Thursday, 29 July 2021

Strengthening Europol’s Mandate: An Appraisal of the Commission’s Proposal to Amend Regulation (EU) 2016/784 (Europol Regulation)


Niovi Vavoula and Valsamis Mitsilegas, Queen Mary University




The European Union Agency for Law Enforcement Cooperation (Europol), the legal basis of which is Regulation (EU) 2016/794 (Europol Regulation), has a key role in supporting EU Member States on cross-border police cooperation. Europol is described as the EU’s ‘criminal information hub’, as it facilitates information exchange between Member States, Europol, other EU bodies, international organisations and third countries, and produces criminal intelligence on the basis of information acquired from various sources, including Member States and its partners. Amongst its many tasks, Europol also supports and coordinates cooperation on cross-border police work and produces regular assessments that offer comprehensive, forward-looking analyses of crime and terrorism in the EU.


On 9 December 2020, the Commission presented a proposal for a Regulation amending the Europol Regulation, accompanied by a two-part Impact Assessment, aiming at enhancing the Agency’s mandate in numerous respects. From the outset, it must be emphasised that the timing of the proposal is dubious, as the Europol Regulation has not been subject to an evaluation yet and according to Article 68, such evaluation was due in May 2022. Instead, scarce information is included in the Impact Assessment accompanying the proposal and some EU documentation, which, however, cannot replace the lack of a proper evaluation. As a result, the effectiveness and impact of the agency cannot be fully and properly assessed.

The proposal encompasses widespread reforms to Europol’s tasks, which may be divided in nine themes, as follows:


(1) Enabling Europol to cooperate effectively with private parties;

(2) Enabling Europol to process large and complex datasets;

(3) Strengthening Europol’s role on research and innovation;

(4) Enabling Europol to enter data into the Schengen Information System (SIS);

(5) Strengthening Europol’s cooperation with third countries;

(6) Strengthening Europol’s cooperation with the European Public Prosecutor’s Office (EPPO);

(7) Enabling Europol to request the initiation of an investigation of a crime affecting a common interest covered by an EU policy;

(8) Strengthening the data protection framework applicable to Europol; and

(9) Other provisions, including enhancing political accountability and parliamentary scrutiny.


This blog post aims to provide a snapshot of the proposal and highlight key privacy and data protection concerns by looking in turn into the thematic blocks. It is based on a study commissioned by the LIBE Committee of the European Parliament published on 27th May 2021, which argues that the proposed Regulation, as it stands, will radically transform the nature and powers of Europol and its relationship with key stakeholders without introducing adequate safeguards.


Enabling Europol to cooperate effectively with private parties


A first set of revisions concerns the enhancement of cooperation between Europol and private parties in countering criminal offences committed in abuse of the cross-border services of private parties. Currently, Europol is allowed to exchange personal data with private parties, but Article 26 of the Europol Regulation provides a series of restrictions: the traditional way for the agency to receive personal data from private parties is indirectly via competent intermediaries and Europol is prohibited from transferring personal data directly to private parties, unless one of the three exceptions applies. The proposal aims to establish the agency as a central point of contact in cases of multi-jurisdictional or non-attributable datasets,. Europol will be enabled to: (a) receive personal data directly from private parties on a more regular basis; (b) inform such private parties of missing information; and (c) ask Member States to request private parties to share further information. Additionally, Europol will be able to provide its infrastructure for the exchange of data between national authorities and private parties and support Member States in preventing large scale dissemination of terrorist content or violent extremism, on which Regulation (EU) 2021/784 was recently published.


These changes constitute a considerable paradigm shift for the agency, which is line with the emergence of the trend in past years, exemplified by the e-evidence legislative package, to establish direct channels of communication between law enforcement and private parties and foster a public/private partnership. Questions about the ability of private parties to undertake the role of law enforcement authorities in scrutinising fully and effectively the fundamental rights implications of transfer of personal data held by them for the purposes of law enforcement emerge, as Europol will be enabled to forward requests on behalf of Member States and proactively request information. Private parties do not enjoy equality with public authorities in terms of cooperation and the same will also apply in the case of Europol.


Therefore, they may find themselves in a subordinate position, being ‘cornered’ by both Europol and Member States to hand over the personal data requested. Important safeguards, in particular obtaining prior judicial authorisation and scrutiny of compliance with fundamental rights, risk being bypassed. Applying this approach to the case of Europol requires detailed rules on the duties of Europol, Member States and the private sector, e.g. when the private parties may refuse to cooperate, as well as provisions on independent authorisation of transfers and remedies for individuals, which are missing from the proposal. Even the concept of ‘private parties’ is open-ended and there are no limitations as to their nature. Whereas certain safeguards are included, e.g. the requirement for ‘absolute’ or ‘strict’ necessity, there are additional safeguards that are mentioned in the Impact Assessment, but not explicitly stated in the proposal. It is further argued that the European Data Protection Supervisor (EDPS) could be involved before the agency makes such transfers. In addition, whereas the proposal proscribes systematic, massive or structural transfers in cases where the private party is outside the EU, this is not extended to those private parties within the EU. Finally, it must be ensured that Europol’s role in supporting Member States to prevent the dissemination of online content related to terrorism and violent extremism conforms with the Europol’s role as foreseen in Regulation (EU) 2021/784 on preventing the dissemination of terrorist content online.


(2) Enabling Europol to process large and complex datasets


This reform aims to address the so-called ‘big data challenge’ following the admonishment of the agency by the EDPS on 17 September 2020. The proposal aims to enable Europol to conduct ‘pre-analyses’ of large and complex datasets received and identify whether these concern individuals whose personal data may be processed by Europol in line with Annex II of the Europol Regulation. Another proposed provision aims to enable the pre-analysis in support of a criminal investigation following transmission of an investigative case file to Europol.


