Crises are of growing interest to researchers focusing on the complex issues at the business and society nexus. Some scholars have examined existential crises such as accidents, emergencies, or disasters (Williams & Shepherd, 2016), whereas others have looked at more wide-spanning crises such as the economic crisis of 2008 (Munir, 2011) and the so-called refugee crisis of 2015 (Guo et al., 2020). This scholarly attention is more than warranted in the crisis-ridden world we live in today. However, researchers seem to have neglected two aspects which should be central for our understanding of how we address crises effectively: Most of the crisis literature has an instrumental focus (Bundy et al., 2017) rather than one on normative questions and the moral character of crises, which would help us understand the role of responsible and ethically guided action in meeting them (Islam & Greenwood, 2021). The literature is also focused on individual organizations (Bundy et al., 2017) rather than on how institutions are dealing with the adversity that crises produce, which may help us derive implications for more systemic responses to crises (Williams et al., 2019).

The concept of institutional resilience refers explicitly to the interplay between crises and institutions as “the process whereby institutions recover after having undergone a significant disruption” (Barin Cruz et al., 2016, p. 971). Unfortunately, and much like a mirror of the literature on crises, while research on resilience is increasingly present in organizational research (Hillmann & Guenther, 2020), we lack investigations of resilience at the institutional level (Williams et al., 2017). Institutions capture the regulations, norms, and rules in a field (Hinings et al., 2017; Scott, 2001). They are therefore especially relevant for understanding reactions to the moral character of crises. Previous research furthermore suggests that multistakeholder collaborations are of major importance in view of complex problems, which crises inarguably represent (Gray & Purdy, 2018). Such collaborations could thus play a major role in institutional responses to crises. The research on collaborations has helped us understand the instrumental motives for organizations to collaborate, e.g., for partners to join alliances. What we need in addition, however, are deeper insights into the moral motives that organizations may have to engage in collaborative problem solving (Bakker et al., 2019). Such an angle would help us better understand what type of actor might take what kinds of action to contribute to institutional resilience and address crises, and why organizations choose to act in these ways (Wang et al., 2022). To address the overlapping blind spots across the three literatures, which center on a coinciding lack of attention to the institutional level, multistakeholder engagement and aspects of morality, I ask: How do existing institutions cope with the adversity created by moral crises? And how and why do organizations engage in collaborations to increase institutional resilience, that is institutions’ ability to deal with such adversity?

In the work integration fields of three European countries, that is the institutional arrangements meant to integrate disadvantaged groups into the labor market, I examine how multistakeholder partnerships (MSPs) contributed to increasing institutional resilience in response to the economic crisis and the refugee crisis. The international work integration experts who were consulted in the research identified MSPs as a novel response to work integration challenges, because MSPs seek to provide an integrated response to skilling, educating, and employing people from disadvantaged backgrounds, whereas these steps were traditionally separated (Gardin et al., 2012). I define the considered crises foremost as moral crises, where public, private, and nonprofit actors choose to engage or not engage out of a sense of responsibility. Moral crises offer the possibility to advance the business ethics discourse as to when, why, and how organizations choose to act on challenges where these challenges do not inhibit them in a direct way, but where through acting organizations contribute to institutional resilience. The comparative process tracing research was performed over a period of two years within a large project involving a research consortium. In addition to 44 expert interviews, the analysis builds on secondary data, for instance from published reports or vignettes of exemplary partnerships. I use this breadth of data to reconstruct and compare processes at the nexus of moral crises, institutional resilience, and multistakeholder partnerships and develop a framework and research propositions for future research.

I contribute in the following ways. First, in contrast to instances of deinstitutionalization or struggles for organizational survival that are typically studied as consequences of crises (e.g., DesJardine et al., 2017), I suggest moral crises may act in an enabling fashion by galvanizing fragmented approaches into joint action. Moral crises may create pull factors for joint action, when actors reinterpret crises as opportunities, and push factors, which force previously disengaged actors to take a moral stand. Second, adding to research that encourages more theory on the procedural dynamics between crises and resilience (Williams et al., 2017), I highlight nested contingencies that determine pathways to institutional resilience. My article shows that reactions to crises will depend on interaction between (1) the capacity of existing institutions and the level of adversity produced by crises, but also on (2) how much new actor constellations can build on existing institutional precursors and whether crises challenge institutions directly or indirectly. Third, while we see more research on how multistakeholder collaboration is beneficial in view of wide-spanning, longer-term crises (e.g., Fehsenfeld & Levinsen, 2019), we lack understanding about which type of actors will take the lead in contributing to institutional resilience, and under which conditions (Wang et al., 2022). I suggest that civil society organizations will act on moral motives early on and play a more important role in contributing to institutional resilience in contexts less severely hit by crises, while businesses are more hesitant to take over responsibility but take the lead when adversity is very high. My contributions open up new directions for investigating how actors may collaborate to increase institutional resilience that is impeded by crises. These contributions are pertinent not only because of the COVID-19 pandemic, but relevant in view of other major crises that society has been, is, and will be facing.

Crises, Institutional Resilience, and Multistakeholder Collaboration

Lack of Moral and Institutional Approaches to Crises

Crises, that is unexpected events that have high impact and present a threat, are a major factor provoking change in organizations and institutions (Pearson & Clair, 1998). They create adversity, a concept that comprises a variety of meanings, ranging from stress, to disturbance, to shock, to disruption (Williams et al., 2017; Hillmann & Guenther, 2020). Some streams of literature have focused on crises of, for, and in organizations, for instance, in governance (Maitlis & Sonenshein, 2010), reputational management (Desai, 2011) or relational management (Kahn et al., 2013) and how organizations cope with these challenges. Other streams have turned to crises with wide societal consequences and how organizations reacted to them. Part of this research has dealt with what we might refer to as existential crises such as accidents, emergencies or disasters, which organizations must react to immediately because entire livelihoods are at risk (Williams & Shepherd, 2016). Another part has looked at more wide-spanning crises such as the economic crisis (Munir, 2011) and the refugee crisis (Guo et al., 2020), which I argue especially in advanced economies foremost represent moral crises.

The idea of moral crises has its roots in the debate about businesses’ social responsibility and refers to situations when organizations struggle to decide on a course of action because they need to choose between profit maximization and ethical behavior (e.g., Zenisek, 1979 in reference to Petit, 1967). In contrast to this organizational level focus, I define moral crises at the institutional level as situations where (disadvantaged) groups in society rather than organizations themselves are under high and material adversity, whereas organizations may or may not feel responsible to act and meet the adversity. This definition is not meant to negate the profound consequences of, e.g., the refugee or economic crisis for the people directly affected. Instead it highlights, from the perspective of institutions or organizations, that they themselves are not under severe strain and therefore have some discretion as to whether and how they choose to become active. This is different from when there is, e.g., a major corporate scandal which the organization must address to survive. What follows is that moral crises are an ideal setting for probing the ethical foundations of organizational activities, with a particular emphasis on moral considerations in embracing or rejecting responsible and morally charged action (Islam & Greenwood, 2021).

