Defence organizations in emergency networks: the early response to COVID-19 in Europe

2.1 The pandemic as a slow-onset disaster

Pandemics and epidemics are categorized as slow-onset disasters (Adamo, 2011; Yamori and Goltz, 2021) due to the time it takes for them to develop. In addition to epidemics, other examples of slow-onset disasters include famines, droughts (Yamori and Goltz, 2021), sea-level rise, desertification and salinization (Human Rights Council, 2018). Such disasters often receive sporadic media attention (Yamori and Goltz, 2021) compared to sudden-onset disasters. One characteristic of slow-onset disasters is that they are often invisible to authorities and local populations before they grow into a full-blown crisis. As a result, affected people are ill-prepared when responding (Glantz, 1999). Therefore, Nguyen-Trung et al. (2020, p. 3) remind us that “less awareness of the crisis could impinge on responses or reaction and could extend to recovery in the post-disaster context” and that “it is believed that more visible impacts could lead to stronger reactions from affected communities” as well as other actors. The Ebola epidemic and the COVID-19 pandemic are good examples of this phenomenon.

According to Coppola (2015, p. 33), “disasters occur when a hazard risk is realized” and “to be considered disastrous, the realized hazard must overwhelm the response capability of a community”. If a slow-onset disaster is not managed properly, it may trigger a complex emergency, especially when the affected area is prone to other hazards. The term emergency is often related to urgency and is used in cases of unforeseen disasters and threats. According to Keen (2008), complex emergencies could involve a combination of armed and civil tensions, widespread damage to society and the economy, food shortages and often armed groups and/or government policies that contribute to insecurity, resulting in a need for large-scale humanitarian assistance. The Inter-Agency Standing Committee (IASC) defines complex emergencies as “a humanitarian crisis in a country, region or society where there is a total or considerable breakdown of authority resulting from internal or external conflict and which requires an international response that goes beyond the mandate or capacity of any single and/or ongoing UN country program” (IASC, 1994).

Although the term complex emergency may be used when the world faces pandemics and/or epidemics, for example, the Spanish flu between 1918 and 1919, which killed an estimated 25–50 million people, the swine flu between 2009 and 2010 (Coppola, 2015) and the Ebola epidemics in West Africa between 2014 and 2016 (CDC, 2019), not every pandemic qualifies as a complex emergency. Additionally, crises play out differently in different parts of the world. Our focus is on European countries. As in the rest of the world, the pandemic had a major impact on societies and resulted in complex national responses. However, neither of the countries covered in our study experienced a humanitarian crisis or a breakdown of authority. Hence, although the pandemic could add to the complexity of emergencies in other parts of the world, we will not define the pandemic in Europe as a complex emergency. By clarifying our research context through this distinction, we intend to identify context-independent factors that can improve our understanding of how societies react to major disasters.

2.2 Civil-military interaction in humanitarian operations

DOs have a long history of assisting in epidemics and pandemics. For example, when the plague hit Bari (Italy) in the 17th century, military efforts and related expenditures were motivated by urban programmes that sought to save people’s lives (Koch, 2014). When responding to the dengue epidemic in Brazil in 2008, the military quickly set up military field hospitals to ease the shortage of hospital beds (Coppola, 2015). In the Ebola pandemic in West Africa, military troops were deployed to close internal roads in Sierra Leone (Koch, 2015). Arie (2014) even stated that “only the military can get the Ebola epidemic under control”. US defence forces responded to that crisis by sending personnel and equipment to Liberia in 2014. However, significant delays occurred due to the slow arrival of both equipment and skilled personnel. The whole response came “far too slowly to have contributed directly to the management of the rapidly attenuating disaster” (Nevin and Anderson, 2016, p. 52).

Many humanitarian actors find themselves increasingly inhabiting spaces with a strong DO presence (Abiew, 2012). This is the case for both national DOs within a crisis-affected country (Madiwale and Virk, 2011) and international DOs intervening in foreign nations (Auerbach et al., 2010). DOs are often no longer limited to the role of providing logistics support as originally described under the Oslo guidelines, but they are increasingly engaged in direct humanitarian action (Thiessen, 2011; Hall and Deinla, 2021). DOs, in many crisis responses, provide services that are identical or very similar to those of humanitarian organizations, and such services can be integrated into broader strategic objectives (Bell et al., 2019; Egnell, 2013). Humanitarian actions by DOs can be used to engage influential civilians (Ankersen, 2013) and to win over the civilian population and garner their support (Hall and Deinla, 2021) in direct contravention of traditional humanitarian principles. This dissolution of boundaries has been perceived critically as it may make humanitarian organizations targets for attacks (Hall and Deinla, 2021) and could lead to inefficiencies (Thiessen, 2011).

