Does Evidence-Based EU Law Survive the Covid-19 Pandemic? Considering the Status in EU Law of Lockdown Measures Which Affect Free Movement

When Member States restrict free movement on public health grounds they must show that their measures have a sound scientific basis. However, during the pandemic Member States have imposed a wide variety of restrictions, at the border, and internally. While Member State governments have invariably had local scientific advice, the variety of their measures suggests that their actions have also been driven, to some extent, by public opinion, contrary to what EU law generally allows. This situation could be seen as a defeat for EU law as traditionally conceived, and the triumph of local preferences over scientific standards. Perhaps we learn that in a crisis, local desires for symbolic security and closure trump both law and science. Alternatively, it can be argued that the Court of Justice's emphasis on exclusively objective justifications for measures is unrealistic and over-strict. The pandemic responses show that (i) science is often neither clear nor determinative, and (ii) policy is invariably a mix of science and values, even in apparently technical fields. In either case, the absence of legal challenges to Member State actions leaves free movement in an uncertain state. Have we entered a new phase, where national fears are a more legitimate justification for restricting movement, or will the pandemic be treated as so exceptional as to be beyond law, and thus not a precedent?


INTRODUCTION
EU law traditionally has an evidence-based bias, at least in theory (Dawson, 2016) and EU legislation must satisfy objective tests of effectiveness and necessity 1 . Analogous constraints are imposed on Member States when they restrict free movement (Chalmers et al., 2019, p. 828-837). However, during the pandemic Member States have imposed a wide variety of restrictions, at the border, and internally (OxCGRT., 2020). These have had significant social effects, impacting most harshly on the less wealthy, urban, those at risk of domestic abuse, and those without residence rights, who are particularly imprisoned by their fear of controls. While Member State governments have invariably had local scientific advice, the variety of their measures suggests that they cannot all be seen as purely scientific responses to public health needs. They have also involved a degree of political responsiveness to local majoritarian fears and desires.
Usually in EU law the Court of Justice rejects apparently scientific arguments tailored to public wishes, and seeks to separate objective public health needs from local preferences 2 . However, despite some criticism from EU Commissioners there has been no significant legal challenge to national measures. This could be seen as a defeat for EU law doctrine, and the triumph of local preferences over scientific standards. Perhaps, in a crisis, local desires for symbolic security and closure trump both law and science. Alternatively, it can be argued that the traditional line in EU law is unrealistic and over-strict. The pandemic responses show that (i) science is often neither clear nor determinative, and (ii) policy is invariably a mix of science and values, even in apparently technical fields.
In either case, the absence of legal challenges leaves free movement law in an uncertain state. Have we entered a new phase, where national fears are a more legitimate justification for restricting movement, or will the pandemic be treated as so exceptional that it has no precedential value?

