key: cord-326220-uosjmht0 authors: Gaskell, Jen; Stoker, Gerry; Jennings, Will; Devine, Daniel title: Covid‐19 and the Blunders of our Governments: Long‐run System Failings Aggravated by Political Choices date: 2020-08-11 journal: Polit Q DOI: 10.1111/1467-923x.12894 sha: doc_id: 326220 cord_uid: uosjmht0 More urgently than ever we need an answer to the question posed by the late Mick Moran in The Political Quarterly nearly two decades ago: ‘if government now invests huge resources in trying to be smart why does it often act so dumb?’. We reflect on this question in the context of governmental responses to Covid‐19 in four steps. First, we argue that blunders occur because of systemic weaknesses that stimulate poor policy choices. Second, we review and assess the performance of governments on Covid‐19 across a range of advanced democracies. Third, in the light of these comparisons we argue that the UK system of governance has proved itself vulnerable to failure at the time when its citizens most needed it. Finally, we outline an agenda of reform that seeks to rectify structural weaknesses of that governance capacity. THE COVID-19 PANDEMIC has confronted every government in the world with a wide-ranging set of urgent challenges and policy dilemmas. Some governments are considered to have handled the crisis better than others, prompting debate over the reasons for variation in performance. This article does not claim to provide a definitive answer to why some countries have fared better (so far) than others during the crisis and has no crystal ball as to what the future might hold for the virus. Rather, it argues that to assess the performance of governments we should focus on how the structural features of governance systems interact with, and amplify, missteps in policy choices made by political leaders and their advisors. To develop our argument, we proceed as follows. Firstly, we develop the idea that governance failures reflect not only policy choice errors but also structural features of governance systems that make shortcomings more likely. Secondly, we briefly assess the performance of national governments on Covid-19 in a range of European democracies and show that other comparable countries, with similar or less warning of the dangers of the virus than UK, achieved better outcomes. We argue that outcome was possible in part because those countries had governance systems with a greater depth and spread of decision-making points, resources, and the capacity to use them. No governance system is perfect, but some perform better than others and some are better able to mitigate the ever-present possibilities of deficient political choicesespecially in times of crisis. The UK population has suffered an excess mortality rate owing to the Covid-19 pandemic in the first half of 2020 that was among the highest for comparable developed nations. Collectively, those deaths represent a governance failure that demands we break from a history where policy disasters, failure, catastrophes, fiascos, and blunders have been a recurrent theme in assessments of British government. 1 Our evidence and argument might be used to apportion blame, but it would be better deployed to encourage thinking and action about a radical reform of our system of governance The Political Quarterly and we conclude with suggestions for that process. A framework for understanding governance failure We need an answer to the question posed by the late Mick Moran in his essay in The Political Quarterly nearly two decades ago: 'if government now invests huge resources in trying to be smart why does it often act so dumb?'. 2 A useful framework based on an extensive reading of the comparative literature on policy failure is provided by Jennings, Lodge and Ryan. 3 They identify factors that account for the propensity of governments to commit policy blunders and draw a distinction between those based on the intentional choices of individuals or groups of individuals and factors that reflect underlying structural features of systems of governance. Blunders often occur because politicians and other policy makers take the wrong decisions (for example through over-or under-reaction to problems, or owing to preoccupation with ideological and symbolic concerns) or because the wrong policy tool is adopted to meet the challenge. Sometimes policy choices are made with good faith and in the light of the limited evidence available, but turn out with the benefit of hindsight to have been mistaken. Sometimes decision makers persist with courses of action at odds with available evidence and advice, against even their own interests, in the hope of turning failure into success. These choice-based causes of failure are widely recognised and often the subject of formal inquiries or media investigations, and are a focus for a politics of blame or, more positively, accountability at the ballot box. The focus on structures, however, is equally important and may hold the key to answering the question as to why blunders seem to be repeated regularly. How could structures or systems of governance affect the propensity to blunder? Structures of governance frame the way that decision makers work and operate, what they can see and hear, what tools they think they have available to them, what resources they view as at their disposal, and how they view other actors in the system. Different kinds of institutions allow individuals to think different kinds of thoughts and may indeed construct 'thought styles' that are self-reinforcing, blocking out potential challenges or collaborations. 