Introduction

The events are now well known. In December 2019, the city of Wuhan (Eastern China) went under lockdown to contain the spread of a fast-growing epidemic caused by a coronavirus. Shortly after the images of the sealed-off city reached European television screens and newspapers, the first cases were reported in Italy. After Asia, Europe became the second continent to be affected, with the first cases recorded in Italy on 31 January 2020 (Ritchie, 2020). On 13 March, the number of cases in Europe exceeded those in Asia, leading the World Health Organisation (WHO) to declare Europe the epicentre of the crisis, with 40% of the world’s cases and 68% of the deaths (WHO, 2020). In less than two months, several European governments were overwhelmed by the spread of the pandemic. Unlike their Asian counterparts, European decision-makers had little recent experience with coronavirus-induced respiratory infections. The last severe influenza pandemic occurred in 1968 and affected European countries mostly unevenly (Viboud et al., 2005). In addition, at the end of January 2020 there was little scientific knowledge about the virus, while in March 2020 expert advice was highly conflicting. The spread of the virus among children, its airborne nature, the mechanisms of immunity and its duration were among the major unknowns.

The COVID-19 pandemic presented European policymakers with a dilemma. They had to balance the need to protect the health of the population while maintaining the normal functioning of institutions. Giving absolute priority to protecting health can lead to restrictions on individual freedoms, including freedom of movement, privacy, trade and religion. Such a priority can also negatively impact political rights through the cancellation or postponement of elections, referendums and the restriction of petitions or demonstrations. Finally, the balance of powers can be altered, as the need for quick and coordinated decisions increases the power of the executive to the detriment of parliament, the judiciary and regional authorities. More specifically, exceptional crisis-management policies can have a negative impact on the economy and on the level of trust in the institutions, as well as creating opportunities for democratic backsliding. In contrast, giving priority to the normal functioning of institutions can lead to an out-of-control pandemic with dramatic consequences in terms of human lives.

In deciding on the best course of action, policymakers had little certainty and evidence, at least during the first wave of the pandemic. Faced with the same constraints, European policymakers responded in very different ways. Some activated constitutional emergency provisions, while others relied on crisis-management legislation. Several countries took a very centralised approach to pandemic management, while in others, policies were adopted at the regional or local level. In some countries, restrictions were very strict (including a curfew, a ban on leaving one’s home for more than one kilometre, and strict police controls), whereas in others, people were called upon to exercise their sense of civic duty and solidarity by restricting their travel and daily activities. Neighbouring countries that share the same history, political culture, legal systems and institutions reacted in contrasting ways. For example, while the Dutch cabinet implemented one of the softest approaches to crisis-management in the early stages of the pandemic (January–March 2020), Belgium introduced a very strict lockdown early on (Massart et al., 2021).

Such different policy responses were however associated with a high degree of semantic proximity, making the analysis of legal responses even more arduous and possibly confusing. Although nearly all European countries—to the notable exception of Sweden—based their crisis-management strategy on a “lockdown,” the amount of the restrictions decided upon as well as the level of coercion used in their implementation considerably vary (Engler et al., 2021; Egger et al., 2021).

This diversity of policy responses is even more striking as European nations are closely connected with one other and that pathways of policy diffusion and learning abound. In particular, the European Union (EU) offers coordination mechanisms in public health that go beyond the general guidelines of the WHO and drive from the 2014 EU Agenda on Health Systems (European Commission, 2020b). In addition, joint crisis-management mechanisms—through the EU civil protection instrument—are long established among European countries (European Commission, 2020a). Nevertheless, governments reacted to the pandemic in multiple and sometimes contrasting ways—and continued to do so well into autumn 2021. As a result, the cross-national variation in COVID-19 policy responses in Europe is very high and, as such, has sparked scholarly interest (Egger et al., 2021; Engler et al., 2021; Rausis & Hoffmeyer-Zlotnik, 2021).

