Dr Clotilde Pegorier, Lecturer in Law, University of Essex
Note: the hyperlinks to the relevant webpages are in their original languages – French and German.
On 13 May 2020, the French Parliament passed a new bill geared to combatting online hate speech. Disputed from the outset, the bill was, on 18 June 2020, subsequently ruled by the Conseil constitutionnel – the Court that reviews legislation to ensure compliance with the French Constitution – to be partially, even largely, unconstitutional. Indeed, the ruling effectively quashed seven of the bill’s provisions, and made substantial amendments to several others, notably paragraphs I and II of Article 1. Small wonder that Bruno Retailleau, Vendéan Senator and president of ‘Les Républicains’, spoke of the ruling – in fitting French manner – as having “totally decapitated” the bill.
What is afoot here? And what does this mean for the French government’s efforts to regulate online content?
What was in the Original Bill?
Before reviewing the Conseil constitutionnel ruling, let us first consider the rationale and content of the original bill.
Named for its main sponsor, MP Laetitia Avia of Emmanuel Macron’s ‘La République en Marche’ party, the law was largely inspired by the German Netzwerkdurchsetzungsgesetz (NetzDG), which came into effect in October 2018 and which foresees significant fines for online platforms that do not remove “manifestly illicit” content within a stipulated timeframe of 24 hours after it being reported.
The ‘Loi Avia’ was designed in the light of the NetzDG to update the current legislative framework supplied by the Law on Confidence in the Digital Economy (Loi pour la confiance dans l’économie numérique, LCEN) of 2004, notably by reinforcing the contribution of digital providers and platforms to the struggle against online hate. Its central provision, set out in Article 1, was to command online platforms falling under the purview of the bill “to render inaccessible, within 24 hours of notification by one or more persons, any content manifestly constituting of the offences” stipulated in this and other laws – that is, content that violates France’s hate speech provisions. According to the bill, platforms were also obliged to adopt “appropriate resources to prevent the redistribution” of content deemed manifestly illegal (article 2). The scope of the law was to extend to “operators of online platforms […] offering an online public communication service based on connecting multiple parties for the purpose of sharing public content or based on classifying or referencing content by means of computer algorithms, which is offered or placed online by third parties, where this activity on French territory exceeds a threshold, determined by decree” (article 1). Where, precisely, this threshold lay was to be decided subsequently. Notably, the bill covered social media platforms and search engines, but not internet service providers. Failure to comply with the new law would incur a criminal fine of up to 250’000 euros for individuals and 1’250’000 euros for corporations. In addition, an administrative penalty of up to 20 million euros or 4% of a company’s global annual turnover could be imposed for “serious and recurrent” failures.
The Process of Adoption
It is worth reflecting for a moment on the particular process by which the bill was first adopted. In May 2019, the Government decided to apply the ‘procedure accelerée’ (accelerated procedure) foreseen in Article 45 of the French Constitution. This decrees that, after a reading by each of the two chambers of Parliament – the Assemblée nationale (roughly equivalent to the House of Commons) and the Sénat (House of Lords) – and in the case of no agreement being reached on a common text, the Prime Minister or the Presidents of the two Houses can convene a joint committee, comprising equal members from each House, to propose a compromise text on debated issues. This is what occurred here: the two parliamentary chambers could not find an accord on the text of the bill and a commission was constituted. This failed, however, to yield a compromise text acceptable to both sides, and so the ‘normal’ legislative procedure resumed – the original text as amended and adopted by the Sénat went back to the Assemblée nationale, which made its own modifications, and this new text was then returned to the Sénat for further amendment.
As the process stalled in this back and forth between the chambers, the Government eventually decided to give a final reading before the Assemblée nationale – again in line with constitutional provisions – and the bill was adopted in May 2020. All of which is to say that, as a consequence of such wrangling, the bill was passed only by one of the two parliamentary chambers, albeit it the more ‘democratic’ one. Given the nature of the bill, and the current “state of health emergency” in place in France, one can readily question how well- or ill-advised this move was on the part of the Government. What seems clear, though, is that it lent an air of almost inevitability to subsequent challenge and dispute. Following adoption, on 18 May, 60 members of the senate submitted an appeal to the Conseil constitutionnel to contest the constitutionality of the bill.
The Conseil Constitutionnel Ruling
The arguments put forward by the challengers to the bill – and those upheld by the Conseil constitutionnel – were, unsurprisingly, connected to the matter of legitimate and illegitimate restrictions to freedom of expression. Unsurprising, as these concerns were already at the forefront of jurisprudential and public debates and discussions during the bill’s drafting.
Citing the 1789 “Declaration of the Rights of Man and of the Citizen”, the Conseil constitutionnel determined in its ruling that both Paragraph 1 (demanding the removal of content relating to terrorism and child pornography within the hour) and Paragraph 2 (requiring the removal of hateful content within 24 hours) of Article 1 constitute an “infringement on the exercise of freedom of expression and communication that is unnecessary, inappropriate and disproportionate”. The follow-through from this determination on Article 1 was to render an entire raft of subsequent provisions unconstitutional. The removal window in both scenarios was, the council held, “particularly brief”, and the severity of the proposed sanctions would “only incite online platform operators to remove flagged content, whether obviously unlawful or not,” especially in the absence of specific cause that exonerates from responsibility. With no judicial intervention foreseen, it would be for platform administrators (as private actors) to determine whether or not particular content is unlawful – a situation that would, in the verdict of the council, likely encourage an excessively censorious approach and the removal of materials that are in fact lawful.
What remains of the bill after the ruling is modest. Perhaps most notable is the acceptance of a proposal to create an official online hate speech watchdog (article 16). While by no means inconsequential, this and other minor provisions represent a meagre return when set in the context of the bill’s ambitious aim to overhaul the legislative landscape for dealing with online hate speech.
So where does this leave the government’s efforts to police online content? Clearly, this is a substantial setback. While the bill was officially enacted, the ruling of the council stripped it of almost all meaningful impact. If not quite in tatters, the government’s strategy is tarnished, and there is obvious need for a rethink. Not that there is any sign of submission – in a statement following the ruling, Laetitia Avia vowed not to give up the fight, and asserted that the judgement offered a “roadmap to improve a plan that we knew to be unprecedented and therefore perfectible.” Thus the show will go on. But the implications of the ruling should not be downplayed. These may also extend beyond national borders – the government had hoped that the new bill might provide a template for the European Commission’s Digital Services Act, scheduled to be put forward by the end of the year. The Commission said that it “took note” of the council’s ruling.
Both of itself and as part of France’s extended efforts to regulate speech across diverse contexts, this recent chapter is variously revealing of the idiosyncrasies of our jurisprudence and constitutional arrangements, of the relationship the French state maintains with its citizens, and of its approach to balancing free speech with anti-discrimination concerns and the fight against harmful content (which differs markedly from the US, for example). It has also proved another flashpoint in ongoing debates on possible limits to freedom of expression and the dilemma of hate speech. That this is a fraught and thorny issue barely needs restating. Nor does its importance. The question of where to set the line between permissible and impermissible speech is contentious, daunting and potentially confusing – most reflective minds would probably admit to being pulled in different directions at different times and in different contexts. Just as we bristle at attempts to muzzle freedom of expression, so we do at the harms caused by hateful speech. Marking that boundary was, is, and will likely always remain, a tightrope walk. How the French government opts to move forward in the coming months will be an interesting watch.