Centralizing and decentralizing governance in the COVID-19 pandemic: The politics of credit and blame

Abstract

COVID-19 led to significant and dynamic shifts in power relations within and between governments, teaching us how governments make health policies and how health crises affect government. We focus on centralization and decentralization within and between governments: within government, meaning the extent to which the head of government controls policy; and between governments, meaning the extent to which the central government pre-empts or controls local and regional government. Political science literature suggests that shifting patterns of centralization and decentralization can be explained by leading politicians' efforts to gain credit for popular actions and outcomes and deflect blame for unpopular ones. We test this hypothesis in two ways: by coding the Health Systems Response Monitor's data on government responses, and through case studies of the governance of COVID-19 in Austria, Czechia and France. We find that credit and blame do substantially explain the timing and direction of changes in centralization and decentralization. In the first wave, spring 2020, heads of government centralized and raised their profile in order to gain credit for decisive action, but they subsequently tried to decentralize in order to avoid blame for repeated restrictions on life or surges of infection. These findings should shape advice on governance for pandemic response

1. Introduction

In 2020, the COVID-19 pandemic led to something like a whole-of-government approach focused on public health in most of the world. Understanding how politics and power worked in those unusual months can provide insight into both what it takes to direct state action, and why politicians make the public health decisions that they do.

Who took charge of the pandemic, when, and why? The question matters because it locates power and responsibility for public health and other measures, which is necessary for influencing policy and holding leaders accountable. We argue that COVID-19 led to significant and dynamic shifts in power relations within and between governments. We have two aims. The first is to define and operationalize centralization and decentralization in the context of the pandemic. In particular, we emphasize that centralization is not only territorial; it can also apply within governments and change the amount of autonomy ministers and agencies have. These constitute the phenomenon we wish to explain. The second aim is to develop and test our explanatory framework: the politics of credit-claiming and blame-shifting. The theory of credit and blame politics builds on the fact that politicians, ones in democratic political systems, have enormous incentive to claim credit for popular developments (such as low infection numbers or effective vaccination campaigns) and shift blame onto others for unpopular developments (such as repeated restrictions on public life or a resurgence of cases). We posit that changes in the power of central governments during the pandemic can be explained by understanding the politics of credit and blame.

1.1. Centralizing and decentralizing

Centralization in government has two dimensions of importance: within and between government. Centralization within government means an increase in the power of the core executive- the head of government and their officials, and in some cases the finance minister or other key generalist central ministries whose task is allocation of priority and money [1,2]. The easiest indicator is usually the extent to which the head of government took charge of communications and decision-making, typically through some combination of ad hoc committees and emergency legislation.

Centralization between governments means an increase in the power of the central government vis-a-vis other “subnational” governments such as regions, states, provinces, or municipalities. "Command and control" is a common recommendation in public health emergencies [3] and central governments do often take powers over or away from subnational governments in crises. This is most politically contentious in federal states such Spain, Canada, or Germany, but can happen even in countries where there is a history of only local government (such as Ireland, Portugal, or the Nordic states).

Centralization between governments may vary widely between countries. Central governments may use their additional power to impose policies on regions and municipalities. For example, the central government may set legally binding criteria for introducing lockdown measures based on incidence or hospital occupation rates. Regional and local governments would still oversee implementation. Or, central government might empower itself to directly give orders to organizations such as police or hospitals that are normally overseen by subnational governments.

Centralization of power may not always yield the expected results and often may be nominal rather than substantial. It may turn out that an ambitious government or prime minister lacks the political influence, legal authority, resources or an adequate strategy to actually centralize and hold power. In general, it makes methodological sense to treat centralization as an intention to claim authority; its empirical success should be studied as a separate problem since it is easier to claim authority than to grasp and wield it.

1.2. Credit and blame

It is almost axiomatic in political science that politicians seek credit and avoid blame [4,5]. If something seems popular, they try to take credit for it; and if there is something unpopular, they will try to avoid blame and, if possible, cast the blame onto opponents. If good or bad outcomes cannot be traced to their actions, they will opt for "position-taking" in which they declare their fidelity to what they see as popular positions [6].

The details of this basic calculus will vary by not just political system but also the individual party and politician. For example, some parties are not generally equally associated with the issue of health policy, and so their politicians had the problem of finding something to say that would redirect the conversation onto issues where they performed better in public opinion. Many politicians of the populist radical right had problems working out how to position themselves early in the pandemic, opting to try and relate it to issues that worked well for them such as immigration, though as the pandemic wore on the ones out of power increasingly focused on blaming incumbent governments for public health measures [7].