Our edited volume aims to contribute to such a literature by providing a comparative account of COVID-19 policies across Europe. While most existing handbooks and edited volumes compare national case studies, mostly focusing on containment policies (see, inter alia, Hale et al., 2021; Cheng et al., 2020; Grogan & Donald, 2022), this volume brings together scholars of crisis-management policy, comparative politics and law to analyse the variation in the design and implementation of a wide range of COVID-19 containment policies at the national and subnational levels. To achieve this goal, we make use of new comparative data—collected during the EXCEPTIUS project—mapping the diversity of COVID-19 policies in 32 European countries based on the computational analysis of legal archives.

Mapping COVID-19 Containment Policies in the European Economic Area: The EXCEPTIUS Project

Empirically, the contributions to this volume draw on data collected by the EXCEPTIUS project (Egger et al., 2022). During the course of the pandemic, a group of 17 scholars representing five disciplines from the social sciences and humanities, coordinated by the University of Groningen (the Netherlands) and the University of Grenoble (France), came together to collect data on the modalities of exceptional decision-making and policy-making during the COVID-19 pandemic. Simply defined, exceptional measures impose extraordinary constraints on democratic governance, fundamental rights and everyday freedoms compared to pre-crisis business as usual. To achieve this goal, we have taken a three-pronged approach. First, we trace the measures back to their origins, creating a corpus of COVID-19 legislation that uniquely allows researchers to analyse the diversity of legal instruments used to contain the COVID-19 pandemic. Press releases or expert surveys, which are commonly used in competing projects (Hale et al., 2021; Cheng et al., 2020), capture such dimensions only indirectly and imperfectly. In contrast, legal sources detail the types of containment measures adopted, its modalities of implementation and the changes they introduce to pre-crisis governance. Analysing legal data allows to go beyond the semantic proximity of COVID-19 measures to uncover how a single measure (e.g., a lockdown) was implemented in a more or less stringent manner over time and space.

Second, our project defines the most comprehensive taxonomy of exceptional measures in the area of democratic governance, the rule of law and fundamental rights and freedoms. Four meetings took place among consortium members between May and September 2020 to draw up the project taxonomy. Classes and subclasses were identified by applying both top-down and bottom-up approaches. They are presented in Table 1.1 below.

Table 1.1 EXCEPTIUS Taxonomy of COVID-19 containment measures
Full size table

The scientific board first did an extensive literature search to identify modalities of exceptional crisis-management measures in their relevant field. A sizeable literature has especially focused on the modalities and typologies of emergency powers (Rooney, 2019; Ferejohn & Pasquino, 2004; Gross & Aoláin, 2006; Bjørnskov & Voigt, 2018) allowing us to identify two first classes of events respectively focusing on: (1) state of emergency provisions—based on constitutional or ordinary law; (2) derogations to fundamental rights and freedom—as listed in the European Convention on Human Rights. To identify countries that introduced exceptional measures based on crisis-management legal frameworks, we introduced a last class (10—emergency decision-making).

We added to these two seminal classes, a third one, that derives from the ability of governments to derogate from international human right treaties in times of emergencies (Hafner-Burton et al., 2011) and covers broader suspension to international agreements and commitments. We then reviewed the literature on pandemic management approaches as well as similar text analysis initiatives focusing on COVID-19 (Cheng et al., 2020; Porcher, 2020; Hale et al., 2021) to define two other classes. The first covers restrictions of daily liberties targeted at preventing the spread of the COVID-19. This includes restrictions placed to daily activities such as sport, different forms of private and public gathering but also obligation to maintain social distances and wear a mask. The second focuses on closure and lockdowns in various sectors such as education, commercial activities or culture but also restrictions to national and international travels. Two classes—mobilization of the army and of the police—were added by board members specialized of security public policies to identify exceptional implementation modalities. Collecting data on implementation modalities is central to have a more fine-grained view of the stringency of the exceptional measures adopted. In some countries, such as Denmark, the state of emergency was declared but not implemented. In others, such as initially the Netherlands, the police largely played a mediation role and raised awareness among citizens about the importance of complying with exceptional measures. Other countries—such as France, or Poland—have seen a large deployment of the police and the army to support public hospitals and sanction non-compliance.

