I. Introduction
As the second wave of the SARS-CoV-2 pandemic is flaring up in the European Union (EU), Footnote 1 the debate regarding the EU’s role in these situations continues. The threat and consequences of COVID-19 are expected to affect the daily lives of European citizens for many years, Footnote 2 and the economic effects will be widespread, Footnote 3 permanently transforming many industries. Footnote 4 From a legal perspective, it is interesting to see that the EU’s obstacles to providing leadership in health crises such as the SARS-CoV-2 pandemic do not appear to be linked to a lack of competence; the Treaties already provide a strong basis for EU action under such circumstances. Footnote 5 Nevertheless, the response to COVID-19 was initially primarily national, with the EU having to defend its seat at the table.
The regulatory response to the pandemic also reignited discussion regarding the future role of the EU in health law and policy. As is often the case in debates on EU competence, Footnote 6 the choice appears to be a dichotomous one between a return to Member State-based health politics and policies or the creation of a clear EU mandate for EU legislation and coordination of health policy, including, but also beyond, future pandemics. Footnote 7 Those who believe that a larger health-related role for the EU is preferable consider the establishment of a European Health Union (EHU) – modelled after initiatives such as the Energy Union – to be a desirable next step. Footnote 8
This special issue engages with the emerging debate on an EHU by offering insights pertaining to Treaty changes that may or may not be necessary to facilitate it and the sectoral reforms that these may entail. Our contribution engages with this topic by highlighting how an EHU, if established, will have to play a role in the EU’s difficult balancing act between protecting individual fundamental rights and collective societal interests, in particular between the individual right to privacy and the collective interest of public health. We take as our starting point that the EU already plays a role in balancing fundamental rights with health law and policy, Footnote 9 and we consider the regulatory frameworks and privacy safeguards that an EHU will have to abide by. Footnote 10
The motivation for this focus is twofold. First, during the current pandemic, Member States have adopted different types of measures that curtail the right to privacy to various degrees. Footnote 11 For example, the Slovakian parliament passed an amendment – justified under a state of emergency – that gives the Public Health Office access to location data from mobile phones in order to track individuals who reported positive for the virus.Footnote 12 In France, a similar amendment was rejected, despite France also having declared a state of emergency.Footnote 13 As we will show in Section III, this disparity between Member States extends to the use of certain smartphone technologies that collect, store and process personal data.
Second, we expect that new public health-related technical innovations will continue to be developed to curtail the current pandemic and to prepare for future epidemiological crises. Specifically, we foresee a push towards the integration of data-driven “persuasive technologies”Footnote 14 into smartphone apps to persuade or “nudge” individuals into conforming with public health measures. The use of apps in this way may negatively affect the right to non-interference with one’s decisions (“decisional privacy”), especially when it is difficult for individuals to observe how they are being nudged. This dimension of privacy is increasingly relevant in relation to public health regulation, as public health measures that target people’s decision architectures are becoming more common.Footnote 15 Legally, there is recognition of this dimension of privacy, especially in case law, but at the same time gaps in regulation persist.