The classic role of a DO is to secure humanitarian corridors (Harris and Dombrowski, 2002) and to provide a secure environment for humanitarian operations (Bell et al., 2019). However, DOs also provide expertise, equipment and surge capacity for tasks as diverse as constructing hospitals, delivering supplies, establishing transport links and building refugee camps (Heaslip and Barber, 2016). This necessitates a nuanced approach, with military involvement being considered a much larger issue, with the potential to jeopardize humanitarian operations in areas embroiled in conflict, while relatively peaceful areas are more tolerant (Madiwale and Virk, 2011). As the role played by DOs expands to incorporate a wider scope of activities, a range of civil-military interactions can be initiated, from mere coexistence to full collaboration in crisis operations (Cook and Yogendran, 2020).

The tensions over humanitarian space between humanitarian organizations and DOs are not a major concern in the context of this study. In European countries, both civil and military responses in the initial phase of the COVID-19 pandemic occurred under the auspices of the relevant national governments. Thus, the focus was on collaboration and co-operation as part of a national crisis response. The study, therefore, explores interactions in the initial response phase to a crisis where concern about the potential violation of the humanitarian space was not the defining element. Rather, it was a case where DOs were regarded as integral parts of government-led response networks in the face of a crisis that affected all aspects of society globally.

2.3 National response systems and the dual co-ordination challenges of defence organizations

When DOs become involved in a civil emergency response, their tasks differ markedly from their armed assignments. Greet (2008) claims that DOs take on a range of roles to participate in different types of operations. Such roles include domestic humanitarian operations; humanitarian operations abroad; and humanitarian operations in failed states; with tasks that include services in co-ordination with civil actors; the provision of accommodations; reconstruction; and the delivery of critical supplies, logistics and medical interventions.

The primary task of DOs is the safety and security of their country (Kaldor, 2013). During societal crises, this role can present dual co-ordination challenges (Listou, 2019). Firstly, since defence logistics largely rely on civilian commercial suppliers, there is a vertical co-ordination challenge in designing and managing supply chains that include both commercial suppliers and military units (Ekström, 2012; Listou, 2019). Resources located outside of the defence hierarchy must be governed in ways other than military command and control. Therefore, DOs should be regarded as customers of logistics support, with civilian (commercial) actors as suppliers. Secondly, there is a horizontal co-ordination challenge related to the cross-sectoral support of DOs within the state. In this horizontal dimension, DOs play the role of providers of logistical support to society, as suggested by van Wassenhove (2006), Heaslip (2011), Kaneberg (2017) and Listou (2019).

Although DOs possess military-specific assets that are difficult, if not impossible, to replicate, they also have resources that are of value for civil society. This encompasses tangible resources such as transportation capacity, construction equipment, medical supplies and communication means, as well as intangible resources such as their command-and-control structure, traffic control skills and medical competence. Rietjens et al. (2007) and Bollen (2002) claim that the military has the capability to benefit society because DOs can add extra capacity and create flexibility. For example, personal protective equipment (PPE) can be owned by suppliers or by national emergency agencies but distributed by the military. Thus, DOs act as suppliers or intermediaries in supply chains. The tighter the integration between DOs and their suppliers, the higher the value of these relationships will be for the DO. However, such integration might reduce the availability of these resources for society during complex emergencies (Listou, 2019), particularly if commercial actors hold stock on behalf of the military.

2.4 Displaying response structures as networks

The pandemic responses include a myriad of actors, both public and private. Different actors contribute using their resources to perform activities and chains of activities to reduce the consequences of the disease. There are different approaches to the study of network structures, such as social network analysis, which was proposed by Carter et al. (2007) for mapping the role of the individual in supply chains. Steen (2010) describes actor-network theory as a tool for understanding the resource dimension in business strategy, and Ford et al. (2008) apply the industrial network approach (INA) to study inter-organizational networks. In this research, we choose to build on the INA because its systematic structure is suitable for analysing networks at both the individual and inter-organizational levels.

Within the INA, networks consist of actors, resources and activities (the ARA model); see Figure 1. Actors can be individuals, groups, organizations or firms that perform activities and control resources (Håkansson and Snehota, 1995). Actors are goal-directed, transforming their goals into specific intentions (Lenney and Easton, 2009). Resources are categorized as tangible and intangible (Lenney and Easton, 2009). Tangible resources include production units and products, while intangible resources include business units and business relationships (Gadde and Snehota, 2000). Activities occur when one or several actors combine, develop, exchange or create resources (Håkansson and Snehota, 1995).