Understanding the political likelihood and sustainability of different policy options means understanding the likelihood that incumbent governments can take credit and avoid blame for them, and the likelihood that opponents will be able to portray the policy options as somehow blameworthy (e.g. ineffective, authoritarian, unscientific). As many in public health know, it is difficult to persuade politicians to risk blame, which is why public health advice to react early and forcefully to communicable disease risks so often goes ignored. Persuading politicians to do something which risks blame involves finding ways to diffuse the blame so that it is not traceable to anybody, find a way to blame others, persuade them that the people who will blame them would not vote for them anyway or turn it into credit (e.g. portraying harsh measures as a sign of responsible statesmanship). Early studies of the politics of COVID-19 response have indeed found that credit and blame shed light on governments' decisions [8], [9], [10].

We understand the pandemic politics of credit and blame in light of both the epidemiology of the COVID-19 pandemic and the well-established trajectory of disaster response. Europe's pandemic began with a spring first wave followed by a largely low-prevalence summer and a serious second wave in autumn, with a spring third wave underway at the time of writing. The "phases of disaster" approach from psychology, buttressed by much historiography, finds a consistent pattern in which a disaster leads to first a "honeymoon" period of solidarity, creativity, and excitement (such as nightly applause for health care workers) followed by a period of disillusion and resentment [11]. In Europe, the honeymoon largely corresponded to the first wave, and the period of disillusion to the second and third. (From the start of 2021, vaccination drives created new potential policies and options for credit and blame, which is why we end the analysis in March 2021).

In the first wave / honeymoon period there was a lot of credit to be had for decisive action. In the summer, there was a lot of credit to be had for declaring victory and opening up and blame to be had for maintaining unpleasant lockdowns. In the autumn, there was mostly blame: blame for rising case counts and mortality and blame for new public health measures Table 1. maps out the conceptual relationship discussed in our three case studies.

Table 1.

The interaction of credit, blame, centralization and decentralization.

		Independent variable
		Blame-avoidance	Credit-claiming
Dependent variable	Centralization	AT (autumn), FR (autumn)	AT (spring), FR (spring) CZ (spring)
	Decentralization	CZ (autumn)	AT (summer) FR (summer) CZ (summer)

Source: Authors. Spring runs until roughly mid-May 2020, summer to the start of September; autumn to the end of January 2021.

The politics of credit and blame would suggest that during the honeymoon, when visible heroism could lead to credit, heads of government mostly made claims to central authority and took on very important roles at the expense of ministers, autonomous agencies, and subnational governments. It was, so to speak, time to be a hero [12]. During the summer, when there would be ample credit for low case counts and re-openings, heads of government would also take that credit, maintaining their claims to control but starting to decentralize possible future blameworthy activities such as localized business restriction. In the autumn, when blame abounded, they would not just delay public health measures but also try to decentralize, giving subnational governments and ministers prominence if not power as unpopular public health measures coincided with a painful second wave. Bad press surrounding Europe's vaccination drives in early 2021 would increase politicians' interest in avoiding blame for the pandemic.

2. Methods and materials

Political power can be hard to track in a comparable way across countries for a variety of reasons. It is difficult to establish similar metrics of centralization and decentralization when countries differ in formal institutions (e.g. legal traditions, judicial oversight, administrative structure) and informal political arrangements (e.g. the relationship between the executive and legislature, or the power of the head of government over their own party). The result is that instead of specifying metrics a priori, it is safer to combine the inductive HSRM approach with case studies.

We therefore test our framework in two parts. For the first, we report a broad cross-country comparison of centralizing measures (N=52). We coded responses in the Health Services Response Monitor (HSRM) produced by the European Observatory on Health Systems and Policies covering policy between 1 March 2020 and 1 March 2021 [13]. Two authors coded all the country responses by time and whether they centralized or decentralized within the central government or between the central government, cross-checked by a third author. Table 2 summarizes the results.

Table 2.

Austria, Czechia and France in comparative perspective.

	Austria	Czechia	France
Formal presidentialism	No	No (but president informally powerful)	Yes
Formal federalism	Yes	No	No
Head of government's party	Christian Democrat	ANO 2011 (centrist / “technocratic populist”)	La République en Marche
Coalition government?	Yes (Christian Democratic/Green)	Yes (ANO / Social Democrats)	No
Population	8.8m	10.6m	67m
GDP per capita (US$, PPP, most recent)	50,137	23,494	40,439

Source: authors; World Bank for population and GDP.

