Responding to the COVID-19 crisis: a principled or pragmatist approach?

ABSTRACT

Uncertainties run deep during a crisis. Yet, leaders will have to make critical decisions in the absence of information they would like to have. How do political leaders cope with this challenge? One way to deal with crisis-induced uncertainty is to base all decisions on a core principle or value. This is what we call a principled approach. The pragmatist approach offers an alternative: an experimental, trial-and-error strategy based on quick feedback. In this paper, we consider both approaches in light of the COVID-19 experience in four European countries. We conclude that the pragmatic approach may be superior, in theory, but is hard to effectuate in practice. We discuss implications for the practice of strategic crisis management.

KEYWORDS:

• Strategic crisis management
• political leadership
• COVID-19
• pragmatism

Introduction: categorizing strategic approaches to crisis management

The COVID-19 crisis has presented political leaders everywhere with extreme governance challenges. They had to contain a mysterious virus, with limited data, widely varying estimates, unexpected capacity shortfalls, and intense contestation about interventions whose intended and unintended consequences were unknown.

The uncertainties ran deep. There was uncertainty with regard to the main channels of transmission, whether successful recovery would lead to immunity from future infections, if and how the virus might mutate. There was uncertainty about the number of people infected, the number of people who would require intensive care, the number of potential fatalities or the long-term effects of the virus on recovered patients (‘long-COVID’), the available capacities in healthcare facilities, or the economic and social consequences that various policy options would imply. Policy-makers were faced with indicators (such as the now infamous reproductive R number) and new policy ideas (‘flatten the curve’, ‘hammer and dance’) that seemed simple but were easy to misunderstand.

The existing playbook on how to manage pandemics was soon found wanting. Yet, critical decisions had to be made. Crisis leaders had to decide whether borders, schools and restaurants should be closed, whether people could go to work, whether businesses should be financially supported. Tough dilemmas had to be resolved: limit hospital access to COVID-19 patients or scale regular care down (sending cancer patients home and postponing operations)? Should leaders go in ‘hard and early’ or try to control the outbreak and keep the economy going? Such decisions had to be made without recourse to accurate information, let alone policy-relevant evidence. Political leaders were navigating the biggest crisis of our times in what crisis researchers call the ‘fog of war’.

We know quite a lot about the decision-making processes that play out under normal conditions in the offices and hallways of executive buildings, giving rise to fateful decisions. While we have quite a few hypotheses about crisis decision-making, we don’t really know how crisis leaders approach this challenge in practice. The COVID-19 crisis provides a unique opportunity to compare and explore how national crisis leaders responded to the same megacrisis, operating under very similar levels of uncertainty.

In this paper, we explore the analytical merits of two ideal-typical approaches to decision-making under conditions of uncertainty: the principled and the pragmatic approach (Ansell & Boin, 2019). The principled approach adopts a guiding principle (‘minimise harm’, ‘minimise restrictions on economic and social life’) and applies it consequently to every decision that forces itself on the agenda. The pragmatist approach rejects the idea that a principle should shape decision-making in times of uncertainty. It proposes an experimental, trial-and-error strategy that relies on a mixture of reasoning and feedback: try something that appears likely to work, study the consequences, and adjust where necessary.¹

To deepen our understanding of these two approaches, this paper explores leadership approaches to COVID-19 in four European states: the UK, Germany, the Netherlands and Sweden. The leaders of the UK and Sweden adopted a principled approach (but applied very different guiding principles). The Netherlands and Germany started out with a more pragmatist-oriented approach, but migrated towards a principled approach. To be sure, this paper does not seek to evaluate different national responses (the available data is still scarce and problematic). The contribution of this paper is to explore whether a theoretically derived distinction has analytical purchase in understanding how national crisis leaders design response trajectories under conditions of uncertainty.

We begin by outlining the two ideal-typical approaches to crisis decision-making. We then describe the responses of the four countries in more detail. We explore the differences between the approaches and introduce three factors that may help to explain the initial adoption of the approaches and the subsequent shifts (or lack thereof) as the crisis developed. We end by asking whether either approach has demonstrated advantages that leaders could take into account when confronting their next crisis.

A key leadership challenge in crisis: taming uncertainty

How do leaders cope with uncertainty in a highly dynamic and threat-filled environment? In Roman times, it was normal to have an esteemed official dissect a bird and ordain certainty based on a reading of the bird’s entrails. The role of priests and shamans has given way to an era of expertise in which rationality is the key norm. Policymakers are trained to collect as much information as possible, analyse that information using ‘evidence-based’ models and theories, and draw inferences. Such ambitions are reflected in the investment in data collection and data analysts, probability-handlers and modelers, as well as the ubiquity of cost–benefit analysis.

Such claims may sound reassuringly technocratic in front of rolling cameras and the finger-thumping twitterazzi. But they have always been at odds with reality. Elaborate information-gathering tools and high-powered processing technologies cannot solve ‘wicked problems’ (Lodge & Hood, 2010). In a world of wicked policy problems, few certainties exist with regard to causal relations, probabilities and potential impacts of policy solutions (Daviter, 2019; Peters, 2015; Rittel & Webber, 1973). In a world of bounded rationality at the individual and collective level, the unquestioned reliance on rational approaches tends to generate its own blind spots (Bach & Wegrich, 2019) and unintended consequences (Merton, 1936; cf. Lodge & Wegrich, 2016).

