Chinese Investments in the EU, National Security, and Investor-State Arbitration

By Dr. Szilárd Gáspár-Szilágyi, Birmingham University School of Law

Disclaimer:

This blogpost is based on a recent publication by the author in the Journal of International Dispute Settlement (Oxford University Press), entitled ‘When the Dragon Comes Home to Roost. Chinese Investments in the EU, National Security, and Investor-State Arbitration.’ The data for the publication was gathered between September-December 2022.

1. Context

The gradual rise of China as an economic, normative, and lending power has resulted in more protectionist measures in areas of the world that traditionally championed economic liberalization, such as the USA, the UK, and the EU. Currently, 21 out of 27 EU Member States have national laws or measures in place for the screening or review of foreign investments and the EU’s 2019 Investment Screening Regulation is under reform. However, the proliferation of national screening mechanisms and the adoption of investment control measures could trigger the international responsibility of EU States under the bilateral investment treaties (BITs) in force between China and 26 EU Member States (except for Ireland). This is evidenced by the recent Huawei v. Sweden request for ICSID arbitration, following the exclusion of Huawei from the development of 5G mobile networks in Sweden (see also here).

In light of the above, the aim of this research was to look at the BIT-level and other variables that can influence Chinese investors in bringing investment treaty-based arbitration (ITA) claims against the investment screening measures of EU Member States.

2. The BIT-level Variables and the Findings

Whether a Chinese investor can bring an ITA claim against an EU Member State because of the latter’s review or screening measures, and whether the EU State can defend itself against such a claim, will depend on a set of BIT-level and other variables. China has so far concluded 163 investment treaties (both BITs and Preferential Trade Agreements [PTAs] with investment chapters) and has BITs in force with all EU Member States, except Ireland. The oldest Chinese BIT in force with EU States was concluded with Sweden in 1982 (amended in 2004) while the newest one was signed in 2009 with Malta.

The paper organizes the BIT-level variables into more general variables and Non-Precluded Measures (NPM) clauses. Among the more general variables one can name:

• The presence and type of ITA clauses
• The IIA’s coverage of pre-and post-establishment
• The presence and scope of carve-outs.

NPM clauses needed special attention, as security considerations are also becoming more prominent in investment law and investment relations. Such clauses are important for a State’s defence as they limit ‘the applicability of investor protections under the BIT in exceptional circumstances’. However, only a minority of treaties include such exceptions. Even among these, the wording and scope of NPM clauses will differ, some being similar to the Article XXI GATT exception, while others use a different wording. Among the more specific NPM variable, the following were included:

• The presence of an NPM clause in the BIT
• The type of NPM clause if present: Article XXI GATT-type or ‘typical’
• The wording of the ‘nexus’ element
• Whether the NPM clause applies to the whole treaty or only certain standards, such as the national treatment clause
• Whether the NPM clause is self-judging.

The most important BIT-level variable is the presence and the type of ITA clause, as without one a Chinese investor cannot bring an international arbitral claim against an EU host State. The agreements with Italy, Denmark, Austria, and Bulgaria have no ITA clauses, all these agreements having been signed in the 1980s when China did not or included very limited ITA clauses. Therefore, these states will not face international treaty-based arbitration, unless an unlikely broad interpretation of the Maffezini doctrine, which allows to import more favourable jurisdictional conditions, would occur. Surprisingly, the original agreement with Sweden (1982) also did not include ITA, but this was modified in the 2004 Amendment Protocol, which in Article 6 bis now provides for ICSID or UNCITRAL arbitration. The post-2000 agreements are much more liberal and provide mainly for ICSID, and sometimes ad hoc arbitration under UNCITRAL and other arbitration rules agreed by the parties. Thus, countries like the Netherlands, Germany, Latvia, Finland, Belgium, Luxembourg, Spain, Czechia, Portugal, France, and Malta are at an increased risk of ITA against them. Furthermore, it must also be noted that these treaties require an EU investor to first exhaust the local Chinese administrative procedures before initiating arbitral proceedings against China (but not the other way around!). Therefore, there is a lower threshold for Chinese investors to commence arbitration than vice versa.

We must also consider the pre-establishment variable, since many screening measures only target investors that have yet to enter the host State and have yet to make the investment. Of the BITs analysed, only one, Article 3.3. of the BIT with Finland (2004) provides for protection ‘with respect to the establishment’ of the investment, and not just protection during the operation, management, and winding up of the investment. Two BITs, the 1985 one with Italy and the 2009 with Malta, do not cover the pre-establishment phase per se, but they do include clauses meant to facilitate the travel of the foreign investors for travels connected to the investment. In other words, except for Finland, the other EU states should not be overly concerned that their measures affecting the pre-entry phase of Chinese investments will result in an international arbitral claim against them. The same, of course, is not true for post-establishment measures.

