The Elephant in the Room

Let's Talk About Hate Speech

The podcast “The Elephant” is an online video series that tackles the conversations New Zealanders often avoid.

It dives into big, uncomfortable questions, looking beyond the echo chambers in search of a fearless and honest debate. In episode 10, hosts recently released, Miriama Kamo and Mark Crysell ask ‘When does free speech become hate speech?’

The promotional material for the programme states:

New Zealand’s long-running fight over where free speech ends and hate speech begins has reignited, after the Law Commission recommended stronger legal protections for transgender, non-binary and intersex people.

The Ia Tangata report, which means “each and every person”, calls for gender identity and expression to be written into the Human Rights Act.

While the report itself doesn’t review hate provisions in the law, it has revived debate over how New Zealand regulates speech, discrimination, and protections for minority communities.

Participants in the programme were Awa Puna described as a film-maker and Shortland Street’s first trans actor – or should that be actress.

Puna’s contribution to the debate was more about the abuse that she personally suffered –

Slurs shouted from passing cars, hostility at school and online harassment. “Someone rolls down their window and says, ‘get away from that thing – it’s not a woman’. When someone is facing that kind of hate all the time, it affects you.”

Is that a manifestation of hate or of ill-mannered abuse?

The fear intensified at Auckland’s Posie Parker protest. “I was spat on … my hair was pulled … it really hurt,” she says. She believes stronger protections would have changed her life. “When I think about my younger self, it won’t be the same for kids if those things are in place.”

Other participants were Nathan Seuli, formerly of the Free Speech Union but now of the recently formed group Pillar - Protecting Individual Life, Liberty, And Rights.

Seuli’s position was in favour of free speech and as much of it as possible. He argued that the current law, which bans incitement to violence, defamation and libel, draws the line where it should.

“Those things are objective standards of legally restricted speech. Everything else is fair play.”

He rejects expanding the law to protect groups. “Groups don’t have rights … individuals have rights. The law protects individuals. We don’t need to create caveats that protect a group.”

Seiuli warns that regulating misgendering or dead-naming (calling a transgender person by the name they used before they transitioned instead of the name they use now) could infringe on freedom of conscience and religion. “That would restrict people’s beliefs,” he says.

On the other side of the argument was Paul Thistoll of Rights Aotearoa who brought a more relativistic approach to free speech.

“Hate speech is corrosive to the dignity of both individuals and groups,” he says. “It says ‘you don’t belong here. Your place in this society is worth less than others’.”

Thistoll dismisses the idea that hate is too subjective to legislate against. “Courts already interpret hate as an aggravating factor in crime,” he says. “The law (The Human Rights Act) hasn’t been updated since 1993. It needs to be.”

The Government has not confirmed whether it will adopt the recommendations, but the debate is unlikely to go away regardless.

Thistoll is in error when he says that hate is an aggravating feature in crime. Section 9(h) of the Sentencing Act when listing aggravating features that may be taken into account at sentencing reads as follows:

h) that the offender committed the offence partly or wholly because of hostility towards a group of persons who have an enduring common characteristic such as race, colour, nationality, religion, gender identity, sexual orientation, age, or disability; and

(i) the hostility is because of the common characteristic; and

(ii) the offender believed that the victim has that characteristic:

At no time does the word “hate” appear. The word is “hostility” as I have emphasised. And this probably is helpful in considering much of the debate between Seuli and Thistoll.

Seuli argued that “hate speech” is very much a subjective concept. Hatred itself – an emotion – is subjective. Thistoll argued to the contrary – that it is speech that is corrosive and suggests that a group does not belong.

An academic and measured perspective was brought to the debate by Professor Simon Kellar from Victoria University who focussed more on the quality of the speech and that there had to be very good grounds to limit the freedom of expression.

But there was still a very large elephant in the room. It was this. The term “hate speech” was not defined. At one stage Mark Crysell put to Nathan Seuli how would he feel if someone said something hateful to him.

I assume Crysell was being a devils advocate because he also at one stage advanced the straw man argument of Brenton Tarrant as an example of where hate speech could lead. Or I hope he was being a devil’s advocate.

Thistoll didn’t define hate speech but stated that a number of European countries had hate speech laws and the one he preferred was that of Germany.

Germany’s hate speech laws prohibit incitement to hatred, Holocaust denial, glorification of Nazism, public insults, and violent threats, with punishments including heavy fines and, for repeat offenders, jail time.