4 All governance systems are multilevel and involve a mix of institutions and a distribution of power that can be either more centralised or decentralised. There is no perfect system of governance, but failure is more likely to occur when a structural weakness is compounded by a thought style that is problematic. Table 1 highlights some prominent causes of governance failure that reflect these connected dysfunctional elements that in turn lead to policy blunders. Let us examine the dynamic of each of these causes in more detail. There is a structural difference between governance arrangements that decentralise power within the system and those that concentrate power at the centre, as the UK tends to do. Governance systems can be highly centralised and one advantage of this, it is claimed, is that they can make decisive and timely interventions. Strong coordinating capacity is highlighted by many as essential to managing the Covid-19 crisis. Yet if coordination is replaced by either over-confident or ineffectual central direction these systems can be prone to blunders when combined with decision making affected by group think and defensiveness when challenged. Other problems include over-confidence of governing elites driven by a lack of understanding of operational issues and limited access to operational know-how. These issues can be compounded by losses of institutional memory caused by perpetual administrative reorganisations and shrinkage of the central state machinery, which creates a recipe for policy blunders. A second structural feature rests on the distinction made by Hooghe and Marks between 'Type I' and 'Type II' multilevel governance. 5 The first type reflects federalist thinking and disperses authority to a limited number of discrete, multipurpose governing units. The second type has different kinds of decentralised units constructed around functions and tasks, creating a complex mix of responsibilities and overlapping jurisdictions. Most governance systems involve some mix of Type I and Type II divisions. The emerging post-devolution settlement system in the UK creates an especially complex hybrid of these two multilevel governance models, with a strong emphasis on functional division combined with territorial autonomy granted in different forms to Scotland, Wales and Northern Ireland. Such systems are characterised by functional divisions where some powers are 'devolved' (differently across the three nations in the case of the UK) while some are 'reserved' at the centre. What are the strengths and weaknesses of the two intergovernmental systems? Type I systems, given the power vested in autonomous decentralised units, can be vulnerable to inertia and gridlock owing to obstruction by 'veto players' at different levels of government. Pride in local, regional, or national autonomy, and a desire to express difference is the thought style commonly associated with Type I systems. But a relatively small number of decentralised units and potentially effective integration does offer capacity for dialogue and negotiation between government levels, if participants are willing. Type II systems can be flexible and provide an opportunity for matching task to scale in a more varied way. The comparative advantage of these institutions is their technical expertise and concentrated focus. But the system in which they operate can suffer from complexity and proliferation of the number of institutions that might be involved in a cross-cutting task. Type II systems tend to encourage silo thinking in that they are set up to operate within their functional domain rather than to reach out beyond it. Coordination problems and weak communication between different parts of government are common in Type II systems. Hybrid systems such as that of the UK are even more vulnerable to failure where policy issues cut across both devolved and reserved functions, such as Brexit for example and the Covid-19 crisis. 6 Connected to this last point is the argument that all systems of governance need effective joining up, and a third structural cause of failure is attributable to deficiencies in the connecting of separate spheres of governance. The division essential in any governance system is best when matched with collaborative capacities to share and incorporate knowledge relevant to policy making. 7 But collaboration is not easy and requires both partnership institutions, resources, and commitment and capacity for learning and adaptability. It also requires trust between levels of government as well as between government and citizens and sectoral interests such as businesses or wider civil society organisations. The relevant competences and technical capacities to support collaboration require shared values and mutual respect. Failures of collaboration can arise because of breakdowns in information sharing and can be compounded by a thinking style reflecting a lack of trust and respect between actors, all of which can lead to governance failure and blunders. The fourth and final structural weakness is weak capacity for policy experimentation and adaption in response to heterogeneous local contexts. Decision making benefits from a diverse mix of participants, thereby reducing group think. Lu Hong and Page use mathematical modelling to show that when facing complex challenges, the best strategy is to not to rely on a team of 'the brightest and the best', but instead pick as diverse a group of problem solvers as possible. 