A more bottom-up inductive approach was followed to complement the initial identification of measures. This allowed us to compensate for literature gaps as well as to mitigate an initial biased focus on restrictive measures. Documenting exceptional measures aiming at granting exceptional rights to citizens—such as the possibility of attending court hearings online or the extension of legal and administrative deadlines—is equally important to uncover how governments navigate the dilemma between protecting the life of their citizens and preserving the functioning of democracies in pandemic times. To do so, the scientific board, supported by a researcher specifically hired for the project manually analysed a random sample of 50 legal documents focusing as much as possible on general COVID-19 law and containing at least one document per country. This process not only enabled us to refine the taxonomy of the subclasses identified but also to come up with a new class, focusing on government oversight.

Third, we leveraged computational text analysis methods to apply this taxonomy to our legal corpus. We experiment with a large array of supervised classifier models together with the support of a team of computational linguists (Tziafas et al., 2021). The automated pre-processing of the corpus performed by the most performant multilingual classifier was then provided to a multinational team of coders that further described the measures (duration, target groups, authority). The combination of both computational and close reading techniques allowed us to classify exceptional measures on a daily basis from January 2020 until the end of April 2021 in 32 countries of the European Economic Area at both national and subnational levels of governments.

The final EXCEPTIUS dataset (Egger et al., 2022) is structured as a longitudinal and cross-sectional dataset. Each entry in the dataset is an exceptional measure codified in a legal act and adopted on a specific day in a specific place (the national territory of a country or a subnational region based on the NUTS 2 typology). Each of the measures is allocated to a class and subclass based on the EXCEPTIUS taxonomy of measures (presented in Table 1.1 above). In addition, for each subclass, the modalities of enforcement of the measure, its duration, target groups as well as the sanctions foreseen for non-compliances are specified. Additional descriptive data include the authority which decided on the adoption of the measure, its legal source (type and URL) as well as the geographical coverage of the measure. The coverage of the dataset concerns 32 countries of the European Economic Area over 15 months (from January 2020 to April 2021).Each contribution to this volume leverages the EXCPETIUS dataset and sheds new comparative light on the patterns of variation of COVID-19 containment measures in Europe.

Facts First

Ideally, social scientists should be able to assess how effective various public policies are in containing a crisis. This is key not only to ensure that political accountability is informed by scientific evidence but also to trigger effective forms of policy learning. The COVID-19 pandemic offers a unique opportunity. The disease struck a large range of countries in a short and comparable timeframe and triggered very different policy responses. This context—akin to a natural experiment—provides counterfactuals, i.e., the opportunity to capture what would have happened if different policies had been implemented.

However, this unique opportunity is hampered by a lack of quality data. Current evaluations go in two different directions. Some rely on comparative data to analyse the impacts of COVID-19 policies but don’t take the policies themselves as a central focus. For example, the study by Altiparmakis et al. (2021) uses surveys in 11 countries to predict citizens’ positive or negative evaluations towards government responses as a whole. Other comparative studies are based on the Oxford COVID-19 Government Response Tracker (see Aknin et al., 2022; see Mendez-Brito et al., 2021 for a systematic review) which focuses on three core categories of containment policies (closures and lockdowns, economic and health response) but neglect changes compared to pre-crisis governance patterns. Yet, the way policies are decided upon is key to their legitimacy which in turn shape citizen compliance and ultimately the effectiveness of crisis response (Bargain & Aminjonov, 2020).