THE EVIDENTIAL REQUIREMENTS IMPOSED BY EU FREE MOVEMENT LAW
The European Union endorses an evidence-based vision of law. EU legislation must comply with procedural and substantive requirements aiming to ensure it effectively addresses genuine needs and does not go beyond this (European Commission, 2015;Meuwese, 2017). The mere fact of popular desire for a measure does not make it legal.
Proposed harmonization measures therefore undergo an impact assessment, which is a structured and technocratic assessment, rather than a political judge of desirability (Meuwese, 2017). A measure may even be annulled after adoption it if does not comply with requirements of objective necessity and effectiveness.
These constraints follow from the nature of EU competences: it is typically granted powers to achieve a specific goal, rather than merely to take measures of a certain type or within a certain field. The exercise of its powers is only legitimate where it actually contributes to that goal (Davies, 2015).
This purpose-limitation arises because the EU is a creature of conferred powers 3 . It may only act within the range of its purposes because the alternative, to give it an open-ended democratic mandate such as a state enjoys, would eliminate any legal constraint on full federal integration of the Member States into a unitary legal order (Davies, 2018). The day may come when this is seen as desirable, but for the moment, constraints on EU power remain prominent in the EU Treaties.
The emphasis on goal-oriented, evidence-based law is also found in Court of Justice caselaw reviewing national measures, particularly those restricting free movement between the Member States. The caselaw now contains a number of principles which subject national law to the same "objective" and "scientific" scrutiny to which EU law is accustomed.
These principles apply whenever Member States take measures which impact, directly or indirectly, on free movement 4 . Not only explicit restrictions on crossing borders are caught, but also other measures which in practice make it harder for individuals or companies to engage in cross-border activities (Tryfonidou, 2014).
These principles aim to ensure that restrictions on movement are genuinely necessary, and not covert protectionism or xenophobia. However, the law goes beyond uncovering disreputable motives. It also examines domestic laws to see whether their goals can be achieved in ways that are less restrictive of movement and more compatible with openness (Tryfonidou, 2014) 5 . Consequence for "outsiders" are often neglected when national law is made. Free movement law re-opens insular regulatory processes (Maduro, 1998).
As this implies, measures restricting free movement cannot be legally justified merely by popular support. Many national measures enjoying widespread domestic support and sometimes deep cultural roots have been ruled incompatible with EU law because they were not justified in sufficiently objective and scientific terms 6 . That the Rheinheitsgebot was a part of German history was not only insufficient, it was not even apparently relevant, when that measure was successfully challenged as a restriction on imported beer 7 .
The public mood was addressed in Commission v France, where French farmers sought to exclude foreign agricultural produce 8 . When France was finally-after years of problemssued for failing to remove protesters and clear the roads for foreign trucks, it claimed that such measures could lead to popular uprisings and a breakdown in public order. The Court of Justice found that such justifications could only be accepted in extreme cases, and the burden was on the state to show that it was incapable of maintaining order without the measures in question. In this case, France lost. The public desire for strong action, and for closure, could, in theory, legitimate a measure but only if the evidence is convincing that without that measure public order would break down.
Even where a measure is justified by reference to science, that does not mean it escapes scrutiny. Restrictions imposed in the name of public health must be based on demonstrable scientific arguments, corresponding to international scientific consensus 9 . Where states rely on a domestic scientific consensus that is out of tune with the international scientific community, the Court has ruled this unlawful 10 . Science is not, in principle, nationally specific.
Where, however, science is not yet conclusive, Member States are free to differ in their understandings, and may take any position that can be objectively defended-even if other approaches are possible too. In the absence of a mainstream consensus on the safety of vitamins added to food, the Court found that it was for each Member State to decide its own position, provided that their choice was defensible on scientific grounds and not just an expression of political will (Nic Shuibhne and Maci, 2013) 11 . Also, where a risk is in issue, Member States may choose to tolerate different degrees of that risk 12 . The degree of safety or security that a population demands in different contexts can vary and such national differences are generally respected. That is provided, however that the measures they take correspond objectively to the chosen level of risk. The burden is on Member States to demonstrate this, using substantive or procedural evidence or both (De Witte, 2013), or to cease applying their measures. The legal frame applied here is the proportionality principle-the requirement that the measure has a legitimate goal, is an effective way of pursuing this goal, and does not go beyond what is necessary or effective.
These principles are applied with varying degrees of intensity. Where constitutional or national values are put forward to justify restrictions, the Court is often surprisingly, perhaps problematically, deferential (Kochenov, 2018) 13 . However, the more severe the restrictions on movement involved, the less likely such deference becomes. Moreover, and most relevant here, where an apparently objective and science-based justification is put forward, such as the protection of public health, evidential requirements will generally be applied strictly (Chalmers et al., 2019, p. 842). If a state claims to be following the science, it will have to show that this is the case.

MEASURES TAKEN BY MEMBER STATES TO PREVENT SPREAD OF COVID-19
The Existence of Variation in Approach Between Member States All Member States of the EU have taken measures to limit the pandemic. Common features include hygiene and social distancing, restrictions on cross-border and domestic movement and on time spent outside, and closures of non-essential businesses. However, the particular combinations of measures, and the severity of restrictions, have varied greatly from state to state (OxCGRT., 2020). There has also been considerable variation in how and when states have begun to relax their restrictions (Pacces and Weimer, 2020;Van Elsuwege, 2020).
The question here is how this variation fits the framework of the law. One might expect that a response based on scientific best practice would be largely uniform from country to country. It may seem that at least some Member States must have been not really following the science (Van Elsuwege, 2020). raise the question whether their measures were compatible with EU law (Carrera and Chun Luk, 2020).
However, it is suggested here that policy variations can be explained in ways that fit existing law. Four explanations are considered below. Just one, the last, raises uncomfortable questions about free movement doctrine and its future. The explanations are: (i) Scientific advice varies between states (ii) Populations differ in their preferences regarding risk and liberty (iii) Policies are tailored to local circumstances and attitudes (iv) Governments are responding to varying public fears and desires (even when these have no scientific basis).

Variation in Scientific Advice
In the early months of the pandemic there was uncertainty about many basic questions-mechanisms of transmission, effectiveness of masks, and lockdowns, the role of children as spreaders, and the fatality risk, for example. Expert policy prescriptions varied significantly. In such circumstances, Member States, who invariably turned to their domestic expert institutions, may well have received differing advice. This was a "vitamins" situation: where the science cannot provide certain answers, any standpoint which enjoys a reasonable scientific basis can be accepted 14 .
The causes of that variation in advice will have been complex, recommendations being influenced by institutional and professional factors, the character of key individuals, and, quite possibly, factors to do with the culture and habits of that state.

Different Risk Tolerance/Liberty Attachment
This brings us to the second explanation: that some populations-majorities-may have a greater appetite for risk than others, or a greater distaste for limitations on their freedom. Such differences are to be expected between states, and are observable in other situations of threat, such as wars, terrorism or economic crises. They need not be stable features of national culture-although they could be-but can be products of the prevailing majority of the moment, and of the resulting government in power. In either case, the law should not have a problem with this: Member State are generally entitled to choose the level of protection from risk that they wish to establish 15 . The mere fact that they take stricter, or less strict, measures than other Member States does not make their actions unlawful provided the measures are appropriate to the level of risk they have chosen.