Moral and ethical issues have undoubtedly been covered in research on crises. For example, in case of the economic crisis a lot of research focused on how violations of moral and ethical behavior gave rise to the crisis itself (Graafland & van de Ven, 2011; Kvalnes & Nordal, 2019), or how changing professional behavior may prevent such crises in future (Aldohni, 2018). In the context of the refugee crisis, morality was tied more closely to organizational actions, for instance as an important mechanism that spurred collective action (Kornberger et al., 2017). However, the lion share of the crisis literature “takes an instrumental approach to crisis management, focusing on efficiency and effectiveness as opposed to moral and normative obligations” (Bundy et al., 2017, p. 1682). What is more, the crisis literature tends to zoom in on individual organizations and focuses “on internal dynamics of a crisis” (Bundy et al., 2017, p. 1661), instead of considering how crises affect institutions and strain their resilience (Barin Cruz et al., 2016).

As I further unpack below, what I do here is different. I adopt a normative focus on the effects and dynamics of moral crises. This concerns for example how some actors are morally pro-active in tackling crises whereas others act passively and out of a sense of coerced duty or urgency, and what this means for building institutional resilience. This lens helps me to identify when and where moral responsibility mobilizes anticipatory action, or where by contrast moral pressures need to grow strong enough to provoke interventions. This may happen when dominant institutions are incapable of acting or when societal stakeholders express high moral demand for action.

Lack of Institutional Approaches to Resilience

There is an intricate connection between crises as a source of adversity and resilience as the capacity for dealing with the adversity (Williams et al., 2017). Institutional resilience as a concept (Barin Cruz et al., 2016) has been quite present in the healthcare field (Carthey et al., 2001), political science (Lowndes & McCaughie, 2013), or dedicated crisis research (Hills, 2002). In these literatures institutional resilience often refers to entire systems or regimes and how they cope with fundamental challenges of ensuring encompassing service provision under scarce resources, or in the absence of regulatory stability. Research on resilience has a long tradition in organizational research too, but much in the same way as the literature on crises, previous research mostly focuses on resilience at the level of the organization rather than that of institutions (Williams et al., 2017; Hillmann & Guenther, 2020). In consequence we know a lot about the structures and practices that enable organizations and their members to figuratively bounce back when they are under strain or process high levels of task complexity and uncertainty when crises occur (Powley et al., 2020).

However, we know much less about how wider institutions, that is a collective of diverse actors (public, private, and/or nonprofit) that constitute for example a field or industry, react to or manage to cope with crises. In the rare occasions when previous studies have examined resilience at the intersection of organizations and institutions, they focused, for example, on how organizations perform maintenance and repair work to withstand institutional change that is imposed on them (Micelotta & Washington, 2013). Institutional resilience instead refers to the capacity of wider institutions to cope with adversity, “a bracketed stability […] making it easier for actors and organizations [that are part of the wider institutional setting] to absorb disturbance” (Barin Cruz et al., 2016, p. 975; my addition). The concept has been proposed and continues to be used in the realm of extreme operating environments that are risky and marked by disruption (Hällgren et al., 2018). It has also seen wider uptake, for example as a relevant dimension for studying the temporary organizing in project-based management and the change such projects may trigger (Naderpajouh et al., 2020). However, it has not been applied in the context of wider and long-term crises and rarely in more systems-oriented studies of resilience (Williams et al., 2019). Institutions, in contrast to individual organizations, enforce regulatory, normative, and cultural rules in a field (Hinings et al., 2017; Scott, 2001). I assert that thereby institutional resilience becomes a concept which refers to the aggregate capacity of those elements of a field as well as the actors populating it to deal with the consequences of different kinds of adversity. Its focus on rules, norms, and multiple actors makes institutional resilience a relevant reference frame to examine responses to the moral crises discussed above.

Given the increasing importance of crises that affect whole fields, regions, nations, or even the globe and that challenge existing institutions profoundly, such as COVID-19 (Brammer et al., 2020), more research on the relations between institutional resilience and moral crises is warranted. Scholars have also stressed that we need a better understanding of the process of increasing resilience, as opposed to merely reinstalling resilience to some previous status quo (Linnenluecke, 2017). This is of major importance, because moral crises may demand novel responses that deviate from that status quo.

Lack of Moral Grounding to Collaboration in Crises

Multistakeholder collaboration has been highlighted as an effective response to social problems that are growing more complex, whereby crises are inarguably a driving force of that complexity (Gray & Purdy, 2018). To date, the work on multistakeholder collaborations often looks at how these collaborations create superior value for the partners themselves or for society, by way of unusual combinations of activities and viewpoints (Cabral et al., 2019; Quélin et al., 2017). In terms of motives, a recent review shows that the majority of studies in the area looks at collaborations from a utilitarian and strategic perspective and highlights instrumental motives for partners to collaborate, whereas the moral dynamics which might underlie multistakeholder collaboration remain neglected (Bakker et al., 2019). In terms of remit, research has mainly had two focuses: First, previous work has looked at how collaboration leads to gradual improvements against wide-spanning challenges, such as sustainability in supply chains, where the goals to be achieved are very broad (The Annual Review of Social Partnerships, 2018). Second, scholars have stressed that diverse actors when collaborating, bring in different regulatory, participatory, or resource competencies that can help address crises (Pinkse & Kolk, 2012). The benefits of such a combination of competencies have often been probed in relation to emergency situations, when crises are existential because chaos and confusion reign and partners get together to address temporary institutional voids at a fast pace (Boin et al., 2017).

As opposed to these temporary and spontaneous acts of collaboration, scholars have called for more attention to institutionalized collaborative constellations between a wide range of stakeholders from different sectors (Di Domenico et al., 2009). In contrast to collaboration on more generic challenges, we are recently seeing more attention to how collaborations may address moral crises which affect entire fields over a longer period of time. Interestingly, this latter research has probed the interplay of different logics in collaborations (Hesse et al., 2019), the promotion of values as an outcome (Daskalaki et al., 2019), or the relations between volunteer and professional work in meeting the crises (Fehsenfeld & Levinsen, 2019). However, the role of morality and ethics in driving responsible and values-oriented behavior that may or may not lead organizations to enter and propel such collaborations is left underexplored. This circumstance in turn limits our knowledge about what type of actor might drive the collaboration under what conditions (Wang et al., 2022). This question matters especially in moral crises because contestation of actor motives and benefits from collaboration will be high and consequences at the level of institutions uncertain (Nohrstedt et al., 2018).

