Creativity caged in translation : a neo-institutional perspective on crisis communication

Crisis communication research has primarily focused on universal models guiding managers of various organisations in times of crisis. Even though this is about to change, a tendency remains for research in the field to overlook the impact of structural conditions on organisation’s crisis communication. In order to add to the emergent discussion on new theoretical and empirical venues within the field of crisis communication, this paper proposes a framework based on new institutional theory for analysing crisis communication practices as a societal phenomenon. New institutionalism is advocated due to its ability to shift the focus from agency to structure and in doing so emphasise the social preconditions for organisational activities. In line with this, this conceptual paper discusses crisis communication as an institution, i.e., as a set of more or less conscious ideas about formats (the organisational structures developed for crisis communication work), contents (the content of organisations’ communication in times of crisis) and contexts (the situations during which organisations are expected to perform crisis communication). Moreover, we discuss how these ideas become translated (i.e., modified) as they travel (i.e., become legitimate, popular and get widely spread) across organisational and institutional contexts. In order to illustrate the framework described above, the Swedish authorities’ communication in connection to the A/H1N1 outbreak is used as a case study. 1 Senior lecturer at Department of Journalism, Media and Communication (JMG), Faculty of Social Sciences, University of Gothenburg, Sweden. 2 Assistant professor at Crismart (Crisis Management Research and Training) at the Swedish National Defence College. She holds a PhD from the Department of Journalism, Media and Communication at Stockholm University. 3 Senior lecturer at Department of Business Studies, Uppsala University, Sweden. REVISTA INTERNACIONAL DE RELACIONES PÚBLICAS, No 8, VOL. IV [Páginas 65-84] 2014 66 ISSN: 2174-3681

Keywords: crisis communication, A/H1N1, neo-institutional theory, Sweden, translation Resumen La investigación de la comunicación de crisis se ha centrado principalmente en modelos universales en los que se basan los gerentes de las organizaciones en tiempos de crisis. A pesar de que esto vaya a cambiar, la tendencia sigue siendo en la investigación en este campo pasar por alto el impacto que las condiciones estructurales de la organización tienen en la comunicación de crisis. A fin de contribuir en la discusión emergente sobre los nuevos escenarios teóricos y empíricos en el campo de la comunicación de crisis, este documento propone un marco basado en la nueva teoría institucional para el análisis de las prácticas de la comunicación de crisis como un fenómeno social. La nueva institucionalidad aboga por su capacidad de cambiar el enfoque de la agencia a la estructura y al hacerlo destacar las condiciones sociales de las actividades de la organización. En línea con esto, este trabajo conceptual analiza la comunicación de crisis como institución, es decir, como un conjunto de ideas más o menos conscientes sobre los formatos (las estructuras organizativas desarrolladas para el trabajo de comunicación de crisis), el contenido (el contenido de la comunicación de las organizaciones en tiempos de crisis) y contextos (las situaciones en que se espera que las organizaciones llevan a cabo la comunicación de crisis). Por otra parte, se discute cómo estas ideas se convierten (es decir, se modifican) a medida que viajan (es decir, se convierten en legítimas, populares y son ampliamente extendidas) a través de contextos organizacionales e institucionales. Con el fin de ilustrar el marco descrito anteriormente, se analiza el caso de la comunicación de las autoridades suecas en relación con el brote de A / H1N1.

