Agency and resilience in the time of regional economic crisis

ABSTRACT

Using a comparative case study on the closure of the research and development facilities of a pharmaceutical company in two regions in Sweden, this paper investigates the emergence of regional economic resilience from an agency perspective. Findings include a networked region engendering agency from non-state actors whilst substantial state intervention is needed to facilitate agency from a less networked region. The paper finds broad confirmation that interactions of actors are due to how a region is organized and the contingencies of its institutional context. Collective agency depends on the ability of actors to align interests and coalesce agendas which, as shown by the two cases, can be based on shared regional affinity as in the case of Södertälje or innovation strategies as in the case of Lund. Regional economic crises bring forth conditions and impetus for temporary modes of cooperation that mobilize resources to launch adaptive resilience strategies. Conflicts in resource distributions or operational complexity make developing agency-based resilience challenging. Emergent regional outcomes from agency-based resilience from these two cases exhibit adaptation with the potential for adaptability in Lund, and adaptability with degrees of adaptation in Södertälje.

KEYWORDS:

• Regional resilience
• agency
• agency-based adaptation
• adaptability
• institutions

1. Introduction

Different industries are undergoing large-scale restructuring, terminating multiple plants across the world and ceasing their operations in regions dependent on their commerce. This has beset regional economies with problems of unemployment, abandoned facilities, and out migration of labour. Regional actors are increasingly recognized to have key roles in organizing diverse responses to these challenges including but not limited to sector-specific state support and policies encouraging resilience (Fromhold-Eisebith, 2015; Pike, Dawley, & Tomaney, 2010). These efforts vary in terms of effectiveness but they are believed to have long-term consequences on regional development. This has inspired policy and scholarly interest in resilience (Pendall, Foster, & Cowell, 2010; Pike et al., 2010)

Resilience has several orientations, to which scholars from different disciplines, subscribe. Engineering-oriented resilience refers to the ability of a system to go back to its pre-shock state (Pendall et al., 2010). Socio-ecological resilience focuses on the ability of resource-dependent communities to recover from natural disasters (Adger, 2000; Folke, 2006; Holling, 1973). Resilience informed by complexity science, on the other hand, emphasizes adaptation, adaptability and adaptive capacity as forms of resilience (Martin, 2012). There is a preference for this complex view in economic geography because it moves away from equilibrist thinking and views resilience as a dynamic process of responding to shocks and perturbations (Hassink, 2010b; Martin, 2012; Martin, Sunley, Gardiner, & Tyler, 2016; Simmie & Martin, 2010).

Whilst extant literature on resilience has provided profound insights into the dynamics of economies in crisis at the macro level, it has been criticized as inadequate in explaining uneven resilience capacities across places (Mackinnon & Derickson, 2013). Economies seem to be treated as self-organized systems and the process of change, autonomous (Cretney, 2014; Pike et al., 2010). This seeming preoccupation with the macro level reflects the broader critique that roles actors play in influencing directions for economic resilience have been theoretically and empirically obscured from analyses (Bristow & Healy, 2014). This constitutes a significant gap in the literature because it is at the micro-level where skills and capabilities in adapting to a crisis, actualize.

Table 1. Summary of primary actors.

	Type of Actors	Role
Lund – Medicon Village
FIRS	Research Council – multi-stakeholder membership	Initiator, negotiated with Astra Zeneca, recruited tenants
Lund University	University	Put up the initial capital for rental to enable bank loan, leasing tenant
Mats Paulsson Foundation	Philantrophic Foundation	Bought the whole facility and put up capital to enable bank loans
Region Skåne	Regional Governance Body	Put up capital to enable bank loans; leasing tenant
Astra Zeneca employees	Individuals	Facilitated the handover and the change of management and closure
Södertälje Science Park/Biovation Park
Södertälje Municipality	Municipal Government	Initiator, negotiated with Astra Zeneca; leasing tenant; faciliated scaling up of the park
Scania	Business in Large automobiles	Tenancy at the facility and lobbying the government
Acturum	Joint Venture between FAM and PEAB – Hybrid	Bought the facility as a joint venture, manages the facility.
FAM (Acquisition of arm of Wallenberg Foundation)	Investing	Bought the facility as a joint venture.
PEAB	Business	Bought the facility as a joint venture.
Ministry of Education	National Government	Expanded KTH by building a new facility in the former Astra Zeneca facility and increasing student population.
Karolinska Institute	University	Employed personnel from the animal testing facility arm in Snäckviken,and formed SWETOX.
Astra Zeneca management and employees	Individuals	Facilitated the handover and the change of management and closure.

Table 2. List of interviewees.

Actors	Number
Former Astra Zeneca, Lund management and employees	3
Former Astra Zeneca Sweden	1
Former Astra Zeneca Södertölje	1
Public commentators on the closure	2
Trygghetsrådet Lund	1
Current Astra Zeneca	1
Södertälje Science Park	1
Södertälje municipality	2
Medicon Village	1
FIRS	2
Lund University	2
Total	17

A promising direction in the literature unpacks some of the micro-level dynamics involving regional resilience. Though still in its nascent stage, with relatively few articles focusing on this in economic geography, an agency-based view of resilience is emerging and explicates how actors adapt to regional economic crises and the mechanisms by which this feeds back into emergent regional capacities for resilience. The aim is to contribute to this research track and articulate further the link between agency and resilience by extending on and empirically testing key parameters laid out by Bristow and Healy (2014) in their seminal article examining resilience through an agency perspective. This paper, then, examines actors or agents (used interchangeably here), referring to individuals and organizations, with particular adaptive agencies in the region, pointing to their potential to alter and shape regional resilience directions. The main research question to be interrogated is: how does agency emerge and function in the context of regional economic resilience?