In this article I seek to bring together the three strands of research to address the blind spots located at their intersection by asking: How do existing institutions cope with the adversity created by moral crises? And how and why do organizations engage in collaborations to increase institutional resilience, that is institutions’ ability to deal with such adversity?

Methods

Background and Focus of the Research

Work integration is considered one of the key building blocks of European welfare states, since it is meant to tackle social inequalities and help disadvantaged groups of the population enter the labor market (Battilana et al., 2015; Bode et al., 2006). Institutional resilience in the work integration field has been strained by both, the economic and the refugee crises. In contrast to droughts or natural disasters, the breakdown of entire economic regimes, or in fact the war that made people flee their home countries, these crises represent mainly moral crises for organizations and institutions in European countries. While straining national governments and their budgets substantially, few of the crises’ effects threatened established institutions themselves. However, the crises undoubtedly increased division and mistrust among citizens and made public opinion call into question the resilience of the existing system in place (Kerr, 2013; Morgan et al., 2011). A lot of the action that ensued to meet the consequences of the crises was driven by those public perceptions and media reporting, including moral evaluations (Hesse et al., 2019; Vaara, 2014). Reactions were often based on solidarity and other means of ethically driven, communal coping that involved a variety of stakeholders (Daskalaki et al., 2019; Mensink et al., 2019).

Against this background, a research team conducted a cross-country and multi-step consultation of academic, practitioner and policy experts in work integration to find out where they saw new approaches for increasing the field’s institutional resilience. Experts outlined that formalized multistakeholder partnerships (MSPs) represented a new approach in work integration that was contributing significantly to increasing institutional resilience. Experts and extant research highlighted that the challenge of work integration had previously been addressed by two largely separate spheres of intervention. On the one side, training and qualification were performed by nonprofit actors, in some European countries primarily by so-called work integration social enterprises (WISEs) (Kerlin, 2006). Whereby a substantial share of employment was generated in quasi-public, second labor markets instead of the regular first labor market (Göler von Ravensburg et al., 2018). On the other side, when actual integration into the first labor market took place, it was financially promoted by the state and ultimately realized by the business sector, with a relatively stark disconnect from the training and qualification work (Gardin et al., 2012).

In contrast to this previous status quo, MSPs tackle the work integration challenge in an integrated fashion: in a joint effort, civil society organizations, the state, and firms reach out to disadvantaged target groups, perform multi-modal training activities to secure easier transition from education to practice and to enable long-term employment for disadvantaged target groups. The MSPs that experts suggested are defined by five core traits (see also Leca et al., 2019). MSPs: (1) involve partners from at least two, but ideally all three sectors: the private (commercial), the public, and civil society sector; (2) are formalized, not necessarily as a separate entity, but as a clearly identifiable organizational form with a name and contracts governing the partners’ work; (3) entail active resource contributions from all partners in the operation of the partnership, be they financial investment or expertise; (4) ensure the representation of all partners in the governance of the partnership, and (5) are built on reciprocity, that is joint action toward a common goal.

Country Settings

The research was conducted over two years from 2015 to 2017 within a major EU-funded research project. Table 1 summarizes the seven steps that were performed and three selective decisions that were made during the process. The decisions concerned the countries and the phenomenon to be studied as well as the exemplary MSPs to be analyzed. It also displays the methods applied and the amount of data analyzed to move from one step to the next. First, three countries were selected from a bigger research consortium of nine project countries. The decision was based on country vignettes of the work integration fields in each country, including data from various reports. The eventual choice of countries was guided by two criteria to increase the likelihood of detecting major changes in the fields and differences between countries: First, the research looked at the importance of work integration in the respective countries. In this, the research considered the field’s significance judged by its degree of institutionalization and long-standing traditions as well as its perceived importance in view of societal challenges (Spear & Bidet, 2005). Second, the team considered dynamism in the field, including the salience of new developments, major reforms or controversies in political or public debates.

Table 1 Identification and analysis of MSPs
Full size table

A gradual process of comparison, led to Decision 1 to select France, Germany, and Spain for further investigation. Despite similar features, France and Germany present quite different fields of work integration, especially as regards state involvement. Even though the nonprofit sector is an important provider of work integration programs in Germany, the field is strongly state dominated. Besides, nonprofit actors often do not cooperate directly with the state but, in contrast to France, they largely depend on state subsidies to operate. In France, cooperation between the state and the nonprofit sector has traditionally been high. Spain contrasts strongly with these two countries. The Spanish state has a relatively low involvement in active employment policies and lets the market shape the field, while the nonprofit sector is traditionally strong in this realm to compensate for cases of market failure.

Selecting and Process Tracing the Phenomenon

Once the countries had been selected, a consultation of work integration experts was initiated in the three countries for a more comprehensive field description (Step 2, 11 interviews). Then a second step of consultation followed, with a specific focus on sensing new approaches for dealing with work integration challenges (Step 3, 8 interviews). Interviewees in these two steps were from a variety of backgrounds (see Appendix Table 6 for the full list of interviews). The interviews, building on material gathered and analyzed in the previous step, marked the beginning of a comparative process tracing that led to the identification of MSPs as the phenomenon to study and gradually deepened our understanding of their emergence and their contributions to institutional resilience across countries.

Process tracing is a method often applied in political science and especially suited to perform a systematic tracking of complex processes across contexts (Bennett & Checkel, 2017). It shares some traits with process research in organization studies (Langley, 1999). However, in contrast to most process studies which often build on subjective viewpoints and start with traces of a process (Langley & Tsoukas, 2016), process tracing typically starts with a materialized outcome (here: MSPs) and seeks to deepen insights on the emergence of that outcome while maintaining comparability across contexts (Bengtsson & Ruonavaara, 2017). In order to go analytically deep while upholding the comparative dimension, process tracing employs a relatively rigid set of questions (Collier, 2011). These questions serve as a scaffolding and are oriented at different phases of the process (see Beach & Pedersen, 2019): (1) What are the core traits of a phenomenon today? (2) What were critical junctures and influencing factors in its emergence? (3) What were the roles and motivations for actors to engage in the process and steer it into one direction or the other? While the analytic repertoire is predefined and process tracing follows a top–down logic (starting from the phenomenological outcome and systematically tracing it back) rather than a bottom-up logic (picking up an interesting thread and trying to craft the process), it allows for identifying and probing aspects that unexpectedly emerge from the analysis (such as the central role of crises; see below). A key strength of process tracing is that it reconstructs a process by means of forming an integrated perspective based on various kinds of evidence that can be unified due to a shared methodological framework (George & Bennett, 2005).