An approach that is based on complete information is particularly ill-suited for the world of crisis (Boin et al., 2016). Uncertainty with regard to causes, dynamics and potential consequences is, by definition, a key problem in a crisis (Rosenthal et al., 1989). The information that decision-makers would like to have is simply not available. There is no time to collect more information, to consult widely or to ‘sit out the problem’. Moreover, it is not immediately clear when information will become available nor is it clear whether new information is accurate. This places decision-makers in an unenviable position: they have to make quick decisions without much indication as to the effectiveness of their decisions and possible side-effects.

Yehezkel Dror (2001) described decision-making under conditions of uncertainty as ‘fuzzy gambling’ (without understanding the odds). The predicament for crisis decision-makers is worsened by the knowledge that their decisions may have deadly consequences. They are fully aware that they will be held accountable for their decisions after the immediate crisis has been overcome (Boin et al., 2008; Hood, 2011). Yet, they must make a decision, now.

To explore how crisis leaders approach this ‘fuzzy gambling’, we introduce two ideal-typical approaches: the principled and the pragmatist approach (Ansell & Bartenberger, 2019; Ansell & Boin, 2019). In theory, both approaches have political and administrative advantages. Both approaches come with their own distinct Achilles’ heel that are likely to be exploited by opponents (Bach & Wegrich, 2019; Lodge, 2019). Let’s see how these approaches could be used in a crisis and consider their pros and cons from a political-administrative perspective.

Coping with uncertainty: principled versus pragmatist approaches

One way to deal with crisis-induced uncertainty is to base all decisions on a core principle or value. We call this a principled approach (Ansell & Bartenberger, 2019). For instance, a decision-maker may use ‘freedom’ or ‘efficiency’ as a guiding principle when facing a conundrum. During the COVID-19 crisis, we saw two types of crisis response that resembled a principled approach: one that prioritized public health (protect as many people as possible from the virus) and one that prioritized economic well-being. In both types of responses, other values played a role but there was one overriding value.

The principled approach corresponds with a theoretical perspective on crisis management that was long popular among social scientists who studied political leadership during international crises. They viewed political leadership in terms of making critical decisions – ‘do or die’ decisions that determined the outcome of a conflict situation (cf. Hermann, 1963). This decisionist approach to crisis management has gradually given rise to a more contextualized approach (Allison, 1971; Ansell et al., 2014; Brecher, 1993; Rosenthal, Charles and 't Hart, 1987).

The principled approach offers some distinct attractions for crisis leaders. The embrace of a dominant value removes all complexities and vagaries that come with an implicit set of undefined values. There is no need to trade one value off against another. It facilitates, even invites, the use of sweeping rhetoric that is often expected from leaders in times of crises. Moreover, the adoption of a single or dominant value fits well with a so-called command-and-control model, in which unity of purpose, emanating from the top, serves to relieve implementing bureaucracies of the need to consider multiple values and interests (cf. Parker et al., 2019). In sum, the principled approach feeds the image of a ‘war machine’ – strategy and implementation fused into an all-out response to the crisis at hand.

The principled approach has some disadvantages. A principled approach is a binary approach. An all-out choice for one value (public safety) immediately implies a (at least temporary) disregard for, or downplaying of, other values (economy, freedom, efficiency). A principled approach implies a certain degree of inflexibility. But a prolonged crisis is likely to shift in nature and render formulated strategies invalid. It may, then, be politically infeasible to perform a sudden U-turn. This is especially the case when leaders issue sweeping statements that underscore the all-important nature of the chosen value.

The pragmatist approach to dealing with crisis-induced uncertainty is very different from the principled approach. In fact, it was formulated as an alternative to precisely that approach. A small band of American thinkers (William James, Charles Sanders Peirce, George Herbert Mead and later John Dewey) saw the unquestioned adherence to ‘principles’ as a driver of the Civil War they had lived through (Menand, 2001). The Pragmatists formulated an alternative approach to deal with upheaval and uncertainty (see Ansell [2011] for an engaging analysis). To be sure, the Pragmatists had nothing to say about political crisis management. But their approach had a lot to say about the challenges of protracted uncertainty and the ‘quest for certainty’ that policymakers so often pursue (Ansell & Boin, 2019; Dewey, 1929; Farjoun et al., 2015).

In essence, Pragmatists believe that reality cannot be known by collecting and analysing ever-more information. They put a premium on discovery through action (for which they use the term ‘enactment’). In a situation of uncertainty, you can learn about the environment you seek to negotiate by acting. It helps if you start by forming a mental picture of the possible state of play (a working hypothesis). Doing something will inevitably trigger a reaction (or an unsuspected non-reaction), which tells you something about that environment. The working hypothesis is confirmed or needs to be adjusted. By acting and reacting, you can learn and adjust. The Pragmatist approach thus resembles a scientific approach to solving a research puzzle. Feedback is very important in this approach: you need to understand what the reaction is to your action. Uncertainty is tamed by constant interaction with the environment.

