CSDD and PIL: Some Remarks on the Directive Proposal


by Rui Dias


On 23 February 2022, the European Commission published its proposal of a Directive on Corporate Sustainability Due Diligence (CSDD) in respect to human rights and the environment. For those interested, there are many contributions available online, namely in the Oxford Business Law Blog, which dedicates a whole series to it (here). As to the private international law aspects, apart from earlier contributions on the previous European Parliament resolution of March 2021 (info and other links here), some first thoughts have been shared e.g. by Geert von Calster and Marion Ho-Dac.

Building on that, here are some more brief remarks for further thought:

Article 2 defines the personal scope of application. European companies are covered by Article 2(1), as the ones «formed in accordance with the legislation of a Member-State», whereas those of a «third country» are covered by Article 2(2). While other options could have been taken, this criterium of incorporation is not unknown in the context of the freedom of establishment of companies, as we can see in Article 54 TFEU (basis for EU legal action is here Article 50(1) and (2)(g), along with Article 114 TFEU).

There are general, non PIL-specific inconsistencies in the adopted criteria, in light of the relative, not absolutethresholds of the Directive, which as currently drafted aims at also covering medium-sized enterprises only if more than half of the turnover is generated in one of the high-impact sectors. As recently pointed out by Hübner/Habrich/Weller, an EU company with e.g. 41M EUR turnover, 21M of which in a high impact sector such as e.g. textiles is covered; whilst a 140M one, having «only» 69M in high-impact sectors, is not covered, even though it is more than three times bigger, including in that specific sector.

Article 2(4) deserves some further attention, by stating:

«As regards the companies referred to in paragraph 1, the Member State competent to regulate matters covered in this Directive shall be the Member State in which the company has its registered office.»

So, the adopted connecting factor as to EU companies is the registered office. This is in line with many proposals of choice-of-law uniformization for companies in the EU. But apparently there is no answer to the question of which national law of a Member-State applies to third-country companies covered by Article 2(2): let us not forget that it is a proposed Directive, to be transposed through national laws. And as it stands, the Directive may open room for differing civil liability national regimes: for example, in an often-criticised option, Recital 58 expressly excludes the burden of proof (as to the company’s action) from the material scope of the Directive proposal.

Registered office is of course unfit for third country-incorporated companies, but Articles 16 and 17 make reference to other connecting factors. In particular, Article 17 deals with the public enforcement side of the Directive, mandating the designation of authorities to supervise compliance with the due diligence obligations, and it uses the location of a branch as the primarily relevant connection. It then opens other options also fit as subsidiary connections: «If the company does not have a branch in any Member State, or has branches located in different Member States, the competent supervisory authority shall be the supervisory authority of the Member State in which the company generated most of its net turnover in the Union» in the previous year. Proximity is further guaranteed as follows: «Companies referred to in Article 2(2) may, on the basis of a change in circumstances leading to it generating most of its turnover in the Union in a different Member State, make a duly reasoned request to change the supervisory authority that is competent to regulate matters covered in this Directive in respect of that company».

Making a parallel to Article 17 could be a legislative option, so that, in respect to third-country companies, applicable law and powers for public enforcement would coincide. It could also be extended to jurisdiction, if an intention arises to act in that front: currently, the general jurisdiction rule of Brussels Ia (Article 4) is a basis for the amenability to suit of companies domiciled (i.e., with statutory seat, central administration, or principal place of business – Article 63) in the EU. In order to sue third country-domiciled companies, national rules on jurisdiction have to be invoked, whereby many Member-States include some form of forum necessitatis in their national civil procedure laws (for an overview, see here).The Directive proposal includes no rules on jurisdiction: it follows the option also taken by the EP resolution, unlike suggested in the previous JURI Committee draft report, which had proposed new rules, through amendments to Brusselas Ia, on connected claims (in a new Art. 8, Nr. 5) and on forum necessitatis (through a new Art. 26a), along with a new rule on applicable law to be included in Rome II (Art. 6a) – a pathway which had also been recommended by GEDIP in October 2021 (here).

As to the applicable law in general, in the absence of a specific choice-of-law rule, Article 22(5) states:

«Member States shall ensure that the liability provided for in provisions of national law transposing this Article is of overriding mandatory application in cases where the law applicable to claims to that effect is not the law of a Member State.»

So, literally, it is «the liability provided for» in national transposing laws, and not the provisions of national law themselves, that are to be «of overriding mandatory application». This may be poor drafting, but there is apparently no material consequence arising out of it.

Also, the final part («in cases where the law applicable to claims to that effect is not the law of a Member State») does not appear to make much sense. It is at best redundant, as Geert van Calster points out, suggesting it to be struck out of the proposal. Instead of that text, it could be useful to add «irrespective of the law otherwise applicable under the relevant choice-of-law rules», miming what Rome I and II Regulations state in Articles 9 and 16.

A further question raised by this drafting option of avoiding intervention in Rome II or other choice-of-law regulations, instead transforming the new law into a big set of lois de police, is that it apparently does not leave room for the application of foreign, non-EU law more favourable to the victims. If a more classical conflicts approach would have been followed, for example mirrored in Article 7 of Rome II, the favor laesi approach could be extended to the whole scope of application of the Directive, so that the national law of the Member-State where the event giving rise to damage occurred could be invoked under general rules (Article 4(1) of Rome II), but a more favourable lex locus damni would still remain accessible. Instead, by labelling national transposing laws as overridingly mandatory, that option seems to disappear, in a way that appears paradoxical vis-à-vis other rules of the Directive proposal that safeguard more favourable, existing solutions, such as in Article 1(2) and Article 22(4). If there is a political option of not allowing the application of third-country, more favourable law, that should probably be made clear.