Other studies analyse a wider range of policies but limit their analysis to one country (see Knaul et al., 2021; in Mexico or Cramer et al., 2022, in the United States; Askim Christensen, 2022, in Norway). To the best of our knowledge, this volume is the first to provide a comparative and descriptive account of the way various governments handled the COVID-19 pandemic focusing on a large range of policies and countries.

The simple yet important aim of this book is, first and foremost, to precisely document the diversity of measures adopted across the European continent. Before evaluating the effectiveness of crisis governance, our ambition is to present in detail how each authority—at a national or subnational level—has governed. What everyone knows as lockdowns, compulsory vaccinations, wearing of masks, states of emergency or exceptional measures are in fact a set of rules that considerably vary from one country to another. Lockdowns were, for example, stringent in Europe and the depth of control used in enforcing them differed.

Each chapter focuses on a specific policy event to provide a detailed account of the differences across countries or so-called pandemic waves. This mapping comes with complementary data and attempts to assess the level of exceptionalism characterizing each country.

Prioritizing the establishment of the facts does not mean that the EXCEPTIUS dataset is immune from errors. Despite the attention and efforts of the many researchers involved in the project, the database will need to be corrected and expanded in the future. Some regulations may have been forgotten and the coding process, based on complex legislative texts, is not as mechanical as we would like it to be. Nonetheless, our project is the first to systematically record national and subnational regulations based on legal data and during the first three waves of the pandemic. The overall picture of pandemic management we draw from the data offers new and somewhat surprising insights.

First of all, the extent to which individual rights have been curtailed varies. ‘Lockdowns’ have been declared in almost all European countries, with the notable exception of Sweden. However, the experience of European citizens living under lockdown differed depending on the country they lived in. Chazel’s contribution in this volume shows that the most severe lockdowns are reported in France, Italy, Romania, Austria and Spain. Such strict lockdowns persist over time, across the different waves of the pandemic. Some Northern European countries, such as the Netherlands, relied on a rather light lockdown compared to their European counterparts. In addition to lockdowns, freedom of movement was also curtailed by limitations on international and domestic travel. Conde Belmonte, Huesca González and Villacián Goncer confirm that Spain and Romania have taken a particularly radical approach, banning almost all forms of travel, including for asylum seekers. Italy stands as a notable exception as the country maintained a certain degree of openness during the different waves. On the other hand, Finland and Switzerland closed their external borders but allowed internal travel.

The management of the COVID-19 has raised concerns about freedom of speech, freedom of the press and freedom of religion. The spread of critical opinions against the lockdown or COVID-19 vaccines, especially if supported by disinformation strategies, could have a negative impact on the management of the pandemic by reducing compliance with COVID-19 measures and trust in the government. Restricting the expression of these opinions during crisis times could increase the effectiveness of crisis-management strategies but also create a disproportionate burden on fundamental liberties. Neves, Merrill, Miguel and Forte explain that such blanket have rarely been observed in Europe to the notable exception of Romania. Indirectly, however, the lack of transparency, limitations faced by journalists to access information and restrictions on press conferences have negatively impacted the quality of public debates. While such restrictions have been experienced in many countries, they are particularly prevalent in Southern Europe.

Another typical pattern of the management of the pandemic is the disruption introduced to pre-crisis democratic governance and, in particular, the balance of power. In almost all European countries, the executive played a dominant role during the pandemic. Poyet demonstrates that this trend was almost absent in the Scandinavian countries, Austria and Ireland. In Germany, Norway and Austria, the parliament stayed very active. In contrast, in new and fragile democracies such as Hungary and Poland, members of parliaments have been hardly consulted over crisis-management modalities and these countries experienced the strongest concentration of power in the hands of the executive. Egger shows that the declaration of the state of emergency only partially accounts for such differences as countries that relied on constitutionally or legally enshrined exceptional provisions were not characterised by a stronger role of executive but did derogate more heavily on fundamental rights. Crises are often associated with the centralisation of power because of the need for a rapid and coordinated response. Magni-Berton does not find convincing evidence for such a rationale and shows that many countries have opted for managing the pandemic in a decentralised way. In federal countries, such as Germany, Switzerland and Belgium, much of the decision-making was located at the subnational level, while in Austria, power was highly centralised. In regionalised countries, crisis-management was implemented at regional level (Italy) or regions were granted additional capacities during the second wave of the pandemic (Spain). France and the United Kingdom also increased their local management over the waves. Finally, Portugal and Finland only opted for a more decentralised approach in the third wave. The other (unitary) countries maintained a centralised approach.