Process tracing may be used for theory testing in empirical settings that are quite saturated and where we have a set of competing explanations about a phenomenon. It is useful for theory building to specify the connections between several empirical elements (e.g., crises, institutional resilience, and cross-sector collaborations), where the intricate links and reciprocal influence are unclear, as is the case here (Beach & Pedersen, 2019). The main interest lay in identifying new approaches for dealing with work integration challenges. When the research was initiated MSPs were not predefined as the phenomenon to study, but they were suggested by the interviewed experts. In the same way, although it was clear they would play some role, the research team did not ask explicitly about the influence of the economic or refugee crisis. Instead, the team was asking about the potential role of crises more generally, or took them up as and when interviewees highlighted the role of crises. This combination of analytic rigidity, resulting from a predefined set of questions, and flexibility to account for country specificities, especially as to the role of crises and contributions of MSPs, enabled us to compare three processes, which although resulting in the shared phenomenon of MSPs to increase institutional resilience, had quite different evolutionary paths.

The questions that were applied in the interviews (see Appendix Table 7), but also in relation to the other material we gathered and analyzed (reports, data from websites, etc.), went from more general to more specific. For example, the team moved from asking “what is the status of the work integration field in your country?” in order to understand the general context, to “how did the partnership evolve over time?” to understand dynamics of the MSPs’ evolution, including engaged types of actors and their roles. Decision 2, in particular the identification of MSPs as the phenomenon to understand factors that increased institutional resilience in work integration, was made based on the commentary provided by three advisory board members and discussion at the research project’s mid-term conference.

After the identification of multistakeholder partnerships as the phenomenon to study, a team of 4–5 researchers in each country documented a portfolio of national MSPs that generally adhered to the definition. In a continuous process of discussions within and across the teams, a list of partnerships was compiled that would qualify for exemplary analysis. The turnout of this process was very different across the countries (see Decision 3). The French team had to short-list a handful of partnerships from over 20 identified MSPs that adhered to the definitional criteria I introduced. In Germany in turn the initial list had over 100 examples, of which, however, not even a third were formal partnerships and only two adhered to the definitional MSP criteria at all levels. Finally, in Spain about 10 collaborations were identified, one of which represented a very strong fit; an MSP with a membership of over 1000 organizations—and thus far greater in size and scope than the exemplary MSPs in Germany and France. MSPs were selected based on adherence to the five definitional criteria introduced above. Table 2 characterizes the specific partnerships finally studied as exemplars of the cross-national emergence of MSPs in work integration.

Table 2 Characteristics of exemplary MSPs
Full size table

The research teams performed semi-structured interviews with representatives of the exemplary MSPs, in each case involving decision makers from firms, civil society, and public administration. In combination with the previous ones, 44 interviews were conducted in total (see Appendix Table 6 for country split). All interviews were conducted, transcribed, and coded in the local language, together with the other sources of data. Two researchers, who sought consensus when initial assessments were different, separately coded each document or transcribed interview (Miles & Huberman, 1994). Due to the structured approach of process tracing, rather than looking out for themes to autonomously emerge from the data, the national teams used the scaffolding the guiding questions offered to allocate statements into three broad categories. (1) how MSPs represented a deviation from existing institutions as well as how they contributed to increasing institutional resilience; (2) reasons for the emergence of MSPs, in particular the role of the crises therein; and (3) actors’ roles and motivations in the emergence of MSPs.

Once this systemization was completed, researchers jointly brought all these elements into a chronological order for each country with an explicit view to specifying the role of crises and collaboration, and their consequences for institutional resilience. In the interviews themselves, the team did not employ the terms I now use to conceptualize the findings, but relied on terminology the interviewees were more familiar with. For example, to relate to adversity the team asked for challenges the work integration field was faced with. To sense how MSPs increased institutional resilience the team prompted interviewees to describe how these helped deal with challenges.

As per the previous steps, interviews were only used as one source of information. The team also gathered further information, for example from the MSPs’ or the involved partners’ websites. It is important to restress that the studied MSP were treated as an example of the wider phenomenon rather individual organizational cases. The analytic focus was always on the process of increasing institutional resilience as well as the contributions of various actors to it. All information combined resulted in an individual trajectory of MSPs in each country, which subsequently enabled comparison across countries (see Fig. 1 below). Tracing MSPs back to their origins, resulted in different time spans in each country and returned different roles of crises for institutional resilience. Only because of the comparative setting can I now identify contingent factors on the interplay between crises, institutional resilience and MSPs.

Fig. 1
figure 1

Comparative analysis of MSP emergence to increase institutional resilience

Full size image

Findings

Figure 1 maps the three MSP trajectories over time. I present the figure upfront to illustrate the different trajectories and outline the factors that explain differences in the shared MSP phenomenon. The vertical axis is without unit and only serves to display countries in a comparative fashion rather than independent of each other. The figure highlights key incidents, whereby light circles zoom in on major instances that I refer to as focal points. It also highlights the actor coalitions involved by marking influence from civil society (circled star), businesses (triangle), and the state (cuboid). Dotted lines represent preformalized or fragmented activity, solid lines mark the emergence of coordinated efforts and bold lines the emergence of MSPs according to the definition. I want to stress that the upward slope signals that MSPs have evolved since when they first emerged, it does not mean that the individual MSPs went from strength to strength without any struggles, or that there is no risk the phenomenon may lose relevance for strengthening institutional resilience in future. The vertical and horizontal arrow in the lower right corner signify that the scope and pace of the shifts depended on the interplay between the nested contingencies that I unpack after discussing each country.

As the role of crises got more and more prominent in the data analysis, not only as a source of adversity for institutional resilience, but also as an enabler of the increase in resilience, I treat crises as a central influencing factor for the evolution of MSPs rather than as mere context. Table 3 provides quotes and indicates the influence as well as the moral consequences of crises in each country. I forestall it for the same reason as the figure, namely to set the scene and provide directions. In what follows, I first establish a core explanatory thread for each country and then go into the comparative analysis.