In theory, there are some clear advantages to this Pragmatist approach to political crisis management. First, and most importantly, it seems much more realistic. The Pragmatist approach accepts uncertainty, which is, by definition, a key feature of a large-scale crisis. It saves decision-makers from paralysis, as it does not require a complete picture of the situation before decisions can be made. The Pragmatist approach provides decision-makers with flexibility: they can make U-turns without burning political capital. It protects political leaders against making costly and hard-to-reverse decisions.

If this is such a superior approach, why would political leaders not embrace it more often? One answer is the communication challenge that this approach creates for political leaders. If people look to their leaders in times of crisis, as conventional wisdom has it, they may not want to learn that their leaders are ‘embracing uncertainty’, initiating experiments to ‘discover’ the nature of the crisis, and are ready to change a rule or policy the day after it has been declared. In other words, this approach assumes a societal preference for small incremental steps that simply may not exist. Moreover, it does not fit conceptions of leadership that emphasize strength, stamina, and decisiveness. This brings us to a perhaps unfortunate paradox: while the Pragmatist approach can only work with a daring and strong leader, the principled approach is the one that will make a leader look daring and strong.

It is not just the political presentational costs that stand in the way of a pragmatist approach towards crisis management. The pragmatist approach builds on a very technocratic view of the world in which administrative and political fine-tuning is feasible, even under conditions of deep uncertainty. The reliance on immediate feedback processes requires reliable data-flows, processing-capacity and a speedy adaptive response.

In light of the prerequisites for a pragmatic approach, the attractions of a principled approach come back into view. The latter approach eliminates the costs of extensive information gathering by reducing decision-making to one single rough indicator: ‘will the system cope’ (in the COVID-19 crisis, one dominant indicator related to the number of available IC-beds). In other words, the principled approach explicitly economizes (rationalizes) on rationality (Hood & Lodge, 2005). To be sure, the principled approach is not devoid of learning and adaptation. It may, in fact, provide breathing space so as to enable learning without the need for constant fine-tuning.

Table 1 summarizes the key elements of the two approaches. It highlights political advantages and disadvantages, administrative pre-requisites and the most prominent weakness of these two approaches.

Table 1. Contrasting principled and pragmatist approaches.

	Principled approach	Pragmatist approach
Central theme	Choose central value, adjust policies solely to maximize this value	Sequential experimentation that relies on feedback for escalation/de-escalation
Political advantages	Signals clear leadership and ‘vision’, ‘making big calls’	Avoids the necessity of embarrassing U-turns by enabling incremental updating of policy preferences
Political disadvantages	High ‘sunk costs’ lead to concerns over subsequent U-turns Initial reaction may be qualified as an ‘over-reaction’ in ex-post inquiry	Difficulty to justify ongoing experimentation in view of other approaches Lack of ‘clear message’
Administrative pre-requisites	Capability to support overarching aim	Capability to continuously monitor and respond in view of ongoing analysis
Achilles’ heel	Other values are discounted, leading to opposition and neglect of other concerns	Reliance on feedback mechanisms and fine-tuned administrative capacities

In Table 2, we develop an initial set of indicators to guide our empirical discussion of different national experiences during the COVID-19 crisis. We distinguish between three phases: the initial phase of the crisis, the ‘wait and see’ phase, and the phase that marked the end of the first wave and a search for an exit strategy.

Table 2. Empirical indicators.

	Principled approach – precautionary variant	Principled approach – economic well-being variant	Pragmatist approach
Initial phase	Early decision to reduce social interactions to minimize threat to public health	Limited steps based on WHO guidelines; reliance on individual measures to continue economic and social life	Incremental addition of individual measures
Crisis regime	Strict enforcement and strongly focused communication to emphasize solidarity and rule adherence	Downplaying mortality rates and emphasizing individual responsibility	Continuing calibrations of measures in light of feedback and emerging evidence
Easing	Relaxation only if measures proven to be ‘safe’	Emphasize economic strength and relative gains when compared to other regimes	Relaxation follows experimental path of trial and error

In the following section, we will investigate whether these operationalized ideal-types have any analytical purchase. We also want to identify factors that may affect the adoption of either approach (or cause a shift from one to the other). Building on our theoretical discussion, we identify three types of factors that might help to explain patterns across the four countries: perceived uncertainty about COVID-19, administrative feasibility, and political necessity. Perceived uncertainty relates to the emerging knowledge regarding the properties of COVID-19 and its potential impact on individuals and societies. Administrative feasibility refers to the ability of administrative systems to facilitate either the principled or pragmatic approach. Political necessity refers to the demands on politicians that push for a principled approach or enable the use of a pragmatist approach.

Exploring four national responses to COVID-19

In this section, we briefly describe the trajectory of COVID-19 decision-making in four European countries (building on publicly available information). We use these case vignettes for the sole purpose of exploring the usefulness of the categorization in pragmatist and principled decision-making styles. We ask whether these ideal-typical approaches can help us describe and compare the different leadership approaches witnessed over the course of the COVID-19 crisis. We concentrate on the first wave (March–May 2020) during which uncertainty remained high and new challenges kept emerging.

