Hate speech severity assessment
Derived from the Rabat threshold test, this hate speech level framework can be used for a preliminary assessment of the severity of hate speech. Some additions have been made in relation to forcibly displaced and stateless people.
Although the Rabat framework is helpful in assessing the severity of hate speech, it is important that the person making the assessment has a deep understanding of the context in terms of language, culture, and politics. A 2019 report on online hate speech against human rights defenders in Guatemala emphasised that speech that may be considered low-risk in an environment characterized by democratic institutions and rule of law could be extremely risky in a country plagued by corruption, political violence, and inequality. The report further noted that the Independent International Fact-Finding Mission on Myanmar found that it was difficult to identify some instances of online hate speech against the Rohingya because of subtleties in the Myanmar language and the use of fables and allegories.
As the ICRC note, is challenging to demonstrate and evaluate the scope, severity and frequency of information risk-related consequences and associated vulnerabilities for people who may be affected by them. The lack of certainty regarding the kinds of metrics, indicators and criteria that should be used to help calculate the degree of exposure, and regarding the likelihood or potential severity of adverse digital events further complicates the processes of evaluating the need for a response and deciding on the level of priority to afford it, in comparison with other humanitarian and protection-related needs.
| Criteria of severity | Indicators | Questions to ask |
|---|---|---|
| 1. Context | The legal, political, social and economic context | Is the target group part of (i) a refugee; (ii) asylum seeking; or (iii) stateless population? Is the target group part of an IDP population? Is the target group part of a mixed population? Is the target group not within UNHCR’s mandate, but at risk of being forced to flee due to threats of imminent violence? Are some members of the target group currently part of diaspora populations outside the country? Is there an ongoing conflict or are there incidents of violence against the targeted group? Does the law recognise the targeted individual’s or group’s identity? Is this related to their immigration status, or to their racial or ethnic origin? Is there any national anti-discrimination legislation, is it aligned with international human rights norms and standards, and does it specifically prohibit discrimination on the grounds of refugee or migrant status, and does it define hate speech? How does the media report on the targeted group, if at all? Is the media independent? In what ways are the target group present on social media through posting their own content, and through content posted about them? Which platforms are the most relevant here? Are there upcoming elections, and what is the role of anti-refugee and migrant narratives in the electoral campaigns? Are there any challengers to the hate speech? If so, who are they? Link to the indicators from the UN Framework for Analysis of Atrocity Crimes. |
| 2. Speaker | The position or status of the speaker in society and their authority or influence over their audience | Does the speaker have power or influence in society? Are they a forcibly displaced or stateless person? or an official speaking on behalf of a State? Are they a national leader, politician, public official, religious or faith leader, or social media influencer? What is their reputation and standing in society? What is their relationship with targeted groups? |
| 3. Intent | The state of mind of the speaker | Did the speaker intend to engage in advocacy of hatred against an individual or group on the basis of their refugee or migrant status, or another protected characteristic strongly associated with their refugee or migrant status (e.g. ethnicity or religion depending on the context)?Is there evidence that the speaker worked with a professional agency or otherwise used automated software involving fake social media profiles and trolls? (this is to get to level of intent, scale, and sophistication, but also would point to a different response). Was there a triangular relationship, that is, did the speaker intend to incite the audience (host community) against the target group (refugees or migrants or stateless people)? (In the case of incitement only), was the speaker merely negligent or reckless in their expression? Was the speaker’s communication in poor taste or showing a lack of judgment? |
| 4. Content and form | Nature and style of the expression | To what extent was the speech provocative and direct? What was the form, style and nature of arguments deployed in the speech? Was there any balance of arguments deployed in the expression? Was the expression in the public interest? Was it artistic or academic expression? |
| 5. Extent and magnitude of the expression | Reach of the expression | How public was the expression when it was made? How widely was the expression disseminated? How large was the audience that was exposed to the expression? How influential was the audience that was exposed to the expression? Was the expression disseminated offline and/or online? If it was disseminated online how broad was the reach, and on which platforms? Is the expression repeated over time and over different channels and audiences? |
| 6. Likelihood, including imminence | Degree of risk of harm | Was there a reasonable probability that the speaker’s communication would succeed in inciting actual action by the audience against the target group? (In the case of incitement only, was there a reasonable probability that harm would result from the expression? (e.g. physical and/or psychological harm to an individual or group, or harm to social cohesion) Was there a reasonable probability that the audience would believe and repeat the expression to others? Would the harm have affected particular individuals within the targeted group (e.g. women, children, LGBTIQ people) more than others? Would the harm have a different impact upon women and men? |