Table 3 Crises: influence and moral consequences
Full size table

MSPs in France: Building Out a Long Tradition

In the research, France yielded a larger number of exemplary MSPs than Germany and Spain: in fact, there were too many to relate to all in the analysis, which is why I focus on describing shared trends here rather than focusing on specific partnerships as done in the other national settings. The first partnership started in 1996 (an MSP between Adecco and Groupe Id’ees). In addition to a gradual evolvement there seemed to be an increase in numbers around 2010 (examples are: Reseau Cocagne & Fondation Caritas, or several partnerships between Fondation Veolia and smaller-scale WISEs), possibly as a result of the economic crisis, although no partnership explicitly suggested this. MSPs, with some exceptions, were forged between WISEs of considerable size and with a history often long predating the partnership (see WISE tradition as a prerequisite for MSPs in Fig. 1). Other examples of WISEs involved in MSPs are Ares or Vitamine T. All mentioned organizations have more than 650 employees. The formation of MSPs, although primarily established between business and WISEs, was pushed strongly by the French state with two key incidents: First, the introduction of social clauses in France dates back to the mid-1990s and made work integration a central prerequisite for winning public tenders, favoring WISEs or the collaboration of WISEs and firms. Second, a consultation process initiated and administered by the French state in 2007/08 known as Grenelle de l’Insertion (see focal point in Fig. 1). It was pushed for by nonprofit networks acting on poverty and thus indirectly caused by economic challenges affecting disadvantaged groups. This makes it a reaction in response a latent crisis (see influence and moral consequences in Table 3). The formation of MSPs, promoted by the consultation, explicitly aimed at mobilizing cross-sector and multistakeholder collaboration to address this latent crisis. A representative of the French employer’s union MEDEF described the situation in France as follows:

“I used to get a lot of flak [from employers] for working on these topics…That was how things were ten years ago. I think there has been a real improvement in what that type of organization [a WISE] can offer a private firm. CSR has also become more important over that time period…there are things that cause a lot less debate than before. Today, we’re seeing more and more original partnerships between private firms, the third sector, and WISEs.” (FR 2)

By accounts of several interviewees the consultation process was significant for the development described above. Another interviewee stressed that “it was a time of mutual discovery for private firms and WISEs.” It even led to legal changes such as the drafting of a model agreement of partnerships originally not foreseen in French law. Despite its history and the significant number of high-level engagement with the issue, the experts interviewed suggested that there was not enough activity, given both the scale of the WISE sector and the number of disadvantaged, unemployed people. Several reasons for this were proffered, in particular a fragmented WISE sector, the legal barriers mentioned above and a lack of professionals with hybrid profiles, that is an expertise in and orientation at pursuing social and commercial objectives at the same time. Nevertheless, there was growing recognition of the benefits of integrated approaches and formalized partnerships early on.

MSPs in Germany: Civic and State Action Preceding Firm Engagement

During the first decade of the new century, there was a fairly stable environment in the state-centered field of work integration in Germany, with an established landscape of providers. As interviewees stressed, most pathways to work integration went through separate workshops, located outside the first labor market, or through public–private partnerships predominantly between the state and firms (also Bode, 2011). According to one interviewee, more inclusive and holistic approaches to work integration were promoted by political actors, mainly by the European Commission which in recent years performed a lot of agenda setting as regards new solutions to social challenges. Another important factor triggering new work integration efforts was the steep rise of asylum seekers in 2015, which affected many countries but Germany in particular, both, in terms of the number of refugees seeking permit to stay and political polarization that resulted from what became known as Angela Merkel’s Willkommenskultur (see focal point in Fig. 1). In consequence, after a long stationary period, the field became very dynamic with unprecedented actor constellations appearing on the scene. All interviewees agreed that the refugee crisis mobilized a considerable amount of resources across all sectors.

Two specific initiatives adhered particularly well to the concept of MSPs. Both represent partnerships between a nonprofit organization experienced in providing vocational training to disadvantaged groups, public policy actors or public administration (State Senate of Berlin in one case, the Chamber of Industry & Commerce in the other) and private local businesses. Both MSPs aim at integrating all steps of the work integration process: qualification, often through on the job training; placement in the participating firms; and mentoring once placement was successful to ensure job retention. While interviewees agreed the founding partners had been equally behind the establishment of the MSP, the provided training related to previous programs run by the nonprofit partners. What is more, interviewees agreed that increasing public awareness for the challenges caused by a significant influx of refugees led to a ‘politization’ of the social climate which was a driving factor for small and medium-scale firms to engage in such work integration efforts. As the public discourse progressed, they were forced to take a stand:

“The need is getting bigger and the social discourse is politicized. This led to a new attitude with firms or those governing them, motivated by influences from their private life: They now have an attitude, which is pro-integration and against populist xenophobic tendencies and they want to bring this into their firms. We recognize increasingly that there is readiness and interest and in fact active humanitarian engagement on the side of firms as regards refugees.” (GER 7)

The second reason why firms got engaged was that they saw the crisis as an opportunity to meet a shortage in skilled labor, which had been an issue in Germany for many years (see second influence and moral consequences in Table 3):

“[I]t [the motivation] ranges from meeting a lack of qualified labour and an almost non-existent aspect of civic engagement, to a readiness to engage for society.” (GER 8)

The bottom line of these observations is that firms in Germany were followers of efforts previously initiated mainly by nonprofit actors or the state. Among the partners, there was clearly an impetus for the initiatives to become a blueprint for the wider institutionalization of new work integration practices: “[W]e don’t have to build up a huge [name of the individual initiative]. Our self-conception is to nudge, to initiate something bigger” (GER 9).

MSPs in Spain: Firm-Civil Society Network to Cope for Weakened Institutions

The first WISEs in Spain were created in the 1980s by local neighborhood and church associations (Vidal & Claver, 2005). The public sector in Spain promoted WISEs as a tool of active employment policies, but also fostered the engagement of firms as direct employers of disadvantaged people. Yet firms remained unresponsive for a long time and according to the interviewees, the public sector in Spain played a rather passive role in regulating, funding and supplying targeted initiatives (Rey-Garcia & Mato-Santiso, 2017). Until today nonprofit organizations are seen as the leader in work integration due to their representation of the most vulnerable groups of the population (Gobierno Vasco, 2012). However, corporate foundations, located at the borderlines between the for-profit and nonprofit sectors, and companies have gained in significance over time. They became active within their social responsibility strategies, as recruiters, and also by designing tools for work integration programs.

The economic crisis that hit Spain in 2007 with effects unfolding until 2009 and thereafter led to a surge in unemployment and widened the share of people that were considered socially excluded (see first focal point in Fig. 1). During the first years, economic stimulus policies were implemented, such as Plan E, a program to stimulate economic growth through public funds. However, according to one of the interviewees such initiatives, by focusing on those who had recently lost their jobs, exacerbated rather than improved the situation for the most disadvantaged groups, such as low qualified workers, people in remote areas or immigrants (also MSSSI, 2013). This lifted the precariousness of their situation over a critical threshold (see third set of influence and moral consequences in Table 3):

“Government´s priorities of reducing the global unemployment rates have damaged the work integration possibilities of the most disadvantaged since 2008” (ES 2).