This paper finds that agency from non-state actors emerges from a networked region whereas a region with less interactions create space for substantial state intervention. These interactions are related to how a region is organized and the contingencies of its institutional context. Collective agency depends on the ability of actors to align interests and coalesce agendas which could, primarily, be based on shared regional affinity as in the case of Södertälje or innovation strategies as in the case of Lund. Regional economic crises bring forth conditions and impetus for temporary modes of cooperation that mobilize resources to launch adaptive strategies. In this process, networks are used extensively but conflicts in resource distributions or operational complexity make developing these adaptive strategies challenging. Actors improvise in selecting strategies under conditions of uncertainty. Emergent regional outcomes from agency-based resilience from these two cases exhibit adaptation with the potential for adaptability in Lund, and adaptability with degrees of adaptation in Södertälje.

The paper is organized as follows, Section 2 reviews debates surrounding economic resilience as a contested concept and identifies how resilience can benefit from further development of the agency perspective. Section 3 discusses agency and parameters to examine it, drawn from the framework of Bristow and Healy (2014), integrates ideas mainly from organizational and neo-institutional studies, where necessary, to extend this framework, and expounds on adaptation and adaptability; constituting the conceptual framework. Section 4 delves into the methodology in empirically investigating how coalitions of actors operationalize adaptive strategies in a regional economic crisis, using a comparative case study of the closure of AstraZeneca in Lund and in Södertälje, Sweden. Section 5 discusses findings and finally, Section 6 concludes with policy implications and future research directions.

2. Resilience as a contested concept

Despite being largely regarded as a positive concept in policymaking and academic circles, resilience has inspired debates surrounding its quality and usefulness as a concept in economic geography (Bailey & Turok, 2016). Purported shortcomings of the resilience literature include an excessive macro-level focus and equilibrist thinking that misses the micro-level dynamics of resilience (Bristow & Healy, 2015). Furthermore, critics claim that the literature ignores institutions, politics, power, and failures, as well as privileging spatial sites (Cooke, 2017; Evans, 2011; Mackinnon & Derickson, 2013). Recovery seems to be naturalized in the literature despite evidence showing that this is not inevitable for many struggling places; thus, adjudicating the resilience framework as ill-equipped in explaining disparities between regions (Cumbers, Helms, & Swanson, 2010; Hassink, 2010b; Mackinnon & Derickson, 2013; Mackinnon, Cumbers, Pike, Birch, & Mcmaster, 2009; Pendall et al., 2010; Swanstrom, Chapple, & Immergluck, 2009). More recent literature have taken steps to address different aspects of these deficits (Dawley, Pike, & Tomaney, 2010; Deverteuil & Golubchikov, 2016; Gong & Hassink, 2016; Martin et al., 2016).

One such strand in the economic geography literature features work representing serious attempts at rectifying some of these micro-level shortcomings (Bristow & Healy, 2014; Bristow & Healy, 2015; Cowell, 2013; Cumbers et al., 2010; Doran & Fingleton, 2016). It focuses on the capacity of non-firm actors to shape adaptive strategies and contribute to the skills and capacities of places. By doing so, it looks at more micro-level processes such as agency, along with dimensions of institutions, politics and social factors; critics of which claim have been missing in the resilience literature. It is an eclectic accounting of agency, drawing from shared ontologies from multi-disciplinary sources such as psychology, ecology and disaster studies (Adger, 2003; Berkes & Ross, 2013; Davidson, 2010; Folke, 2006). Bristow and Healy (2014) propose parameters to study agency within resilience and this framework shows promise in what it can potentially uncover. Particular spatial sites, such as regions, are privileged in the analyses but for justifiable reasons. Regions are critical containers of human action and structure where institutions are spatially sticky and social factors, regionally bound (Gertler, 2010).

Exploring the role of agency in resilience more fully, however, is not without conceptual and empirical difficulties. Questions remain such as how agency emerges and functions, from where and whom does it emerge, which regions are able to engender agent-based resilience and why. Moreover, agency, as a concept, comes with its own peril. An inflated account of agency runs the risk of suffering from the paradox of embedded agency which refers to the logical incongruence of agents changing structures by which they themselves are supposedly determined (Battilana, 2006). Developing this view of resilience, therefore, requires a more fastidious treatment of structure alongside agency to explain the enabling conditions under which actors enact changes. Current conceptual developments are neither sufficient in answering these remaining questions nor constitute enough empirical evidence for an accomplished theoretical framework of agency-based resilience. There is further work to be done to fully articulate the link between agency, structure and resilience, and the empirical evidence to show this link is robust. It is to these ends that this paper intends to contribute.