We selected four European countries that struck us as particularly interesting: Germany, the United Kingdom, the Netherlands and Sweden. The leaders of these countries formulated approaches that could be easily classified according to our pragmatist-principled categorization. These countries also displayed a variance in dynamics, which differed from the stringent approaches that most European countries maintained during the first wave (see Table 3).

Table 3. Overview of stringency level from 17 European countries in March 2020.

Country	First case reported	Stringency level
		1st March	15th March	31st March
Austria	25 February	11.11	48.15	81.48
Belgium	4 February	11.11	50.93	81.48
Denmark	27 February	11.11	65.74	72.22
France	24 January	34.72	49.54	87.96
Germany	27 January	25	32.87	76.85
Ireland	29 February	11.11	48.15	85.19
Iceland	28 February	16.67	25	53.70
Italy	31 January	69.91	85.19	91.67
Luxembourg	29 February	0	53.70	79.63
Netherlands	27 February	0	54.63	79.63
Norway	26 February	11.11	51.85	79.63
Poland	4 March	11.11	57.41	81.48
Portugal	2 March	11.11	32.41	82.41
Spain	31 January	11.11	67.13	85.19
Sweden	31 January	0	27.78	53.70
United Kingdom	29 January	11.11	12.96	79.63

Source: Oxford COVID-19 Government Response Tracker.

As we will see, the response strategies of these four countries show some interesting similarities. They also diverged at critical junctures. Each country entered the crisis (we take the first COVID death as the beginning of the crisis) with an institutional structure designed to respond to pandemics. Political leaders quickly discovered, however, that the available expertise – captured in protocols and pandemic planning – did not offer effective prescriptions for situations in which many people had contracted the virus and stayed asymptomatic (WHO, 2006).

Politicians all faced the same challenge at this point: they had to decide what to do – close borders, impose social distancing, reason with the public or do nothing – without having the necessary information to make those decisions. Interestingly, the four countries by and large arrived at the same crisis regime: people were asked to stay home, social congregation was prohibited or curtailed, schools and nursing homes were closed (only in Sweden primary schools did not close and restaurants were allowed to stay open). People were not confined to their homes for indefinite periods of time and some economic activities were allowed to continue. Each country then entered a period of ‘sacrifice and patience’ – waiting for the crisis regime to take effect.

Each country eventually managed to bring ‘the curve’ (reproduction rates, hospitalization, fatalities) down, but the speed by which this was accomplished differed. It follows that the timing of the exit strategies differed as well: Germany and the Netherlands led the way; the United Kingdom trailed behind and Sweden never did exit its crisis regime.

We will very briefly characterize the crisis trajectories of these four countries, roughly distinguishing between three phases: the early crisis phase (the first two weeks after the first death), the ‘crisis regime’ phase (when the country is awaiting the results of the imposed measures) and the exit phase (between the announcement of the first relaxation and the first return to normal). These admittedly very rough distinctions work as a heuristic device to identify shifts in the chosen crisis approach.

The Netherlands²

The Dutch initially responded to the coronavirus in a way that closely resembled a pragmatic approach. The Dutch started out by adhering to WHO guidelines. People who felt sick were admonished to stay home and large-scale events were cancelled. In the first major press conference on COVID-19 (9 March), prime minister Mark Rutte and the chief of the Outbreak Management Team, Jaap van Dissel, told the Dutch to wash their hands, sneeze in their elbow and avoid shaking hands. Walking off the stage, Rutte shook hands with Van Dissel to the general amusement of journalists present.

Within the week, it became clear that the WHO approach was not working. The number of hospital admissions was exploding. Prime minister Rutte assumed control over the national crisis response. He acknowledged that he knew little: ‘we have to make 100% of the decisions with only 50% of the information’. A complete picture of the situation would not emerge anytime soon, as the testing capacity was rather limited and health authorities had all but given up on ‘tracking and tracing’ those who had been in touch with a Corona patient. But action of some sort was clearly required. The Dutch health system ran the risk of being overwhelmed.

On 16 March, Rutte went on TV to explain the situation. He found a complete lockdown as had been imposed in China and Italy unrealistic and not fitting the ‘sober Dutch’. While the health system had to be safeguarded, the same was true for the Dutch economy. So Rutte proposed a set of measures that amounted to a ‘lockdown light’ – people were admonished (but not forced) to stay home, while most stores stayed open and large industries and construction were allowed to continue. Schools, bars and restaurants were closed. Rutte promised flexibility. He offered the analogy of a sick patient: the medicine would be adjusted as the health of the patient declined or improved. If everybody worked together to ‘flatten the curve’, measures could be soon relaxed. But if the Dutch ignored the measures, a full lockdown would be in the offing. The Dutch could ‘earn’ their way out of the crisis by displaying good behaviour, the prime minister seemed to suggest. After a few weeks, the measures were tightened a bit more but a complete lockdown would not happen until the second wave.

The Dutch approach gradually appeared effective in bending the curve. It is no surprise, then, that Rutte continuously communicated the importance of sticking with the imposed crisis regime. The eyes of the nation were trained on the number of corona patients in IC-beds. While the number of patients went down, adherence to the crisis regime became the norm.