We combine our broad comparison, based on the whole HSRM, with country case studies. The case studies perform two functions. First, they allow us to capture changes in governance that are slow to appear in HSRM reports or other sources. Committees that cease to meet frequently, lax central monitoring of enforcement, fewer mentions of the issue by senior politicians, and changes in the frequency and line-up of presenters at press conferences all send clear signals of centralization and decentralization but are informal and sometimes difficult to interpret without contextual knowledge. Even emergency laws or invocation of emergency powers can be difficult to interpret without country context. In short, the case studies allow us to partially compensate for potential underreporting of change.

Second, they allow us to understand how centralizing and decentralizing measures operate in context and their political meaning. It is an established method in comparative politics to test the causal links posited in a broad study with comparative analysis of individual cases that allows more precise identification of mechanisms and shows how the argument works in different contexts [14]. Our three cases are Austria, Czechia, and France. We chose these cases to illustrate these dynamics across countries with different experiences of the disease, different political institutions and party systems, and different politicians. Our case selection method is therefore a variant of Mill's method of Agreement, in which causal inference comes from finding similar outcomes in different situations, and it is buttressed, as in good case study research, by deeper research into the presence or absence of similar mechanisms [15,16]. Other literature identified two key institutions that have shaped responses and which influenced our case selection: presidentialism and federalism [17,18]. By presidentialism we mean cases where an "executive with considerable constitutional powers -generally including full control of the composition of the cabinet and administration- is directly elected by the people for a fixed term and is independent of parliamentary votes of confidence" [19]. By federalism we mean the presence of multiple levels of general-purpose elected governments with constitutional status which cannot abolish each other [20]. In the scope of wealthy and largely comparable European countries, the three differ significantly, as shown in Table 2. Broadly, an argument that applies in all three should have a strong chance of being applicable to any other wealthy democracy and might be applicable beyond them.

3. Results

3.1. Comparative analysis

The summary of our comparative finding is presented in Table 3 , which shows changes reported in the HSRM. There was, indeed, considerable centralization within and between governments in the first wave (roughly March through mid-summer), indicating that heads of government were putting themselves forward as leaders, including in countries so small (San Marino) or with such limited local government (Ireland) as to limit possibilities for intergovernmental centralization.

Table 3.

Centralization and decentralization by country and domain of intervention.

Domain of intervention		Centralization within government (spring/summer)	Centralization between governments (spring/summer)	Centralization within government (autumn/winter)	Centralization between governments (autumn/winter)	Decentralization (any kind) (autumn/winter)
Governance	Interministerial committee, Coordination agency, National security council	AL, AM, AT, BA, BE, BG, CA, CH, CY, CZ, DE, EE, FR, GE, GR, HR, IE, IL, IT, KG, LT, LU, LV, ME, NL, PL, PT, RS, RU, SK, UA, US	BE, DK, FR, KZ, RU	DK, ES	BE, BG, RS	BE, BG, CA, CH, DE
	Expert/Vaccine committee	AL, BE, BG, CA, EE, FR, HR, IE, MK, SI, TR, US	CA	FI, FR, IE, LT, LU, NL, RU
	State of emergency/ Emergency Laws	AM, BA, BE, BG, EE, ES, FR, GE, HU, LU, MD, MK, MT, PL, RO, RS, US	CH	FR, PT, ES
	Centralized governance of the healthcare system		AT, DE, IT, LT		LT
Preventing transmission	Health communication	GR, HU, IT, ME, RS
	Physical distancing		BE, CA, DE			DE, FR, ES
	Contact tracing		ME
	Isolation and quarantine		FI			UA
	Monitoring and surveillance, Contact tracing, Reporting cases and hospital capacity	BE, BG, CY, DK, IL, MK, PL, RO	AT, BG, CA, CH, DE, ES, GB, IT, MK, NO, PL	IE	DE	DK, NO
	Testing	GR, IE, IL, ME	LT, ES		DE	IT
Physical infrastructure and workforce capacity	Physical infrastructure	DK, HR, IE	DE		DE
	Workforce				DE
Providing health services effectively	Planning services	BE	IT, RO	GR		BG
	Managing cases		LT
	Maintaining essential services
Paying for services	Health Financing	BE, ME	CA, DK, FI		BG, CA, CH, ES, RU
	Entitlement and Coverage

Centralization was much less evident in the second wave. This is partly because in many countries centralizing policies had not ended, but also because in many cases heads of government deliberately promoted localizing measures that made subnational governments explicitly responsible for public health measures. Late in the pandemic, only Germany centralized significantly and across policy areas. They did this, in part, because Germany's winter waves were worse than those of spring and summer, which created blame and intersected with a perceived poor performance in vaccination as well as a struggle for succession among ruling Christian Democrats. This undermined previously successful horizontal coordination.