In order to understand the potential role of an EHU regarding these issues, we will first highlight the distinctions between informational and decisional privacy (Section II). We will then consider the implications of the fact that decisional privacy is not (yet) protected by the EU’s privacy legislation to the same extent as informational privacy. If part of the mission of the EHU is to assist EU citizens – and their Member States – in effectively responding to future public health emergencies, the EHU will increasingly have to consider measures that affect both informational and decisional privacy. This raises the question of whether existing EU privacy legislation is equipped to guide the EHU on the issue of decisional privacy. As part of this discussion, we will consider some of the measures that have been taken in the EU to protect the public from the spread of SARS-CoV-2 through the use of apps, and how these apps may evolve in the future (Section III).Footnote 16
II. Legal protection of informational and decisional privacy
The right to privacy has existed for centuries, long before the creation of modern-day technology. In Jewish law, privacy law is built on the doctrine of hezzek re’iyyah (injury caused by seeing), which recognises that one should be protected even from the possibility of observation.Footnote 17 Our understanding of privacy, and the ways in which it may be infringed, continues to develop with society.Footnote 18 The rise of “Big Data”Footnote 19 has been formative for our understanding of privacy as “informational privacy”, which centres on those aspects of privacy that relate to controlling access to one’s personal information.Footnote 20 As a result, the right to data protection is now often seen as an essential prerequisite to the right to privacy.Footnote 21
Since 2018, the EU has regulated informational privacy through the General Data Protection Regulation (GDPR).Footnote 22 The GDPR is widely regarded as representing the state of the art with respect to data protection.Footnote 23 It applies to any data controller or processor that offers goods or services to “data subjects” (the term used to refer to individuals whose data are processed) within the European Economic Area (EEA).Footnote 24 This includes data controllers and processors who are not located in the EEA themselves but whose data subjects are.Footnote 25 Many jurisdictions outside the EU/EEA have used the GDPR as a model for their own legislation.Footnote 26 With respect to informational privacy, the GDPR’s provisions on the rights of the data subject are of central importance; they include transparency,Footnote 27 access,Footnote 28 rectification and erasureFootnote 29 and the right to object,Footnote 30 as well as conditions that are placed on controllers and processors with regards to holding personal data, including pseudonymisation,Footnote 31 securityFootnote 32 and the assignment of a data protection officer.Footnote 33
While the right to data protection is not absolute in the EU, especially in light of the importance of data for national security, the European Data Protection Supervisor (EDPS) has an oversight role in these areas, and the European Court of Justice has found a broad scope of application for the Charter protections in this area.Footnote 34 An important issue in this context, which is especially relevant for the measures adopted during the pandemic,Footnote 35 is the question of “interoperability” between information systems. While it is important for Member States to be able to share information, especially in light of the freedom of movement of EU citizens and the borderless Schengen area, such exchange is subject to strict conditions.Footnote 36
The issue of decisional privacy, which, as mentioned, relates to non-interference in one’s decisions and actions,Footnote 37 is less pronounced in EU privacy legislation, although it is increasingly underlined in the jurisprudence of the Court of Justice of the European Union (CJEU) and especially that of the European Court of Human Rights (ECtHR).Footnote 38 Already by the 1970s, the ECtHR expanded its interpretation of Article 8 of the European Convention on Human Rights (ECHR) so as to include a positive liberty, in addition to the right to be free from arbitrary interference from authorities.Footnote 39 This Court interpreted this positive liberty as imposing a positive obligation on the state to protect individual privacy in, among others, horizontal relationships (ie between private parties), including specifically in the area of data protection.Footnote 40 Extending this to the medical sphere,Footnote 41 the concept of “decisional privacy” and its protection may be said to already be part of the ECtHR’s jurisprudence, given the centrality of personal autonomy within Article 8 ECHR.Footnote 42 However, similar safeguards have not yet been written into legislation, specifically with respect to the EU’s GDPR.Footnote 43 This means that the regulatory framework for decisional privacy, such as it is, has not (yet) benefitted from parliamentary debate as to the scope and form of regulation, or as to its relationship with other societal aims.