Finally, Ehin and Talving document that many local elections were postponed because of the pandemic. The only exception was the Polish presidential election, which was postponed by 49 days. When they were held on time, the turnout was much lower than usual. Throughout Europe, referendums were suspended, although in Switzerland people were allowed to vote in two suspensive referendums on pandemic preparedness, with high turnout and a recorded positive effect on social cohesion. The integration of special voting arrangements into regular electoral law before the onset of the pandemic increased the opportunities for people to cast their votes other than in person at the polling station.

Towards New Avenues to Explain and Evaluate Pandemic Policy-Making in Europe

The priority given to describing the public policies adopted in Europe is followed, in some chapters, by exploratory efforts to explain and evaluate these policies.

Explaining the options followed by governments is important to understand the institutional, cultural and ideological context in which the COVID-19 pandemic was handled. Past large-scale crises—such as the jihadist attacks in Paris in November 2015 or the L’Aquila earthquake in 2009—often draw public attention to government (Boin et al., 2016). This attention is associated with high expectations about the ability of leaders to manage the crisis effectively and strong demands for accountability. The visibility on public authorities has fuelled academic interest in the patterns and dynamics of crisis leadership, such as charisma (Pillai, 1996; Halverson et al., 2004) or personality traits (Fredrickson et al., 2003).

Personality traits were also used as predictors of specific policy responses in COVID-19. Medeiros et al. (2022) found that world leaders scoring high on ‘plasticity’ (extraversion, openness) were quicker to implement travel restrictions and provide financial relief, and generally offered a stronger response. In contrast, leaders who scored high on “stability” (conscientiousness, agreeableness, emotional stability) offered both faster and greater financial relief. On the other hand, some studies have focused on the strategic role played by leaders in times of crisis. Boin et al. (2016: 15) identify five central tasks that leaders perform for an effective crisis-management strategy in democratic settings: (1) sense-making: gathering and processing information that helps to identify a crisis and understand its significance; (2) decision-making and coordination; (3) meaning-making: proposing a narrative that mobilises crisis managers and citizens; (4) accounting for the way the crisis has been managed in public forums; and (5) learning from the crisis and its resolution. In analysing these tasks, the authors note that crises often trigger “situational leadership,” i.e., the emergence of individuals who take actions, make decisions and assume authority roles that are not necessarily foreseen. In a similar vein, Demiroz and Kapucu (2012) provide an analytical framework that allows for the comparison of the competencies of leaders in a crisis, linking the tasks identified by Boin et al. (2016) to concrete skills.

Although such a literature provides relevant insights into different levels of leaders’ performance in crises, leaders’ personality traits and skills can be considered as an endogenous feature of crisis-management institutions. As Janis (1989) points out, leadership traits and skills are determined by the specific procedures of crisis decision-making. A more cynical interpretation sees crises as perfect times for leaders—usually constrained by democratic checks and balances—to seize absolute power (Voigt, 2022). In this volume, we move away from a focus on leadership in crisis to one on the institutional configurations of decision-making and policy-making in crises.

A pattern of concentration of power in the hands of the executive was observed in many countries during the COVID-19 pandemic (Dostal, 2020; Kühn et al., 2021; Turnbull & Bernier, 2022). The chapters by Egger and Kuebler in this volume document a similar trend but show variations in the intensity of such concentration of power based on emergency legislation (Egger) and pre-crisis levels of democratic quality (Kuebler). At the same time, our volume also highlights minority tendencies to decentralize crisis decision-making to the benefit of parliaments (Poyet) and local authorities (Magni-Berton). Such an important—but neglected—role of decentralization in crisis-management was already observed by Hart and his co-authors in 1993.