3. The role of agency and structure in regional economic resilience

Agency is the capacity for human action to conjecture, foresee and adjust attitudes, improvise plans and behaviour in order to forestall or allay a crisis (Bristow & Healy, 2014). It refers to temporally nested processes of social engagements constituted by actions and reactions intended to acquire new competencies, technologies and strategies geared towards making changes or adjustments to the social and economic environment (Battilana, 2004). Agency encompasses diverse human capacities for social and human actions such as in the way actors ascribe meaning to shocks and in the manner they respond to them (Bristow & Healy, 2014). Agendas, underpinning these actions, operate alongside structural aspects of economies to shape emergent outcomes manifest in regional experiences connected to local disturbances. Actors are products of their structural environment but their actions produce these environments as well (Bandura, 2000). As such, individuals and organizations can be active and reflexive agents of resilience (Brown & Westaway, 2011). Because resilience is the capacity to adapt, which can be deliberately obtained and exercised, it is in these adaptive strategies that agency in human and social action can be located and observed.

Agency is interdependent and indivisible with and from structure (Battilana, 2006), that is to say, the economic, political, social and institutional context of time and place. Adaptive behaviour, then, is dynamic and contingent on internal and external conditions, historical legacies and associated economic realities (Martin, 2012). Specifically, this can mean that regional resilience is a function of particular resources or capitals that are bound locally. Regional assets include spatially accessible financial capital, social, political and cultural capital in communities, as well its built-environment (Bristow & Healy, 2014). These assets buttress the material relations formed in regions.

3.1. How agents are organized

Due to finite resources and interdependent conditions, actors tend to collectivize and organize in order to reduce transaction costs in gathering information, and pool in resources in mitigating risks (Williamson, 1981). Whilst it is important to look at firms and firm-related actors, regions have a variety of organizations such as the state, universities, local governance units and civil society, which can play key roles in regions (Berkes & Ross, 2013). Actors can have layered and overlapping connections with other actors through communications, common tasks, and transacting on market agreements. As such, regions need to be examined as networked and multi-scalar entities, emerging from interactions between actors in a connected system (Bristow & Healy, 2014). Regions benefit from novel ideas gestated from networks that have space for experimentation which can enable conditions for the emergence of new solutions to crises (Bristow & Healy, 2015). These, coupled with the dynamic interplay of constituent actors, are critical for the emergence of agency-based resilience.

3.2. Forming coalitions and collective agency

Interactions have the potential to coalesce similar cognitive frames for a collectively meaningful interpretation of their environment (Bandura, 2000). Collective agency emerges as interactions build trust, align interests, influence behaviour and enable social learning (Putnam, 2004). This process helps coalitions form in order to adapt to changing circumstances in their environment (Adger, 2003). The implication of which is that a lack or few channels for interactions could have detrimental effects on the potential for agency in a region. Collective agency fosters expectations for the future because actors devise plans and reinforce ideas within groups, including how risks are perceived (Moore & Westley, 2011). Risks are threats of harm to people and to other more abstract ideas they value like community, security, and freedom (Kasperson et al., 1988). Perceptions of risk are amplified and attenuated through social means (Kasperson & Kasperson, 1996), contextually curated within cultural beliefs, myths, narratives of shared experiences, and often particular to groups of actors and places (Dake, 1992). In turn, these inform both the impetus for action and choices of actors for adaptive strategies for resilience (Bristow & Healy, 2015).

3.3. Agendas: politics and resource distribution

Political-economic ideologies like neoliberalism often underpin policy decisions through economic crises but often overlooked in resilience studies (Mackinnon & Derickson, 2013; Swanstrom et al., 2009). In a fundamental sense, politics can be accounted for in the agency perspective of resilience by looking at agendas of actors (Bristow & Healy, 2014). Agendas represent manifold human endeavours and interests demarcated by different groups across and within organizations in places (Birkland, 2007). These agendas can be conflictual because it can limit and order the resources of actors (Sheingate, 2009). Allocating finite resources inevitably favour some agendas over others with resulting allocations as products of contested and negotiated processes (Birkland, 2007; Mahoney & Thelen, 2009). A part of these processes is maintaining prioritization of specific agendas which are not automatic and often requires continuous resource mobilization (Mahoney & Thelen, 2009). Thus, instances of resource distribution towards agency-based adaptive strategies and the structures that facilitate, reinforce, and allocate are important dimensions of regional resilience.

3.4. Institutional context

Institutions are mechanisms such as rules, norms, expectations, heuristics and social codes of interactions, devised by humans to influence actions. They structure and reduce the complexities of daily life (North, 1990). Institutions are spatially sticky and conform to the history of a region and eventually become embedded as, for example, cognitive frames, identities and norms (Gertler, 2010). They constitute the multi-scalar structures in which actors are situated, and incentivise and constrain behavioural outcomes (North, 1990). Specifically, rules within organizations determine the ability to set agendas which refers to the access and power to allocate and mobilize resources to particular groups (Battilana, 2006; Bourdieu, 1989; Mahoney & Thelen, 2009). Since a region is a confluence of diverse agendas and groups, a complex interplay of local institutions, capital and material relations, and networks constitute the capacities of individuals and organizations for regional resilience (Bristow & Healy, 2014). Therefore, to understand the role of agency in resilience is to understand how individuals and collective agents behave in relation to varying contingencies, underpinned by factors and institutional mechanisms by which decisions are made and adaptive capabilities, obtained.