In public health practice, decisional privacy may be said to be under pressure due to two co-emerging developments. The first development is the interest in the use of “nudges”Footnote 44 to promote public health.Footnote 45 The idea of nudging is to alter choice situations in ways that predictably change people’s behaviour without reducing the available choice options or altering people’s economic incentives. “Good” nudges should be transparent to the person who is being nudged, easily opted out of and encourage behaviour for which there is good reason to believe it will benefit the individual.Footnote 46
The use of nudging in public and private health has been hotly debated, as it can be argued that nudging constitutes a form of manipulation, or at least presents a risk of manipulation.Footnote 47 Despite this and other criticisms, the practice of nudging is becoming mainstream and increasingly accepted by policymakers.Footnote 48 Consider, for example, the debate around obesity and healthy eating and the suggestion to feature healthy food items more prominently in supermarkets. This subtle change in the presentation of food choices is expected to nudge people to buy healthier food items without actually restricting individuals in their food choices. Similarly, placing footprints or arrows on the floor leading to a staircase instead of an elevator has been demonstrated to be an effective nudge towards taking the stairs.Footnote 49
Alongside a general move towards nudging in various areas of public policy, the second relevant development is the rise of data-driven technologies that use persuasive techniques such as personalisation and tailoring.Footnote 50 These kinds of technologies can provide powerful “hypernudges” by personalising choice architectures in real time, using behavioural, preference and contextual data from the individual as well as from groups that are deemed relevant to the individual.Footnote 51 Examples of hypernudging include personalised advertisements that Facebook presents to its users and the way in which Google varies its search results ranking depending on who is performing the search. In addition, hypernudges are increasingly implemented in self-tracking and self-improvement apps, where choice situations may be framed differently for different individuals, based partly on a person’s preferences and prior behaviour and partly on contextual information about the person’s present situation (eg time of day, weather, social environment and so on), in the hopes of steering that person towards performing a certain behaviour.Footnote 52 Thus far, these kinds of apps typically try to promote behaviours such as healthy eating, exercising and recycling, but it is easy to imagine their application extending to the promotion of personal hygiene, social distancing or home quarantining in the fight against SARS-CoV-2.
III. Futureproofing EU regulations for decisional privacy
1. Use of apps during the SARS-CoV-2 pandemic
The use of apps during the SARS-CoV-2 pandemic started during the first wave and has since become widespread. Broadly speaking, the use of apps has been aimed at three public health goals. Footnote 53 First, there are contact tracing apps aimed at quickly detecting potentially infected individuals so that infections can be controlled and contained. These apps have raised informational privacy concerns in many Member States. In the Netherlands, for example, the Dutch Data Protection Authority (DPA) heavily criticised a first initiative by the Ministry of Health, Welfare and Sport for developing such an app, questioning its necessity, the legal foundation for its use and the extent to which GDPR safeguards would be met.Footnote 54 Similarly, there were public debates in France about privacy in relation to their “StopCovid France” app, which obtains contact data through the use of low-energy Bluetooth technology and stores this data on a central server, despite the European Data Protection Board recommending a decentralised solution where the data are only stored on the devices themselves. Footnote 55
A second category of apps focuses on information dissemination with the aim of supporting self-diagnosis. The Spanish “AsistenciaCOVID-19” is an example of such an app. With the user’s consent, the app can also collect geolocation data. According to officials, these data will only be used to ensure that individuals get the right information for the region they are residing in, but these data will nevertheless be stored on a centralised server until the end of the health crisis (after which the data will be anonymised and made ready for statistical analysis). Footnote 56
The third use for apps concerns surveillance aimed at enforcing quarantine orders for those already infected or for those at risk of infection. An example of this is the “Home Quarantine” app launched by the Polish Ministry of Digital Affairs on 24 March 2020.Footnote 57 The app is meant for people required to go into home quarantine after returning from abroad or having been diagnosed with COVID-19, Footnote 58 and it monitors their compliance with the home quarantine order by frequently requesting selfies to be uploaded with geolocation data from home. Footnote 59 This app invariably also triggered informational privacy concerns as it collects personal data, including Global Positioning System (GPS) location data, and stores them in a centralised data store, where these data can be accessed, processed and analysed by third parties, potentially even for different purposes than originally intended.Footnote 60
Together, the examples above show a considerable eagerness from Member States to use data-collecting apps for various purposes during the SARS-CoV-2 pandemic, and they highlight the importance of the GDPR in protecting EU citizens’ informational privacy. As the pandemic continues and the willingness of individuals to comply with restrictive measures can be observed to dwindle, we foresee that Member States will look for new ways to increase compliance, in particular through the use of apps that utilise persuasive techniques to nudge individuals to exhibit the desired behaviours.