Controlling Transmission Using Methods Adapted to Local Circumstances
However, it may also be that differing packages of measures reflect differences in what is acceptable to the public, and what is likely to be effective in that particular national context. Some countries take more naturally to social distancing than others, and the infrastructure of cities and public spaces affects how easy this is to maintain. Facemasks are seen by some as comforting, and as minor infringements on liberty, by others as a dehumanizing last resort. Similar remarks can be made about strict lockdowns, in particular for children; are these examples of precaution, or of panic? The closure of businesses, its acceptability, is also dependent to some extent on the economic structure of society and on government. Thus, different packages of measures can, to some extent, be seen as different ways of achieving essentially the same end. All states seek to reduce transmission, but do so in ways that reflect the particular socioeconomic, institutional, physical and cultural circumstances of that state (Pacces and Weimer, 2020), as well as the particular stage and progress of the epidemic in that state.
Differing attitudes toward the state and the individual play a role here. Some majorities may expect a strong role for the state, and experience strict measures and formalized enforcement (forms to be filled, uniformed checks) as comforting-the state as parent is showing its presence. In others, this kind of behavior is seen as threatening, and a sign of breakdown of order, a symptom of impending collapse. In more liberal Member States the emphasis has been on individual responsibility, and minimizing restrictions on freedom has been understood to show that the state is still functioning normally, that matters are under control, and that the core principles of social order are not threatened. The freedom to go shopping, or spend the day in a park, can give the same feeling of safety in one context as the obligation to remain inside can in another.
These differences can be accommodated within the law. The fact that measures must be effective does not imply a uniformity of approach. The Court accepts that it is for Member States to decide how to protect interests such public health, for which they have primary competence, so long as they respect EU law principles such as proportionality 16 . Given variations in epidemiological circumstances, economy, infrastructure and public attitudes, varied approaches could show Member States are taking this seriously: considering what works in their state. Uniform measures across the EU would almost certainly have differing degrees of success in different Member States, and not in fact be as effective as locally-tailored approaches.

Responses to the Public Mood
Despite the above, it seems inevitable that Member States have sometimes taken measures because the public mood demanded it, not on the basis of expert advice, and perhaps even contrary to it. The Dutch prime minister, Mark Rutte, acknowledged this when announcing the closure of schools, saying that the Dutch Public Health Institute had advised him that this was neither necessary nor effective, but he could see that parents and teachers were nervous and he wanted to avoid public unrest 17 Whether recognized in public or not, similar concessions to the public will have been made in other Member States.
Examples may include restrictions on internal travel, on being outside, and, above all, the closure of borders. Expert opinion on these has been divided, but what they have is a strong political, symbolic, aspect-they cultivate a sense of emergency which helps to keep the public obedient. In particular, the closure of borders may satisfy an instinctive desire to create safety by excluding, to lock the door to strangers. There may also have been expert support, but that may not have been the primary reason for the measures (Carrera and Chun Luk, 2020).
However, the French Farmers logic-a public desire for closure only justifies restrictions if not meeting that desire would lead to catastrophic consequences-while it failed in that case, might legitimately succeed here. Controlling the pandemic required public support and acquiescence, and compliance with restrictions. While in general EU law expects Member States to resist, and reform, the public mood, in this situation it may genuinely be the case that they have a justification for conceding to it. The late spring and summer brought an increase in resistance to lockdown measures and public demonstrations. Acceptance is clearly not self-evident. However, were there widespread rejection of the measures, then even a well-designed pandemic policy would fail. Thus, pandering to unfounded fears, and adopting symbolic measures, may have been politically necessary in order to make the genuinely scientific measures accepted and successful.

CONCLUSION
EU law allows for variation in responses to public health, and other, threats, and allows the taking into account of particular national circumstances, including behavioral and cultural factors. However, it does not usually allow for political responses to public calls for action, when those calls are not themselves scientifically founded.
To the extent that pandemic measures have gone beyond science, and also been aimed at calming public fears-an extent which is still unknown, but will perhaps become apparent in coming years-they may seem to challenge existing law. The need for public support may seem to have entered the law as a justification for restricting free movement. On the other hand, the pandemic clearly is exceptional, which limits its precedential value.
Absent legal challenges, we are left in a degree of uncertainty. However, this may itself revitalize discussion about when national measures derogating from free movement are justified, and about the meaning of ideas such as necessity and effectiveness. An overly political interpretation of these threatens to make EU law ineffective and undermine its mission to transform national societies. Yet an overly scientific interpretation can undermine its legitimacy, for as the pandemic measures show, science often speaks with many voices, and good policy decisions take account of science, value choices and the need for support. Treating the public interest as a largely technical matter is an adequate, even optimal, approximation to the truth in many situations. In more serious ones, a more complete engagement with reality may be required. The future challenge for the law is to incorporate this distinction.

AUTHOR CONTRIBUTIONS
The author confirms being the sole contributor of this work and has approved it for publication.