INTRODUCTION
Flexibility, creativity and improvisation have become central concepts in research on crisis communication. After early attempts to create universal models guiding managers in times of crisis -often with focus on tactics and re-action(s) -more recent attempts have focused on the abilities of managers to adapt to situational factors (cf. Eriksson, 2014;Falkheimer & Heide, 2010;Gilpin & Murphy, 2008). Albeit this shift, research in the field most often understands social actors as more or less isolated entities with extensive liberty of action.
Structural conditions tend to be overlooked with legal aspects as the only exception.
In order to address these shortcomings, we suggest a neo-institutional framework to the study of crisis communication as it offers a focus on social and institutional aspects for explaining organisations' activities. In this stream of research, organisations are regarded as embedded in institutional environments, imposed on them by rules, norms and ideas. As a result, practices are seen as a result of institutional expectations and taken-for-grantedness rather than their expected effects. From this we have learned that social activities tend to be performed in conventional ways that are socially accepted, a tendency which becomes even stronger in situations with high levels of uncertainties and extensive unpredictability i.e., crisis. However, this is not to say that institutions are deterministic in the way they influence organisations and their behaviour. On the contrary, institutions are open for interpretation, negotiations and innovations (Czarniawska & Joerges, 1996;Oliver, 1991). In analysing the practice of crisis communication from a neo-institutional perspective, we can show how organisations take on an active role in the way they relate to and handle institutional pressure without losing track of the necessary constraints institutions have on how these activities are and can be performed.
In this paper we will illustrate how crisis communication can be understood as an institution (Fredriksson, Pallas & Wehmeier, 2013). That is to say, the way organisations communicate is based on a set of more or less conscious ideas about the formats (the organisational structures developed for crisis communication work), contents (the content of organisations' communication in times of crisis) and contexts (the situations during which organisations are expected to perform crisis communication) and how these ideas become translated (i.e., modified) as they travel (i.e., become legitimate, popular and get widely spread) across different organisational and institutional contexts (Czarniawska & Joerges, 1996;Sahlin-Andersson, 1996;Sahlin-Andersson & Engwall, 2002;Sevón & Czarniawska, 1996). Hereby, we want to enrich the field by providing a framework that opens a more advanced analyses of crisis communication practices and its conditions and consequences; an ambition we share with a growing numbers of scholars pointing towards the need for more analytical perspectives on public relations in general (cf. Edwards, 2012;Fredriksson, Pallas, & Wehmeier, 2013;Ihlen, Ruler, & Fredriksson, 2009;Ihlen & Verhoeven, 2012) and crisis communication more specifically (cf. Frandsen & Johansen, 2009;Fredriksson, 2014).
The rest of the paper is organised as follows: first we present what others have described as the "old" and the "new" perspectives on crisis communication, and what we see as limitations in these perspectives when it comes to their abilities to explain why crisis communication is practiced as it is. Our main argument is that both perspectives disregard the social conditions under which organisations functions. Thereafter, we describe how institutional conditions are expressed in rules, norms and beliefs and how they govern organisational activities. In the next section, we present translation as a way of explaining how organisations actively relate to and handle these conditions by transforming various forms of institutional ideas into their activities in general, and more specifically when it comes to crisis communication. The neo-institutional approach and its consequences for research on crisis communication are then discussed in the concluding part of the paper.
Throughout the paper we illustrate our theoretical standpoints by using the A/H1N1 pandemic case and how it was communicated in a Swedish context. Here we rely on three previous reports analysing how Swedish governmental agencies and County Councils carried out their communication activities during the pandemic (Fredriksson, 2011;Ghersetti & Odén, 2010;Olsson, 2011).

APPROACHES
Crises communication is thought of as being required in situations characterised by unexpected events that cause high levels of uncertainty among organisations as well as stakeholders (Ulmer, Seeger, & Sellnow, 2007). Research in the field has traditionally been devoted to examining and offering general models for the planning and performance of crisis communication aimed at supporting managers in their work. It has been argued that crisis communication -as with all other forms of public relations -can be prepared and anticipated by the use of standardised measures for issues management, planning, decision making and performance leading to predicable effects (Gilpin & Murphy, 2008).
More recently, this approach has been criticised as being unrealistic. Based on theoretical support from chaos and sense-making theory, scholars point out that contemporary organisations should instead develop their unique capacity to function in highly complex environments based on the principles of adaptability, improvisation and flexibility (cf. Deverell, 2010;Eriksson, 2012Eriksson, , 2014Falkheimer & Heide, 2010;Finch & Welker, 2004;Gilpin & Murphy, 2008Lagadec, 2009;Pang, Jin, & Cameron, 2010;Roux-Dufort & Vidaillet, 2003). A returning argument to this approach is that communicators should perceive themselves as jazz musicians improvising in acute situations based on creativity and earlier experiences (cf. Eriksson, 2014;Falkheimer & Heide, 2010). Despite its different axioms compared to earlier research, in terms of how to understand organisations, individuals and communication, this "new" perspective does not offer any reliable explanation for why crisis communication is practiced as it is. The number of empirical analyses are limited but the ones offered confirm rather than falsify what we already know; crisis communication is to a large extent practiced according to conventional ideas where organisations try to gain control of the situation by the use of centralised communication.
When scholars find examples of creativity or improvisation it is often limited to tactical matters, as when an organisation starts to use social media instead of conventional media (cf. Eriksson, 2014).