One way to go about this would be to extend existing apps with persuasive techniques. For example, consider how contact tracing apps could actively adapt the language they use to be more authoritative (eg about having to go into home quarantine) when an individual continues to receive messages about being exposed to infected individuals over time. Or consider how the Polish Home Quarantine app could tailor its language and framing based on the information they already have of the individual, such as age, sex and place of residence.
Another strategy could be to require commercial parties to include public health measures in their existing apps. Here, one may think of requiring Google Maps to suggest walking routes aimed at promoting social distancing rather than efficiency. Or having dining apps restrict options for the elderly to the exclusion of restaurants known to have hosted an infected individual in the recent past. Or having fitness apps recommend workouts with bodyweight instead of with free weights or machines when there is a known local outbreak of the virus in the vicinity, in the hopes of nudging people to exercise at home instead of at the gym. If we understand the right to decisional privacy as both the negative freedom to make personal decisions and the state’s positive duty to ensure that citizens can make meaningful decisions,Footnote 61 we can see how these types of hypernudges could affect people’s decisional privacy through subtle interference with their decision-making. This raises the question as to whether the existing legal framework would be sufficient to guide Member States, and potentially the EHU, in the use of these apps.Footnote 62
2. The future of decisional privacy in public health
As witnessed during the first wave of the SARS-CoV-2 pandemic, Member States may diverge in how they approach the regulation and use of apps.Footnote 63 Despite the EU’s active role in data protection, most Member States took action without consultation of, or coordination by, the EU. Footnote 64 Only in the last weeks of June 2020, once some national borders started to reopen, did the Member States, with the support of the Commission, agree on a framework that would allow for the inoperability of national contact tracing apps in line with EU data protection law. Footnote 65
Moving forward, we envisage a role for the EHU in coordinating the use of smartphone apps in order to promote a more unified approach to taking public health measures in the EU, especially with an eye on health crises such as the current SARS-CoV-2 pandemic. Insofar as these measures will involve apps that incorporate forms of hypernudging, as we have suggested they might, the EHU will need a regulatory framework governing decisional privacy that can guide its decisions and recommendations, in parallel to the way in which the GDPR would guide the EHU with regard to informational privacy.
Although a full exploration of the various safeguards that such a framework should include is beyond the scope of this contribution, we want to point to two key concerns that currently are not sufficiently dealt with through existing EU privacy legislation and should be addressed.
First, there is the question of the extent to which individuals need to be made aware of how new technologies, including smartphone apps, may change their decision-making through nudging. This is particularly pertinent in relation to health, where the protection of personal autonomy has been accepted as being fundamentally important.Footnote 66 If future realities incorporate a variety of decisional steers, legislation on the transparency of these mechanisms and their use should develop accordingly.
Second, the control of public and privately sourced apps differs insofar as the development and deployment of governmental apps, particularly mandatory ones, is often subject to parliamentary checks. Privately sourced apps do not benefit from such scrutiny, which may be explained by the fact that they more clearly include the option to “opt out”. However, once in use, the decisional privacy concerns may prove very similar. Notably, the ECtHR has interpreted the right to (decisional) privacy as imposing an obligation on the government to also protect individuals in horizontal relationships. This suggests that there is an active obligation to ensure that the use of hypernudges by corporations is scrutinised and regulated. This could be done through changes to the GDPR, which already provides such horizontal protection vis-à-vis data collection and informational privacy more generally, or through the adoption of dedicated legislation.
In summary, we envisage a role for the EHU in coordinating the use of technologies such as smartphone apps, including those involving hypernudging, in the domain of EU public health. Footnote 67 In doing so, the EHU is guided, and supported, by existing EU privacy regulation, which – increasingly explicitly – incorporates decisional privacy. However, the jurisprudential basis for decisional privacy protection means that a broader legislative debate on questions such as justifiable infringements, types of remedies and procedural protections has thus far been sidestepped. In better equipping the EHU – and Member States – for these questions, such a debate in the Council and Parliament would be welcomed.