Inter-relationships are formed among these three layers of ARA in the form of activity links, resource ties and actor bonds (Håkansson and Snehota, 1995). Links relate to how the activity structures of two or more units connect as a means of co-ordination by mutually adapting these activities, which in some cases, are further developed into activity chains and patterns. Ties refer to how different resource elements are connected such that resources are exchanged, transferred, accessed or reciprocally used. Finally, bonds concern the different factors that connect actors, help them build each other’s identity and develop trust and commitment. Koporcic (2017) reminds us that these layers are all interlinked and mutually affect each other. Actors have the potential to benefit by sharing resources and activities with other actors (Hakansson and Waluszewski, 2003). Aspects such as trust, mutual understanding, learning and a co-operative atmosphere (Abrahamsen and Håkansson, 2016, p. 199), as well as a shared understanding of resource usage and the activities to be performed (Abrahamsen et al., 2011), are crucial for this co-operation.

The study emphasizes the networks of crisis response of European countries and the role played by DOs during the early response phase of the COVID-19 pandemic, here set to encompass 75 days. The roles and positions occupied by the DOs in the early phase of the pandemic response thus need to be understood based on the activities they performed, the resources they brought in, and their relationships with other public services, private actors and voluntary actors.

3.1 Data analysis

The data were organized by observation period, dating from the first reported cases of infection, the actors involved in the initial response and the requested military support in each of the 13 countries (as shown in Table 1). In some cases, the dates when resources were used are indicative, i.e. in some instances, the decision to use military capabilities was made on one day, while the actual use occurred later; or, in other cases, resources were ready for use, but there was no need to deploy them.

Our data analysis builds on coding and the search for common themes and categories. The coding was performed by several members of the research team. Our study is grounded in the notion of civilian networks and military assets being used in national emergency responses. Some of these networks existed before the outbreak, and some came into being during the early response phase.

The ARA model served as a tool to code the elements of our analysis addressing categories of ARA and their activity links, resource ties and actor bonds, as outlined, e.g. by Håkansson and Snehota (1995). The data analysis was carried out by coding the data, following the recommendations of Corbin and Strauss (2008) and Saldana (2013), whereby first-level codes were based on the extant literature, with further codes developing during the data analysis. The elements of ARA served as first-level codes in the data analysis (i.e. as activities, actors or resources). Then, second-level codes were established based on the content of the data. For example, actors were categorized as NGOs and the voluntary sector, governmental authorities, medical structures, infrastructure operators, international bodies and educational establishments. Within a third level, “individual actors” were coded, and international bodies were broken down into foreign governments, North Atlantic Treaty Organization (NATO), World Health Organization (WHO) and United Nations Office for the Coordination of Humanitarian Affairs (OCHA). This approach was also applied for coded “activities” and allowed for comparison within and across nations; even when actors were named differently, coded “activities” were used to obtain different descriptions.

3.2 Trustworthiness

In qualitative studies, research reliability is usually expressed as confirmability and dependability. Dependability is obtained using an auditing approach with high importance placed on keeping complete records of the research process (Guba, 1981; Lincoln and Guba, 1985; Lincoln and Guba, 1986). Dependability was achieved here by maintaining a shared database among the researchers that contained all data and notes documenting the data analysis processes, thereby allowing any future researchers to repeat the analytical procedures (Stuart et al., 2002). As coding is a decision-making process (Elliott, 2018), the research team discussed the decisions made frequently during the data analysis phase.

Confirmability involves ensuring that the researchers have acted in good faith and have not overtly allowed personal bias to sway the conduct of the research and reporting of the findings (Guba, 1981; Lincoln and Guba, 1985; Lincoln and Guba, 1986). In this study, a clear chain of evidence was established, and the data analysis was extensively documented to increase transparency (Ellram, 1996). The close co-operation of the research team during the research process exposed the decisions made by each researcher to immediate scrutiny and helped minimize any existing biases.

Transferability refers to external validity (Guba, 1981; Lincoln and Guba, 1985; Lincoln and Guba, 1986), which is concerned with “the domain to which a study’s findings or presumed causal relationships may be generalized” (Stuart et al., 2002, p. 430). Generalization is not the aim of this study, as it strives to explore a context of some novelty compared to pre-existing knowledge about civil-military interactions in response networks. Following Lincoln and Guba, the approach taken here is to focus on the contextual uniqueness of the national response networks under investigation but with the aim of obtaining a detailed account that can enable researchers to make an informed judgement about the possible transferability of the findings. This is achieved here through an extensive and detailed findings section that expresses the richness of the primary data and a detailed understanding of the national response networks through the lens of ARA.