This situation provoked solidarity action. Initial exchange between business and nonprofits culminated in 2012 in the formation of the MSP “Together for the employment of the most vulnerable people” (Juntos por el empleo de los más vulnerables). Interviewees reported that this happened in particular out of increasing awareness that fragmented efforts would not be able to have sufficient impact and that the government’s capacity in helping people who had been disadvantaged for a long time was severely limited:

“There was a time in which isolated initiatives came together, (…) in some way, there was a confluence, and an agenda and other activities were set, and somehow something collective was created.” (ES 2)

The MSP brings together over 1000 nonprofits, around 70 businesses and 13 public agencies. Accenture Foundation plays the most pronounced role in governing the MSP, while the network of members is more reactive and selective in their contributions. Among the main activities are: training in basic and transversal skills, documentation of the situation of disadvantaged groups and corresponding advocacy efforts, or the provision of microcredits. Despite the magnitude of this new form of organizing, and the realization that “it was an effective response to the economic crisis” (ES 4), almost all interviewees from Spain saw MSPs as being in an experimental stage and noted that they emerged late, in particular when held against to the magnitude of the work integration challenges in the country:

“It started not many years ago, and also the development is still very small. I think that we started talking about the need of that collaboration many years ago, but effectively I think the first experiences with more development and more strategic visions took place probably since 2008.” (ES 4)

A galvanizing factor in addition to the unresponsiveness of government was the legal pillarization of problem areas in Spain, which also led to significant variance in the pace of the different activities and services provided by the MSP:

“In Spain we divide [the challenge] into segments. Therefore, because of segmentation that depends on the type of vulnerability, there are processes or projects that are more advanced than others” (ES 3).

These factors in combination led to a significant reinterpretation of the responsibilities but also of capacities of businesses in dealing with the crisis.

Nested Contingencies of Institutional Resilience

The differences in increasing institutional resilience through MSPs in reaction to the crises depended on nested contingencies, which I conceptualize in Table 4 and exemplify along the empirical observations in the countries in Table 5. The first nested contingency exists between (1a) the capacity of existing institutions, and (1b) the level of adversity provoked by the crises (causing, e.g., momentary inhibitions of institutions). The second nested contingency exists between (2a) traditions in the institutions of the field (in particular institutional precursors), and (2b) the link between the crises and existing institutions (direct or indirect influence). Table 4 indicates a potential range for each of these factors and shows when it would promote or not promote substantial shifts in institutional resilience through MSPs (marked by + or -, or ± where the direction is not clear). Only when both nested contingencies are considered together, do we understand their consequences for the MSPs and their contributions to institutional resilience. This is symbolized by the two three-branch scaffolds and their interaction (displayed as arrows) which lead to three different levels of shifts. A strong shift occurs where entirely new actors take the lead in MSPs as a result of crises. A weak shift occurs when there is continuity in previous actor constellations.

Table 4 Nested contingencies of institutional resilience
Full size table
Table 5 Country examples
Full size table

I define institutional precursors as elements of existing institutions, which play an important role for a new phenomenon to emerge. WISEs represent such an institutional precursor. They themselves represented a novelty in the work integration field at some point, because they stressed explicitly a dual social and economic mission, which distinguished them from purer nonprofit or public providers of work integration (Battilana et al., 2015). Later on, this made them natural allies for other actors in the process of forming MSPs.

I define direct or indirect influence of the crises, respectively, in terms of whether they objectively exacerbated the problem to be addressed (that is whether the crisis took on a material character that went beyond its moral character), or whether they created issues on another level that affected institutions indirectly, such as controversial public discourse.

Each country had distinct existing institutions, which were affected by the crises in different ways. France shared with Germany a pronounced role of the state in work integration and thus a high degree of state capacity in regulating the field, whereas state capacity was generally lower in Spain and work integration was promoted primarily by nonprofit and civic actors (1a, capacity of institutions). Also, France had a strong WISE tradition in common with Spain, whereas it was not present in Germany (2a, prominence of institutional precursor). In France, the adversity created by the crises was relatively low, if compared to Spain (substantial economic downturn) or Germany (politization) (see 1b). At the same time, the effect on the existing WISE or partnership activities was direct in France, just as was the case in Spain, whereas in Germany the primary adversity arose not because the moral crisis exacerbated the work integration challenges to a degree that the state could not handle them, but mainly because it strongly polarized political and public discourse (2b, indirect relation of crises to institutions).

The nested contingencies, in France, made existing institutions stay in the default position, because the system was resilient enough to also deal with the new challenges. I conceptualize this as continuity in Table 4, exemplified by state-catalyzed partnerships in Table 5 (blue three-branch scaffold and corresponding low shift). This observation and pathway is non-trivial since the few changes in France were not because there was no crisis, but because the equipment of the institutions to deal with adversity was anticipatory and predating the major crises. Actors, in particular the state, took responsible action in promoting multistakeholder partnerships as a more effective pathway to integration early on. This happened as a consequence to the more latent crisis of long-term unemployment of disadvantaged population groups and fortified institutional resilience ex ante. The institutions in Germany and Spain instead, for quite different reasons, departed from the status quo and embraced MSPs to increase institutional resilience as a consequence of the crises.

In Spain, the economic crisis did not only raise the general unemployment level (2b, direct influence of the crisis), it also increased the problems of particularly vulnerable groups (1b, severe crisis). The crisis lifted the precariousness among disadvantaged groups above a tipping point—so far that it was perceived as a threat to social cohesion at a very fundamental level. This circumstance did not change through government action primarily targeting the newly unemployed. Finding adequate responses was not only made difficult because of the comparatively low level of state involvement to begin with, but also because policies and legislation divided action by different forms of vulnerability which blocked holistic approaches (1a, low institutional capacity). While previous efforts provided a link to deal with the crises (2a, institutional precursor of WISEs), the inhibited capacity of the government to steer and motivate these activities led to fundamental moral considerations which redefined the role of business. This resulted in what I conceptualize as a new actor lead in Table 4, and specify as a business–civil society driven network in Table 5 (red three-branch scaffold and corresponding high shift). While the majority of organizations was civic, the governance of the MSP was more strongly in the hands of firms. This was supposedly not only because of a new sense of corporate responsibility, but also because of firms’ high endowments, not only financially but also in terms of reputation and management capacity.