3.5. Emergent outcomes

Agency-based adaptive strategies influence regions toward paths of adaptation or adaptability with policy choices and decisions enacted by actors (Bristow & Healy, 2014). Adaptation refers to responses of agents geared towards mitigating immediate and short-term effects of an economic perturbation by adjusting to changes in the economic landscape using previously successful and pre-conceived strategies prior to the shock but with improvements (Pike et al., 2010). This introduces path dependence into the resilience process which refers to institutional legacies shaping future prospects (Martin & Sunley, 2006). Adaptation can lead to lock-in because institutions become homogenous and insular with the same set of actors interacting with one another (Grabher, 1993; Hassink, 2010a)). This lack of variety in the constitution of agents impairs the ability to generate new ideas and find novel solutions to problems (Grabher, 1993).

Adaptability, on the other hand, refers to regions creating novel paths, related or completely different from previous trajectories (Pike et al., 2010). It is defined as the dynamic capacity to trigger potentially diverse paths through spatially lax connections between agents, making regions and their systems more nimble in responding to unforeseen changes (Grabher, 1993; Hassink, 2015). Resilience, through adaptability, requires absconding from formerly lucrative paths in favour of an alternative strategy (Martin & Sunley, 2006). A nuanced view of regional resilience posits that aspects of adaptation and adaptability can be present in the resilience process and both can be utilized and maneuvred by different actors to forward their agendas (Pike et al., 2010). This implies that both can be examined as contrasting but also potentially complimentary views on how to understand ostensibly disparate trajectories of agency-based resilience; thus, some advocate moving away from this dualism altogether (Hassink, 2015).

4. Methodology

This paper explores empirical evidence on the emergence and contingencies of agency located within adaptive strategies for regional economic crises. It utilizes a comparative case study method in analyzing two instances of regional economic crises, the closures of the Astra Zeneca research and development facility in Lund, and Södertälje, respectively (Tables 1). Material is gleaned from 17 semi-structured interviews with varied actors conducted from years 2015–2017 (Table 2). Desk research from annual reports, web pages and policy-related documents provide supplemental materials to the interviews conducted. Specifically, it examines material regarding the roles of actors and their interactions, activities, choices, behaviour in coalition building, and use of networks as well as components of the regional setting at the time of the crisis. This is not an exhaustive exploration of crisis response surrounding the life science industry in Sweden but does contribute empirical evidence towards agency-based resilience from these two specific regions. It offers a qualitatively oriented case to the literature on the local phenomenon of agency-based adaptation of specific actors. The case study approach is selected because it allows for an exposition of the sources and mechanisms of regional adaptation to economic crises (Yamamoto, 2011). In this case, quantitative methods would be ineffectual at capturing agency, specific responses of regional actors to local economic shocks, and attempts to reorient their industrial structures (Cowell, 2013). As such, resilience epistemologies of this nature require social scientific inquiry that is both localized and qualitative (Evans, 2011).

4.1. Background on AstraZeneca

AstraZeneca is a Swedish-British merged pharmaceutical company operating globally. It employs 59,700 employees worldwide, 8,400 of which are in R&D, 12,200 in manufacturing, and the rest in sales and marketing (AstraZeneca, 2016). Some of its most successful products turned over of a billion dollars for the company in the areas of cardiovascular, gastrointestinal, infectious, respiratory, neuroscience, oncology and chronic obstructive pulmonary diseases (AstraZeneca, 2010; AstraZeneca, 2012). However, the life science industry is undergoing daunting market changes with products coming off patent and thereby losing exclusivity of popular drugs; such as Crestor, Nexium and Seroquel for AstraZeneca between the periods of 2011 to 2015 (AstraZeneca, 2014). As a result, pharmaceutical companies stand to face increasing competition from generic drug companies worldwide. These difficulties are exacerbated by high production costs and a developmental pipeline for new medicine that can take up to 15 years if it succeeds at all (AstraZeneca, 2011; AstraZeneca, 2016).

These industry-wide changes have compelled AstraZeneca to undergo a continuous process of restructuring and efforts to rebuild their pipeline to improve competitiveness (AstraZeneca, 2016). Their strategy has been to streamline their in-house R&D to focus on three main broad therapy platforms: oncology, cardiovascular and metabolic diseases, and respiratory with corresponding strategic research centres in Cambridge in England as their headquarters, Gaithersburg, Maryland, USA and Mölndal in Gothenburg, Sweden (AstraZeneca, 2013; AstraZeneca, 2016). This has meant closing down some of its research facilities, namely, in Charnwood, England, Reims in France, Mountain View, California in the US, Montreal in Canada, Lund and Södertälje in Sweden (AstraZeneca, 2011; AstraZeneca, 2012). The decision was made in 2009 to close the R&D facility in Lund over a two-year timeframe, affecting almost 900 jobs. 90% of employees were offered their jobs in Mölndal, Gothenburg with 40% of those, deciding to move (Severinsson, Turunen, Fröjd, Andréasson, & Bengtsson, 2012). Coverage of relocation costs and other financial incentives were offered to encourage mobility. The closure in Södertälje in both locations, Snäckviken and Gärtuna, was announced in 2012 and had a one-year timeframe, affecting 1200 jobs (Trygghetsrådet, 2015). A manufacturing plant in Södertälje continues to employ 2,200 people (AstraZeneca, 2016).