Overall, the picture in autumn and winter was mixed, with relatively tardy, sporadic, and sometimes incoherent changes in the authority and responsibility of different actors within and between governments, reflecting heads of government and others' efforts to deflect blame for public health measures, economic damage, popular frustration and rising case counts.

3.2. Case studies: Austria, Czechia, and France

This section tests our argument at the necessary detail in three different, diverse, political contexts. Their policy stringency scores and excess all-cause mortality over prior years (2015- [19]) are presented in Figs. 1 and 2 for reference.

Fig. 1.

Policy stringency over time, Austria, Czechia and France. Source: Hale et al. 2021.

Fig. 2.

Excess mortality, all ages (% change over 2015-19). Source: Our world in data.

3.2.1. Time to be a hero: the first wave

In the first wave, the "honeymoon" period of disaster phases coincided with a poorly understood and dramatic period of rising cases in many countries. Heads of government saw the risk of massive blame- for failure to respond- and the chance of gaining credit for forceful responses.

3.2.1.1. Austria

Austria is a federal country where health competencies are split between the Federal government and the country's nine Länder. The Federal government has the exclusive task of passing and executing laws concerning public health, which includes managing the prevention of epidemics and pandemics. The one area the Federal government has no say over is hospital care. The nine Länder are each responsible for their own hospital and municipal sanitation organization.

On the 25^th of February 2020, the first COVID-19 case arrived in Austria. In a matter of days, the Minister of Health, Rudolph Anschober (Green), the Minister of Interior, Karl Nehammer (new Austrian People's Party - ÖVP) and most importantly, Chancellor Sebastian Kurz (new ÖVP) united as a central force and led the country through the first wave. These men, including, at times, vice Chancellor Werner Kogler (Green), were the faces of the crisis management, giving almost daily press conferences and announcing measures that would protect the country during the 1^st wave [21]. By the evening of the 25^th, Health Minister Anschober had announced a “Corona Hotline” as well as a general “Health Hotline” and that 59 hospitals were prepared to treat patients with the virus [22,23]. Also, on the 25^th Minister Nehammer assured the public that all ministries were working together around the clock to ensure the safety of the citizens. Two days later, Chancellor Kurz presented the contact tracing strategy that officials would follow. On March 13^th, the trio led by Chancellor Kurz held a press conference announcing the 1^st lockdown, which would go into effect on the 16^th of March [24]. The very next day, the Social Partners, Chancellor Kurz and several other Ministers announced that €4 billion would be set aside to help businesses. Two days after the commencement of the hard lockdown on March 18^th, the government presented a €38 billion aid package to “save workplaces”. Chancellor Kurz stated, “no matter the cost, we will do whatever it takes to secure jobs” and vice Chancellor Werner Kogler (Green) assured citizens that “no one would be left behind” [25]. The government's quick actions and assurances garnered it much praise from citizens and other governments alike [26].

3.2.1.2. Czechia

Similar to other Central and Eastern European countries, Czechia's number of cases and deaths per capita during the first wave were some of the lowest in Europe [27]. The Czech Prime Minister, Andrej Babiš (of the centrist populist ANO) was in a good position to take credit for these results [28]. Together with his minister of health, Adam Vojtěch, and deputy minister of health, colonel Roman Prymula, an epidemiologist, Babiš fronted the decision to introduce a comprehensive lockdown, as well as to mandate the use of facemasks, at a time when there was only a limited number of known infections in the country compared to Western Europe, and no deaths [29].

The PM followed a centralizing impulse, both at the national level and within government. Public health is an exception to the unitary, centralized nature of the Czech state. Communicable disease surveillance and management had been the prerogative of a network of 14 regional public health offices, which were, however, in March quickly superseded by a country-wide response coordinated by the ministry of health and the government. The initial lockdown decision, as well as subsequent measures, was taken at central government level, without notable involvement of the regional public health offices, the Central Epidemic Committee or other institutions. Babiš interfered with established processes: he hesitated to convene the Central Crisis Team, a governmental emergency task force, and when he did so, he interfered with staffing issues, sidelining the interior minister, who had been instrumental in sounding alarm bells in early March [30,31]. Established processes and institutions were not suspended or dismantled formally; instead, they became increasingly irrelevant as decision-making concentrated at the higher government level, aided by the state of emergency, declared on March 12.

While this level of centralization of power attracted some criticism, the unprecedented context as well as the undeniable results of the timely intervention translated into genuine political credit. The government's measures enjoyed overwhelming popular support: an April survey found 76% of respondents saw them as “adequate” (and 16% as too lenient) [32]. This support rubbed off on Babiš and ANO, as well as on Prymula and Jan Hamáček, interior minister and leader of the junior coalition partner (Czech Social Democratic Party) [32,33].