When the moment came to decide on whether and how to loosen restrictions, the pragmatic approach had all but disappeared. The prime minister now embraced the precautionary principle: a loosening of the crisis regime would only be possible if that would not have any negative effects. But when people became impatient – the hospitals emptying out, businesses bleeding money, and, importantly, the weather finally turning nice – pressure to relax the regime increased rapidly. Rutte grudgingly turned pragmatic again: many restrictions were lifted (including the opening of bars, brothels and gyms), but festivals were still not allowed. The mayors of the big cities were allowed to experiment with measures such as the (mandatory) wearing of face masks. Rutte issued a clear warning: a new spike in cases would lead to a return of crisis measures (as happened in the late fall).

Germany

The German response to COVID-19 was fairly similar to the Dutch approach. Despite an early cluster of cases in Munich in late January and a lockdown of one regional council in early March (Heinsberg), the response until 12 March aimed to continue social and economic activities whilst relying on administrative capacities to trace and isolate those infected with the virus. By early March a growing number of infections increased political pressure to ensure that health system capacity would be maintained. In response, German states agreed on a range of measures without full knowledge as to which measures (and their combined effect) would reduce transmission rates (Desson et al., 2020).

On 8 March, large events (1000 guests or more) were advised to be cancelled and working from home was encouraged. On 10 March, Health minister Spahn announced that hospitals would have to delay planned operations to create extra capacity for COVID patients. On 11 March, Chancellor Merkel and Federal President Steinmeier appealed for citizens to pursue social distancing. On 13 March, schools were closed in the majority of Germany’s federal states. On 16 March, Germany closed its borders, bars, clubs and theatres. Social gatherings were banned and conditions were imposed on canteens and restaurants (Robinet-Borgomano, 2020).

The German approach required a federal orchestration of the Länder governments’ decisions (who, constitutionally, were in charge of public health). The Länder established their measures somewhat inconsistently and at different times. Merkel negotiated with the prime ministers of the Länder to formulate a unified approach, which was announced on 22 March. Federal legislation of 25 March enabled the federal government to take exceptional public health matters normally reserved for Land governments (Robinet-Borgomano, 2020). When the crisis regime was firmly in place, a principled approach reigned.

The ‘easing’ of crisis measures, as in the Dutch case, was characterized by pragmatism. Soon after the declaration of the lockdown, debates about easing of measures emerged. These were, in part, driven by Länder governments that had seen low infection rates and therefore regarded the measures as too stringent and disproportionate. These pressures, and the lack of constitutional authority, meant that the federal government had to compromise with those Länder eager to proceed with easing. An intergovernmental agreement on 15 April allowed for a partial reopening of social and economic life after the 4th of May (e.g., reopening of schools and shops, and mandatory facemasks) with gradual ‘easing’ of measures subsequently (announced on 30 April and 6 May) (Robinet-Borgomano, 2020).

The United Kingdom

Prime Minister Johnson entered the crisis with a principled approach that was premised on the critical importance of continued economic activity (Gaskell et al., 2020). A lockdown was out of the question. Policymakers warned that undue haste in responding to COVID-19 would represent panicked knee-jerk responses.

Johnson publicly downplayed the risk and did not take charge of the response. Johnson let his Secretary of Health, Matt Hancock, chair the first five COBRA meetings. Johnson visited hospitals with COVID patients and shook their hands after reiterating the importance of hand-washing (for as long as singing Happy Birthday twice). A festival with more than 240,000 guests took place whilst 6 people had already died because of COVID. An economic support package was established before the first (real) measures were implemented (Morales et al., 2020). In the meantime, schools reported that parents were withdrawing their children from classes and the British population voiced its discontent on Twitter, using the hashtag ‘#WhereisBoris’ (Ellyatt, 2020).

On 16 March, Johnson suddenly changed course and announced the imminence of far-reaching measures. The Government subsequently shut down the schools (18 March), pubs and other social activities (20 March). On 23 March, Johnson announced a lockdown, including the shutting of non-essential economic activities, limits on leaving one’s home as well as a ban on all social contacts (Embury-Dennis, 2020). It marked a shift from one principled approach (economy as key value) to another (public health as key value) – within the course of one week.

When the new crisis regime was in place, Johnson was hospitalized (in intensive care) with the virus (other cabinet ministers and experts were also tested positive for the virus) (Morales et al., 2020). Subsequent crisis communication, especially following the prime minister’s recovery, was undermined when his key advisor, Cummings, was shown to have violated the rules of movement without resigning (unlike others who were shown to have violated lockdown rules).

A pseudo-pragmatic approach was chosen in the form of five indicators to guide the easing of the lockdown conditions. An ongoing tension existed between those in the government who believed in the need for a rapid return of economic activity and those warning against an early easing (the prime minister himself was said to be in the latter camp following his experience in intensive care as COVID-19 patient) (Morales et al., 2020). In pragmatic vain, Johnson suggested that the subsequent handling of local outbreaks would resemble a ‘whack a mole’ exercise (Sky News, 2020). A partial re-opening was permitted as of 4 July, although subsequent adjustments in terms of travel-related quarantine rules as well as a ‘re-tightening’ of some rules accompanied continued concern with the possibility of a second wave.