Overall, it is welcome that the prior processing is limited to a maximum period of one year, which can be extended following authorisation by the EDPS. One suggestion is to define the terms ‘large datasets’ and ‘digital forensics’ and explicitly delimit processing when there is an objective necessity, which is not mentioned, so as to ensure that the derogation of Article 18(5a) does not become the rule. Clear criteria to determine that it is justified to extend the maximum period of pre-analysis must be laid down and it could be useful to consider that prior to each pre-analysis the EDPS must be at least informed and that the Europol Data Protection Officer must provide authorisation. The relationship between the new rules and the existing derogation under Article 18(6) of the Europol Regulation must also be clarified, as well as the relationship between the two new provisions foreseen. As these rules constitute an exception, their application must be strict and the existence of a link to an on-going investigation is crucial. In addition, the Regulation should lay down certain conditions and/or thresholds, such as scale, complexity, type or importance of investigations. Finally, the involvement of the EDPS not only in cases where an investigative case file is submitted by a third country, but in general in supervising the processing of large and complex datasets should be maintained and enhanced.


(3) Strengthening Europol’s role on research and innovation


The proposal foresees a greater role for Europol as regards processing of personal data for research and innovation matters for the development of tools, including the use of AI for law enforcement. One must be mindful though that when developing new technologies extensive processing of large quantities of personal data may be required, for example to create and test algorithms or for encryption. Therefore, the potential impact of such processing for research and innovation purposes to the principle of non-discrimination and the rights to respect for private life and protection of personal data must be guaranteed. The processing of personal data for research and innovation should take place only if needed in order to reach the objectives of the project. Furthermore, the processing of synthetic, anonymised or pseudo-anonymised personal data, as opposed to real operational data must be preferred, where possible, and the processing of special categories of personal data must be explicitly excluded or accompanied by additional safeguards. Moreover, principles of data protection law—in particular the principles of data minimisation, data quality and privacy by design and by default—must be taken into account.


(4) Enabling Europol to enter data into the Schengen Information System (SIS)


One of the thorniest aspects of the Europol reform concerns the possibility of enabling the agency to enter alerts into SIS. Currently, Europol has ‘read-only’ access to all types of alerts stored in SIS, both immigration and law enforcement related. The proposal creates a new alert category that Europol can use to enter alerts into SIS following consultation with the Member States and after authorisation by its Executive Director. A detailed process for the issuance of so-called ‘information alerts’ is foreseen in a separate proposal amending Regulation (EU) 2018/1862.


However, whether this power, which to an extent equates Europol with Member States, fits within Europol’s mandate, as laid down in Article 88 TFEU, is doubtful.  It is also questionable whether Europol will be able to conduct a proper quality check before issuing alerts into SIS. Importantly, the operational value of such alerts is also questionable, as the alerts will provide significant discretion to national authorities to follow up and wide divergences may arise in practice. The impact on individuals whose personal data will be inserted in SIS is significant and potential liability issues may also arise if the quality of data contained in the alert is not high. In light of the concerns voiced by a number of Member States within the Council, the Portuguese Presidency proposed an alternative to delimit these alerts to those concerning terrorism. However, it is feared that merely opening up Europol to SIS will become the gateway through which in the future Europol may acquire further powers to enter other types of alerts into the system (e.g. on missing persons).


(5) Strengthening Europol’s cooperation with third countries


Another important reform of the proposal concerns cooperation with third countries. Under the current legal framework, as laid down in Article 25(1) of the Europol Regulation, the agency may receive personal data from third countries based on: a) adequacy decisions under Directive (EU) 2016/680; b) international agreements under the current Regulation concluded in accordance with Article 218 TFEU; and c) cooperation agreements concluded between Europol and third countries under the previous Europol Council Decision (for the agreements, see here). Finally, the Executive Director can authorise the transfer of personal data to third countries and international organisations on a case-by-case basis for certain exceptional––but arguably broadly worded––reasons. With no adequacy decisions adopted and the negotiations for eight international agreements stalled, the calls for a less cumbersome regime for the exchange of personal data with third countries have proliferated. To that end, the proposal foresees a (seemingly minor) change enabling the Executive Director to authorise not only transfers, but also categories of transfers of personal data to third countries or international organisations in specific situations and on a case-by-case basis. However, it is not clear what exactly is meant by ‘categories of transfers’ and this reform may broaden the remit of such transfers from criminal investigations on specific suspects to surveillance activities in general, thus changing Europol’s powers.


That said, within the Council Member States have expressed their wish to further expand Europol’s capabilities to exchange personal data with third countries by transplanting the wording of Directive (EU) 2016/680 (Law Enforcement Directive) and Regulation (EU) 2018/1727 (Eurojust Regulation) to the Europol legal framework, and creating a new legal ground for exchanges of personal data on the basis of appropriate safeguards outside the three already prescribed grounds. This reform poses significant legal challenges as it bypasses existing institutional safeguards and undermines the importance of an adequacy decision, the procedure for assessing the data protection framework of a third country as adequate in violation of the constitutional limits placed by the Court of Justice of the EU (CJEU) in Schrems, as well as the institutional framework for adopting international agreements. The possibility of using bilateral agreements as the legal basis for such exchanges may result in divergences and different standards applied.


(6) Strengthening Europol’s cooperation with the European Public Prosecutor’s Office (EPPO)


This reform concerns the reinforcement of Europol’s cooperation with the EPPO in the aftermath of the adoption of Regulation (EU) 2017/1939 (EPPO Regulation) on the establishment of the EPPO. However, the proposal is not fully aligned with the rules of the EPPO Regulation and minor modifications to the text are necessary.


(7) Enabling Europol to request the initiation of an investigation of a crime affecting a common interest covered by an EU policy


The proposal aims to enable Europol to request competent authorities of a Member State to initiate, conduct or coordinate an investigation of a crime which affects a common interest covered by an EU policy regardless of the cross-border nature of the crime, for example in high profile sensitive cases such as the murder of Daphne Caruana Galizia in Malta. However, the necessity of this reform has not been substantiated and effectively removes control from judicial authorities over the opening of their investigations in cases affecting one Member State only.


(8) Strengthening the data protection framework applicable to Europol


A positive development of the proposal is the enhancement of Europol’s data protection framework by extending the reach of Article 3 and Chapter IX of Regulation (EU) 2018/1725 concerning the data protection framework applicable to the processing of personal data by EU institutions, bodies and agencies, to the work of Europol and explicitly adding biometric data within special categories of personal data, which was not the case. Whereas this is a welcomed reform, further alignment is necessary, not least because the EDPS’s general powers are still not aligned with the prescriptions of Article 58 of Regulation (EU) 2018/1725.