3.2.1.3. France

Public health policies are traditionally determined by the French government and then implemented by national, regional, and municipal authorities. At the national level, within the Ministry of Health, the General Directorate of Health (acronym in French, DGS) and the General Directorate of Health Care Services (acronym in French, DGOS) are responsible for defining health policies and performing health security tasks. Expert agencies, such as Public Health France, issue recommendations and conduct scientific studies on behalf of the executive branch. At the regional level, Regional Health Agencies (acronym in French, ARS) and municipalities carry out prevention, public health, and planning missions [34].

Regional and local authorities are responsible for tailoring these policies to the specific needs of their constituents. One can therefore expect a degree of variation in the design and implementation of public health measures. During the first months of the Covid-19 pandemic, however, the government centralized the crisis governance and embraced a “one size fits all” approach, adopting a national lockdown and deploying a containment and mitigation strategy that applied uniformly to continental France and its overseas territories. In March 2020, although France had significantly invested in public health and could point to a relatively recent but robust network of health security institutions, the President confined the management of the crisis to a small circle of trusted advisors. An emergency health law adopted on March 23 gave broad powers to the government, including the ability to rule by decrees for several months. Originally led by the Ministry of Health, the pandemic response promptly shifted to the Prime Minister and the President in February 2020. Most strategic decisions were taken within a “Defense council”, composed of experts discretionarily handpicked by Macron, with limited accountability. The Defense council's proceedings were kept confidential [34]. The “state of emergency” allowed the government to rule via decrees and regulatory documents – as opposed to legislative initiatives – for most of the Covid-19 pandemic, on measures restricting the freedom to come and go, the freedom to conduct business and the freedom of assembly as well as on measures leading to the requisition all goods and services necessary to put an end to the health crisis.

Additionally, the government created its own sources of expertise, establishing two scientific councils on March 10 and March 24, staffed by handpicked members [35] . Macron invoked the council's recommendations to legitimize controversial measures, including the upholding of the March municipal elections and the first national lockdown. The government ensured income replacement for those who had no choice but to temporarily stop working and sought to protect vulnerable individuals through financial assistance and housing protections [34].

The government was however criticized for its delayed response to the pandemic due to limited pandemic preparation, surveillance capacity, and supplies such as PPE. President Macron was also criticized for concentrating power and failing to acknowledge the territorial dimension of the crisis. During the first wave, national authorities deployed a “one fit for all” strategy and enacted a national lockdown, even though the virus was not actively circulating in several parts of the country [35]. The opaque and highly centralized nature of the government's decision-making process rankled many local and regional officials. The city of Marseille, for instance, considered creating its own scientific committee in response.

The French government thus attempted to take credit by centralizing the crisis governance, but soon faced mounting criticism from regional health authorities, the population, health care professionals as well as members of the political opposition. These initial missteps prompted legislators to set up two parliamentary inquiry commissions in the summer of 2020. Complaints were filed against several members of government, including former Prime Minister Edouard Philippe and former Health Minister Agnès Buzyn. In July, Phillipe was replaced by Jean Castex, who was tasked with designing France's reopening strategy with greater attention to regional needs and powers.

3.2.2. Declaring victory: the summer

In the summer of 2020 case counts across Europe dropped to very low levels and pressure on health systems eased. As a result, Europe exited both a health care crisis and the solidaristic honeymoon phase in June. The temptation for policymakers was to declare a job well done and lift public health measures despite the ongoing risks (as can be seen in Figs. 1 and 2). In particular, this could mean heads of government taking credit for lifting restrictions and shifting public health responsibilities back to ministries, agencies, and subnational governments.

3.2.2.1. Austria

Due to the decreasing infection rate, Austria was able to begin relaxing measures as early as mid-April 2020. On April 6^th, the trio announced a prospective plan for lifting public health restrictions, beginning on April 14^th. While masks would have to be worn in all stores and in public transportation, the country managed to exit the lockdown much earlier than most countries. By the summer of 2020, the Austrian government had received a trust score of 59%, the highest in years [36]. This increase in trust was likely the result of quick and efficient handling of the lockdown situation in Austria followed by a step-by-step re-opening of the country by the end of May. Worldwide media attention praising the government for their low case numbers and effective management strategy surely helped as well [37]. Chancellor Kurz used the positive media attention to reinforce a sense that Austria handled the COVID-19 crisis well. Along with Health Minister Anschober, high confidence index ratings in February and March [38]. With the beginning of the summer, the confidence index ratings in politicians returned to normal. Kurz and Anschober, whose popularity had spiked in spring, became less visible and their popularity returned to earlier levels [39].