These laws exist alongside constitutional protections for free speech, which do not extend to speech that incites violence, degrades human dignity, or denies historical atrocities.

Online, platforms are required to swiftly remove illegal content and report it to authorities, while individuals can report offending material.

Incitement to hatred is defined in the following way:

“Whosoever, in a manner capable of disturbing the public peace:

1. incites hatred against a national, racial, religious group or a group defined by their ethnic origins, against segments of the population or individuals because of their belonging to one of the aforementioned groups or segments of the population or calls for violent or arbitrary measures against them; or

2. assaults the human dignity of others by insulting, maliciously maligning an aforementioned group, segments of the population or individuals because of their belonging to one of the aforementioned groups or segments of the population, or defaming segments of the population,

shall be liable to imprisonment from three months to five years”

Sadly, Thistoll did not cite chapter and verse, so I assume this is the section of the German Criminal Code to which he is referring.

I found Thistoll’s approach interesting. Although he criticised Seuli for what he suggested was an absolutist approach to free speech (Seuli doesn’t hold that position) Thistoll in his own way is as dogmatic and absolutist himself.

Dane Giraud wrote a Substack post on the programme – The Compassion Racket – which was critical of the approach and in which he (uncharitably) described Thistoll as –

“ serial litigant, caller of venues demanding events be shut down, and, it would be fair to say, staunch anti-feminist”

I thought it more profitable to examine some of Thistoll’s public work where there are some clear thought lines in his Spinoff columns, Rights Aotearoa essays and media comments – especially on gender identity, hate speech, digital regulation and free expression.

Thistoll’s main themes appear to number five and I discuss each as follows

1. International human rights as trump card over domestic debate

Thistoll consistently treats UN human-rights “soft law” (General Comments, treaty-body interpretations, Rabat Plan, etc) as setting a binding minimum floor that New Zealand must meet, and portrays current law as clearly non-compliant.

For example, in arguing to add “gender identity or expression” to the Human Rights Act, he relies heavily on UN committees’ view that “sex” discrimination implicitly includes gender identity and says Parliament should simply “make the implicit explicit”.

In his hate-speech framework he again begins from ICCPR art 20(2) and General Comment 34, claiming NZ is “far below” the required standard and must expand protected categories and sanctions.

This approach is flawed. UN committee interpretations are influential but not binding in the way he implies. They are one interpretive aid among several, not a constitutional override of Parliament or the courts. Treating them as a hard “floor” sidelines NZ’s own democratic choices and Bill of Rights balancing.

In addition Thistoll adopts the most maximalist UN interpretations (e.g. that “sex” already includes gender identity, and that states have a “positive duty” to criminalise speech far beyond race-based incitement) as if they were uncontested. In reality they’re politically and jurisprudentially controversial, including in other liberal democracies.

Thistoll makes a shift from “UN body thinks X is desirable” to “the state is legally obliged to legislate X” but never properly argues the shift. That elides the distinction between international aspiration and domestic legal duty.

2. Law as identity-affirmation and feelings-protection

According to Thistoll, law is repeatedly framed as primarily about signalling who belongs and affirming the “dignity” of particular groups – especially transgender, non-binary and intersex people.

He describes explicit gender-identity protections and Ia Tangata’s proposals as modest, technical steps that “do not alter the rights or freedoms of any other group”, but are crucial because “law does more than regulate; it signals who

Hate speech and digital-harm laws are justified largely in terms of protecting people from speech that tells them they “don’t belong here” or that their place is “worth less than others”.

This approach has problems.

Subjective hurt becomes justiciable harm – “Feeling excluded” or “feeling that the law doesn’t see me” are morally important, but are not stable legal tests. If “dignity” and “belonging” are defined by those who feel aggrieved, any disagreement about identity, sex or gender can be cast as a rights violation.

This position also denies real conflicts of rights. Thistoll insists changes “don’t alter the rights or freedoms of any other group”, yet Ia Tangata itself acknowledges implications for employment, education, accommodation, single-sex services and sport that must be worked through.

Those are real conflicts between sex-based protections and gender-identity claims; brushing them off as cost-free is simply inaccurate.

Thistoll also values group-based protection vs individual equality – The more the law becomes a tool to affirm particular categories (trans, non-binary, etc.) rather than to protect all individuals equally, the easier it becomes to justify differential treatment – including restrictions on others’ speech – in the name of one group’s “dignity”.