5. Discussion

5.1. Regional assets in Lund

Situated in the Skåne region in southern Sweden, Lund has a population of 116,000 and hosts Lund University, the largest research university in Sweden (Lund Municipality, 2017). It provides the region with high levels of human capital with 41,000 students and 7,500 employees. The university turns over 808 million euros, two thirds of which are from research and one third from education (Lund University, 2017). It is the third largest employer in Lund, behind only the municipal government and Region Skåne, the regional governing body (Lund Municipality, 2017). Lund has a large university hospital network, innovative capacity, and an active civic community. It is in close proximity to Malmö, the 3rd largest city in Sweden and connected to Copenhagen, Denmark via the Oresund Bridge, allowing daily commuting for work. A large food industry is situated in Skåne due to fertile farmlands. Resident food manufacturing companies include Scan and Skånemejeriet. ICT companies such as Axis and Qliktech are also in Lund. Skåne enjoys a certain level of autonomy from the Stockholm capital because of its distance but this also means that it does not have levels of political capital that proximity might bring.

5.2. How actors are organized

In this specific case, primary actors in this region include non-firm actors like Lund Municipality, Region Skåne, and Lund University. According to interviews, key members of these three organizations were highly networked and interacted both professionally and socially prior to the announced closure of AstraZeneca in Lund. A source of these interactions stem from shared regional responsibilities. Lund municipality is in charge of lands, schools and amenities, whilst Region Skåne is responsible for public transportation and healthcare which includes managing the network of hospitals (Lund Municipality, 2017; Region Skåne, 2017). Lund University owns a substantial amount of assets in the region such as lands and buildings. The delineated but interdependent nature of their positions and responsibilities bring opportunities for dense interactions. These administrative tasks and joint projects require levels of routine cooperation, allowing actors interested in building social capital and trust, the ability to do so.

5.3. Formation of coalitions and collective agency

One coalition formed in 2010 from these regional interactions is Förskning och Innovationrådet (FIRS), a research innovation council in Skåne made up of members primarily from Lund University, Region Skåne, and Lund Municipality. FIRS advocates a triple-helix framework of cooperation, referring to the importance of close ties between governments, universities and businesses in spurring innovation (Leydesdorff & Etzkowitz, 1996). FIRS was initiated and led by the vice chancellor and chairman of the board of Lund University of this period. The vice chancellor was a director general of Vinnova, the innovation arm of the Swedish government. The chairman of the board of the university was a former minister of finance of Sweden, former director general of the European Commission in Brussels and chief negotiator for bringing the European Spallation Source (ESS) to Lund. These previous appointments suggest the strategic positions of these actors in the regional network and their respective abilities in forming coalitions with key actors.

When AstraZeneca publicly announced imminent closure of its R&D facility in Lund in 2010, FIRS initiated an attempt to purchase the facility but failed to do so. Using their personal connections, FIRS then asked a philanthropist, who made his fortune in the construction company, PEAB, to negotiate the purchase instead. AstraZeneca acquiesced to selling the facility as well as leaving all laboratory equipment, to the newly created foundation of this philanthropist, the eponymous Mats Paulsson Foundation. This demonstrates the efficacy of coalitions with members occupying strategic positions as well as the significance of connections obtained through personal and professional networks.

5.4. Agendas: politics and resource distribution

One difficulty, however, was raising the financing for this purchase. Bank loans needed to be guaranteed with long term leasing contracts of the facility to prove financial viability. The Mats Paulsson Foundation and Region Skåne put up their share of the financing needed. In this case, Region Skåne was more active than Lund Municipality, which did not contribute any financing. This is primarily due to the political mandate of Region Skåne in managing healthcare. But it may also reflect an attempt of Region Skåne to assert a bigger role in the region which has been seen as somewhat subordinate to Lund Municipality because of its relative lack of resources. Additionally, when the vice chancellor of Lund University, sought the procedural approval from the council of deans to put up a share of the necessary financing, this substantial resource allocation was denied. The vice chancellor had to lobby the board of directors of the university to overturn this decision. The chairman of the board, at the time, was also one of the founders of FIRS and supported the creation of Medicon Village. The decision was successfully appealed and Lund University approved the financing of leasing contracts necessary to procure a bank loan for the purchase of the AstraZeneca facility. Connections with ESS helped recruit them to lease space from Medicon Village which helped shore up the last amount needed.

Examining agendas of actors locate some of the conflicts in resource distribution within organizations involved in establishing Medicon Village, impacting long-term prospects of agency-based resilience projects. The decision to guarantee rental contracts directs resources to groups which can be relocated to Medicon Village; inevitably privileging some groups over others within the university. When the term of the vice chancellor and chairman expired, oppositional agendas sidestepped previously by appealing to the board, resurfaced with the succeeding administration, perceived to lack the same level of commitment to Medicon Village relative to its predecessor. Thus, because adaptive strategies are subject to differing agendas of groups and the dynamics resulting from conflicts in resource distribution, agency-based resilience strategies can be temporal and its long-term support structure, tenuous.