Credibility is measured by internal validity in the evaluation of quantitative research (Guba, 1981; Lincoln and Guba, 1985; Lincoln and Guba, 1986) and was achieved through validation by seeking corroboration of the findings through feedback from those involved in the response networks. In addition to data collection, some members of the research team simultaneously participated in the pandemic response as affiliates of NGOs or military organizations in various European nations, which increased our understanding of the different response conditions within Europe. Their insights as experts on the ground served as an essential sense-check to ensure accurate comprehension, for example, of military terminology or of the complex links between various actors.

3.3 Limitations

Our methodological approach has limitations. Firstly, by relying on secondary data, we do not obtain deep knowledge about decision-making processes and assessments of the different measures undertaken. Secondly, our timeframe of 75 days from the first confirmed case of COVID-19 in each country restricts us from investigating the path dependencies of the response networks, just as it restricts us from understanding how the outcomes of these networks influenced response measures taken after this initial phase. Thirdly, given the data sources available to us, we could not access the details of the responses in each country. To capture the initial response in all countries, there was a need for a more generic overview instead of an in-depth country-by-country analysis. Linking early research on civil-military co-ordination was essential for identifying nationwide differences in the use of DOs as part of the national response networks. Fourthly, we used the ARA network approach to categorize our findings; however, the ARA frame did not help provide in-depth knowledge about the adaptation processes in the dynamic networks created, as our study represents only a snapshot of the national response networks in the early response. Instead, the ARA frame in this study was used to identify actors, activities and resources and to show how DOs in diverse roles could provide defence resources ad hock in the initial response, how resources from different actors were combined and how activity structures evolved.

4.1 Response networks

The pandemic response played out differently in each country. There are, however, many similarities in how resources were combined, which actors were involved and what activities were performed.

Regarding relationships, our analysis shows that several relationships evolved during the first days of the response. This quick alignment seems to have been motivated by the criticality of gaining access to medical, transportation, personnel and other resources.

Our data provide interesting examples in which co-operation with the DOs was intensified, especially by non-governmental, business and voluntary actors and educational establishments. The DOs reinforced the civil authorities’ capacities with regard to specific services (ambulance, transport, logistics skills and personnel) when resources were scarce in the initial response. On average, it took the DOs 12 days to become fully involved in the response, with some exceptions. For example, the Spanish DOs were involved from the very first day. Most of the DOs made their resources available for the national response operation in less than 20 days.

The actors’ resources were tied together to provide access to and allow allocation of facilities and personnel. For example, temporary field hospitals (e.g. FR, DE, IT, Poland [PL], Spain [ES] and the UK) were deployed, including support services, transport modes, expert units, medical items, essential medical equipment, PPE, respirators, oxygen (e.g. DE, IT, PL) and operational support (e.g. ES, Sweden [SE]).

During the first 75 days, the DOs assisted in the repatriation of citizens by connecting transportation; administrative processing; triage; and care provision (e.g. FR, DE and PL) and preparing infrastructure (airports and ports) for the reception of patients (e.g. IT and NO). As the situation progressed, the national response networks increasingly focused on logistics activities, such as connecting warehousing; handling; consolidation; and distribution activities with the national and international transportation of goods (e.g. CZ, PL and Slovakia [SK]) and, in some cases, the procurement and acquisition of medical equipment and medicines (e.g. DE and UK). Subsequently, the DOs’ support of the response networks encompassed population-related activities, such as arranging and managing testing (e.g. DE, PL and SE); providing support to local populations (e.g. DE, PL and ES); operating information hotlines (e.g. DK); and supporting care homes (e.g. PL and ES).

The DOs co-operated with local governments and municipalities, authorities and services, medical institutions, local governments, NGOs, educational establishments and international bodies. The only exception was Sweden, which did not demonstrate any specific co-operation with international bodies related to the above-listed tasks.

In only five of the countries mapped in this study, the DOs co-operated with non-governmental and voluntary sectors. In DE, the DOs provided support to the local vulnerable population, distributed information, answered questions from the population, trained volunteers and participated in the production of medical equipment. In ES, the DOs assisted the voluntary sector in citizen repatriation, international transport of goods, national transport of patients, national transport of goods, disinfection of public areas, support to the local vulnerable population and information support to civil society. The DOs in IT were active in the international transport of goods, the international transport of medical staff, the disinfection of public areas, the provision of support to the local vulnerable population and the provision of personnel support for individual testing. In PL, co-operation involved the international transport of goods; the international transport of medical staff; the national transport of goods; warehousing; the disinfection of public areas and transport equipment; the provision of support to the local vulnerable population; the provision of IT support and operational support to health organizations; and the provision of personnel support for individual testing, information support for civil society and leadership support in the testing process. In contrast, in Portugal, the DOs helped with the international transport of patients, the provision of operational support to health organizations and personnel to civilian hospitals, the preparation of air/seaports for the reception of patients and increasing capacity at civilian hospitals.