3. Expansive hate-speech and digital-harm regimes

This theme was the main thrust of Thistoll’s presentation and argument on “The Elephant”. But what he said on that programme represented only the main points of his apparoach.

He argues NZ’s hate-speech law is “abandonment of communities”, because s 131 Human Rights Act has a narrow ground list and high threshold; he wants a hybrid civil–criminal system with broad protected characteristics (including gender identity and expression) and an Australian-/Canadian-style model.

He champions the Harmful Digital Communications Act (HDCA) as a necessary “digital age” tool and treats the Ms Z prosecution as pure corruption of a good law, not evidence of structural problems.

More generally, he defines hate speech and harmful communications in sociological terms – as structural attacks on marginalised groups – and is openly critical of what he calls “free speech absolutism”.

These arguments have the following problems:

The first is the pre-emptive framing of opponents as bad-faith – By construing existing law as “abandonment” and repeal proposals as “privileging perpetrators over victims”, dissent is framed as morally suspect from the outset rather than as a legitimate concern about overreach.

Secondly he underplays the chilling effect of his approach.

Thistoll stresses high thresholds (e.g. “wilful promotion of hatred”) and Rabat Plan safeguards, but real-world experience in comparable jurisdictions shows that even narrow hate-speech laws can expand via prosecutorial practice and civil processes, encouraging self-censorship in controversial areas such as religion, gender ideology and immigration. His framework barely engages with this risk.

His HDCA example cuts both ways – The Ms Z case does show institutional corruption – but it also shows how a speech-based offence with an inherently subjective standard (“serious emotional distress”) gives officials a convenient lever to turn complaints into prosecutions.

Thistoll treats the law as entirely innocent and only the actors as culpable. In fact, poorly bounded powers plus bad actors is exactly what civil-liberties critics worry about, and the HDCA’s design is part of that problem.

A further problem is that Thistoll fails to understand that the HDCA provides remedies for individuals and is anot available for groups.

Finally, alternative tools are dismissed in that Thistoll asserts that older frameworks (defamation, harassment, privacy) are “demonstrably” inadequate, yet many jurisdictions deal with online abuse largely with those tools plus targeted updates (e.g. specific image-based abuse offences) without HDCA-style speech policing.

4. Algorithmic determinism & “death of the marketplace of ideas”

In his “Death of the Marketplace” essay Thistoll claims the free-speech “marketplace of ideas” metaphor never really worked and is now actively harmful in an algorithmic age.

He relies on the Guillaume Chaslot / YouTube radicalisation story to argue platform algorithms systematically push users toward extremism for profit, and that our information environment must be reconceived as a “discursive ecosystem” requiring active stewardship and regulation, up to “constitutional recognition of information ecology”.

There are difficulties in this approach.

First, Thistoll overstates the empirical case. The particular claims about YouTube recommendations dragging users inexorably to extremism are contested; later empirical work paints a more mixed picture about causation vs selection effects. Thistoll takes early, headline-friendly research as settled truth and generalises it to the entire online environment.

Secondly he relies on a false dichotomy in that he sets up a choice between a naïve, laissez-faire “marketplace” and heavy-handed ecosystem “gardening”. But in practice liberal systems have always combined robust free-speech protections with targeted interventions (fraud, threats, harassment) without treating the whole infosphere as something the state must centrally curate.

Thirdly he employs a vague but powerful regulatory hook – “Discourse health”, “ecosystem stewardship” and “information pollution” are rhetorically potent but legally fuzzy.

Once you constitutionalise those ideas, almost any unpopular speech or viewpoint can be reframed as pollution to be cleaned up. The essay never really grapples with who decides what a “healthy” discourse looks like, and by what limits they are bound.

5. Culture-war framing and legalisation of disagreement

In this final theme in Thistoll’s approach opponents of his preferred reforms are routinely described as engaging in “global culture wars”, “backlash”, “moral panic” or smear campaigns, rather than as people with legitimate competing rights-based concerns (notably women, gender-critical feminists, LGB groups and some clinicians).

His correspondence with professional bodies and regulators (e.g. over the Genspect conference, or gender-affirming care debates) is framed as enforcing ethical and human-rights obligations but his critics argue it functions in practice as an attempt to intimidate and silence dissenting professionals.

This approach too has problems.

First, it pathologises disagreement – Once opposition is rebranded as “culture war” or “hate”, there is no answer substantive arguments about evidence, risk, women’s privacy, child safeguarding, etc. In essence he utilizes veto words to shut down a response.