8 Why? Again, the reason has to do with the way that institutions and contexts make people think. A centralised collection of similar, talented people will generate ideas and strategies of quality, but will tend to believe that one-size-fits-all and be prone to group think. A more diverse group will use different heuristics depending on their backgrounds, search more widely and with greater variety of outcomes for solutions, and will build their searches on the back of local knowledge and understanding that is not available to a group of the brightest and best. More institutionally focussed studies consistently demonstrate the advantages of power sharing and what the Nobel Prize winner Elinor Ostrom refers to as the advantages of decentralised systems nested in a polycentric system. For Ostrom these advantages, established through painstaking research, include use of local and disaggregated knowledge, inclusion of trustworthy participants, and lower enforcement costs. 9 A capacity for local experimentation and a willingness to celebrate difference and diversity would appear to be a system-enhancing feature and their absence a potential cause of system failure. Having established a framework for thinking about the causes of governance failure, we need to establish some assessment of the performance of different systems in the unfolding of the Covid-19 crisis during the first half of 2020. We start, in Figure 1 (a), with data relating to the outbreak in the UK and fifteen other west European countries (France, Germany, Italy, Spain, Austria, the Netherlands, Ireland, Portugal, Finland, Norway, Greece, Belgium, Denmark, Sweden, Switzerland), with the UK's trajectory indicated by the black line and the other countries shown in grey (since our focus is on the comparative performance of the UK). We choose these countries as advanced democracies that are geographically proximate to the UK and broadly comparable in terms of the level of economic development and the public health systems. The figure plots the number of reported cases of Covid-19 in each country from January to June 2020. This reveals that while the outbreak in the UK initially lagged behind three other countries (Italy, Spain and France), by mid-May it had overtaken its European counterparts-with the slope of its curve steeper at the end of the period, indicating that the growth of cases was declining at the slowest rate. It is also possible to compare the timelines and restrictiveness of the containment policies put in place by the governments of these countries, using the Oxford Covid-19 Government Response Tracker (OxCGRT). This systematically collects information on common policy responses and calculates a standardised scale of the stringency of government measures. The index is plotted in Figure 1 (b) and highlights the different timings and levels of stringency of containment measures across western Europe. This tells us how quickly the country shut down, in the context of the severity of the health impact of the virus. What is most striking from this graph is that, with the exception of Sweden, the UK was the last major European country to introduce significant restrictions on social and economic life-waiting until 23 March to do so, when other countries had already enacted substantial containment policies by 18 March at the latest. This delay is more notable if one considers that even by 18 March the UK was sixth in terms of cases (1,950) and fourth in terms of deaths (81). It also highlights that it began to ease its shutdown around the same time as many other countries, despite having a considerably higher number of cases. In broad comparative terms, this evidence suggests that the UK has performed badly during the pandemic, an observation supported by a death toll from Covid-19 that is at the top of those recorded by developed nations, notwithstanding some differences in counting practices. A quick look at other cases suggests how things might have been different, without suggesting that any country met the challenges without problems or stumbles. With one of the largest ageing population in Europe and a healthcare infrastructure weakened by a decade of austerity, Portugal took the threat posed by the virus very seriously and moved quickly to close schools and impose a more general lockdown on 16 March when it had recorded just 245 cases, although an outbreak of cases from late May onwards has damaged its reputation as a 'safe destination' for visitors to some degree. One commentator suggested that strong central direction was crucial for achieving this outcome: 'Although the epidemic is concentrated in the north, around the city of Porto, Portugal's centralized system of government allowed for rapid nationwide measures rather than the piecemeal regional action adopted elsewhere.' 