(9) Other provisions, including enhancing political accountability and parliamentary scrutiny


In addition to other minor reforms further expanding and clarifying Europol’s tasks, the proposal aims to enhance political accountability and parliamentary scrutiny by enabling the Joint Parliamentary Scrutiny Group (JPSG) to receive information regarding the matters falling under themes (1)-(4), as discussed above. Whereas this is a welcome development the proposal is a missed opportunity to further enhance political accountability and parliamentary scrutiny. Despite the establishment of the JPSG and the proposed amendments, parliamentary scrutiny and oversight remain weak. Shortcomings concern the structure and work of the JPSG, including the weak powers of the Group in the participation to and appointment of Europol’s Management Board. With the addition of new tasks to Europol, the need to ensure a better framework for parliamentary oversight and political scrutiny must be emphasised and therefore there is significant scope for improving the proposal in that respect.


Concluding remarks


The analysis above aimed to highlight that the proposal entails widespread reforms to Europol’s mandate, which transform the nature of the agency and its relationship with the Member States. The large majority of these reforms that enhance the data processing capacities of the agency have been met positively by the Council, which refined some provisions when it agreed its position on the proposal in June 2021. This massive expansion of Europol’s powers is explained; Europol’s ability to bring about results is tightly related to Member States’ input and participation and research shows the reluctance of national authorities that were, and are, sometimes not very keen to share their data with the agency. As a result, the proposal wishes to bypass such reluctance, by enabling the agency to directly ‘deal the cards’ and centralise information processing. At the same time, certain operational reforms, particularly the SIS-related ones, have been met with greater scepticism, albeit without an outright dismissal. At the time of writing, the Council is still scrutinising the proposal but a significant amount of work has already been conducted. It remains to be seen whether the Parliament will success in adding further safeguards to circumscribe these additional powers and enhance its own role.


Barnard & Peers: chapter 25

Photo credit: OSeveno, via Wikimedia Commons


No Pain, No Gain? Implementing the European Green Deal and Getting the EU Maritime Transport Sector “Fit for 55”


Killian O’Brien, DAAD Lecturer (Fachlektor), University College London

The EU has set a target of reducing its greenhouse gas emissions by 2030 by at least 55% compared to 1990 levels. This is to be achieved as part of the European Green Deal. With a somewhat unfavourable title reminiscent of a middle-aged sport group, the European Commission rolled out its much-anticipated “Fit for 55” legislative package on 14th July 2021. This bundle of legislative measures aims to secure rapid decarbonisation of the EU economy by transforming the bloc's emissions trajectory and bringing EU policies and legislation in 12 different areas into step. Given the significant impact of the transport sector in contributing to the EU’s overall emissions balance (ca. 25% of total emissions) and that emissions from this sector have been rising in recent years, it is unsurprising that many of the proposals are directed toward transport emissions. Given the sheer breadth and volume of proposals tabled, it is difficult to engage in a thorough analysis so soon, yet this short blog aims to outline some aspect of the proposed measures as they relate specifically to the maritime transport sector.

Four of the 10 proposals announced on 14th July address the maritime transport sector specifically, three being applicable to shipping and the fourth to certain ports within the European Economic Area. The three proposals related to shipping address establishing a revised system for greenhouse gas emission allowance trading (EU ETS proposal, COM(2021) 551 final), the use of renewable and low-carbon fuels in maritime transport (Fuel EU Maritime, COM(2021) 562 final) as well as a proposal restructuring the Union framework for the taxation of energy products and electricity (Energy Taxation Directive, COM(2021) 563 final). The proposed Regulation affecting ports deals with the deployment of alternative fuels infrastructure (AFIR, COM(2021) 559 final).

Arguably, the most significant proposal is for a revised Emissions Trading System Directive (EU ETS), which would include international shipping within the EU ETS framework. Under the broader applicable international legal framework (UNFCCC, UNCLOS, MARPOL etc), shipping, which is not regulated under the Paris Climate Agreement, falls under the auspices of the UN’s International Maritime Organisation (IMO). Despite having a broad mandate to regulate marine pollution, including GHGs (Article 2, IMO Convention), progress within this designated global regulator has been at times so painfully slow that the EU has now decided to take the step towards unilateral regulation and bring shipping within the EU ETS framework. Indeed, the European Commission notes at page 5 of the explanatory memorandum to the revised EU ETS proposal that “[IMO] measures are insufficient to decarbonise international shipping in line with international climate objectives”. This is not particularly novel as the EU has previously taken the lead in unilaterally lowering the sulphur content of marine fuels following a lack of progress at IMO and legally there is no obstacle to this kind of action. UNCLOS permits States to take action to “establish particular requirements for the prevention, reduction and control of pollution of the marine environment as a condition for the entry of foreign vessels into their ports or internal waters” (Art. 211 (3) UNCLOS). Indeed, broader principles of jurisdiction in public international law allow for such unilateral measures where a substantial nexus to the matter in question can be demonstrated. The significant effects of GHGs on the environment and on individuals’ health would seem to satisfy this jurisdictional criterion and the EU is therefore entitled to push forward with the creation of unilateral market-based measures in the form of ETS. Having said that, when considered from a trade policy point of view, there is some merit to calls for the EU to continue to work towards a global solution due to the threat of loss of competitiveness as well as the potential for harmful regime fragmentation and increased costly administrative burden for shipping companies and others due to, for example, multiple monitoring and reporting standards. To that end, the EU ETS Directive proposal contains a review clause to assess progress at IMO (Article 3ge).