The extent to which the role of the executive was strengthened during the pandemic depends on how strong the role of the executive was before the pandemic. The increase in executive power is particularly weak in consensual countries (in Lijphart’s classification) or in countries with many veto players (in Tsebelis’s classification). Similarly, powers are least centralised in federal countries and, more generally, the degree of centralisation of decision-making before the pandemic is associated with the degree of centralisation during the pandemic. However, this institutional inertia is not universal. Austria sits oddly in this list, as it adopted exceptionally stringent measures, given its consensual and federal institutions.

While this could reflect a weak capacity of consensual and federal states to quickly respond to a pandemic (Hegele and Schnabel 2021), this conclusion is not consistent with the data. While the restrictions in these countries were globally less severe, the first measures to combat the pandemic were implemented early and, after two years, they still experienced fewer deaths and higher levels of political trust.

However, this does not mean that consensual and federal institutions are more effective, as other confounding factors may explain these results. These institutions are more prevalent in wealthier countries and in countries where political trust was higher before the pandemic. Many contributions confirm that these factors contribute to a less severe response to the pandemic (see Egger et al., 2021). Prosperity allows governments to find alternative ways to reduce the spread of the virus, such as identifying epidemic outbreaks or implementing rules. Political trust helps to achieve compliance without formal restrictions. In her paper, Vulcano shows that mobility is significantly reduced not only by the stringency of restrictions, but also by trust and the ability to enforce the law. However, EXCEPTIUS data also show that, in the long run and after an initial rally-around-the-flag effect, strict restrictions have a strong negative impact on public trust. This can lead to a vicious circle: in times of crisis, when public trust is low, policies need to be stricter to induce compliance, which in turn exacerbates the lack of trust. In the long run, this vicious circle weakens the ability of societies to respond to crises. More broadly, the observed and robust statistical relationship between the quality of democracy before the pandemic and the level of concentration of power during the pandemic is likely to reveal a reinforcing mechanism in which each factor reduces the other.

However, other exogenous factors should also be considered. For example, Conde Belmonte and co-authors show, in this volume, that interregional domestic travel bans and checkpoints have been introduced both South-Western and some Eastern European countries. Such a high level of restrictions can be understood as a protection against internal and external tourist travel, as they systematically occur in tourist destinations. Similar patterns can be observed in the case of COVID-19 passes that got introduced in the peak touristic season and later on in Northern European countries. Contexts of high affluence typically increase the potential spread of the virus and therefore the need for stringent measures. These factors suggest that some countries need to be more legally prepared for a pandemic than others. Specific institutions can be put in place to minimise changes in governance. An example can be found in the contribution of Ehin and Talving on special voting arrangements such as early voting, postal voting, proxy voting, remote internet voting and the mobile ballot box. Countries that had such arrangements in place before the pandemic were better able to maintain electoral continuity.

Although the study of legal preparedness and crisis-management after COVID-19 is still in its infancy, the EXCEPTIUS data will certainly expand the evidence base of such evaluation exercises. Managing a pandemic should not be a matter of feeling and intuition. Social engineering, based on experience and rigorous evidence, should be able to provide governments and citizens with standard policies capable of effectively combining health protection and political legitimacy. In contrast to long-term policies, which are expected to be driven by debates and values, emergency policies are paradoxically less challenging because they only aim at introducing efficient instrument to tackle a time-bounded emergency without upsetting the status quo. In an emergency, exceptional policy choices should be minimised so as not to interfere with the path chosen by each country in normal times. To minimise these choices, data, and analysis on the effectiveness of crisis-management are crucial. For the pandemic we have recently suffered, this book aims to support such crisis preparedness exercises.