10 Throughout the crisis, the Portuguese government applauded citizens' 'civic spirit' in complying with strict containment measures that were clearly and consistently communicated to them, including the early days of easing restrictions on a regional basis in light of differing requirements. The government also benefitted from the collaborative approach taken by the opposition, with Social Democratic Party (SPD) member of parliament Ricardo Baptista Leite observing: 'We're no longer dealing with the Socialist government; we are dealing with the government of Portugal', further noting 'We are having very candid but supportive meetings to present constructive criticism and also solutions through back channels, showing a sense of unity for the national interest.' 11 In Denmark, Prime Minister Metter Fredriksen delivered her lockdown address to the nation on 11 March when the country had around 500 recorded cases, closing schools against the recommendation of its health authorities. 12 Denmark has a well-developed multilevel governance system with decentralised power to regional and local authorities. In this case, strong central direction was combined with a reflective, consensual style of leadership. In her announcement, Fredriksen urged citizens to 'stand together by keeping apart', a message that struck a chord of national solidarity among Danes. She also acknowledged that the country was entering uncharted territory and took responsibility for 'undoubtedly making mistakes along the way'. 13 Neighbouring Sweden is an interesting case because its light-touch response has been at odds with other European countries. Indeed, its schools and industry have remained largely open throughout the pandemic, with central government providing support to furloughed workers and healthcare provisions delivered by regional and local authorities. In this way, Sweden offers a model of coherent strategic direction combined with a coordinating approach by the centre, strongly influenced by its chief epidemiologist Anders Tegnell. However, in contrast to other Scandinavian countries its Covid-related fatalities passed the 5,000 mark in mid-June, and its strategy has been widely criticised as having failed to manage down the number of cases while experiencing the same economic downturn as countries that pursued stricter lockdown measures. While Jon Pierre argues that the strategy was premised on erroneous assumptions regarding asymptomatic transmission, he also describes a failure of governance arrangements in the lack of capacity at regional and local levels in terms of healthcare staff training and the absence of communication channels between different governance levels which were traditionally perceived to be successful within their functional remits. 14 When this perceived competence broke down in the crisis, the lack of mechanisms for collaboration and mutual learning, according to Pierre, resulted in 'poor performance of nursing homes and other elements of elderly care' which only became apparent to the government months into the pandemic. 15 Some countries who did not benefit from a strong central direction and its perceived ability to provide swift, decisive actions were able to leverage other aspects of their more decentralised systems. Germany for example suffered early delays in implementing any nationwide policy. Unlike Denmark and Portugal, which implemented strict measures before any Covid-related deaths, Germany only closed schools three days after its third death was recorded on 12 March. It then suspended public events and closed its land borders four and eight days later, respectively. 16 Germany's relative success so far has been credited to two main factors: Chancellor Angela Merkel's leadership style and the country's decentralised response to the crisis. 17 The country's leading coronavirus expert has stated that besides being extremely well-informed, Merkel's successful steering of competing regional and local interests through the crisis can be attributed to her leadership style of 'thoughtfulness and ability to reassure', as well as not seeking to use the crisis as a political opportunity. 18 A spokesperson for the German federal government also attributed the country's low mortality rate to 'a confluence of many other factors. These include the country's federal system of government, which means there are hundreds of health officials overseeing the pandemic response across the 16 states, rather than one centralized response from the country's national Health Ministry'. 19 Switzerland suffered similar delays to Germany at the very start of the crisis as a result of its highly decentralised federal system. However, as the pandemic developed, it benefitted from inclusion of a broader range of stakeholders in decision making. Indeed, all three levels of governance-the federal, cantonal (regional) and communal (equivalent to county or city councils)-mobilised resources to tackle the outbreak. Each commune has the power to issue its own specific guidance in line with federal and cantonal directives. According to the Swiss press, communes are at the frontline of the response effort. 