5.5. Institutional context

Rules regarding transfers of agenda-setting powers are rigid within Lund University. The term limit for a vice chancellor is set to six years. Moreover, vice chancellors interpret their roles and positions and set their own agendas during their term, based on their own cognitive frames, priorities and expectations. Continuity of projects from previous administrations is not guaranteed or even expected in some cases. The term limit of the vice chancellor is just one of the many administrative turnovers of key actors involved with the creation of Medicon Village, with most of its key proponents, having since left their respective posts. These organizational changes present challenges for the continued development of Medicon Village in two main ways. Firstly, successors of top level management from Lund University and Region Skåne no longer sit on the board of Medicon Village as they had previously, which is perceived as a decrease in commitment. In turn, this makes it difficult for Medicon Village to efficiently and quickly coordinate between complex and procedurally rigid organizations such as academia, industry and governments. This has impacted the management and operations of the science park in terms of having less access to decision making and resource allocation processes from these partners. The previous coalition, through frequent interactions, built trust, and aligned their interests and cognitive frames to support an inherently demanding process of collaboration with parties having their own unique institutional cultures. A collaboration such as this is now significantly more difficult to manage without the necessary conditions and dynamics that enabled the previous coalition. Secondly, without agenda setting powers, previous proponents and advocates, become less effective, themselves, at influencing the maintenance of resource allocations to Medicon Village, despite continued support. Therefore, agency-based strategies must be cognizant of how rules and dynamics of administrative succession and subsequent transfers of agenda setting powers affect the trajectory of resilience projects.

5.6. Emergent outcome – adaptation

Medicon Village accepts member companies from the life sciences and health care sector. Renting space at the science park entails receiving associated benefits and advantages, including use of laboratory equipment, access to networking events, intranet and a multifarious working environment allowing social proximity to other companies. It contributes to reorienting the life science sector in the region in line with the industry-wide changes mentioned. The creation of Medicon Village has enabled small companies to set up shop and use lab equipment that would otherwise be unaffordable at an early stage. These companies are responding to the demand for potential late-stage development projects and external collaborations that big pharmaceutical companies are increasingly pursuing (AstraZeneca, 2016). Pharmaceutical companies like AstraZeneca have rebuilt their pipeline so that they are spending less on in-house R&D and on discovery/early stage projects (AstraZeneca, 2015; AstraZeneca, 2016). The park has 1600 employees working in the facility, an increase from when AstraZeneca was still in operation in Lund. It has successfully reoriented the former AstraZeneca facility without straying too far from the original purpose of the facility with adaptation being the main strategy for resilience.

Start-ups not working to develop medicine directly but have intersecting interests in healthcare are also accepted as leasing members at Medicon Village such as an ICT company developing e-health software. Companies providing support services for businesses relating to recruitment, accounting, legal advice on patenting, and printing are also members at the park. Moreover, its location within the larger fabric of the university town of Lund means it has access to diverse actors across different fields. This suggests that lock-in does not seem to be a risk, for the moment, since there seems to be a deliberate attempt to foster diversity amongst resident companies and wider collaborations outside the park. As such, possibilities for adaptability remain, considering substantial resources are devoted to linking actors in the region and pursuing collaborations through a complex web of events and projects. This is an operationally demanding but necessary strategy in order to generate a continuous income stream for the park and optimize its resources.

5.7. Regional assets in Södertälje and Stockholm region

Södertälje municipality, with about 71,774 inhabitants, is situated in the south of Stockholm region (Södertälje Municipality, 2017). Large companies such as Scania, were founded in Södertälje 1911, and Astra, in 1913. Astra later became AstraZeneca from a merger with a British firm. Stockholm region hosts the capital and the centre for governance, financial and cultural life in Sweden. Södertälje municipality has good transport links to central Stockholm and is part of the commuting belt in the region. Proximity brings levels of access to political and financial capital. There is, however, no big university presence in Södertälje. At the time of the closure, The Royal Institute for Technology in Stockholm (KTH), had only 600 students in Södertälje and planned to phase out that branch as its main campus is in central Stockholm. There is relatively low human capital in the municipality since many workers and students commute to the area but do not live there.

5.8. How agents are organized

The main governance unit in Södertälje is its municipal government. It is in charge of schools, hospitals, roads and other amenities in the region. There is no formal regional governing body in the Stockholm region, like that in Skåne. AstraZeneca announced plans to close their R&D plant in Södertälje in 2012 over a one-year interim period. Unlike in Lund, these jobs were not being moved to Mölndal but ceased almost completely with only a select group of individuals offered to continue their research in Boston, USA. Non-state actors did not immediately emerge with novel ideas on how to resolve this crisis as it did in Lund. A new term mayor had just been elected and lacked a network with other actors in the municipality and the region. This suggests the regional organization in Södertälje initially fostered less interactions and connections between its constituent actors relative to Lund.

5.9. Formation of coalitions and collective agency

The municipality led initiatives in responding to the crisis and made arrangements with different stakeholders in the community to build coalitions. Södertälje does not have a big research university presence like Lund. Improvising for this lack of regional asset, it leaned on a strong actor present in the region, Scania, and like AstraZeneca, part of the regional and industrial heritage of Södertälje. Scania was one of the first actors to make a strong commitment of support and involvement to the municipal government partly in the name of regional solidarity. Eventually, the construction company, PEAB, and FAM, which is the acquisition arm of the philanthropic Wallenberg Foundation, formed a consortium, Acturum, to buy the facility, and create the science park, Biovation Park. Meanwhile, the Snäckviken part of this facility where animal testing was formerly conducted by AstraZeneca, was taken over by SWETOX, leased from Acturum, a consortium of 11 universities initiated and funded by Karolinska Institutet, providing advanced toxicology services to the wider public (Acturum, 2017).

