What proportion of social-media-post arrests in the UK are for crimes like harassment, hate speech, or terrorism?
Executive summary
Available reporting shows UK police made roughly 12,000 arrests a year for “offensive” online communications around 2023, equivalent to more than 30 arrests per day, but those figures cover a broad mix of communications offences rather than being broken down by motive or specific offence types such as harassment, hate speech or terrorism [1] [2] [3]. Detailed public data that attributes arrests specifically to harassment, hate speech, or terrorism is not present in the cited sources; several commentators and parliamentary briefings warn that many arrests under communications laws do not lead to prosecution or imprisonment [4] [5].
1. What the headline numbers actually are — and where they come from
The most-cited headline is “more than 30 arrests a day” and “over 12,000 arrests in 2023” for offensive online messages; that figure was reported by The Times and repeated in parliamentary material and advocacy coverage [1] [6] [3]. Aggregators and commentators—AllSides, House of Lords Library and civil-liberties groups—use the same custody/court data to reach an average of about 12,000 arrests a year for social-media-related communications offences between 2021–2023 [2] [4].
2. What those arrests cover — broad legal categories, not tidy labels
The arrests cited are under communications offences such as Section 1 of the Malicious Communications Act 1988 and Section 127 of the Communications Act 2003, which criminalise sending indecent, grossly offensive or otherwise harmful messages [4]. Those statutory descriptions are wide and can encompass harassment, racist or religiously aggravated abuse, threats, incitement and other conduct; the sources make clear the categories are legal and broad rather than precise behavioral labels like “hate speech” versus “harassment” [4].
3. Missing: a reliable breakdown by offence type (harassment, hate, terrorism)
None of the provided sources offers a clear proportion or table that separates the 12,000-ish arrests into harassment, hate speech, terrorism or other offence buckets. House of Lords research and FOI replies note forces record and report differently and that extracting “social media” arrests requires manual case review, undermining any clean national breakdown in the available reporting [4] [7]. In short: available sources do not mention a definitive proportion by those crime types [7] [4].
4. What happens after arrest — few converts to prison sentences
Reporting and fact-check summaries point out that while arrests are numerous, prosecutions and custodial sentences are much rarer. MythDetector and other outlets highlight that most arrests do not end in imprisonment; for example, only a small fraction of cases under the relevant provisions led to immediate imprisonment in recent years [5]. The House of Lords library explains that common reasons for low conviction rates include evidential difficulties and victims not supporting further action [4].
5. Geographic and force-level variation — the Metropolitan and others lead numerically
The Times-derived data show the Metropolitan Police accounted for the largest number of arrests in the dataset , followed by West Yorkshire and Thames Valley; adjusted-per-capita rates differ, with some smaller forces showing higher rates per 100,000 people [3] [4]. Reporting also notes gaps in responses: some forces did not provide full FOI data, so national totals in media stories may under- or overstate complete coverage [8].
6. Competing perspectives and possible agendas in the reporting
Civil liberties groups and some parliamentarians frame the numbers as evidence of over-policing of speech and a free-speech emergency [6] [3]. Proponents of strict enforcement point to real harms from incitement, threats or calls to violence—cases the sources reference as legitimate reasons to use communications statutes [5]. Media outlets and advocacy organizations repeating the “30 a day” figure have different editorial stances; The Times supplied the custody data, while advocacy outlets amplify the civil-liberties angle [1] [3].
7. What this means for your question and where to look next
If you need the precise proportion of arrests for harassment vs hate speech vs terrorism, the current sources do not provide that breakdown; retrieving it would require either force-level FOIs with offence-type filters or CPS/Home Office datasets that map arrest codes to specific charges [7] [4]. For policy context, follow House of Lords Library work and FOI disclosures from individual forces—the sources above identify those as the places that most reliably illuminate methodology and force variation [4] [7].
Limitations and transparency: this analysis uses only the supplied reporting; the sources converge on headline arrest totals and on the lack of a neat public breakdown by specific offence types [1] [4] [7].