The carve-outs for the various substantive standards of protection are also important. If for example, a specific sensitive sector of the economy is excluded from the BIT’s coverage, then the tribunal should decline its jurisdiction. There are no real carve-outs, however, for specific sensitive economic or industrial sectors in the BITs under discussion. This increases the risk of claims by Chinese investors.

When it comes to NPM clauses only the BIT with Finland (2004) in Article 3.5 provides for an Article XXI GATT-type clause and refers to measures ‘necessary to’ protect ‘essential security interests.’ Three cumulative components which are not easy to fulfil. The BITs with Germany (2003) and Portugal (2005) include typical NPM clauses that refer to measures that ‘have to be taken’ to protect public order, security, health, and morals. The Portuguese clause only applies to MFN treatment, whilst the German one covers FET, NT, and MFN.

3. Other Variables and Findings

It is important to recognise that the BIT-level variables do not exist in a vacuum but need to be read in the context of the current geopolitical climate. Whether EU Member States should fear a ‘deluge’ of ITA cases brought by Chinese investors against their investment screening measure will also depend on the volume of Chinese investment in the EU, the treatment of Chinese investors by the host EU States, and the importance of security alliances.

Here, based on the Merics Report, the paper finds that Chinese investments in the EU have been falling from their peak in 2016/2017. Compared to investments from other countries, such as the USA, Canada, or EFTA members, they are much less in number (both in terms of the number of investments and assets) than one might expect. Whilst the numbers are not trivial, less investments should result in a lower likelihood of ITA claims. Some authors also argue that there might be an ‘aversion’ of Chinese investors towards arbitration. Be that as it may, this does not guarantee that Huawei v. Sweden will not be used as a test case, the success of which could result in more ITA cases.

Next, security alliances in the current climate of protectionism cannot be overlooked. For example, Romania adopted various legislative measures (indirectly) against Huawei and other Chinese 5G developers, following an originally secret Memorandum of Understanding with its main NATO ally, the USA. A potential claim could be initiated pursuant to the Romania-China BIT’s FET and MFN clauses. Thus, EU States which are more open to Chinese investments might change their policies due to the pressure from NATO allies, increasing their exposure to ITA claims.

The case of Sweden and Huawei is also telling as the Chinese company had been established in Sweden for over two decades and they were directly named in the state measures excluding them from the bid to develop 5G networks. As the Huawei v Sweden arbitration illustrates, how the foreign investor is treated during a national investment screening or control process can potentially trigger the international responsibility of an EU State under the existing BITs with China.

4. What to do next, if anything?

One must be cognisant of the broader geopolitical shifts that are taking place in the world when writing concluding remarks. The ‘securitization’ of issues previously not in the realm of security concerns, combined with changing policies in many EU states towards Chinese investments mean that the interplay between national security concerns and ITA will be a field worthy to be followed in the upcoming years.

As we have seen, not all EU States block Chinese investments or have a mechanism in place to screen them. Furthermore, the number of EU companies owned by Chinese investors is not as high as one might think and Chinese FDI in EU States has been falling since 2016/2017. Whilst this does not exclude the possibility of Chinese ITA claims pursuant to national security reviews, it does somewhat lower the statistical probability of cases. Moreover, if we combine this with a certain ‘aversion’ of Chinese investors towards international arbitration, then the likelihood of cases falls even more.

Nevertheless, as Huawei v. Sweden shows, some ITA claims are to be expected, especially if Huawei is successful in winning the case or settling it ‘out of court’. Some EU States with older BITs, such as Italy, Denmark, Austria, and Bulgaria will not face claims from Chinese investors, unless a tribunal would be willing to give an extremely broad interpretation to the MFN clause under the Maffezini doctrine. For those States that have a more modern ITA clause, the first line of defence is the absence in most cases of the pre-establishment provisions. However, for established investors, like Huawei in Sweden, the reviews and screenings based on national security grounds could trigger the liability of the EU host State. For this latter group of states, the situation from a legal perspective is somewhat more challenging: (a) the post-establishment phase is covered; (b) the ITA clauses of the post-2000 BITs are quite broad; (c) most of the standards of protection do not include carve-outs for certain sensitive sectors; and (d) most treaties do not include any type of NPM clause.

Amending or terminating their existing BITs with China all pose their own challenges. For now, EU States might want to wait until the Huawei v. Sweden case is concluded as this is probably a test case. If the investor is successful, then similar cases could easily follow. Separating Chinese firms from the Chinese Communist Party (CCP) is often not an easy task. Thus, one could say that future cases could be influenced by the policy the CCP takes. However, it is also worth mentioning that State power in China is often not as concentrated and centralised as one might think, as illustrated by the lending practices of various central, regional, or local Chinese institutions, that often do not know what aid is granted by which institution. Thus, this could lead to a more autonomous decision making by the Chinese investors on whether to pursue ITA claims.