20 In Switzerland, mutual learning is institutionalised in policy making at the cantonal and communal levels which continuously informs and is integrated into central decision making. During the crisis, these consultation mechanisms were accelerated rather than discarded, taking place weekly and feeding into crisis decision making. This seemed to have been further strengthened by a commitment to different approaches and providing local governance centres with the freedom and resources to undertake what they feel is needed for their communities. The Commune of Bovernier, for instance, a small village with 900 inhabitants in the Canton of Valais, decided to call each of the 110 households with a resident over the age of sixty-five to arrange food and medicine deliveries. 21 In Geneva, ORCA's (Organisation in case of catastrophe and extraordinary situation) immediate focus was on job security and support for commercial organisations. In this way, different approaches are encouraged for their localised capacity to innovate rapid, relevant responses, as the realisation emerged among various stakeholders that 'one-size might not fit all'. These examples highlight several points. Firstly, it is not necessarily the formal constitutional arrangements of a governance system (federal or unitary) that matter as the capacity to work through and with the system in place-an argument behind the framework put forward by Jennings, Lodge and Ryan. Secondly, it is not only that some countries adopted better tools than others, but also the case that some proved better at implementation. Crucially, countries needed to avoid the drivers of governance failure identified in Table 1 : not relying too much on central direction or allowing rigidity in structures to undermine dialogue. They need also needed to be capable of effective collaboration and local adaption to circumstances. Most countries partially failed at least one of those hurdles. The UK, as we shall see, failed all four. In this section we outline how structural failings and dysfunctional policy choices have combined to produce such a flawed response to the Covid-19 crisis in the UK. Following the framework presented in Table 1 , the UK suffered in part because of over-weening and ineffectual central direction. An initial list of where these issues were to the fore would include: a sluggish approach to implementing lockdown measures; catastrophic shortage of personal protective equipment (PPE) for frontline medical staff; failure of the government's 'Ventilator Challenge' to deliver new machines to meet expected increases in demand; confusing communications that have led to a chaotic experience of easing lockdown; no clear plan in place for the reopening of schools to all pupils in England; and the continued failure to deliver an operational 'test, track and trace' system that is widely deemed essential for safe easing of lockdown measures. The centre of British government lacks operational understanding or links to learn from those at street level and in other centres of government. Perhaps the most consequential decision reflecting that structural weakness at the early stages of the crisis was the discharging of elderly patients from hospitals to care homes without mandatory Covid-19 tests-until the requirement was introduced on 15 April. In March, NHS England and NHS Improvement had advised hospital trusts to expand critical care capacity (a major concern based on the Italian experience) by freeing up hospital beds-which contributed to increased discharge of patients to care homes who were potentially carrying the disease. Between 9 March and 13 April, one in five care homes (around 3,500 in total) reported at least one outbreak of Covid-19 to Public Health England (PHE). There were also signs of over-confidence and group think at the heart of British government. On 3 March, as Italy reached seventy-nine deaths, the Prime Minister reported 'our country remains extremely well prepared', and (contrary to the guidance from scientific advisors) declared 'I'm shaking hands'. This proved symbolic of the casual approach taken by the government (which may have been shaken when multiple members of the government fell ill with the virus), and Covid-19 was subsequently optimistically characterised by the Prime Minister as 'overwhelmingly a disease that is moderate in its effect'. 22 The group think problem was arguably evident in three key areas: an over-reliance on data-driven modelling in making key policy decisions; a reluctance by the Scientific Advisory Group for Emergencies (SAGE) to consider lockdown measures deemed politically out of scope; and the consistent drive to centralise implementation. The government's reliance on epidemiological modelling to inform its strategy meant it was only on 16 March, after new data suggested that critical care capacity in the NHS could be overwhelmed, that SAGE advised more stringent restrictions be introduced as soon as possible. This was despite evidence being available from Italy at the start of March that one in ten people infected with the virus required intensive care. Well before that, the unusually rapid construction of sixteen temporary hospitals in Wuhan in January should have provided a clear warning of the intense pressures that health services would soon face. Even at this point-before the UK's shutdown was implemented-it was known that authorities in South Korea, Hong