Under the EU ETS scheme, companies buy allowances based on one allowance being equal to 1 tonne of CO2 emitted. The proposal makes it clear that shipping companies (or the party made responsible for the emissions) will not be entitled to any free allowances. Sufficient allowances must be surrenders to cover the emissions for the year. Allowances can be bought and sold as commodities so that (environmentally friendly) companies with more allowances than they need can sell them to others or keep the unused allowances in reserve. The scheme will apply to vessels of 5,000 gross tonnes and above. The applicability of the EU ETS Directive proposal rationae loci relates to emissions from all voyages between ports in the European Economic Area (EEA) and at berth in EEA, i.e. intra-EEA voyages. In addition, extra-EEA voyages, whether inbound or outbound also fall under the scope of the EU ETS proposal, albeit only 50% of the respective inbound and outbound emissions. The considerable impetus with which the European Commission is proceeding is evident in the fact that emission allowances will already have to be purchased from 2023. However, a graduated introduction of allowances is foreseen, starting with 20% of emissions in 2023, rising to 45% in 2024, 70% in 2025 and then to full 100% emission coverage for 2026. Sanctions are foreseen with fines of €100 per tonne CO2 and repeat offenders who have been non-compliant in consecutive years could even be denied entry to EU ports.

The other most significant element within this legislative package can be seen in the imposition of the first greenhouse gas intensity requirements on shipping fuels by way of the Fuel EU Maritime proposal. This proposed Regulation stipulates a limit on the GHG intensity of energy used on-board by a ship arriving at, staying within or departing from ports under the jurisdiction of a Member State (Art. 1). Ship fuels used by these ships must improve their GHG intensity by a prescribed percentage compared to a 2020 baseline (Art. 4). Once again, a graduated approach is envisaged, with ever-more stringent improvements required every 5 years until 2050. This proposed legislation will also apply to vessels of 5,000 gross tonnes and above.

The Fuel EU Maritime proposal could lead to a considerable proportion of the energy used by ships calling at EU ports being LNG and biofuels by 2035. Two primary concerns are pertinent in this respect, namely calculations suggesting that LNG, itself a fossil fuel, in fact offers minimal emissions reductions as well as releasing methane unburned during the energy conversion (combustion) process, and that many biofuels are suspected of having less than proven sustainability criteria, depending in particular on where they are produced. In addition, the requirement that shipping companies source adequate fuels, even when the ship is outside the EU, could be particularly burdensome to the point that it is excessive or disproportionate.

A significant development contained in the EU ETS and Fuel EU Maritime proposals is the strengthened recognition of the polluter pays principle in the sector. Typically, shipping companies could have been assumed to be responsible for emissions caused and the GHG intensity of a particular vessel. However, in what would appear to be a late inclusion in the draft proposals, the EU ETS and Fuel EU Maritime proposals include provisions recognising the role of the commercial operator. Under this regime, the shipping company would be permitted to make commercial operators, i.e. charterers who determine the choice of fuel, route and speed of the ship, responsible for costs of compliance with the two draft legislative proposals in question by way of an appropriate “commercial agreement”.

The third Fit for 55 legislative proposal aimed at the shipping industry (Energy Taxation Directive, ETD) relates to the taxation of energy products and discourages the use of fossil fuels by setting higher rates of taxation for fossil fuels and lower rates for renewables products thereby decreasing the relative price advantage of fossil fuels over less polluting alternatives. The material scope of the proposed ETD therefore imposes tax on bunker fuels sold in the EEA for voyages within the EEA, although Member States can choose to extend this to extra-EU voyages. As this is a piece of taxation rather than environmental legislation, unanimous approval in Council is required. It is therefore not unthinkable that certain Member States such as Cyprus, Greece and Malta, with their particular interest in the shipping industry, could block proposals or that proposals could be considerably amended to appease certain vested interests.

In respect of ports, the scope of the AFIR proposal is limited to those ports within the Trans-European Transport Network (TEN-T). Some 329 ports make up this network, which are determined on the basis of a set of quantitative thresholds. Hence, the largest ports in terms of freight volumes and passenger volumes are affected. These ports will be responsible for providing an “appropriate” LNG refuelling infrastructure as of 1.1.2025 (Art. 11) and must also provide a minimum shoreside electricity supply for containerships and passenger ships as of 1.1.2030 (Art. 9). The proposed Regulation contains mandatory national targets for the deployment of “sufficient alternative fuels infrastructure” in relation to vessels as well as road vehicles and stationary aircraft. Common technical mechanisms are foreseen (Art. 19), in addition to national policy frameworks (Art. 13) and robust reporting mechanisms (Art. 14). The European Commission’s choice of LNG as the preferred fuel-source is interesting in that full alignment between the Alternative Fuels Infrastructure proposal and Fuel EU Maritime proposal must be maintained throughout what could be a lengthy legislative process. This is illustrative of the inter-related nature of the various proposals. In any event, considering the aforementioned criticism of LNG, its largely recognised medium-term stop-gap character and the clear preference of some significant carriers for alternate fuel sources such as ammonia, it remains to be seen whether this proposal will have sufficient longevity to have the desired effect. Further discussions will almost certainly centre on avoiding carbon leakage, future proofing, clarifying the addressees of reporting obligations and, almost inevitably, the issue of funding for capital infrastructure investment projects.


This preliminary assessment has shown the generally far-reaching approach adopted by the European Commission with these proposals. The unilateral nature of EU action has been identified as legally acceptable but not entirely unproblematic from a political and trade policy point of view. At the same time, the EU has enjoyed success in driving the international regulatory agenda with a similar past strategy in relation to fuel sulphur-content limits and there is every reason to think that these measures could have the incidental effect of increasing the pace of negotiations at IMO. Of course, it goes without saying that all proposals will now be subjected to considerable scrutiny and negotiation within the Parliament and Council of the EU. It remains to be seen whether and with which measures the EU maritime transport sector can be whipped into shape to make it truly Fit for 55.

Photo credit: Łukasz Golowanow via Wikimedia commons

Barnard & Peers: chapter 22

Monday, 26 July 2021

I never promised you a rose garden: a response to Dominic Cummings


Professor Steve Peers, University of Essex

Last night Dominic Cummings, after a long day of asserting his own brilliance, tweeted about me as one of the ‘charlatan campaigning lawyers on Twitter’, along with Jolyon Maugham, David Allen Green and Jessica Simor. In his view, we were ‘*really* bad at politics’ – his ‘proof’ being a tweet from David Allen Green about the likelihood of the revocation of the decision to leave the EU, and an unsubstantiated assertion about us claiming that Dominic Grieve, the former Attorney-General, would “beat” Vote Leave in 2019.