Secondly law is used as a weapon in moral disputes – Writing to regulators, professional bodies and media councils to depict attendance at certain conferences or publication of certain viewpoints as potential ethical breaches or even criminal intimidation exemplifies “using the law as rhetoric” in exactly the way he accuses others of doing.

Finally his approach is polarisation by other means – Ironically, while claiming to reduce harm and polarisation, this style of advocacy entrenches it: people with mainstream concerns about rapid legal change around gender, speech and children’s medicine are pushed into the same rhetorical box as genuine bigots.

To sum up, the common threads in Paul Thistoll’s pronouncements are:

1. A maximalist, UN-driven human-rights frame that is presented as legally obligatory rather than politically contestable.

2. An identity-centric view of law, where symbolic affirmation and protection from offence/dignitary harm take centre stage.

3. Broad endorsement of hate-speech and digital-harm regulation as essential to protect vulnerable groups, with relatively little weight on free-expression costs.

4. An algorithmic-determinist critique of free-speech liberalism, urging an “ecosystem” model of regulated discourse.

5. A culture-war narrative in which resistance to these moves is treated less as reasoned disagreement and more as a regressive backlash.

Those themes are flawed not because concern for marginalised people is misplaced, but because they rest on:

• shaky claims about what international law requires

• an unstable move from subjective feelings to enforceable legal duties

• underestimation of the chilling and weaponisable nature of speech-restrictive regimes

• and a tendency to delegitimise opponents rather than engage with the difficult trade-offs between competing rights.

But despite all this the Elephant is still in the room. And in fact there may be two.

I say this because although the programme discussed hate speech it should be remembered that the Law Commission was charged with examining the law on this topic and coming up with recommendations. This direction, unsurprisingly, came from the Labour Government.

When the Coalition came into power the Law Commission was instructed to stop its work and for a moment it seemed that the issue might have died the death. But the term finds its way into the debate via an interpretation of the recommendations of the Ia Tangata Report – another doubtful contribution by the Law Commission – and indeed the “hate” issue has been perpetuated by the present Government in another form.

The Law Commission has been asked to look into “hate crimes” and there was a consultation about this in which I participated and a Discussion Paper is expected soon.

So how to define hate speech or indeed hate crime. The first thing is to take the word “hate” out of the equation. It muddies the waters, is intensely subjective and can mean a number of different things to a number of different people. A well known academic talking head to my horror said at a Conference that “hate speech is speech I hate to hear”. You can’t get more subjective than that.

My position is that there should be a recognition that some speech can be seriously harmful to the community. Rather like the Sentencing Act provision in which the word “hostility” is used, I prefer the use of the word “dangerous” because it lends to the speech in question a quality of imminence of harm.

I wrote on this at length in 2021 in a paper entitled “Hate Speech Proposals – A Consideration”. The paper is available on the Social Science Research Network.

I also discussed what I thought would be a satisfactory proposal which would provide a solution to the issue of “hate” speech or “dangerous” speech as I term it.

The language of the proposal is technology neutral. It could cover the use of words or communication either orally, in writing, electronically or otherwise. Although I dislike the use of the words “for the avoidance of doubt” in legislation for they imply a deficiency of clarity of language in the first place, there could be a definition of words or communication to include the use of electronic media.

The language of the proposal is as follows:

It is an offence to use words or communication that advocates or intends to incite actual physical injury against an individual or group of individuals based upon, in the case of a group, identifiable particular characteristics of that group

This proposal would achieve a number of objectives. It would capture speech or communications that cause or threaten to cause harm of a lesser nature than grievous bodily harm stated in section 306 of the Crimes Act 1961.

The proposal is based upon ascertaining an identifiable harm caused by the speech or communicative act. This enables the nature of the speech to be crystallised in an objective manner rather than the unclear, imprecise and potentially inconsistent use of the umbrella term “hate speech.”

The proposal would cover speech, words or communication across all media. It would establish a common threshold for words or communication below which an offence would be committed.

I suggest that this proposal would deal with one of the elephants in the room.

The other one, in my view, is a phantom. There is no need for there to be specifically defined “hate” crime if indeed there is such a thing. Better in my view for the issue to remain where it is – in the Sentencing Act. If hostility as set out in 9(h) is an element of the offending it can be treated as an aggravating factor and an uplift recognizing that could be imposed.