What both the traditional and new turn in crisis communication research have in common is
the disregard of the social conditions in which organisations are embedded. As many other social activities, crisis communication has become institutionalised and as such, it is underpinned by normative systems and cognitive understandings that prescribe how it ought to be practiced, when it should be practiced and why it should be practiced. Thereby organisations' structures, processes, as well as performances in crisis situations are to be understood as responses to social requirements and expectations rather than isolated improvised actions made by organisational members (DiMaggio & Powell, 1983;Fredriksson, 2014;Meyer & Rowan, 1977).
In general, organisations tend to act in similar ways as other organisations and their actions are therefore to be understood as a search for acceptance and a way to minimise risks for disturbances (Meyer & Rowan, 1977;Scott, 2008). These types of alignments increase in times of crises as they amplify uncertainties among organisations as well as their constituencies. When under pressure, organisations act to reduce these uncertainties by the incorporation of solutions, units, models and patterns of behaviour that are perceived as rational, efficient, necessary, and morally correct. The goal is very much to confirm established stakeholder expectations that are intensified in times of crisis and thereby achieve legitimacy, credibility and authority (Power, 1997(Power, , 2007. As a consequence, already established strategies, models and practices tend to be confirmed and reproduced in times of crisis whereas the ability to improvise is a precondition inherent, or not, in actors' institutional context (Boltanski & Thévenot, 2006).
The external demands on organisations are expressed in three different types of institutional conditions -rules, norms and beliefs (Scott, 2008). In the following we address these beliefs and how they come to expression if we understand crisis communication as an institution.

Regulated crisis communication
Laws, regulations and codes constitute a major part of social conditions, which influence organisational activities. As such, they are mandatory and prescribe specific activities, systems, chains of command, etc. (Djelic & Sahlin-Andersson, 2006;Levi-Faur, 2005). To ensure compliance, there are various types of control and auditing mechanisms -such as fines and license issuance as well as sanctioning behaviour that deviate or violate the regulatory framework (Jonsson, Greve, & Fujiwara-Greve, 2009;Power, 1997).
Governmental agencies and authorities usually ascribe the responsibility for designing and administrating different control mechanisms. The existence of regulatory processes is based on the intention of the rule-makers to create rules that ensure predictability and stability in organisational fields, networks or markets (Hoffman, 1999). Or differently, the more the uncertainty increases the more predictable organisational behaviour becomes.
Within the crisis communication field, regulations are well documented and often described as constraining organisations' communication since legal aspects are seen as amplifying the tendency of organisations to give prominence to their own concerns rather than to stakeholder's by denying responsibility and transparency (Brinson & Benoit, 1996;Fink, 1986;Marcus & Goodman, 1991;Patel & Reinsch, 2003;Tyler, 1997;Ulmer, 2001).
In the case of the A/H1N1 pandemic, the Infectious Diseases Act (2004:168) had a strong impact on how communication was organised, managed and performed in Sweden.
According to the act, the National Board of Health and Welfare is responsible for the coordination of Communicable Disease Control at the national level. It is also stated that the county council and the County Medical Officers of Communicable Disease Control are responsible for the execution of all efforts aimed at preventing distribution of viruses before and during a pandemic. This includes, communication and the public's abilities to inform themselves about issues affecting their health and lives. Moreover, it is clearly stated that the medical officer is responsible for this work. This means that the chain of command puts the communicator in a subordinate position, that of a "performer" rather than a "director."