Sweden³

The Swedish response closely followed WHO guidelines. The Swedes adopted a limited set of rules that relied on a mixture of voluntary adherence and a sense of civic responsibility. Gatherings of more than 50 people were discouraged, secondary schools and universities were closed, bars and restaurants had to maintain social distancing measures, and visits to nursing homes were banned. The government asked the Swedes to work from home and act sensibly. By and large, the Swedes could live their lives unimpeded.

Other countries also began with this approach, but Sweden was one of the few countries that stuck with this approach (up to November 2020). Sweden was also one of the few countries that did not initiate a governmental crisis regime marked by centralization. The Swedes had little choice, as their constitution forbids a centralized crisis regime (cf. Pierre, 2020). The crisis response is placed in the hands of the health agency and local administration. State epidemiologist Anders Tegnell became the international face of the Swedish approach.

The Swedish approach resulted in a (comparatively) high number of fatalities. Yet, Tegnell stuck with the chosen regime. Even after admitting that the regime had not worked as expected, he did not adapt. Politicians and science colleagues asked for an adaptation of the regime, as the situation kept deteriorating, especially in comparison to other Nordic countries. The initial approach remained beyond reproach (Claeson & Hanson, 2020).

In hindsight, we can see that Sweden maintained a principled approach throughout the first wave of the COVID-19 crisis. The Swedish approach initially seemed to betray an evidence-based inclination, as the set of prescriptions followed the long-standing WHO guidelines. But whereas the WHO adapted its guidelines, the Swedes did not. They did not fine-tune measures in light of signals that the approach was not working (until much later in the year when the second wave hit). The Swedes stuck with the initial approach, refusing to adopt measures that were widely adopted elsewhere (such as the wearing of facemasks). It gradually became clear that the Swedes refused to prioritize public health over other policy values (Claeson & Hanson, 2020).

Exploring similarities and differences

When we interpret the country vignettes from a ‘pragmatic-principled’ point of view, we see that the crisis responses display some variance (Table 4). Two countries began with a pragmatic approach (the Netherlands and Germamy). Both the UK and Sweden began with a principled stance in favour of maintaining economic and social life to the largest extent possible. The UK performed a U-turn to an equally principled stance in favour of a lockdown when the earlier stance was no longer seen as viable. Three of the countries (the exception being Sweden) pivoted towards a more principled policy emphasis on the value of public health.

Table 4. Overview of national responses.

	Early crisis	Crisis regime	Exit strategy
Sweden	Principled	Principled	Principled
Netherlands	Pragmatic	Principled	Pragmatic
Germany	Pragmatic	Principled	Pragmatic
UK	Principled	Principled	Pragmatic

Once the three countries had decided to move out of lockdown, the approach taken was pragmatist at heart: a continuous evaluation of (local) trends and a fine-tuning of nation-wide measures. The ‘dance’ with the virus was pursued with a dominant strategy of only ‘easing’ when previous easements had been shown not to have caused negative effects.

In short, we see a continuum rather than pure ideal-types. We have learned that countries can change tack, which prompts intriguing questions about the drivers of these crisis responses. Intriguingly, we see that pragmatist approaches have underlying decision rules that can shift over time as well. In the early phase, further economic damage was only tolerated when public health concerns touched a certain (and invisible) threshold. The exit strategy reversed that decision rule: economic gains were only tolerated if public health would not suffer as a result. It reveals the importance for pragmatists to have a decision-rule as to whether to err on the side of Type 1 or Type 2 errors when making decisions in view of imminent feedback processes.

Towards understanding dynamics

The interesting question that emerges from our brief exploration of national approaches to the COVID-19 crisis, in our view, pertains to the shifts in these approaches. With the exception of Sweden, each country employed different decision modes. Here we consider whether the three factors introduced earlier – perceived uncertainty of health risk, administrative feasibility and political necessity – might point to a possible explanation of these dynamics. We organize our discussion in three phases: the initial approach, the regime phase, and the exit phase. We should emphasize that this is a preliminary discussion, which is in need of more detailed empirics (that will no doubt materialize in the near future).

#1 Initial approach: wait and see

Each of the four countries were initially committed to a policy that did not include lockdowns. This is interesting, especially in view of the Chinese and Italian templates that many other countries followed. Looking across the three factors, we can see the lack of conditions that would attract politicians to call for immediate lockdowns, or, at least, forewarn publics of such a likely scenario. This initially created room for a cautious approach (a pragmatic approach) or a principled embrace of economic well-being as a policy driver.

Perceived uncertainty: how bad is this, really? Early 2020, experts in the four countries did not consider the risk of the virus spreading as particularly high and the mortality rates were seen as not particular worrying, even though some Chinese scientists (and subsequently Italian experts) were warning that the virus was highly contagious and many patients were asymptomatic. But ‘hard’ evidence was initially in short supply. When evidence became available, it came in the form of stories from overflowing hospitals. Scientific modelling suggested high mortality numbers if nothing was done to curb community transmission.