3.2.2.2. Czechia

Like Austria, the Czech government relaxed measures relatively quickly, precipitated by judicial challenges to the procedural aspects of lockdown measures [40]. Thanks to low prevalence, rapid reopening did not lead to dramatic increases in infections. Between June and September, there were virtually no restrictions in Czechia [27]. The government did not fail to claim credit for the country's success. On August 31, 2020, Babiš proudly described Czechia as “best in covid”, linking the country's success to his experience as a political leader with a business background [41].

Once the state of emergency expired, in May, the ministry of health returned responsibilities back to the regional public health structures. This led to a minor, but textbook, tug-of-war over decentralized competences during the COVID-19 crisis in Czechia: in July 2020, the public health office in Northern Moravia, as well as regional politicians (from governmental ANO), accused the health minister of forcing them to impose unpopular new restrictions without due notice, disrupting a large music festival [42]. The ministry had manifestly tried to avoid blame by not giving the regional office an overt order, which to some degree worked: protests were aimed at the regional public health officers instead of the national executive. Pushback from influential local politicians, however, led to the minister apologizing. Subsequent regional measures were taken with the official blessing of the health ministry. Regional differentiation soon became irrelevant due to widespread transmission.

In July, the PM created a new advisory group for health risks and appointed himself as its head. His hands-on approach could at times lead to credit claiming, at least in the short term. In August 2020, health minister Vojtěch reinstated mandatory use of facemasks in public places and in schools. The measure quickly turned out to be unpopular with the public. Babiš met with Vojtěch, who immediately exempted schools and businesses from the new measures and tried to shift blame on his advisors [43]. For Babiš, however, this incident meant he could, even as contact tracing failed and cases soared in September, claim victory and insist that the population should “stop spreading fear” and that the rising cases were due to the country's high testing rates [44].

3.2.2.3. France

The French government's response to Covid-19 significantly changed during the summer. As France transitioned out of the first lockdown, the government worked in partnership with regional and local authorities to implement measures tailored to their needs, shift blame, and diffuse the criticisms prompted by its initial approach. Jean Castex became Prime Minister on July 3^rd and committed to pay greater attention to local actors. They were more frequently consulted by the government during the transition phase (May-August), participated in press conferences co-hosted by the government while regional health agencies were increasingly involved in surveillance efforts, in the organization of Covid-19 patient care, in the supply of medical equipment, and in testing and contact-tracing efforts at the regional level.

During the summer, the Ministry of Health also held several consultations with medical organizations, a summit known as “Ségur de la Santé”, in an attempt to reward healthcare professionals. Public hospital workers were offered financial bonuses ranging from EUR 500 to EUR 1,500. Stakeholders unsuccessfully advocated for “decentralizing” the French healthcare system by strengthening the role of local officials within the Regional Health Agencies [45]. The Ségur de la Santé’s limited impact rankled healthcare professionals nationwide.

Bu the end of the summer, the government was grappling with persistent difficulties, including a lack of legitimacy and preoccupying mortality rates. France's contact tracing strategy yielded limited results. The government's “Stop Covid” contact tracing app was only downloaded 2.6 million times between June and October 2020. The Prime Minister himself failed to download the app, which he mistakenly called “TéléCovid”. Several weeks after restrictions were lifted, infection rates increased from 66,000 per week in early September to 122,000 in early October [45]. The government's failure to put in place a critical triptych of testing, tracing and isolation and to contain the spread of the virus would lead to a recentralization of the crisis governance and a second national lockdown in late October.

3.2.3. Blame and more blame: the second wave

In the second wave, which hit most of Europe in October and spiked in November and December, blame was easier to find than credit. The honeymoon period of the pandemic was long over, and disillusionment set in. Rapidly increasing mortality (Fig. 2) looked like a policy failure, and reinstated public health restrictions ran a higher risk of looking blameworthy. The result was incentive to take decisions tardily- avoiding blame as long as possible- and to try and shift blame onto others, notably by decentralizing responsibility.

3.2.3.1. Austria

By August the trio of leaders seemed to unravel. Minister Nehammer's attempt to increase border controls resulted in confusion at the borders, with politicians and agencies trying to evade responsibility for implementing controls. Disputes between the Länder and the national government increased culminating in a fight surrounding Minister Anschober's proposed traffic light system. Party politics became increasingly salient in intergovernmental relations. The Social Democrat (SPÖ) controlled Länder (Vienna, Burgenland and Carinthia) accused the national government of agreeing to regional diversification regarding corona measures to benefit the Länder controlled by the ÖVP. Vienna's government blamed the national government and demanded a national approach as it became clear that the case numbers were increasing in early September 2020 [46]. The strategy that the government followed in the Spring seemed to have disappeared along with the population's approval.

