Olivera Boskovic is a professor of law at the University of Orléans, France.
Many recent studies on defamation and violations of rights relating to personality assert that both jurisdiction and choice of law rules in this area are problematic. The following observations will mainly focus on choice of law.
However, it is worth saying that jurisdiction rules, laid down by the Brussels regulation (articles 2 and 5-3) seem globally satisfactory, even though one has to recognise that they need to be adapted to torts committed via the internet. The mere possibility to access a website from the forum State should not be considered sufficient to found jurisdiction under article 5-3. Closer connection with the forum (through the idea of targeting) should definitely be required. This adaptation does not require legislative intervention, the ECJ can do it. However one problem remains. Under article 5-3 ( as interpreted in Shevill) when jurisdiction is based on the place of damage, the remedy must be limited to damages arising in the forum State. The problem is that for some remedies, it is impossible or at least difficult to limit the remedy so that it does not have an impact in other countries (it is possible for damages, less so for injunctions). However the French Yahoo case (TGI Paris 20 nov. 2000, JCP 2000, Act, p. 2214) shows that it can be done.
Concerning choice of law, the situation is different. The working document of the European Parliament questions the necessity of legislative intervention and envisages the option of maintaining the status quo. It is submitted that this would be an unsatisfactory solution from the point of view of legal certainty. Whatever one thinks of the Rome II regulation and the rules it lays down, it can not be denied that its main objective, that is improving legal certainty, has been attained. The same reasons justify legislative intervention in the area of defamation, area in which conflict of law rules in the member States vary considerably.
Having said that, the main question is obviously what is the appropriate choice of law rule?
Several options had been envisaged during the elaboration of the Rome II regulation. Basically these were the law of the habitual residence of the victim, the law of the place of damage subject to certain exceptions and the law of the country to which the publication is principally directed. The first two were perceived as being more claimant-friendly and the last one as being more favourable to the media.
Actually the country to which the publication is principally directed is not as such, necessarily, more favourable to the media. What explained that perception was that the European Parliament proposed to apply the law of the country in which editorial control is exercised whenever it was not apparent to which country the publication was principally directed. This is definitely favourable to the media and in contradiction with the general orientation of the regulation which chose to give relevance to the law of the place of damage as opposed to the law of the place of acting. The law of the country to which the publication is principally directed is a variant of the law of the place of damage and shall be discussed as such.
As for the law of the habitual residence of the harmed person, apart from the general criticism of being too favourable to the claimant three other criticisms were to be found. The first was uncertainty, based on the fact that celebrities’ habitual residence is difficult to determine. This is very unconvincing. The second and third are linked. The idea is that this connecting factor makes it possible for a media to be held liable for behaviour perfectly legal in the place of acting and hence constitutes a danger for freedom of speech. The first part of the argument is correct, but this is true of any connecting factor other than place of acting, which precisely was rejected by EU authorities. Does the fact that the harmful act involves exercise of a fundamental right change something? Proponents of this argument think so. They take the example of foreign dictators who would become impossible to criticise under the law of their residence, which probably considers any criticism ipso facto defamatory. This would endanger freedom of speech. The argument seems slightly excessive. Surely, in such cases the public policy exception (ordre public) could apply and constitute a sufficient barrier against such laws.
However, there is one argument against the law of the habitual residence of the victim that seems valid. Defamation and violations of rights relating to the personality involve two fundamental rights: freedom of speech and the right to privacy. The way nations all over the world strike a balance between these rights is very different. Hence, it appears that each State should remain in charge of striking that balance for its own territory. This consideration points to the law of the place of distribution, that is the law of the place of damage. Of course this connecting factor needs adaptation in the context of the internet (distribution, as a positive action has no sense in this context). Mere accessibility of a website should not be considered as distribution. Some targeting should definitely be required (this problem would be avoided with the law of the habitual residence of the victim, rejected for aforementioned reasons).
So it appears that the general rule (article 4-1) could perfectly apply to defamation. This is not necessarily true for article 4§2. Initially, one could think that there is no reason to treat defamation and violation of rights relating to personality differently than other non contractual obligations. This would mean that article 4§2 should apply. On second thought, several reasons come to mind. First of all, applying article 4§2 would hinder the possibility of each State striking the aforementioned balance as it thinks fit. Secondly, the general justification of the exception in favour of the parties’ common habitual residence is that this law has closer ties with the case than the law of the place of the damage which is often fortuitous. But precisely, the place of damage in cases we are concerned with is not fortuitous (the media know where the defamatory article, for example, will be distributed), provided that place of damage in the context of internet be defined in a more demanding way.
However, this does not mean that common habitual residence would have no relevance whatsoever. It could certainly be taken into account by the court within the general “closest ties” exception. This exception provides for flexibility and allows for the application of several laws (of places of distribution) or one unique law (possibly of the parties’ common residence) according to the circumstances.
This possible application of multiple laws is often seen as a serious disadvantage of the law of the place of damage rule. However, one may wonder why this is considered to be such a problem in this area, while it is accepted in others, such as unfair competition. In any case the existence of the general closest ties exception would allow to limit the negative effects of the place of the damage rules in extreme cases.
So at the end of the day, the only real problem with the place of damage rule is the internet and defining the place of damage in its context. It appears that it is probably preferable to leave this question to the courts and not lay down a final rule at this stage (although one can say that some targeting must be required).
In any case the public policy exception (ordre public) should apply and should be a sufficient barrier against laws which do not respect the requirement of the European Convention on human rights. No specific exception is needed.