Professional crisis communication
Apart from rules, practices in organisations are to a large extent governed by norms expressed in certain ideals, models, methods and strategies, guiding the way the work of practitioners are organised and performed (Scott, 2008). These normative elements stipulate a desirable, ideal or morally correct organisational behaviour, which often resonate with ambitions of professionalization. They also give rise to routines, conventions, Another difference is that norms are created, proliferated and disseminated by the work of professional carriers such as management consultants, universities and media, rather than ISSN: 2174-3681 legal bodies. However, even if there are no formal sanctions attached to norms, they are still very difficult to disregard (Jonsson, 2002;Jonsson et al., 2009). Declining or violating the prescriptive requirements may, for example, lead organisations to social isolation or punishment (Suchman, 1995). Norms and standards define -but are also defined byrelations and interactions between actors in terms of their responsibilities towards each other and the surrounding community. As such, they constitute a more or less explicit frame of reference for achieving a desirable position in a given social context (Boström, Forssell, Jacobsson, & Tamm Hallström, 2004). In other words -standards and norms not only describe social obligations, they also create opportunities and privileges (Brunsson & Jacobsson, 2000).
As in other fields of practice there are a number of norms surrounding crisis communication.
One the strongest is the notion of crisis communication as foremost a planned and controlled activity. The idea of control and planning can be seen both in the academic literature (Coombs, 2011) as well as in advice provided by professional bodies as well as consultants. For example, the construction of crisis communication plans is a returning moment in the preparation for crisis communication even though the plans tend to be generic regardless of the type of organisation (Nikolaev, 2010). Taken together it provides an example of the strong normative power in the field in terms of what is considered 'good crisis communication.' If we take a closer look on how the communication work was organised during the A/H1N1 crisis it becomes obvious that "The National Plan for Communication during Pandemics," was widely recognised and adopted amongst the governmental agencies responsible for national efforts in times of pandemics. The plan guides the distribution of work between national governmental agencies and county councils and it distributes responsibilities in times of crisis and suggests certain goals, strategies and efforts. As such the plan is a rather conventional product reducing the preconditions for crisis communication to something strategic and manageable. Hence, one can see the plan in itself to be a manifestation of the norms enclosing crisis communication; this also applies to ideas represented in the plan.
Many of the ideas are mobilised by strong normative incentives, one example is the idea of "one organisation -one voice;" an aspect that is seen as essential in most strategic planning.
Much attention is given to the ideas in the plan and it is also clear that the Swedish authorities made major efforts to uphold the plan when the A/H1N1 campaign was carried out. Much attention is given to the coordination of actions and messages to avoid inconsistencies or contradictions even in situations when the circumstances urged other approaches. In some cases, this was due to conflicting interests in others due to extensive insecurity about facts. The suppression of eventual side effects of the vaccine is one example of the latter.

Cognitive structures -making sense of crisis communication
Rules and norms are deliberate and -in many cases -explicit. Practitioners are expected to know laws and regulations governing their activities and they are expected to use the knowledge and skills they have gained via education and experience. However, their activities are also governed by implicit assumptions about the work, its preconditions and accomplishments. These assumptions gain a status of taken-for-grantedness and function as a frame of reference when practitioners set goals, choose aspirations and evaluate results.
The assumptions dictate what is defined as important, how meaning is created and what action is expected as appropriate in different contexts (Scott, 2008).
How communication is conceptualised has a profound impact on the practice of crisis communication. For example, communication as a strategic means tends to be conceptualised as a pipeline which connects a sender with a receiver via a medium transporting information and knowledge (Reddy, 1979). Communication is thus described as "The National Plan for Communication during Pandemics" is an example of where a number of implicit ideas and models about how to communicate become articulated. In the plan, the governmental agency is described as the governing part in the communication process. The process is seen as initiated by the governmental agency and until the crisis ends the agency is expected to be in full control of the communication process. Communication is defined in ISSN: 2174-3681 terms of the construction of the right messages, transmitting them via the right medium, to the right receiver to reach the desired effect(s). Thus, communication is understood as a process, including the distribution of responsibilities, situation analysis, goals, target groups, strategies, and messages. Consequently it is the abilities and competencies of the agencies that are seen as the determinants of the results.