In an effort to disperse the negative press, the government included the Länder and the social partners in their October 31^stannouncement of a second hard lockdown lasting from November 3^rd until the end of November. This blame dispersal tactic gave the appearance that measures were not just being created federally, but that the Länder were also in accord. While all parties (including those in the opposition) supported the first lockdown, the Freedom Party of Austria (FPÖ) and the New Austria and Liberal Forum Party (NEOS) voted against the second lockdown, insisting that the economic damages would be excessive [47]. Not surprisingly, the population's trust in government began to decrease further [48]. It is important to note that no state of emergency was ever declared in Austria because of the COVID-19 pandemic. Such a declaration would have given the country's Federal President, Alexander van der Bellen, limited power to enact urgent measures, which, at this point, might have been useful.

The descent into a politics of blame avoidance accelerated when it was established (end of January 2021) that the government invested in the Astra-Zeneca vaccine, which suffered not only from negative press associated with the side effects of the vaccine but also from production difficulties. The result was that the Austrian government had to change their vaccination strategy several times to compensate for missing vaccines, for which Chancellor Kurz blamed the EU. In culmination, Chancellor Kurz, his party (new ÖVP) and several ÖVP ministers are, as of April 2021, being accused of corruption, party book economics and perfidy. What started out as a centrally managed success story descended into a typically Austrian Fleckerlteppich (hotchpotch) of measures and finally deteriorated into a narrative about the destruction of Chancellor Kurz’ well-polished image, increasing cases, the resignation of the Health Minister and corruption scandals within the governmental party.

3.2.3.2. Czechia

In Czechia, the second wave started early and was left to spread with minimal management. By November-December 2020, Czechia had the highest numbers of weekly new cases and deaths per capita worldwide [29]. Babiš’s public response oscillated between denial and blame-shifting, sometimes interspersed with rhetorical apologies and sporadic efforts to shift responsibility for decision-making onto individual ministries. In September, the PM admitted the extent to which restrictions had been eased during the summer was a mistake, but explained his decisions by “societal demand” and prioritization of the economy[49,50]. In another move to deflect blame away from the highest levels of government, he replaced health minister Vojtěch with the popular epidemiologist Prymula in late September. New restrictions, however, were not taken, despite increasingly urgent recommendations of the Institute of Health Information and Statistics.

The government had a strong incentive to delay restrictions: local administration and Senate elections took place on 2-3 October, and with the honeymoon period over, renewing restrictions could have been reasonably expected to be unpopular. Eventually, a new state of emergency was approved by Parliament on 30 September, but even moderate restrictions only came into force on 9 October. In a clear blame-avoiding logic, they were introduced by Prymula, rather than Babiš, who had been avoiding government press conferences since late September [42]. In the following months Babiš blamed the health ministry [51], unnamed experts who, he claimed, declared the coronavirus weaker [52], experts who “failed to come” with their predictions [29], and finally, the population, who he accused of not respecting restrictions [53].

Babiš continuously underplayed the gravity of the situation. Denial might have taken away some of the need for deflection. In October, he refused to say there had been any “mistakes” or “problems” and re-emphasized his personal, daily involvement in the management of the pandemic [54]. At the same time, intra-governmental disagreements became publicly discussed (typically between the ministry of education and economic portfolios) [54]. This presented an opening to disperse responsibility within government, but, in December, the PM personally took on the role of the national coordinator for vaccination and again announced he would take full responsibility for its success [55], despite the job's significant potential for failure. He ended up blaming the European Union for not approving vaccines faster and scolded a regional administration leader for complaining about inequitable distribution of vaccines[ 56,57]. Blanket denial of the problem or failed solutions in addition to occasional blame-shifting, then, was ANO's approach. In February 2021, 76% of Czechs were dissatisfied with the government's handling of the pandemic [58].

Explicit blame avoidance through decentralization of responsibility between governments came only one year into the pandemic. In February, opposition parties refused to prolong the state of emergency. The government then turned to regional governors, who would become responsible for pandemic management together with the regional public health offices. The governors (mostly belonging to opposition parties) requested that the government declare a new state of emergency [59]. In addition to an alleged lack of legal competences and organizational capacity to deal with a public health threat of such scale, the governors explained their decision by the financial responsibility for potential damages regions might face if they had to individually declare a “state of danger” as per the Crisis Act [60]. The government obliged and, unconstitutionally, declared a new state of emergency [61], hoping the governors’ demand would attenuate some of the political, if not legal, blame for disregarding lawful processes. A year after near-total centralization of responsibility for pandemic management, regional governors were co-opted into the government's blame game that crossed party lines, ahead of the October 2021 parliamentary elections. Clearly, regional leaders did not see much credit claiming opportunity amidst case and death tolls topping global charts and the government considered potential consequences of inaction, including blame, more costly than the risk of blame for an unlawful power grab with some signs of legitimacy.