While the need to act became clear, it remained unclear what the effects of interventions in societal life would be on the propagation of the virus. Three countries (the UK being the exception) built on the WHO guidelines to chart a course, strengthening measures as those in place proved insufficient. Uncertainty still reigned, but a response was found that seemed to work. Emerging certainty pointed to the need for a principled adherence to the ‘discovered’ regime.

Administrative feasibility: we can handle this. The uncertainty surrounding the virus met with robust trust among both experts and policymakers in the state of preparedness of their public health systems. Playbooks (including track and trace procedures) and organizational structures (such as the Outbreak Management Team and the SAGE group) were activated. Previous outbreaks – SARS, MERS, Ebola and Swine flu – had all been controlled and seemed to affirm the soundness of institutional arrangements that were in place. Preventive measures – cancel Carnival activities, prohibit mass events such as football competitions – were regarded as unnecessary. This collective perception of impregnability cancelled the need for a principled approach predicated on the protection of public health and facilitated the idea that a pragmatist approach could work.

Political necessity: no need to kill our economy. In February, when the first cases were just materializing in our four countries, economists and politicians in international forums were widely discussing the economic consequences of the total lockdown in China and Italy. The initial inclination in all four countries was to avoid a total lockdown and rely on a set of milder interventions. Existing WHO (2006) guidelines advised against far-reaching measures on the grounds that they were unproven and the modern citizen would not tolerate them for even a brief time period. The adoption of a science-based approach provided politicians with a story line explaining why a complete lockdown – widely imposed in surrounding countries – was not necessary (at least not yet).

The United Kingdom pivoted towards a lockdown when the Imperial College report caused a sensational scare, predicting a wave of deaths and thus pressuring prime minister Johnson to explain the lack of measures in the UK.⁴ His initial principled approach could only be met with a dramatic U-turn, towards another principled approach. The other countries had much more room to manoeuvre.

#2 Sticking it out

We noted that all countries moved to a principled adherence of the set of policy measures that appeared to work. The Netherlands, Germany and the UK migrated towards a full adoption of a ‘principled precautionary’ approach so as to reduce transmission channels. The protection of the population at large from the virus became the primary objective or value. Sweden was a bit of an exception: even though its approach did not prove effective immediately, the country stuck with the principled belief in the chosen approach. Let’s see if the three factors can help shed light on the embrace of a principled approach in this phase.

Persistent uncertainties, growing effectiveness. While little was learned about the exact causal relations between initiated measures and intended outcomes, the curve began to flatten in each country. This gave the impression that the set of imposed measures – different as they might be in each country – was working. The temptation to experiment and see if the flattening might be ‘sped up’ was curtailed by the realization that the unintended consequences might well outweigh the limited gains (how fast could a curve be flattened, after all?). Crucially, few countries in Europe were doing a quicker job. There was thus a natural tendency to stick with what seemed to work.

The limits of administration. With the possible exception of Germany, the countries lacked the means to initiate experiments and collect the fine-grained feedback needed to assess the results of such experiments. In Sweden, the Netherlands and the UK, ‘tracking and tracing’ capacities were overwhelmed by the growing number of cases. The one crisis indicator that could be easily and accurately measured was the number of IC beds occupied by COVID-patients. That indicator was too crude to measure the effects of any experiments that might be considered. Moreover, looming shortages of health system capacity (IC beds and ventilators) and the lack of supplies (such as protective equipment) appeared on the radar as a threat to the continuity of the health system.

Political necessity: the rising need for a consistent approach. Social distancing measures require a high degree of voluntary compliance by individuals and companies alike. To reinforce social norms of solidarity, crisis communication is essential. Leaders needed a simple yet compelling story to persuade the public. Calls for solidarity do not sit well with the inconsistencies that come with ‘experimentalism’. Imposing measures in one jurisdiction raise the spectre of social activities moving towards those jurisdictions that are ‘still open’. Politically, such ‘racing’ among jurisdictions is highly problematic and is resolved by formulating a unified national response.

#3 Towards the exit

In three countries, we identified a Pragmatic turn when it was time to design an exit strategy. The easing was cautious. Gradual steps were taken to assess potential impacts on public health, enabling both gradual loosening and immediate tightening-up of measures in view of infection rate developments. There was also a growing call to decentralize decision-making, which, in turn, gave rise to criticism concerning inconsistencies in approach (such as school opening guidelines).

Uncertainty: the virus might return. Demands on the health system were declining, but experts had learned how contagious the virus was. They warned that a quick return to normal would give the virus an opportunity to strike again (causing a ‘second wave’). It was clear that heightened virus infection rates would follow if caution was cast in the wind. A relaxation of measures was possible, but it had to be slow and with an eye on negative effects. A pragmatist approach fitted the bill perfectly.

Administrative feasibility: growing capacity to be pragmatic. The new administrative challenge was to design and enforce the post-lockdown regime. While this would require additional street-level capacities, at least it was a familiar challenge that could be decentralized to regional and municipal administrations. It helped that testing and tracing capacities (despite all imperfections) were ramped up, protective gear became available and the number of IC-beds had been expanded. Allowing these lower-level administrations an enhanced degree of discretion was a nice fit with a pragmatic approach of experimentation.