In Germany, in particular the refugee crisis affected work integration efforts by creating a highly contested political climate (2b, indirect influence of the crisis). Although the number of people seeking asylum in Germany was very high, the strongly state-centered system (1a, high institutional capacity) would likely have sufficed to deal with the challenge (1b, medium severity relative to the existing institutions). Because of this, at the beginning firms demonstrated a relative reluctance to engage actively. Only when the discourse became massively politicized, did firms feel a moral urgency to take a stand against far-right political positions and xenophobia. Previous institutional practice was marked by a relatively pillarized system of state contracted nonprofit service provision, so that MSPs driven primarily by a civil society-state alliance represented a variation of the existing actor coalition (see Table 4). It was a mix of medium severity and indirect influence of the crisis which pushed Germany toward a shift instead of staying in the default like France—at the same time, the shift was less significant than in Spain (see Table 5; green three-branch scaffold and corresponding medium shift). The key nonprofit organizations in Germany were not WISEs in the conventional sense (2a, absence of institutional precursor), but came from the more general field of education and youth work.

When it comes to lasting influence by the MSPs on resilience, I would like to mention that as of 2022, all of the studied MSPs remain in operation and as far as could be verified through desktop research were thriving. For instance, the Spanish MSP is currently reporting current positive impact figures as regards increasing numbers of people reached. The German MSP from Berlin has grown in engaged partners and is now under the official patronage of the federal state’s minister of integration, labor, and social affairs.

Discussion

My conceptual link to institutional resilience opens new avenues for future studies of crises at the level of institutions, which is in contrast to the strong focus on individual organizations in the literature (Williams et al., 2017). More such institutional-level research will help us understand the dynamics which occur when actors that make an institution act collaboratively—on purpose or otherwise prompted or enforced—to address a crisis, especially when this crisis primarily has a moral instead of a material character. This analytic angle is different from looking at how organizations manage external stakeholders in crises (Bundy et al., 2017) where each stakeholder connection may be seen as a dyadic relationship that will exhibit a relatively low level of behavioral, inter-organizational interaction as compared to more wide-spanning institutional analyses. When moral crises are analyzed from an institutional perspective, some actors will be more or less affected. Or they will be affected by the adversity of crises at different points in time with different intensity so that behavioral and motivational diversity will be high. Adding a moral angle of analysis helps us understand actors’ motivations and interaction patterns, including their reasons for acting responsibly, or how actors ethically interpret the crisis as well as their own contributions to institutional resilience. Such dynamics remain hidden when we consider the effects of crises from an instrumental perspective. I specify my article’s contributions by formulating testable research propositions.

The Enabling Effects of Moral Crises on Institutions

This article sheds light on normative aspects as opposed to the instrumental approaches to crises in the literature (Bundy et al., 2017). Looking at both, the economic and refugee crises as well as actors’ responses from the normative perspective of moral crises has helped uncover new mechanisms that influence institutional responses. More specifically, it enabled me to uncover a new link between crises and institutions that makes us aware reactions will not only depend on the adversity created by crises (Hillmann & Guenther, 2020), but also on the directionality of the crises. For instance, firms in Germany were primarily prompted to act indirectly when the moral character of the refugee crisis intensified, whereas they were reluctant to act when the moral character was weaker and material adversity not considered severe enough to feel a responsibility to respond. By contrast in Spain, the moral character of the economic crisis in combination with the additional momentary inhibition of the government, and the concomitant material aggravation of the situation for vulnerable groups, spurred solidarity action at scale. Looking at the crises from an instrumental perspective and with a mere focus on the organizational instead of the institutional level, would have prevented us from understanding how and why actors changed their behavior and how MSPs emerged as an institutional-level response to the crises.

Following on from the above, my research revealed some unexpected enabling effects of moral crises. Crises and the adversity they produce, are typically seen as events that actors have difficulties coping with and are struggling to overcome (DesJardine et al., 2017). The widely held expectation is that the typical consequence of crises is organizational and institutional weakening or failure. Only few scholars have pointed out that we need more positive research on crises (Guo et al., 2020) as well as ways of understanding resilience to future adversity (Linnenluecke, 2017) as opposed to research that focuses on how to cope with momentary turmoil (Hillmann & Guenther, 2020). I suggest, moral crises can have enabling as opposed to destructive effects through promoting push and pull factors. The processes in Germany and Spain are examples of push factors that galvanized dispersed into collective action. This happens either when very high moral pressures force unengaged actors to become active, or when moral and material pressures in combination spur collective action. This link between pressures and enabling effects is remarkable, since pressures from outside and high expectations from stakeholders often paralyze actors and render them incapable of strategizing and finding effective responses (Putnam et al., 2016). I propose:

Proposition 1a

Moral crises produce indirect pressures (politically polarizing the public discourse on the crisis) that may force previously disengaged actors to take on moral responsibility and address the crises at the institutional level.

Proposition 1b

Moral crises produce direct pressures (exacerbating the problem to be addressed) that may prompt loose and pre-existing actor constellations to transform into collective responsible action to address the crises at the institutional level.

Crises may also act as pull factors. In Germany, some employers conceptualized the high number of refugees as a potential means for addressing skills shortage, a supposedly bigger, long-term challenge. They interpreted the crisis as an opportunity rather than as a problem. So, crises opened some business opportunities for pro-active organizations, who took on a positive interpretation. In such cases it is hard to judge whether the motivation for action is ethical or instrumental. Yet, such types of action may also be encouraged by a sense of long-term responsibility, especially when an actor’s engagement predates a major crisis. The French state, for instance, acted on the latent crisis of long-term unemployment by brokering and crafting new partnerships early on, which mitigated crises later on.

Proposition 1c

Moral crises open opportunities for positive interpretations that lead to pro-active engagement by actors (potentially out of instrumental motives when the pro-activity coincides with major crises, or out or moral responsibility when pro-activity predates crises), which may fortify institutional resilience for future crises.

Very seldom are crises seen as an opportunity for new approaches to emerge and if so research often focuses on how entrepreneurs do so (e.g., Hjorth, 2013; Martí & Fernández, 2015). I suggest more attention should be given to the opportunities that in particular moral crises may present for institutions to increase their resilience in view of wider and long-term crises and other challenges.

The Nested Contingencies of Institutional Resilience

Some scholars have referred to resilience as a process rather than a property (Williams et al., 2017). The focus on resilience at the institutional level instead of the organizational level opens up conceptual and empirical space to advance this perspective, because it allows us to observe the interplay of a variety of factors (Barin Cruz et al., 2016). From previous research we know that context factors matter and affect the scope and speed of institutional change processes, as I have also stressed in relation to the evolution of MSPs (Micelotta et al., 2017). However, it is often unclear what the constraining or enabling factors of such processes are, and what determines which paths will evolve (Bothello & Salles-Djelic, 2017). I extend previous research in the area by suggesting that path dependence is a real issue for increasing institutional resilience. Thereby I move beyond the weaker argument that history matters, but also defy deterministic arguments of lock-in which suggests once you’ve taken a path, there is no way back (Vergne & Durand, 2010). The national MSP models developed independently of each other, and followed different trajectories, but came to operate on shared principles across countries. So, different situations at baseline, can lead to similar outfits in future. I argue nested contingencies help us better understand the interaction of contextual factors that lead to institutional change (Cobb et al., 2016).