The researchers identified activity links that connected logistics activities and services directed towards the population. Such links were established in close collaboration among local governments and municipalities, authorities, medical institutions and infrastructure operators.

Regarding co-operation with educational establishments, our findings revealed that the CZ, FI, PL and Portugal launched such co-operation. In FI, universities participated in testing the potential usage of massive respiratory cleaning protocols. A Portuguese university assisted with surveying citizens regarding their trust in institutions. In the CZ, students from the University of Defence answered calls from citizens concerning the quarantine. In PL, universities participated in tasks related to COVID testing and information sharing.

The researchers also found that in some countries, the DOs supported civil society in more untraditional ways. In the case of CZ and SK, the DOs sewed face masks. In SK, the DOs also supported the families of their members who were deployed abroad. The Portuguese DOs distributed food to homeless people, and the German DOs delivered food to stranded drivers. Finally, in five countries, the DOs were involved in providing psychological support. Of these, the DOs of PL, CZ and SK started early in the initial phase (days 14, 20 and 22, respectively), whereas PT and DE established such services on days 47 and 70, respectively.

6.1 Further research and implications

Further research is needed to assess whether the new constellations formed during the pandemic response evolved and formalized as adapted response preparedness patterns and whether such patterns are optimal for the next pandemic response. Since this study focuses exclusively on the first 75 days after the confirmation of the first COVID-19 case in each country, managerial implications are related to the countries’ response systems’ inability to react to urgent demands with the aim of obtaining a more comprehensive understanding of the activities, the actors and the swiftly changing relationships that evolved. It is expected, and in many cases already beginning to become evident in the data, that the need for the surge capacity provided by DO personnel receded as routines were established. However, future research should investigate this process, including the handover of activities to other actors and the potential reactivation of DO actors in later stages of the pandemic response, such as during periods of mass testing or the eventual roll-out of the COVID-19 vaccinations.

The pandemic has been followed by other regional and global crises. Here, managerial implications are related to failing global supply chains in part because of the Chinese lockdowns of cities to cope with COVID-19 outbursts and due to energy and food shortages as a result of the Russian war in Ukraine. These crises actualize a debate about societal preparedness and what role the state and the private market should play. The organization of many European countries is based on variations of new public management (NPM) principles (Mascio et al., 2020; Pollitt and Dan, 2011) and the division of tasks and responsibilities between the public and private sectors. According to the industrial network perspective, the movement of resources from one actor to another alters actor bonds, resource ties and activity links in the networks. The effects of activity internalization were recently seen, e.g. when the German state nationalised the gas importer Uniper to secure its energy supply [1], on permanent and ad hoc response networks. How does a reconceptualization of NPM mechanisms affect cross-sectorial response networks?

6.2 Limitations

This study is not without its own limitations regarding the use of the ARA network approach to the categorization of the findings. It does not fully provide in-depth data regarding the adaptation processes in civil-military dynamic networks. The ARA network approach was useful in this study in identifying essential ARA. However, the approach was limited in terms of its inability to show how useful defence resources were related to the various activities of the initial response. A further complication was the support sought in the early literature, as there was a substantial difference between the findings of the early research and those of this study with regard to the use of the resources of different actors in countries with less developed response infrastructures. For example, combining the research on activity structures in less developed countries was not fruitful in supporting the findings of this study, as the networks and path dependencies within a country’s preparedness systems in this study differed from those of countries that are less structured and developed.

While this study includes the effects of different resource combinations that were employed in the initial response to a large range of activities, the effects on the efficiency and effectiveness of response operations were not country-specific, which constitutes an additional limitation regarding the achievement of the objectives of collaboration and interdependencies among governmental, private and voluntary actors. For example, in the ad hoc settings of the initial response to the pandemic, the experiences of existing networks of military and civil actors in Europe were rather unsuccessful in terms of transferring crisis responses to other parts of the world. Another limitation of the study is, to a large extent, the need for consideration of the history, culture, governance and level of trust in the military in some nations, which are preconditions for integrating civil-military collaboration in crisis response.