I don’t know why Mr Cummings paused from contemplating his own greatness to have a dig at those of us who point out that he often gets the law wrong. It’s as if Thanos paused mid-snap to pull the wings off a fly. Be that as it may, here’s my response. I can only speak for myself – although you can see David Allen Green’s response here. Furthermore, this is a fresh blog post: unlike Mr Cummings, I haven’t gone back and amended an old blog post to pretend that I’ve said all this months ago already.

First, the general point: judging me on whether I’m ‘bad’ at politics is as relevant as judging me for my skills at pole vaulting or astral projection. I tweet political views (among other views), but I’m not a politician, a SpAd, a spin doctor, or a political scientist. I’m already aware that my tweets don’t swing millions of votes (whose tweets do?).  My main focus, as it says in my profile, is tweets about the law – particularly EU law, human rights law, and world trade law. In that context, for instance, I have pointed out falsehoods or dubious claims from Mr Cummings about Vote Leave’s breach of electoral spending law, not triggering Article 50, disinterest in international trade, not challenging David Cameron’s claims, and the ‘Turkey is joining the EU’ slogan.

I thought about going into politics when I was younger, but there’s a good reason I didn’t: because I knew I might end up having to tout the equivalent of some innumerate slogan on the side of a bus, or backstabbing EU citizens, in order to further my political career.  Most of my political tweets are about what I think should happen, but I recognise that my views will not necessarily win votes. A more utilitarian approach would be to jettison some less popular policies in order to increase the chances of victory so that some policies could be implemented. I can see the force of that argument, although there comes a point when a political party could end up throwing its raison d’etre out with the bathwater; and the tactic might fail on its own terms, as core voters desert the party, with intended converts recognising a party uncomfortable in its own skin, and voting for a party that actually believes those things instead. Reasonable people can differ in how to deal with this tension in relation to individual policies.

Next, the specific point: as I’m not a political scientist, I don’t generally make forecasts about political events. On the particular issue of staying in the EU, though, far from predicting that the UK would revoke the notification to withdraw from the EU, I never thought that was very likely once the referendum vote went for Leave (for evidence, see my tweets here, here, and here). On Dominic Grieve in particular, none of my tweets mentioning him talk of him ‘beating Vote Leave’ (and note my tweet saying it was an open question whether he would win his own constituency). I did point out that the Benn Act on requesting an extension of EU membership was valid – because it was, and I note that the Prime Minister indeed ultimately complied with it. But still, it’s possible that somewhere in the second half of 2019 I made some political predictions on Twitter that didn’t turn out. Knock yourself out finding them, Dom.

This brings me to the relative importance of law vs politics. It’s a common theme of Cummings that politicians should just do as they choose, ignoring ‘yapping’ lawyers. In fact, I agree that on essentially political questions, politics will take precedence over law. I’ve often said that it was a category error to assume that somehow a court would rule that Brexit was ‘illegal’: as long as the political will existed, it would find a way around court judgments (such as Parliament passing an Act to trigger Article 50, after the Supreme Court ruled that the government needed Parliament’s assent). Deus ex curia is not a Brexit thing.

But law can still impact upon how decisions get made, and in any event to my (legal) mind there’s intrinsic value to getting the law right – in much the same way that Mr Cummings often does not. That’s not only my day job, but it’s also one of the key reasons why I am on Twitter: because I believe there is social value to bolstering public understanding of the law.

On the particular issue of what would have happened under different scenarios, going by Cummings’ fanfic about what would have happened if he had been negotiating all along, his negotiating prowess would have ended up with Varadkar babysitting his son, Barnier driving him to Barnard Castle, and every Irish village setting up a shrine to him – next to the new statue of Oliver Cromwell.

We’ll never know. But I have a less narcissistic view of my own abilities than Cummings. (Who doesn’t?) And I have a more sanguine view of how easily nexus events can lead to branches of the sacred timeline.  If I had been a Mandelsonian spectre haunting British politics for the last twenty years, the UK would not have entered the Iraq War. But would that have saved the Labour Party? It won in 2005, after all; so the Iraq War did not immediately destroy it. Perhaps I would have ordered Gordon Brown to go back and glower at Gillian Duffy, not apologise to her. But would that really have helped Labour?  There could have been a transitional period for workers from Poland etc coming to the UK. But that would have just kicked the can on this issue to 2011.

The LibDems could have handled the coalition differently; but their previous left and anti-politics voters might have rejected them nevertheless just for entering it. Cameron could have promised a referendum on Turkey joining the EU; voters might not have believed him. He might have refused to hold a referendum in the first place; but Leavers would not have just shrugged and gone away. (If Cummings thinks I’m bad at politics, he should meet my Reply Guys who think that Cameron could have stayed on and refused to implement the referendum result in June 2016). Corbyn could have done more referendum events; but no one was paying him that much attention at the time. The Remain campaign could (correctly) have argued more that austerity stemmed from Tory government policy, not the EU; but how on earth could Cameron and Osborne, leading the campaign, credibly have delivered that message?

Cummings says that he would not have triggered Article 50, but tried to negotiate with the EU beforehand, even though the EU policy was to refuse to do so. He says that he would have implemented a withdrawal via domestic law, but he would have had to convince Parliament. Doing so without Parliament might have led to litigation: it’s tempting to imagine Gina Miller Variants frustrating Cummings in multiple timelines. And declining to send the Article 50 notification might have scuppered Conservative party plans to hoover up UKIP votes.

A final point about being ‘really bad at politics’. One of the few things that Mr Cummings and I agree about is that the government’s response to the pandemic has largely been poor. But only one of us played a key role in putting that government in power – complete with a Prime Minister he calls the ‘trolley’, whom he immediately wanted to depose. If that’s being good at politics, I wonder what being really bad at it is.   

Photo credit: MOtty, via Wikimedia Commons

Saturday, 17 July 2021

Headscarves at work: the Court of Justice clarifies when employers can ban them


Professor Ronan McCrea, University College London


Four years ago in the cases of G4S and Micropole, the Court of Justice issued its first major rulings on the compatibility of workplace policies that restrict the wearing of religious symbols by employees with the Framework Directive on discrimination in employment (Directive 2000/78).