Kong, Singapore, and Taiwan had successfully brought the virus under control (with effective test, track and trace operations, and widespread wearing of masks by citizens), yet this did not figure prominently in deliberations of policy responses among UK experts. The 'data-driven' approach taken by SAGE, on behalf of the government, meant that these information signals from more obvious sources were ignored. Coordination problems and low-grade conflicts between levels of government were also constant issues. The devolved governments complained about lack of consultation, as did other public bodies including local authorities, hospitals, and GP services. A constant flow of changing centrally-driven directives to agencies of all types created a sense of confusion and frustration during the initial phase of lockdown. This issue is further highlighted by the way containment measures have been eased across England. After the government announced a relaxation of guidelines on 2 June, the mayors of Greater Manchester and Liverpool held a joint press conference urging a revision of guidance based on localised R estimates. 23 These examples suggest a lack of commitment to sustained consultation, with key stakeholders seemingly excluded from the planning phases of both these strategies. The government's plan that schools would reopen again for all pupils on 1 June offers a notable example, when it was scrapped following opposition from local councils and teachers' unions who argued it to be impossible while maintaining social distancing guidelines. This was further compounded by the ensuing blame game between government on the one hand, and local authorities and teachers' unions on the other, over who was at fault for the failure to get pupils back to the classroom before the summer holidays. Throughout its response, the UK government has been plagued by inconsistent (and some might say incoherent) messaging and implementation. Much of this is the result of a mixture of conflictual relationships at different governance levels, which systematically results in a lack of integration, collaboration and mutual learning, and a weak capacity to experiment and tailor policy to meet the diverse needs of different locations with different demographic makeups, cultures and dominant industries. The ways in which these structural elements of the UK's governance system hampered an effective response to the pandemic were apparent as soon as lockdown guidelines were issued. For a time in March, the NHS 111 service provided conflicting messages to UK travellers returning from Italy to that of ministers. Also, the government guidelines for the construction sector were published and then immediately withdrawn as they were deemed unworkable on construction sites by industry specialists. 24 More directly, early tracking of the spread of the virus encountered difficulties, as despite a reliable test having been developed for Covid-19, PHE had limited capacity to carry out the testing (and that capacity was initially assigned to tests for seasonal flu), and was slow to allow the expansion of testing through labs in hospital trusts, universities and private settings. 25 These issues were compounded by a strong preference for onesize-fits-all solutions which is often contrasted to the successful and decentralised German approach to testing. The UK's governance system's weak capacity to experiment and tailor policy to meet diversity is illustrated in its failure to stockpile PPE for health workers, instead sticking to 'just in time' procurement principles until April-leaving staff in hospitals and care homes to work without protection. 26 Little exemplifies more the confluence of structural failures which led to a national inability to track and trace Covid-19 cases than Ceredigion Council's development of its own in-house test, track and trace system, which contributed to one of the lowest death rates across the whole of the UK. 27 This example shows that local centres of governance can possess a large amount of knowledge and resourcefulness which, if leveraged, can enable responses that benefit local communities and highlight best practice or innovative measures that could be replicated elsewhere. These have been neglected in the UK's response to Covid-19. So far, the UK experience has seen little attempt at allowing local capacity and resources to energise and innovate in responding to the pandemic. Local authorities and the primary health care teams based around GP surgeries have been largely overlooked, despite their local knowledge and contacts, exemplifying the UK's tendency to centralised, one-sizefits-all policy solutions. We have provided only a partial explanation of the weaknesses of the UK's policy response to Covid-19 because we want to focus attention not so much on the complexity of policy choices made in the context of high uncertainty and crisis, but how failings reflect broader structural weaknesses and associated thought styles within the UK system of governance. There is a more detailed, nuanced, and complicated narrative to tell about what went wrong (and perhaps what went right) in the UK government's handling of Covid-19. Certainly, how the government's response was influenced by its ideological instincts and interests, and