METHODOLOGY AND FINDINGS: TRANSLATION OF CRISES COMMUNICATION
The discussion of the three institutional pillars above provides an understanding of crisis communication as an institution -i.e., an activity that is embedded in the structural preconditions and contexts in which it is organised and performed. However, that does not mean that institutions -albeit they constitute a strong behavioural reference frame -are determinant or totalitarian. That is, institutions do not prescribe certain behaviours per se, but rather limit the alternatives to a very small number of conscious options and in doing so they create frames, internalise structures and limit the space of action for organisations. This makes organisations more or less conscious to do as other (successful) organisations do, especially in times of crisis. By adopting -what is seen as -functional solutions, creating particular structures or recruiting professionals with certain competences, organisations search for stability, control and legitimacy. By implementing the concept of crisis communication, creating a crisis communication plan and appointing communicators with competences in crisis handling, organisations strive for security -intentionally as well as unintentionally (Engwall, 2009). This internalisation is more or less conscious, deliberate or done after an evaluation of different alternatives. It is, to a large extent, a form of social learning and imitation of successful concepts where organisations adapt, reshape, and add whole or parts of the concepts they find attractive (Czarniawska, 2005;Sevón & Czarniawska, 1996).
When this is done actively it could be described as translation -a process whereby organisations participate in transforming the various forms of institutional ideas into their activities (cf. Djelic, 1998;Sevón & Czarniawska, 1996;Westney, 1987). This means that organisations and their members are involved in the creation and formation of institutional orders, and not only passive adopters of these orders (Campbell, 2004). Thus, crisis communication is not performed through passive and mechanical adaptation to prevailing institutions but comprises components of meaningful, active construction and interpretations of these institutions (Fredriksson, 2014;Fredriksson et al., 2013;Scott, 2003).
The selection of ideas and models adapted by organisations are often conditioned by perceived resemblance with, or a desire to resemble, those that legitimate and create conditions for successful existence or favourable outcomes (cf. Hedmo, Sahlin-Andersson, & Wedlin, 2005;Sahlin-Andersson, 1996). However, there are other sources or motives. Lost legitimacy for certain organisational forms (cf. Forssell & Jansson, 1996), identity crises (cf. Brunsson & Sahlin-Andersson, 2000) or professionalization of labour forces (cf. Powell, White, Koput, & Owen-Smith, 2005) might also awaken efforts to imitate and transform prevailing institutional models. In line with this absence or inadequate perception of own identity, identification of internal problem, or weak external support also motivates organisation to screen their environment for appropriate and successful models that can be used.
On the other hand, translation processes are not entirely endogenous and motivated by internal identification of the problem at hand; they also arise in relation to external actors like business consultancies, management schools, media, etc. (Hedmo et al., 2005;Sahlin-Andersson & Engwall, 2002). Translation is thereby a concept reflecting the search for solutions to organisational problems that are defined and solved in on-going interactions between different actors sharing common or overlapping institutional environments (Boxenbaum, 2006).