In these rulings, the Grand Chamber of the Court held that the imposition on employees of a policy of visible neutrality of religion or belief amounted to indirect rather than direct discrimination provided that it was consistently applied to all forms of religion or belief. Policies that targeted the symbols of a particular faith, on the other hand, were to be regarded as directly discriminatory.


An obligation of visible neutrality, the Court held could, if sufficiently narrowly tailored, satisfy the test for indirectly discriminatory measures (legitimate aim, pursued by proportionate and necessary means) but was unlikely to be able to meet the test for directly discriminatory measures (genuine and determining occupational requirement).


The judgement was criticised both in the media and by some legal scholars. Eleanor Spaventa worried that the invocation of the right to run a business in Article 16 of the Charter of Fundamental Rights as a reason favouring the compatibility of restrictions on employee dress meant that Member States may not be able to take an approach more protective of religious freedom. She also argued that the Court had paid insufficient attention to the fact that neutrality rule might have a more pronounced effect on those of particular ethnic background or a particular gender. Similarly, Joseph Weiler faulted the Court, inter alia, for failing to provide adequate reasons as to why the goal of neutrality was a sufficiently weighty goal to justify the third limb of the proportionality test.


When these issues came before the Court again this year in the form of Article 267 references in two cases in Germany, the cases were allocated to the Grand Chamber. However, this was not an indication of a change of approach on the part of the Court. This week’s ruling in the (joined) cases, largely reaffirms the approach previously adopted by the Court although the ruling does provide some additional material that, at least partially, addresses some of the concerns of its critics.


Both cases related to employees disciplined for wearing a hijab at work. In IX v WABR eV (‘IX’) WABE, a non-denominational provider of child care services disciplined an employee who worked as a special needs carer for violating its internal rule that required those whose roles involved dealing with parents and children to observe strict political, religious and philosophical neutrality ‘in order to guarantee the children’s individual and free development with regard to religion, belief and politics’. This rule was not applied to those whose work did not bring them into contact with parents and children.


In MH Müller Handels GmbH v MJ (‘MJ’) a woman employed as a cashier and sales assistant was disciplined for wearing a hijab at work in violation of a workplace rule that prohibited the wearing of ‘conspicuous, large-sized signs of any political, philosophical or religious beliefs’.


There was some overlap in relation to the questions submitted by the national court in each case. In IX the Arbeitsgericht Hamburg asked the Court of Justice:


(a) Whether an instruction not to wear visible signs of religion or belief from an employer to an employee who, due to her Muslim faith wears a headscarf, should be regarded as direct or indirect discrimination ‘on grounds of religion and/or gender’ (Confusingly, the English version of the judgment refers to discrimination ‘on grounds of gender’. The French version refers to discrimination ‘fondée sur le sexe’ while the English version of the relevant Directive (Directive 2006/54) refers to ‘less favourable treatment on grounds of sex’. However, the Court refused to engage on the issue of discrimination on this ground on the basis that the referring court had limited its question to Directive 2000/78 which does not address this matter).


(b) Whether indirect discrimination on grounds of religion and/or gender can be justified when the employer is seeks to impose a neutrality requirement in order to meet ‘the subjective wishes of its customers’,



(c) Whether the Directive 2000/78 and Article 16 of the Charter preclude national rules that grant greater protection to freedom of religion or belief (in the light of the provisions of Article 8(1) of the Directive that allows Member States to have provisions more favourable to the principle of equal treatment than those laid down in the Directive.


In addition to asking whether national rules more favourable to equal treatment on grounds of religion were permitted (question c above), in MJ the Bundesarbeitsgericht also asked the Court of Justice:


(d) Whether the requirements of the directive in relation to justifying indirect discrimination are met by a rule that pursues a policy of neutrality by prohibiting only prominent and large-sized symbols of religion or belief (rather than all such signs).


In relation to question (a), the CJEU reiterated its previous ruling that the prohibition on discrimination on grounds of religion or belief applies equally to religious and non-religious philosophical and spiritual beliefs and that ‘since every person may have a religion or belief’ a rule that covers any manifestation of religion or belief without distinction ‘treats all workers of the undertaking in the same way by requiring them, in a general and undifferentiated way, inter alia, to dress neutrally’. Accordingly, such a rule does not constitute direct discrimination notwithstanding that such a rule is capable of causing ‘particular inconvenience’ to workers ‘who observe religious precepts requiring certain clothing to be worn’. In IX the Court noted that an employee wearing a cross had been required to remove it, something that appeared to confirm that the rule was applied across the board (though it noted that verification of this matter was a factual issue for the national court).


Having concluded that the rule in IX was potentially indirectly discriminatory, the Court went on to reiterate its finding in G4S Solutions such the indirect discrimination could be justified in the light of Article 16 of the Charter ‘in particular where the employer involves in its pursuit of [the neutrality] aim, only those workers who are required to come into contact with the employer’s customers.’ However, it added a slight gloss on the G4S ruling by noting that a ‘mere desire of an employer to pursue neutrality’ would be insufficient. An employer must, the Court held, demonstrate a ‘genuine need’ for such a policy, ‘taking into consideration, inter alia, the legitimate wishes off those customers or users and the adverse consequences that the employer would suffer in the absence of that policy.’ In MJ it further clarified that ‘the prevention of social conflicts’ in addition to customer preferences may also form the basis of such a demonstrated need.


Interestingly in relation to question (d), in MJ, the Court came to a slightly different conclusion in relation to the prohibition on conspicuous, large-sized signs. This rule is, the Court noted ‘liable to have a greater effect on people with religious, philosophical or non-denominational beliefs which require the wearing of a large-sized sign such as a head covering.’ Noting that where a rule is based on a criterion that is ‘inextricably linked to a protected ground’ it must be regarded as being directly discriminatory the Court then states that ‘where the criterion of wearing conspicuous, large-sized signs of political, philosophical or religious belief is inextricably linked to one or more specific religions or beliefs, the prohibition will mean that some workers will be treated less favourably than others on grounds of their religion or belief, and that direct discrimination, within the meaning  of [the Directive] may therefore be established’.