Political necessity: pressure to loosen up. Political leaders faced the paradox of success: the more effective the crisis regime, the less likely people were to continue adhering to the rules; eroding adherence undermined the effectiveness of the response. Pressure for relaxation emerged from the business sector, social media personalities and select tabloid newspapers. While public health considerations prohibited a rapid relaxation of measures, the crisis regime was becoming politically problematic, especially as, for example, in Germany, the federal government had no authority over measures taken by the Länder. A Pragmatic approach – feeling your way towards an uncertain future, backed with the perceived reassurance of sufficient administrative capacity – thus became an attractive option. The promise of relaxation as a reward for responsible behaviour: it was an approach that political leaders could sell.

Conclusion: reconsidering the merits of the principled-pragmatic distinction

This paper studies how political crisis managers craft a crisis response under conditions of pervasive uncertainty. It focuses attention on the nexus between crisis management and political leadership, asking how political (and other) decision-makers approach critical dilemmas without the information they might have in ‘normal’ times. More specifically, it aims to explore whether the distinction between pragmatist and principled approaches has any analytical purchase. Moreover, we seek to identify factors that can help to understand why countries adopt pragmatic or principled approaches (or shift from one to the other).

We think that the proposed distinction provides a heuristic device to describe, categorize and compare national responses to crisis events characterized by prolonged uncertainty. The paper demonstrates the differences between these two ideal-typical approaches. It also shows areas of overlap. A principled approach will not completely shun or disregard information feedback processes or rule out the adjustment of policy settings. Similarly, a pragmatist approach cannot operate without some form of basic (principled) policy setting as to how to weigh different objectives. And, indeed, a pragmatist approach may culminate in a strategy that may appear rather principled.

We do think that further work is needed on the refinement of the indicators that we used to identify both approaches. These are, in effect, rough indicators that can be used to describe the nature of a national crisis response. It would be great if indicators could be developed and validated for the measurement of leadership intentions rather than policy responses. Further work is clearly needed.

We conclude that a pragmatist approach may serve political leaders well in a prolonged period of uncertainty. It can only work, however, if highly demanding preconditions are met. In addition, it should allow for the possibility that a more principled approach may be necessary, at least for a period of time. Once a winning formula is found, leaders are very reluctant to let it go (even if there might be a better approach). It takes away any incentive to gamble with something that appears to work. A principled approach makes sense, then, by rationalizing on rationality in precautionary ways when the only thing you know is that the chosen approach appears to be working. Perfection is the enemy of the good, as NASA engineers used to say.

This paper also prompts new questions, especially with regard to the dynamics of crisis approaches that need to be employed for a relatively long period of time. One question in need of further contemplation pertains to the political and administrative benefits of these two approaches. In some cases, a principled approach appears to offer not just political expediency (leaders are able to show that they are taking decisive action), but also implies administrative necessity. The proposed benefits of the pragmatist approach presume certain, highly demanding administrative pre-requisites, given its reliance on feedback processes. Our cases suggest that pragmatism may indeed be hard to maintain in the thick of a crisis. Moreover, we did not see any serious efforts to set up experiments that might lead to a more effective crisis response. It simply may be too difficult to explain the inherent inconsistencies of such an approach. A continuous ‘loosen and tighten’ pattern makes it hard to generate widespread compliance.

In other words, our findings suggest that the distinction between a pragmatist and a principled approach might be a bit too simplistic. The key question that emerges here is whether a principled approach may not be part of a more overarching pragmatist approach. If you can switch from one principled stance to another, the underlying principles of either approach may not be as hardwired as our initial discussion suggested. A principled approach may well be the outcome of a pragmatist process (a lesson learned through trial and error). Perhaps we should consider the possibility that a principled approach is simply a tool in a pragmatist approach.

It would be helpful if we could say something about the effectiveness of either approach. Unfortunately, it is simply too early to make any claims with regard to the effectiveness of national responses. In due course, an avalanche of information with regard to the different COVID-19 regimes in these and other countries will undoubtedly emerge. It is still likely that ‘success’ and ‘failure’ will be viewed through different prisms as the history of this pandemic is being rewritten at different points in the future (cf. Boin et al., 2008). The question, then, may be one of political feasibility: which approach allows for the discovery of a workable and legitimate approach to the threat at hand?

Acknowledgements

We thank the reviewers for their excellent suggestions. We are indebted to Marte Luesink for her assistance in the preparation of this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Arjen Boin

Arjen Boin is a professor of Public Institutions and Governance at the Institute of Political Science, Leiden University. He is also a consultant at Crisisplan BV.

Martin Lodge

Martin Lodge is a professor of Political Science and Public Policy at the LSE's Department of Government and the director of the Centre of Analysis of Risk and Regulation (CARR).

Notes

1 McConnell and ‘t Hart (2019) introduce a third approach: inaction. As our space is limited, we do not discuss it here.

2 The Dutch case description is based on Boin et al. (2020).

3 We based the Swedish case description on Engström et al. (2021) and Pierre (2020).

4 We are referring to the widely cited ‘report 9’ of 16 March 2019 put forwarded by Professor Ian Ferguson and colleagues at Imperial College (https://www.imperial.ac.uk/media/imperial-college/medicine/mrc-gida/2020-03-16-COVID19-Report-9.pdf).