise, "Hate Crime" in this post refers to the racial/religious type. See previous posts for full disclaimers and detail on these points.

Is even Police Recorded Hate Crime rising that fast post-Brexit referendum?

Even taking into account that "perception" based Police recorded Crime doesn't end up in a court or result in a conviction 75%-80% of the time, it isn't rising fast by historical standards either:

  • Racial/Religious Police Recorded Hate Crime increased from 41,554 to 48,917 (+17.72%) - this includes 9 months of post-referendum
  • Police Recorded "Normal Crime" (exact non-aggravated equivalents) also increased from 1,551,174 to 1,798,493 (+15.94%), a very similar amount
  • Hate Crime also rose faster two years prior to Brexit (+19.25%)
  • Normal Crime also rose the fastest in 5 years the year prior to the referendum (+21.11%)

This is shown in these 3 graphs (click to enlarge):

See Hate Crime and Brexit: Exclusive Update for more detail. Note: the above charts only count the specifically categorised racial/religious offences listed, because this is currently the only way of comparing like-for-like offences reliably. Neither the ONS or the Police publish public granular data on all crimes "flagged" as Hate Crimes, only the 5 specific offences listed. The Home Office take the same approach: "increase in hate crime can be seen by using racially or religiously aggravated offence data...the police can record offences as being racially or religiously aggravated. While not covering all hate crime offences, these offences make up over 70 per cent of race and religion hate crime. Therefore, racially or religiously aggravated offences provide a good proxy for race and religious hate crimes" (Home Office 2016-2017 Hate Crime report, Page 5).  

What the (mostly Remain-supporting) media said post-Brexit referendum: all hysterics, no context

Throughout 2016 and 2017 the media unanimously used Police Recorded Hate Crime to show "surges", or "spikes", almost never referring to actual prosecutions or convictions. They also almost never mentioned that these police definitions require no victim, no evidence and no crime - literally. It is still the customary law of the land to be innocent until proven guilty. That didn't stop half the nation being tried in guilt by association, despite even the Police sometimes warning that any rise could not be attributed to Brexit

Here is a sample of some of the main stories:

There were however some honourable mentions: Joanna Williams, education editor at Spiked magazine: “the increase in hate crime might not mean Britain is more hateful after all. In fact, it could easily mean the exact opposite”, Brendan O’Neill (Spiked’s Editor): a “cynical, politically motivated crime panic” and author and commentator Douglas Murray: a “strange claim” and it is more likely that “the official encouragement to report hate-crimes means that more and more hate-crimes are being reported

Okay so from the CPS we now have it definitively - none - zero - of the so-called "huge" increases in police recorded Hate Crime post referendum has ended up in court, In fact, less Hate Crimes are either ending up in court, or being convicted at all, since Brexit, despite enormous public pressure and incentive to do so.....

.....So, what do the media say now, after the spike/surge/explosion has no basis in additional court cases or convictions?

The Independent: "Hate-crime prosecutions fall despite spike in reported attacks after Brexit referendum" ("campaigners called the annual statistics 'unacceptable'").

Yep, they really used "#despite"! Remember - the CPS didn't even mention Brexit in their report!


Racial/Religious Hate Crime Prosecutions are down, Convictions are down, and Referrals from the Police to the CPS are down.

All three are significantly diverging from "perception" based Police recorded Hate Crime. This suggests revisiting these definitions, and returning to something that aligns to that which is used in court. It also suggests revisiting the idea that Hate Crime is 'under-reported" (see below note on this for more). Or, for some unexplained reason, every Police force in England and Wales is recording Hate Crimes, but referring fewer and fewer of these to the CPS.

Much of the media owe millions of people an apology for hyping Hate Crime incessantly over the last 12+ months

Trying to paint the UK as a xenophobic hell on Earth, when in truth the UK remains one of the most tolerant countries in Europe and the World is pretty outrageous.

The media should be much more careful about using Police recorded Hate Crime out of context from CPS data in the future

Or at the very least comparing it to the non-aggravated equivalent versions of these same crimes. I won't hold my breath.

Note: Is Hate Crime actually "under-reported"?

The Director of Public Prosecutions certainly thinks so: "We know hate crime is underreported". Okay, so how does she know? It turns out the evidence base for this statement is pretty flimsy. It traces back to one place: the Crime Survey of England and Wales (CSEW; a census/interview/poll of a representative 35,000 people every two years). This says that there "should" be around 222,000 Hate Crimes recorded by the Police per year, yet there were "only" 62,000 (e.g. in 2015-2016). So Hate Crime is happening, but not being reported or recorded by the police. However, the CSEW "does not ask about Hate Crime directly" as the concept is "not well understood by the public" (according the Home Office) - instead it asks respondents to "speculate" on what the motives of the suspect may be, "which may lead to over-reporting". Even on this broad, vague basis, Hate Crime comes out at under <3% of all crime in the CSEW, making it "[in]sufficiently robust" to even report on an annual basis (5% is the CESW's normal margin of error). Seriously - this is the entire chain of evidence for the statement that 'we know Hate Crime is underreported".  

The BBC Home Affairs correspondent also queries reliance on the CSEW in general: "it's increasingly clear that the [Crime Survey of England and Wales; CSEW), with a relatively limited sample, is not good at gauging emerging crime trends … For now, the police statistics are a more useful guide as to what's really going on". And nobody anywhere else on the entire planet records Hate Crime figures anywhere near what the UK does - e.g. according to the OSCE (which collates figures across the continent), the UK had 9 times the Hate Crime as France, Germany, Italy, Poland, and Spain combined

This determination of "under-reporting" is the reason for "Hate Crime Awareness week(s)", hundreds of stalls/workshops all over the country, millions of leaflets being printed, dedicated SMS numbers, thousands of hours of Police "sensitivity" training, and so on. This has a cost in time/effort/resources for the police and the CPS that if being misallocated, could have real negative effects for other victims of the other 99% of crimes. The reality that prosecutions and convictions for Hate Crime stubbornly refuse to rise is a strong reason to revisit the hypothesis that Hate Crime is "under-reported".