This seems rather confusing. The Court had already noted in its conclusion that the rule in IX was indirectly discriminatory that statistically, the ban applied almost exclusively to Muslim women. Yet in MJ it appears to say that because a ban on large sized signs might have a greater effect on those whose beliefs require the wearing of a head covering it may be ‘inextricably linked to a protected ground’ and therefore may be direct discrimination.


Perhaps what the Court is saying is that a ban on conspicuous, large signs amounts to a hidden targeting of the signs of a particular faith and therefore amounts to direct discrimination. This would be consistent with its previous rulings in G4S and Micropole where an instruction that focused on the symbols of one faith (‘no headscarf next time’) was found to render it more likely that a restriction would be found to be directly discriminatory. Unfortunately, the tortured way it expresses this point is liable to cause confusion in the future.


In relation to the question posed by both referring courts as to whether national provisions more favourable to equal treatment on grounds of religion or belief or freedom of religion and belief are permissible (question (c)), the ruling was clear. The Court stated that the Framework Directive ‘did not itself effect the necessary reconciliation between freedom of thought, conscience and religion and the legitimate aims that may be invoked to justify unequal treatment (…) but left it to the Member States and their courts to achieve that reconciliation’. The Directive has left a ‘margin of discretion to Member States taking into account the place accorded to religion and belief within their respective systems’. While this margin goes ‘hand in hand with supervision by the EU judicature’ the Court noted the lack of consensus on these issues amongst the Member States. Accordingly, ‘national provisions protecting freedom of thought, belief and religion , as a value to which modern democratic societies have attached great importance for many years, may be taken into account as provisions more favourable to the protection of the principle of equal treatment within the meaning of Article 8(1) of the Directive.


 This ruling largely reaffirms the rulings in G4S and Micropole. Despite some confusion, it appears that neutrality requirements, if genuinely and systematically applied to all forms of religion and belief, are in general to be regarded as indirectly, not directly discriminatory. Customer preferences and the need to avoid social conflict may provide the justifications for any indirect discrimination that such neutrality policies involve.


It is disappointing that the Court chose not to address the issue of sex/gender discrimination on the narrow and technical grounds that that referring court had only mentioned Directive 2000/78 which does not cover this issue. With the national court having raised the issue of sex/gender discrimination, the CJEU could have given guidance as to how the relevant legislation in this area should be interpreted.


The Court does seem to have sought to address some of the critics of its previous rulings. This ruling has greater emphasis on the need for employers to demonstrate the necessity of a neutrality policy and the, admittedly confusingly phrased, conclusion that bans that selectively target large symbols are potentially directly discriminatory would seem to be a step towards stricter control of bans that amount to hidden targeting of one faith.


The Court has also addressed the concern that its invocation of Article 16 as a justification for neutrality policies precluded Member States from taking more protective measures.


Indeed, this underlines how the Court has maintained its cautious approach in its area. This seems justifiable. Amending the Framework Directive would need the unanimous consent of all Member States. If the Court were to get its interpretation wrong the chances of remedying this politically are very low.


As I have written before, Europe is in the midst of unprecedented religious change. After centuries of dominance in most Member States, Christianity is in rapid decline. Non-belief is in rapid growth, as is the Muslim population of Europe. Any one of these changes alone would produce endless unanticipated consequences. We do not have a store of precedents that tell us how coexistence is best achieved in this context. There are deep divisions between Member States on these matters.


Religion is hard for the law to regulate as it is both a usually immutable identity and a series of often controversial, chosen beliefs. The treatment that is appropriate for an immutable identity is often inappropriate for a set of beliefs and vice versa. A religious symbol can be worn to express a belief or to reflect one’s identity or simply as a matter of habit yet it is hard to design a law that can distinguish between these different situations. Objection to a religious symbol at work may often be the result of pure bigotry, but it may sometimes be a legitimate reaction to a controversial belief.


The management of religious diversity is equally tricky. For some states, coexistence is best achieved by requiring everyone to hold back on expressing their religious identity in shared contexts and by seeking to stress a shared a-religious citizenship. Other states have felt it best to protect religious expression and identity in a wide range of contexts. Each approach has its fans and its critics and the data as to which approach is better at producing inter-religious friendships, mutual respect and widespread adherence to liberal values is very mixed. In these circumstances it would be immodest in the extreme for the Court of Justice to decide it knew the best approach and to impose it on all 27 states.


That said, values are important and total deference on the part of the Court would be equally inappropriate. Opportunistic embrace of secularist principles by those with exclusionary agendas is a recurrent problem. A duty of neutrality must not be used as a disguised means of excluding minorities such as European Muslims. Both its previous rulings and IX and MJ see the Court taking incremental but significant steps that seek to prevent this by ensuring that any neutrality policy imposed by employers is applied consistently to all forms of religion and belief rather than targeting one faith. In Micropole, open targeting of the symbols of one faith was constrained, in MJ, the Court appears to move towards restricting hidden targeting to a greater degree (though it ought to have expressed itself more clearly on this point).


This will not bring an end to exclusionary practices in the workplace and falls short of the desires of some commentators but in the changing and unpredictable situation that Europe finds itself in, this is probably as big a step as it is wise for a multi-national court to take.


Barnard and Peers: chapter 9, chapter 20

Photo credit:




Saturday, 3 July 2021

Is the Northern Ireland Protocol unlawful? Analysis of the High Court judgment


Sarah Kay, human rights lawyer 

Brexit has not been kind on Northern Ireland. Then again, it was never meant to be. Five years after the referendum, almost to the day, the High Court in Belfast released a decision of intense political nature, with strong echoes of Miller 1, concerning a legal challenge to the Northern Ireland protocol to the withdrawal agreement. Northern Ireland is often described as sitting on a political powder keg, which was on the brink of exploding during the Easter riots. Egged on by political division and the ever so difficult nature of Northern Irish sense of belonging(s), Brexit managed to fulfil the one expectation everyone had predicted: expose every systemic failure.