the nature of scientific advice that it received, seem important factors to understand. But our primary objective has been to illustrate from the experience of other countries that there are better ways to govern (and be governed). Notably, the chief advisor to the Prime Minister has previously made a similar observation about the profound failings of our governing system: One of the most extraordinary aspects of English politics and education is the lack of structured, disciplined thought about what works and doesn't work, and how to build reliable systems that allow improvement in performance . . . Most activity in Whitehall occurs without asking 'who, somewhere in the world, has already solved this problem? ' 28 Maybe it is time to also think about the deep-rooted structural imperfections of our governing system, and the tendency of successive generations of our political class to bandage over them with piecemeal reforms rather than thinking more holistically about why our patchwork of governance lacks coherence and continues to hoard power and resources at the centre. We think, contrary to the arguments of the Chancellor of the Duchy of Lancaster, that neither relocation of the civil service to dispersed geographical locations, nor the fostering of greater intellectual diversity in its makeup by recruiting scientists, data scientists, or mathematicians, provide the answer (though there is merit in both proposals). 29 It is, rather, another example of Mick Moran's charge that by focussing on quick-fix solutions to make government smart (past solutions have included performance management and contracting-out) we run the risk of making it dumber in new ways. The Covid-19 experience already hints at the limits of data science in the context of the absence of good, up-to-date data and subject to conditions of considerable uncertainty. Our agenda for reform, outlined in Table 2 , is both more prosaic and more radical. We identify short-and long-term mitigation strategies. Different approaches and choices by political leaders could provide more consensual leadership, a willingness to share ownership for problems and insights, greater trust, and mutual respect between levels of government and a wider openness to local learning and diversity. These would all represent improvements in the governance arrangements of the UK. Longer-term we need structural reforms that shift both powers and responsibilities, and clearly define and enshrine them on a constitutional basis, rather than muddling through as has long been the British political tradition. We do not offer detailed proposals, but rather map out a general direction of travel. We need more decentralised government with a matching redistribution of powers and resources. That decentralisation must be consistent and comprehensive, rather than uneven, unequal and inchoate. We need much more effective institutions for mutual exchange and consultation between levels and types of government. We need to know much more about what good practice is, what drives it and how it can be replicated. Parallel to that we need to abandon centralised one-size-fits-all thinking and instead recognise the need for local tailored solutions and experimentation. Policy disasters: explaining the UK's record Not steering but drowning: policy catastrophes and the regulatory state Not steering but drowning Comparing blunders in government How Institutions Think Unraveling the central state, but how? Types of multilevel governance Intergovernmental relations in the UK: time for a radical overhaul?', The Political Quarterly Collaborative governance in theory and practice Groups of diverse problem solvers can outperform groups of high-ability problem solvers 281-2; and for further carefully constructed evidence on the advantages of power sharing see P. Norris, Driving Democracy. Do Powersharing Institutions Work? How Portugal became Europe's coronavirus exception First to close-first to reopen: Denmark's gain from virus response', Financial Times Statsminister Mette Frederiksens indledning p a pressemøde i Statsministeriet om corona-virus den 11 Nudges against pandemics: Sweden's Covid-19 containment strategy in perspective How Portugal became Europe's coronavirus exception The Inquiry: Why does Germany have such a low number of deaths from Covid-19? For many, I'm the evil guy crippling the economy"', The Guardian Germany's coronavirus response: separating fact from fiction Les communes en premi ere ligne de la pand emie Prime Minister's statement on coronavirus (COVID-19) North West mayors want to empower communities with weekly publication of R number, to encourage local decision making Construction confusion as UK working instructions published then withdrawn How the UK got coronavirus testing wrong', Financial Times Government ignored warning to stockpile PPE as Covid-19 spread', The Guardian The area of Wales that missed coronavirus-and the simple system it set up The Hollow Men II: some reflections on Westminster and Whitehall dysfunction', Dominic Cummings's blog The privilege of public service', Ditchley Annual Lecture given by Michael Gove This article is part of the work the three-year TrustGov project (https://trustgov.net/) funded by the UK's Economic and Social Research Council (ES/S009809/1).