3.2.3.3. France

Although the first lockdown was deemed “necessary” by most French people, implementation of stringent measures was less accepted by the population after a resurgence of cases at the end of the summer. The President announced a second nationwide lockdown, effective on October 29, as a second wave of Covid-19 threatened to overwhelm hospitals and the number of cases totaled almost 1.3 million since the beginning of the pandemic. Mental health professionals warned about the deteriorating mental states of young adults and urged the government to reopen universities to fight loneliness.

The winter of 2020-2021 was characterized by a hybrid strategy of recentralization and then decentralization of the decision-making process. In contrast to some of France's neighbors, President Macron refused to impose a third national lockdown in late January 2021, despite alarming infection rates, arguing that this “divisive” measure would not be followed by the population. Instead, in partnership with regional authorities, the government enacted territorialized measures such as local curfews, lockdowns, and closures. This approach, which bears witness to the government's attempt to avoid being blamed for a sharp increase in Covid-19 cases in early 2021, ultimately failed to contain the spread of the virus. On March 31^st, however, the government imposed a third national lockdown as Covid-19 surged in several areas, a decision deemed tardy by public health and healthcare professionals and members of the political opposition.

France began its vaccination campaign on December 26th, but it was criticized for slowness. President Macron was blamed for the slow roll out of the vaccine. In addition to France's slow vaccines campaign, the Institut Pasteur and Merck announced their decision to abandon their Covid-19 vaccine based on measles jab on January 25, a significant blow to France's scientific prestige and efforts to contain the pandemic.

As France entered a third lockdown in late March 2021, Macron's popularity rate fell 4 percentage point from a month earlier. A poll released in February 2021 revealed that the government's highly centralized management of the crisis “widened” the gap between citizens and their elected representatives. According to the Cevipof Political Confidence Barometer, 41% of the surveyed population said they experienced “weariness” the fall of 2021 [62].

4. Discussion

Our comparative and case study findings suggest that the politics of centralization and decentralization are important in understanding pandemic response, and that these dynamics can be explained by politicians' understandings of the politics credit and blame. Centralization happens within and between governments, and is a tool that leaders can use to seek credit and avoid blame. As pandemics wear on, and tolerance for public health measures wanes, there is more blame to go around and more incentive for politicians to centralize or decentralize in order to avoid blame and reap what credit there is. Problems in the vaccination campaign of spring 2021 also created more blame, which most member states tried to deflect onto the European Union (and which EU leaders tried to deflect onto Astrazeneca and other countries, pointing out the vaccine nationalism of the US and UK).

Our paper has three key limitations. First of all, we tested our analysis in rich, democratic countries. Politicians in more authoritarian regimes still seek credit and blame, but their strategies are more opaque and the audiences they seek to please are smaller, while politicians in less wealthy countries might lack the range of social and public health policy options that the politicians of these rich countries could deploy. Second, we identify a broad logic of domestic politics, but the idiosyncracies of individual countries and politicians still matter. Politicians can adopt credit-seeking and blame-avoiding strategies that look ill-judged to health policy specialists or even political scientists. Ours is an argument about political incentives but it cannot predict how politicians will interpret or weight those incentives. Third, not all federations are equally resistant to centralization. Another study of COVID-19 in European federations found, in our reading, that while there was centralization and a focus on the head of government everywhere, Belgium and Germany were both more resistant to central coordination and decentralized more quickly; our tables suggest the same [63].

5. Conclusions

Research and policy advice in public health should account for what the pandemic has shown: that centralization and decentralization within and between governments can change sharply in a pandemic, though often temporarily or with unexpected consequences; and that the politics of credit and blame shape politicians' approaches to problems in complex and context-dependent ways. It is a long tradition in public health to lament politicians' interest in credit and blame, or the difficulty of achieving whole of government policies; the pandemic taught us a great deal about how and when politicians' incentives can work to support or undermine public health policies and government coordination.

Declaration of Competing Interest

SLG, SR and MF receive funding support from the European Observatory on Health Systems and Policies. SLG and HJ have acted as consultants for WHO Europe. HJ and SLG receive funding support from the US Army Corps of Engineers, Engineer Research Development Center. OL, NW and MW have no interests to declare.