The oral arguments, presented on behalf of two main applicants - Clifford Peeples, objecting to the establishment of a border in the Irish sea, and unionist members including (now former) First Minister Arlene Foster, opposing the Protocol itself as damaging a unionist position by furthering a separation of Northern Ireland from Great Britain, often were nonsensical. Pushing forward that the European Union was akin to the Vichy regime in its “occupation” of Northern Ireland has no place in a courtroom. Stating that Brexit was never debated in Parliament was simply and painfully inaccurate - and immediately rebuked by Lord Justice Colton (“It was not only debated in Parliament, but outside of Parliament, too, and for some time”, he replied during the hearing). For purposes of time, and clarity, this will focus on two important points of contention.


Acts of Union 1800


As my esteemed colleague Anurag Deb helpfully pointed out in a thread, every single point is politically contentious. Returning to the Acts of Union shows that the core focus would not just be the Withdrawal Agreement and the “Northern Irish backstop”: it was the constitutional existence of Northern Ireland. The question was whether the Protocol overrides the Acts of Union. The decision relied heavily on the very nature of constitutional statute and precedent laid by the Supreme Court. The opposition to articles 6 and 10 of the Acts of Union do not relate to the access of goods and services from Northern Ireland to Great Britain, but from Great Britain to Northern Ireland. The imbalance, that would have been fixed by the Internal Market Bill (in its original form), tips the scale in favour of the EU market. It is worthwhile at this stage to highlight that the implementation is still in progress, therefore leaving some unknowns at this stage of a judicial review. But the decision stands: the Protocol in itself does not create those tariffs, rather provides for payments of tariffs to be made if there is a risk of goods being diverted to the EU (see para. 56 of the judgement)


We then reach the issue of hierarchy of statutes in matters of constitutional standing: the impact of Miller 1, in this situation, can’t be overstated. To declare a constitutional statute unlawful, in the event of conflict, one must understand how said statute came into force. The Protocol was itself negotiated as part of the Withdrawal Agreement 2019, and whilst treaties are not up for review, this is where parliamentary sovereignty comes to assert its fundamental nature: the very approval of the Withdrawal Agreement and all appendices therein by Parliament; those issues were effectively considered, at large, by Parliament, over the course of several votes. Resting on the “forever” qualifier in the Acts of Union omits that the United Kingdom has known various iterations since 1800, due to claims of independence, least of them the emergence of a separate Irish Free State in 1922. There was no loss to be expressed in the fight for self-determination in Ireland. The Acts of Union must then be inclusive of the nature of partition. It has not escaped commentators that this year marks the centennial of said partition.



Good Friday Agreement and fundamental rights


In the very early days following the 2016 referendum, the question as to whether Brexit would breach the Good Friday Agreement (GFA) was raised, and would continue to be raised during negotiations. The European Union reiterated at every given turn that it would never compromise on the Northern Irish peace process, which led to the need for the backstop. Both sides, in this case, argued that they brought their position in order to protect the GFA. The argument over the Article 3 of Protocol No. 1 to the European Convention on Human Rights, helpfully referred to as A3P1, poses the difficult question of how the European Union could continue to implement certain laws and Directives in and to Northern Ireland, without Northern Ireland having access to direct representation in the European Parliament.


The hearing saw many extremely unhelpful parallels. The Vichy analogy notwithstanding, the appellant went as far as to push a US analogy, reproducing the “no taxation without representation” motto. It had been presented several times in Stormont that the Protocol would lead to a diminution of voting rights. The Court considered that the negotiations, especially on future relations between the UK and the EU, would be made in the presence of MPs - that Northern Ireland can vote for in addition to its MLAs. The long debated “democratic deficit” is a political position for which the Court might be later asked to revisit this judgment, should it be appealed. It was considered through the Fourth Recital of the Protocol, highlighting the importance of ensuring there is no diminution of rights arising from the withdrawal agreement, now an interpretation on the interplay between legal and political issues, a perfect summary of this case.


The Court considered that the constitutional nature of Northern Ireland was not changed in the absence of a referendum - which is provided for in the GFA under the principle of consent. The GFA is not incorporated into domestic law. The Court further notes that, generally, the arguments are opposing the arrangements made under the Withdrawal Agreement, rather than the Protocol itself. Perhaps this is the one element that is more widely misunderstood: the Protocol protects the GFA, but does not seek to incorporate it into domestic law. The Protocol is a shield between the Withdrawal Agreement and the GFA. It is therefore very possible this case could be considered a proxy for directly addressing the Withdrawal Agreement. Remorse could be worse than regret, as one MP - Ben Habib (Brexit Party) voted for the Agreement, and is now party to this case. The grievances should then be replaced outside of the courtroom and into the way the arrangements are being negotiated politically, outside of the scope of judicial review, and into the political arena.


Judicial review was refused on all grounds presented to the Court


This decision is not of strictly constitutional interest, but then, when is public law ever just a matter of legal arguments in courtrooms held in remote locations, never to be accessible? This decision - and the motivation behind the application for judicial review - is about identity. In a region where nothing is as contested, conflicted, and tested on a regular basis, it is unsurprising the decision was met with attacks on the judge, Lord Judge Colton; on the barristers presenting the arguments, including former Attorney General for Northern Ireland, John Larkin QC; and by large misinterpretations of the Protocol.


It is July, and protests and riots over the DUP’s interpretation of the role of the Protocol have already led to threats of violence, threats of instability, and threats of attacks against anti-Brexit figures in Northern Ireland. The Protocol, aimed at, by and large, lessening the intense blow Theresa May’s hard Brexit would have inevitably inflicted upon Northern Ireland. The grievances aimed at the Protocol are two-fold: one, they misunderstand that all those concerns were not created by the Protocol, but by Brexit itself. The DUP, having allowed Theresa May to remain in power by entering into a confidence arrangement with the Conservatives, holds little political legitimacy to now oppose the consequences of May’s policies. Two, it is about the very nature of the presence of Northern Ireland as a devolved region, part of the United Kingdom. Whilst not fully phrased as such, the fear of a reunified Ireland - actively discussed and campaigned for since summer 2020 - feels like a death toll for unionism. The Protocol is a protective shield between the Good Friday Agreement and Brexit. The principle of consent will be respected; Article 2 of the Protocol guarantees no diminution of rights. Lest we forget that, in 1998, the DUP refused to vote in favour of the Good Friday Agreement. It is now catching up to them.

Photo credit: via Wikimedia Commons