In the context of electoral campaigns, the main criteria used by the Court to distinguish free expression from unlawful hate speech can be deduced from the case-law mentioned above: a) Content of the message: The Court examines whether the expression gravely insults, demonises or dehumanises a protected group, or calls for its discrimination or harm. Statements portraying a community as inherently inferior, violent, or dangerous are generally deemed incitement to hatred which can be forbidden. By contrast, general – albeit harsh – criticisms of government policies, religions, or ideologies, without inciting hatred against their adherents, may fall within the scope of protected speech; b) Context and scope: It is crucial to assess when and where the speech was delivered, its intended audience, and the likelihood of causing real harm. Thus, a leaflet widely distributed in an electoral campaign exploiting racial prejudice carries a dangerous social impact, whereas extremist words in a controlled debate may be mitigated by the context. The Court evaluates whether the public could reasonably perceive the statement as a serious incitement or merely as political rhetoric, satire, etc. It also considers the position of the speaker: if authors are influential politicians or candidates to elections, their responsibility is greater to avoid rhetoric that inflames hatred; c) Intention (purpose) of the speech: If the apparent aim is to foster intolerance or mobilise negative feelings against a group (for example, “provoking rejection and antagonism toward the population of Islamic origin” in Soulas), the Court leans toward categorising it as hate speech. Conversely, if offensive ideas are expressed as part of a contribution to public debate without intent to incite violence – for instance, a historian discussing a controversial historical event in good faith – the Court tends to grant broader protection, unless it amounts to malicious negationism; d) Severity of interference and State’s margin of appreciation: When reviewing the proportionality of the State’s response, the Court examines the severity of the sanction: Imposing a prison sentence for controversial expressions (a measure the Court views with great suspicion) is very different from applying lighter or administrative sanctions. Indeed, the Court has stated that custodial penalties for speech, except in cases of incitation to violence or extreme hatred, are almost never acceptable as they are disproportionate. States therefore have some margin to determine when speech threatens social coexistence enough to warrant sanction, but that margin is subject to supervision: if a State oversteps its margin of appreciation by punishing opinions that could have been tolerated in the democratic debate, the Court does not hesitate to find a violation of Article 10 ECHR.