The less surprising contingency related to the capacity of the institutions and the level of adversity created through the crises, which represents a dominant focus of research on resilience (Hillmann & Guenther, 2020). The more surprising contingency related to institutional precursors and the immediacy of the link between crises and institutions, which contributed to the emergence of MSPs. For instance, indirect influence as in Germany—when getting big enough—may still provoke a substantial institutional shift in a field with high institutional resilience in the status quo. However, due to the lack of institutional precursors (here: WISEs) and the indirect influence, the institutional shift was neither as profound nor did it have the same outfit as in Spain, where the precursor was present and the link direct as well as the crises more severe. While Germany saw a variation of existing institutions, Spain saw a new institutional configuration emerge. Institutional precursors are different from so-called proto-institutions that designate new practices and rules that go beyond a collaborative relationship but are not yet fully institutionalized (Lawrence et al., 2002, p. 281). Proto-institutions refer to a beta version of an institution in becoming, whereas institutional precursors are elements of existing institutions and build a foundation for entirely new institutional configurations to emerge, e.g., in response to crises. In conclusion, while a research focus on institutional capacity and adversity may be sufficient to understand the magnitude of shifts needed to increase institutional resilience, institutional precursors and the relation of crises to institutions help us understand the outfit, pace, and variations of such institutional shifts. I propose:

Proposition 2a

Indirect influence of moral crises and weak institutional precursors will lead to a variation of existing actor coalitions to increase institutional resilience.

Proposition 2b

Direct influence of moral crises and strong institutional precursors will lead to new actor leads to increase institutional resilience.

Morality and Lead Agency in Collaborations

This article foregrounds mechanisms that spur responsible action in organizations as a reaction to moral crises. This focus is distinct from the predominance of instrumental considerations in much of the collaboration literature (Bakker et al., 2019). My article also goes beyond recent, morally grounded research on multistakeholder collaborations. Such research looks at, e.g., organizations that seek to ensure values alignment between multiple stakeholders (Schormair & Gilbert, 2021) or how market actors try to attain societal legitimacy when engaging with nonmarket actors (Girschik, 2020). I instead highlight the interplay between organizations and institutions in creating moral impetus for collaborative responsible action. In contrast to recent studies on how actors joined forces to address the economic and refugee crises (Daskalaki et al., 2019; Fehsenfeld & Levinsen, 2019; Hesse et al., 2019), I zoom out of specific types of collaborations and abstract to the collaborative arrangement as such (MSPs). This does not only allow me to isolate the factors that may cause similarities and differences in collaborations across contexts, but also helps me contribute to a new strand of research that seeks to specify what types of actors are more likely to lead in collaborations and under which conditions (Wang et al., 2022).

I suggest lasting collaborations between diverse actors may be facilitated by crises. However, this does not primarily happen because engaged actors realize that harnessing synergies may be beneficial for addressing a crisis (Hahn & Pinkse, 2014). It happens when the moral and/or material intensity of crises prompt actors to redefine their role relative to the crisis and to act on rationales of responsibility instead of on instrumental motives. Business actors in Germany and in Spain moved from a passive to an active role in addressing the crises as they unfolded. The Spanish case is in line with suppositions that private actor leadership in crisis will occur because of private actors’ high management capacity (Wang et al., 2022). Such management capacity became particularly important because the dominant field actors (here: nonprofit associations) reached their limits and the Spanish state could not offer adequate support. The German case instead shows that firms are likely to act as followers of pro-active nonprofits, when the crisis is less severe or its influence on institutions indirect. Nonprofits’ agency is not only grounded in deeply engrained moral motives, but motivated by their proximity to target groups that are affected by the crises. My propositions are:

Proposition 3a

Under the condition that dominant actors in the field are too strained to act, and when the adversity produced by moral crises is high, private actors will lead in addressing the crisis because of their high management capacity. Urgency to act may serve to accelerate the redefinition of private actors’ roles and create a new sense of moral responsibility.

Proposition 3b

When the adversity produced by moral crises is rather low, civil society actors will lead in addressing the crisis because of their moral motives and access to target groups.

It is worth noting that there are procedural dynamics involved in actor leads. In the German MSPs, nonprofits played a particular role in initiating the process. However, as partnerships progressed they depended more on the provision of additional resources, reputational but also in terms of management capacity so that state and firm contributions became more important. The French case in contrast to the other two countries demonstrates the role of embedded agency of the dominant institutional actor (Garud et al., 2007). The French state engaged in anticipatory action and fortified future institutional resilience through promoting multistakeholder collaboration early on. It did so through acts of coercion (social clauses) as well as acts of brokering (two high-level networking events). Thus:

Proposition 3c

Future institutional resilience to crises can be increased when the dominant institutional actor engages in embedded agency that promotes multistakeholder collaboration before major crises.

Conclusion

Deepening our understanding of the connections between crises, institutional resilience and multistakeholder collaboration is not only needed due to the current COVID-19 crisis, but warranted in view of today’s grand challenges, which should prove to affect institutions substantially and demand knowledge about the process of increasing institutional resilience. There are lessons to be learnt for future policy and practice from my article. First, in view of multiple impending societal crises, policy makers would do well to continuously encourage multistakeholder collaboration, which they spurred through large-scale events in the context of COVID-19 for instance, instead of committing the mistake of seeing such acts of pro-active brokerage as one-off events within a specific crisis context (Bertello et al., 2021). A steady approach is likely to be benefit institutional resilience. Second, stakeholders from all sectors should more actively explore options for how a pro-active approach to collaboration can spur institutional innovation such as that of the MSPs instead of waiting for these to be coerced as crises occur (Hargrave & van de Ven, 2006). Third, organizations should more consciously acknowledge the benefits of taking on a normative and moral rather than an instrumental view on crises, which can open up entirely new opportunity spaces for how organizations deal with purpose, promote problem solving, and think about partnerships (e.g., Scheidgen et al., 2021). Future research would need to test these implications and my propositions in different fields and different geographic contexts. However, given the comparative, process-oriented and long-term character of the research and a joint methodological repertoire, I have confidence that my conceptualizations will prove of relevance beyond the present application.