The municipality agreed to sign a long-term contract leasing parts of the facility for its own purposes. It also contracted Uppsala Innovation Centre (UIC) to manage the incubator part of the facility for ongoing AstraZeneca R&D projects, at that time, for employees who wanted to stay in Södertälje to continue their research development projects. The municipal government spent more than 3 million on these contracts with further extensions. These are substantial investments and involvement from the municipal government especially compared to Lund Municipality. Having secured a buyer and rental contracts, Scania and the municipal government then successfully lobbied the ministry of education at the national level to expand its KTH student population and allocate resources to build and expand university facilities in Södertälje to increase and develop human capital in the area. This suggests, that contrary to claims that government action crowd out community action, in some cases, local governments have critical roles to fill, particularly in cases where the regional context is not very well networked and agency does not emerge from non-governmental actors during economic crises. These roles include negotiating, lobbying, building coalitions and procurement.

Challenges in the management of Biovation Park have included clarifying areas of responsibilities and expectations between its coalition partners. One reason for this need is that the coalition of the proponents of Biovation Park were not previously networked and with partners perceived to have divergent needs and unspecified agendas. These partners have since scaled up the science park with the creation of Södertälje Science Park (SSP) which launches in early 2018, continuing to use the former AstraZeneca facility but now with a focus on sustainable production (Södertälje Science Park AB, 2017). It also emphasizes a triple-helix environment similar to that of Medicon Village. The chairman of the park is the former CEO of Scania. Biovation Park, the original science park created after the closure, is subsumed under SSP as one of its key components. Partners include Södertälje Municipality, KTH, Acturum, Scania, and with the addition of AstraZeneca. SSP represents an expansion of the ambitions and involvement of Södertälje Municipality with its partners. This adjustment in operational scale is said to be an opportunity for the municipal government to resolve issues, and clarify roles, responsibilities and what it expects to achieve in the region.

5.10. Agendas: politics and resource distribution

The ruling coalition in Södertälje belongs to the Social Democratic party which is left-of-centre oriented in political, social and economic issues in Sweden. Ideologically, it is able to support public ownership of assets and enterprises. The mayor of Södertälje, was advised by a representative of the national government, to have the municipality purchase the facility as the quickest and easiest solution to the crisis. Given the economic stance of the ruling party, it was surprising that this suggestion was immediately cast out, with the mayor saying ‘ … Governments don’t really do these kinds of things anymore.’ In this case, it seems that agency-based resilience strategies, despite political leanings, need to be perceived as socially legitimated, which under a neo-liberal environment, favours non-government ownership of facilities.

Another way agendas are critical to this examination is how it affects the long-term viability of the resilience project in terms of financing and management of the science park. The municipal government achieved broad consensus with other political parties, for its involvement in the crisis, with appeals to regional affinity cutting across party lines and enjoying multipartisan support. This secured the long-term leasing contract from the municipality with the park, helping to make it financially viable. Thus, the ruling political coalition was effective in mobilizing resources in response to the closure and coalescing under an agenda of regional affinity and solidarity.

5.11. Institutional context

The institutional context that might explain relative consensus over the agenda of saving the facility is that the Astra part of AstraZeneca was originally founded in Södertälje over a hundred years ago. It has a manufacturing legacy in this region which has employed thousands of people from different generations. The company is embedded in the industrial and regional history, heritage and identity of Södertälje which helps explain why appeals to regional affinity and solidarity were so effective in forming coalitions and mobilizing resources for this cause. It is interesting to note that those unemployed by the closure have access to other employment possibilities, given the proximity to Stockholm. As well as the fact that many of these employees were not living in Södertälje so were not paying taxes to the municipality. One could argue that the financial loss to the Södertälje municipality would not have been debilitating if it had simply done nothing. But the closure seemed to present risks more abstract and cognitive in nature; the perception that Södertälje might be in crisis once more was something that actors stated they wanted to counteract. Prior to the closure, Södertälje had suffered from incidents of high-profile organized crime and previous economic crises. These experiences seem to have affected the sense of urgency to work towards a path of sustained recovery. Cognitive frames of actors informing their expectations and perceptions of risk in the region, is based on socialized experiences (Bristow & Healy, 2015). Consequently, risks of the closure were perceived to be grave and with serious political and social ramifications, such that the new-term mayor of Södertälje municipality recalls the exact moment when the closure was announced and remembers cancelling every appointment that day and holding an emergency meeting with her staff.

5.12. Emergent outcome – adaptability

Södertälje seems to have the potential to foster adaptability with SSP, hosting companies from three different industries, namely food, sustainable manufacturing and biotechnology. The main strategy is co-location of diverse companies, expansion of KTH to 1200 students, and various activities encouraging sustainable production, innovation and education. Co-location could facilitate new networks and new modes of cooperation that underpin the process of adaptability. However, the amalgamation of three different industries at SSP does not lend itself to a clear framework to anchor the development and operationalization of the park. It is unclear how to deliberately foster interdependencies and how to design organizational structures and conditions in order to further trigger mechanisms of adaptability. Decisions need to be made with regards to the scaled up park; whether the main goal is to facilitate innovation or to optimize the use of the facility or merely to fill the park at maximum capacity. It remains to be seen whether co-location is enough to spur collaboration and foster adaptability.

