Check your content against UK speech law frameworks

Home / Blog / The Threatening Words Threshold: Why Abuse Alone Is Not Enough to Prosecute

← Back to Blog

The Threatening Words Threshold: Why Abuse Alone Is Not Enough to Prosecute

Published 9 July 2026

Under the Public Order Act 1986, you can be prosecuted for inciting racial hatred using words that are "threatening, abusive or insulting." Three categories. But under the Racial and Religious Hatred Act 2006, you can only be prosecuted for inciting religious hatred using words that are "threatening". One category. Abusive and insulting — however extreme, however hurtful — are not enough. The same model applies to sexual orientation hatred under the Criminal Justice and Immigration Act 2008.

That gap — between "abusive" and "threatening" — is not a drafting error. It is the most important free-speech protection in UK hate speech law. And it exists because Parliament could not agree on whether religious beliefs and sexual identity should receive the same protection as racial identity. The compromise created the threshold that defines which cases are prosecuted and which are not.

Disclaimer: General information, not legal advice. If you are under investigation for a hate speech offence, speak to a solicitor.

Jurisdiction: England and Wales. Scotland and Northern Ireland have separate legislation.

The Threshold Spectrum: Where Each Offence Draws the Line INSULTING ABUSIVE THREATENING RACIAL HATRED (POA 1986 Part III): Threatening + Abusive + Insulting — ALL THREE qualify RELIGIOUS HATRED: Threatening ONLY Abusive and insulting = NOT ENOUGH for religious hatred SEXUAL ORIENTATION: Threatening ONLY Same model as religious hatred — CJIA 2008 The gap between "abusive" and "threatening" is where free speech lives — and it was put there deliberately Also relevant: s.127 Comms Act (grossly offensive, 6mo) — can catch abusive/insulting speech even when hatred provisions cannot Sources: POA 1986 Part III; RRHA 2006; CJIA 2008 Part 3B
The threshold spectrum: racial hatred covers all three categories. Religious and sexual orientation hatred cover only "threatening."

Why the Gap Exists

The story of how religious hatred ended up with a higher threshold than racial hatred is a story about free speech, parliamentary arithmetic, and a fundamental disagreement. In 2005, the Labour government introduced a Bill to give religious groups the same protection as racial groups — "threatening, abusive or insulting" words. The House of Lords rebelled. A coalition of peers — led by former Lord Chief Justice Lord Mackay of Clashfern, and supported by Rowan Atkinson, comedians, writers, and free-speech advocates — argued that religious beliefs, unlike racial identity, are chosen and must be open to robust criticism, satire, and debate.

The Lords amended the Bill to restrict the offence to "threatening" words only. They added section 29J — an explicit free-expression proviso protecting "discussion, criticism or expressions of antipathy, dislike, ridicule, insult or abuse" of religions. The government fought the amendment. The Lords insisted. After a protracted standoff, the government conceded. The Racial and Religious Hatred Act 2006 passed with the higher threshold intact.

When Parliament created the sexual orientation hatred offence in 2008, it followed the same model: threatening only, intent required, free-expression proviso. The gap that the Lords created for religion became the template for sexual orientation. The asymmetry in UK hate speech law — racial hatred has the lowest threshold, religious and sexual orientation hatred have higher ones — is the direct result.

What "Threatening" Actually Means

The statute does not define "threatening" — courts have. A statement is threatening if it conveys a threat of harm — not merely that it is aggressive, hostile, or intimidating. The distinction matters. "Your religion is a lie" is abusive but not threatening. "People like you should not exist" is insulting but, arguably, not threatening. "I will burn down your place of worship" is threatening. The line between the second and third statements is where the criminal law draws the boundary.

In practice, this means the CPS must assess whether the words, in context, would cause a reasonable person to fear harm — not just offence, not just anger, not just deep upset. Fear of harm. That is a meaningfully higher bar than many people assume, and it is why prosecutions for religious and sexual orientation hatred are extremely rare compared to prosecutions for racial hatred.

What Qualifies — And What Does Not THREATENING "I will burn your mosque" "We know where you live" All three hatred offences ABUSIVE "Your religion is a lie" Hostile, aggressive language Racial hatred ONLY INSULTING "People like you are disgusting" Scornful, disrespectful language Racial hatred ONLY Important: abusive or insulting speech about religion/sexual orientation may still engage s.127 (6mo) or MCA (2yr) The hatred provisions are not the only laws that apply — just the most serious (7 years, DPP/AG consent) The free-expression provisos (s.29J RRHA; CJIA equivalent) explicitly protect criticism, ridicule, and insult of religion/sexual conduct
Only "threatening" words qualify under all three hatred offences. "Abusive" and "insulting" qualify only for racial hatred.

But Abuse May Still Be Criminal — Under Different Laws

The hatred provisions are not the only statutes that apply to abusive speech about religion or sexual orientation. Two other laws have lower thresholds — and no free-expression proviso like s.29J.

Communications Act 2003, s.127: Grossly offensive or menacing messages. The Collins test (outrage, shock, or disgust) applies. Awareness is enough — no intent required. A message that is abusive or insulting about religion or sexual orientation — but not threatening — could still be prosecuted under s.127, even though the hatred provisions do not apply. Maximum: 6 months.

Malicious Communications Act 1988: Messages sent with purpose to cause distress or anxiety. If the prosecution can prove the sender intended to cause distress, the MCA may apply — regardless of whether the religious or sexual orientation hatred provisions are engaged. Maximum: 2 years on indictment.

The practical effect: s.29J protects you from a religious hatred charge, but not from a s.127 or MCA charge. The free-expression proviso is real — but it is specific to the hatred provisions. A grossly offensive campaign of abuse targeting a specific religious individual could still result in prosecution, even if the hatred threshold is not met.

Human Rights Act 1998, Article 10

Article 10 of the European Convention on Human Rights protects freedom of expression — including speech that "offends, shocks or disturbs" (Handyside v UK [1976]). Every restriction on speech must be prescribed by law, in pursuit of a legitimate aim, and proportionate. Article 10 reinforces the "threatening only" threshold: restricting non-threatening speech about religion or sexual orientation would be disproportionate, and the existence of the free-expression provisos in the hatred provisions would weigh heavily against any such restriction.

Practical Takeaways 1. "Threatening" = fear of harm Not just offensive or aggressive 2. The gap is deliberate Parliament chose this asymmetry 3. s.127 and MCA still apply Free-speech proviso is specific to hatred The gap between abusive and threatening is where free speech lives. It was fought for — and it was defended. UK Content Comply analyses content against 13 UK speech law frameworks. Sources: POA 1986; RRHA 2006; CJIA 2008; HRA 1998 Art.10; CPS Hate Crime Guidance
The gap between "abusive" and "threatening" is deliberate — and other statutes still apply.

Sources