Translation at Work
Translation emphasises processes where behaviours of organisations are exposed to and influenced by ideas and models that permeate the context in which organisations conduct their activities. In this process these ideas and models become modified and equipped with new attributes fitting the specific (local) context (Czarniawska & Joerges, 1996;Sahlin-Andersson, 1998) closely we see that when implementing the Act, the counties also choose to contextualiseand thereby also modify -the legal and professional requirements so they would fit with the local contexts. That is to say, the counties adjusted the Act to their specific needs, expectations, previous experiences and relationships with the media. Facing challenges, for example in regard to different target groups, the journalists' pre-understanding of the gravity of the pandemic, or use and access to different communication channels, the counties designed their own version of the models and strategies describe by the Act.
In sum, we see evidence of both general constrains and localised translation. On a general level the counties followed the legal and normative expectations and requirements that led to striking similarities when it came to concept and overall strategies used. At the same time, the implementation of the regulative and normative models was not static or passive. Rather the counties' crisis communication included redefinition of the prevailing prerequisites that enabled local variations without up-front challenges to the overall norms and rules.
However, translation does not happen autonomously, it follows certain rules limiting and specify what, when, where and how institutional models can be translated (Sahlin-Andersson, 1996). These 'editing rules' allow for an adaptation of ideas so they fit into the local context at the same time as they remain institutionally defined -i.e., they remain recognisable and/or legitimate in the eyes of relevant actors upon which the counties are dependable for moral, legal and financial resources.
One example of this is the translation of the "The National Plan for Communication during Pandemics" in which it is stated that "transparency" is superior to other modes of communication as it is seen as essential to gain and sustain credibility. In line with the plan, the authorities and councils recommended the provision of as much information as possible to help citizens make decisions about issues affecting their health and lives. However, the communicative norm "transparency" was negotiated in relationship to other ideas, models and regulations, many of them advocating "efficiency;" as when The Medical Products Agency distributed information about the potential side effects of the vaccine but strongly recommend people get vaccinated anyway. Thereby, they met the expectations on transparency at the same time as they uphold the idea of efficiency.
A similar case is the decision, at an early stage, not to recommend that pregnant women get the vaccine. According to the authorities, the recommendation was not based on facts but on fear that any miscarriages would be perceived as being connected to the vaccine and regarded as a side effect. According to the responsible authorities, the message was somewhat difficult to communicate since "it can easily be understood as if we recommend not to because we think it is dangerous to get vaccinated during the first phase of the pregnancy" (Ghersetti & Odén, 2010, p. 106). Here, the notion of correct facts is clashing with the idea of mass vaccination in order not to jeopardise any backlash to the original message, that the vaccination is free of risk. In sum, the general norms provided by the plan, in conjunction with the strong message to vaccinate, opened the way for clashes to be handled by the professional having to put strategy into action.

CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH
As argued initially, this paper aims to contribute to the growing debate within the field concerning new theoretical venues in understanding crisis communication both as a practice and as a societal phenomenon. In line with this we have proposed a neo-institutional framework and argued for translation as a concept, thus making it possible to understand the way organisations perform crisis communication in constrained environments. By doing this we assert that the practice of crisis communication is not so much a reflection of organisation or individual abilities to improvise or to be creative, rather it is a witness of institutional embeddedness where general expectations, regulations, ideas and models are actively translated into specific organisational contexts. The ability to improvise in times of crises must then be understood as something that is caged in inherent institutional contexts rather than an activity that is detached from prevailing circumstances and developed by singular and autonomous actors. That said, it is obvious that organisations and their communication activities are not determined by institutional conditions -or restricted to a fixed set of performances -in their responses to crises. The examples presented above shows several instances of situations were actors produce, diffuse and adapt their own versions of ideas and models on behalf of their organisations. Thereby, they contribute to creation and proliferation of new combinations, revisions, accretion and hybridisation of established principles, models and ideas.
In line with this, we argue that research gains from further examinations of how organisations relate and take active part in the production, reproduction and alliteration of the institutional conditions governing crisis communications.
First, as other forms of institutionalised communication practices, crisis communication is expressed in regulations, norms and ideas, models and scripts that are ready to be used in specific contexts or situations. Some of these ideas are well established and have reached extensive distribution and are to be regarded as coercive and unquestionable. Others have not yet reached a position where their abilities to manage the problems they were constructed to solve are fully established and they might never be ready. With this as a starting point, one could ask to what extent crisis communications -as an institutionally defined (and bounded) practice -differ from other communication practices? That is, which aspects are to be considered if we want to understand the rationalities of crisis communication, its ontological and epistemological roots and its motivating factors?
Secondly, when institutional ideas, practices or models are translated into specific organisations they are subjected to a variety of activities and symbolical representations that transform these practices and models into concrete objects (i.e., policies, strategies and