It needs to be noted that degrees of adaptation is observable in the constitution of SSP, particularly with the involvement of Scania and the emphasis of manufacturing at SSP which demonstrates the continued reliance of Södertälje on this company, specifically, and on the manufacturing industry, generally. Södertälje seems to fall back on enabling industrial paths that have proven to be successful for the municipality in the past hundred years since the founding of Scania; but with improvements represented by the ‘sustainable’ component. There is novelty in the inclusion of the focus of the food industry particularly with the project Matlust intended to develop small and medium sized enterprises in sustainable food production, and co-location with manufacturing and biotechnology within SSP. But the selection of industries to co-locate in SSP seems to signal former and existing core competences in the region which demonstrate degrees of adaptation.

6. Conclusion

In order to interrogate the research question ‘how does agency emerge and function in the context of a regional economic crises’, this paper adapts the agency perspective of resilience in analysing empirical evidence from a comparative case study of the closure of AstraZeneca in both Lund and Södertälje, in Sweden. This framework has empirical currency in explaining both cases. Main findings of the paper include the efficacy of agency-based resilience strategies in mobilizing resources to allay local crises in the short-term. In the case of Lund, agency emerged from a coalition of individuals from organizations in the region such as the university, the regional government, a philanthropic organization as well as key affiliates motivated to showcase an innovation concept. Dense networks enabled formation of this coalition and consequently, access to critical assets like financial, social and political capital in order to create Medicon Village. Medicon Village has contributed to the adaptation of the region particularly in answering new demands of the life science industry by enabling the emergence and development of biotech startups. With Lund opting to pursue an adaptation strategy, lock-in is a distinct possibility. However, substantial resources earmarked to fostering collaborations within and beyond Medicon village seems to hold the risk of lock-in at bay, for the time being.

In comparison, agency emerged in Södertälje from a newly-elected municipal government leaning on a heritage company to rally the case for helping Södertälje in the name of regional affinity and solidarity. Improvising on the Lund model, given a lack of strong university presence, Södertälje municipality took the lead in enabling actors to form a consortium, Acturum, to build a science park called Biovation Park. It has since broadened the scope of this coalition by creating a scaled-up version of its initial science park, the larger Södertälje Science Park, set to launch in 2018. SSP focuses on sustainable production and hosts various companies from different industries, namely food, manufacturing and biotechnology. Their co-location at SSP could potentially engender unique cooperation and combinations that could lead to adaptability in the region. It remains to be seen, however, whether co-location is enough to foster these kinds of collaborations given an unclear mechanism on how to further trigger this process.

This paper finds that agendas of actors are critical to resilience projects in two main ways. Firstly, the abilities of actors to coalesce and align their interests and agendas underpin the process of coalition formation. Secondly, resource allocation determines which agendas are privileged. These, in turn, locate conflicts that may have been temporarily sidestepped but potentially leads to recurring opposition to resilience projects, affecting long-term viability. This paper also points to the importance of examining the institutional context such as cognitive frames and perceptions of risk as explanatory factors. Lund municipality was conspicuously and relatively less active than the municipality in Södertälje in terms of responding to the closure of AstraZeneca. This can be explained by the institutional affinity of Södertälje with AstraZeneca as one of its heritage companies as well as past experiences with crises which contributed to the impetus for action and resource allocation. Agency-based adaptive strategies affect the emergent outcomes for adaptation or adaptability, or combinations of both, in regions. Although adaptability is often lauded in the literature over adaptation, in practice, this strategy from an agency perspective seems to be challenging. It is a strategy fraught with complexity and no clear mechanisms for how actors are supposed to deliberately trigger the process of adaptability with respect to their regional context. And although adaptation is at risk of lock-in, it seems to be possible to foster variety within an industry by carefully selecting companies with intersecting interests and going beyond co-location in fostering collaborations. Proximity to regional assets enable this possibility.

Implications of these findings are varied. Bespoke strategies recognizing nuances such as institutional context, varieties in organizational capacity and networks in regions are necessary for sensible policymaking for crises situations. Researchers should be careful, however, in overstating the capacity of agency-based resilience as it is neither a permanent feature nor an automatic process in regions. There may not be existing efforts to counteract potential economic hazards in the region from which actors can quickly and reliably draw. Economic crises bring forth modes of cooperation that the findings of this paper shows to, potentially, be temporary. Long term viability is challenging considering actors are subject to the contingencies of institutional contexts such as cognitive frames, perceptions of risks, agenda setting processes, and resource distribution conflicts that make them limited in terms of their ability to continue allocating resources. Although, these limitations are not insurmountable, it is not clear how they can be systematically overcome by actors. More research on the institutional work needed to make agency based resilience more tenable in the long term is vital. Regional assets of social, political and financial capital are important factors in the ability of these coalitions to come together. Especially since significant resources need to be expended in order for agency-based resilience projects to succeed in the immediate and long term. The adaptive view of resilience has provided analytical space for a micro-perspective and a notion of resilience examining the reflexive behaviour of actors in regions. It brings the human back into the resilience process and by doing so, contributes to an understanding of the varied textures of the economic landscapes of regions.

Acknowledgements

I would like to thank colleagues at CIRCLE and at the Department of Human Geography, Lund University for helpful comments and all interviewees for their time and effort. This author is grateful for constructive feedback on early drafts of this paper from Lars Coenen, Jerker Moodysson, Roman Martin, Martin Andersson, and Josephine Rekers. Any errors in the paper are entirely my own.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

Financial support is generously provided through project funding from FORMAS, VINNOVA and by the Swedish Research Council.