SOCIAL MEDIA IN MANAGEMENT OF THE TERROR CRISIS IN NORWAY: EXPERIENCES AND LESSONS LEARNED

The chapter addresses the question of how crisis and emergency communicators in the justice (police) and health sector in Norway re ﬂ ect on their use (cid:1) or lack of use (cid:1) of social media during the terror crisis on 22 July 2011. We examine how these communicators in the years following the crisis have developed their use of social media to opti-mise their and the public ’ s awareness of similar crises. Our semi-structured interviews with key emergency managers and responders display how the terrorist-induced crisis in 2011 was a wake-up call for communicators in the police and the health sector. They re ﬂ ect on the signi ﬁ cance, strengths and weaknesses of social media in the management of crises such as this one.

This chapter engages with the question of how crisis communicators in two sectors with key roles in managing crises such as 22 July, namely the justice (police) and the health sector, reflect on their use À or lack of use À of social media during the crisis in Norway on 22 July 2011. Especially, we ask communicators about what lessons they have drawn from their experiences in their crisis management and communication. We pose the following research questions: (1) How do crisis managers and responders in the police and health sector in Norway understand and evaluate their use of social media during the terror crisis on 22 July 2011?
(2) How have they developed their use and what do they see as current challenges and possibilities for using social media to optimize their own and the public's awareness of and response to such acute crises?
The questions reflect social media use in the acute phase of a severe crisis (in this study defined as the first hours and days), and the lessons that subsequently can be drawn from possibly inadequate use in order to improve application of social media in emergency preparedness. An objective is, consequently, to illuminate how social media usage possibly may strengthen crisis communicators' role in supporting crisis management of acute crises À in other words, how communicators can apply social media to management tasks such as understanding a crisis as it unfolds, or making and implementing decisions.
Before we present perspectives from research on social media use in crises À and our methodological approach, analysis and discussion À we find it appropriate to include a few words on crisis communication in Norway, its purpose and responsibilities. According to governmental communication policy, the purpose of crisis communication is to provide citizens with rapid and factual information to enable them 'to cope with an unwanted incident in the best possible way' (The Norwegian Ministry of Government Administration and Reform, 2009, p. 16). Crisis communication should convey a division of responsibilities, further coordination between authorities and make clear how crisis-affected persons may obtain additional information or help and support.
The responsibility for crisis management and communication in a particular ministerial domain rests with the agency or service that is responsible for that domain under normal circumstances. For example, in Norway the justice sector has the responsibility for information concerning accidents and disasters, including terror attacks. On a national level the Ministry of Justice, which includes the police, has the overall responsibility both for the handling of the crisis and communication with the general population. The health service is consequently required to coordinate information with the police and be aware of what information should be issued solely by the police.

SOCIAL MEDIA IN CRISIS COMMUNICATION: PERSPECTIVES FROM RESEARCH
Although the effect of social media use on crisis communication processes has not been extensively researched, some significant studies draw attention to how social media change conditions for crisis communication (Coombs, 2012;Hagen, Keller, Neely, DePaula, & Robert-Cooperman, 2017;Simon, Goldberg, Aharonson-Daniel, Leykin, & Adini, 2014;Utz, Schultz, & Glocka, 2013;Veil, Buehner, & Palenchar, 2011). This research points to the increasing spread and use of social media in the population (as noted earlier concerning Norway), and social media's accessibility and immediacy.
Social media can speed up the dissemination of information during crises by linking end users to critical information sources in real-time. Hughes and Palen (2012) and Latonero and Shklovski (2011) are among several studies that show how social media allow public health officials and emergency responders to communicate directly with the public, thus eliminating the time it takes for emergency information and advice to flow through traditional communication channels. Utz et al. (2013) note that whereas classical crisis communication theories such as the Situational Crisis Communication Theory build on a traditional asymmetrical one-to-many model, the arrival of social media has increased the adoption of a many-to-many model. An organisation may now talk directly to its stakeholders, but the stakeholders can also 'respond, forward and discuss the messages with each other' (Utz et al., 2013, p. 41). Simon et al. (2014) emphasise that social media offer a widely available means for a bidirectional flow of information between the public and the authorities. Simon, Goldberg, and Adini (2015) see social media as providing opportunities for engaging citizens in emergency management by both disseminating information to the public and accessing information from them. Coombs (2012) suggests that crisis communicators who use social media need to adjust their traditional communication practices to communicate effectively: they need to listen to what social media users are saying and provide them with access to information, rather than one-sidedly controlling and feeding them with information. Veil et al. (2011) in a similar manner point to how social media allow for interpersonal communication, and they advise crisis communicators to engage in discussions and conversations with publics in daily communication. Strander (2011) calls attention to how web 2.0 technology and social media have led to public expectations that crisis communicators engage in a dialogue with them. According to a survey by the American Red Cross (2013), 70 per cent of Americans believe that emergency responders should regularly monitor the emergency response organisations' websites and social media sites so that they can respond promptly to any requests for help posted there.
Key sub-questions to our overarching research questions arise from the reviewed research: how do the crisis communicators currently reflect on the possibilities of engaging in social media dialogues when an acute terrorist-induced crisis such as 22 July evolves? Is communication with traditional news media worthwhile when social media provide opportunities for distributing information directly to the public during crises?

METHODOLOGICAL APPROACH
We have conducted and analysed semi-structured interviews with professionals who worked within crisis communication in their agencies at the time of the Norway attacks, and who have continued to work with such tasks. Although not all of them used social media in their professional communication during the crisis, all of them have later acquired significant experience with professional use of social media. To secure diverse perspectives on our research questions, we have interviewed employees who work on different levels of the health and police sectors, from the national strategic level to the regional and local level, with different tasks and roles assigned to them during crises. Table 1 displays the roles and functions of interviewed crisis managers, and on which sector level they operate.
Our interviewees (four females, three males) contribute to their organisations' precautionary measures in the face of potential crises. As professional communicators, they are also expected to contribute to processes of understanding and normalising unfolding crises (the last is a major The structured part of our interviews was first presented to our informants before they were interviewed. As our interviewees are skilled communicators and public speakers with the potential ability to control and frame interviews, we prepared follow-up questions to avoid well-rehearsed statements. We found that this approach contributed to getting our subjects to reflect openly around their experiences of their own inadequacies and dilemmas regarding their use of social media. All interviews were digitally recorded and de-identified.

ANALYSIS
In this section, we analyse and elaborate on perspectives that our informants offered on the research questions. Before representing our perspectives of our interviewees in each sector, we briefly summarise how each sector's communication efforts were officially evaluated À emphasising their social media use.

POLICE SECTOR
Official evaluations of the 22 July crisis management were for the most part critical of the police and the justice authorities' efforts. Tweets posted here are also continuously running À together with tweets from other police districts in Norway À on online media pages, such as 'VG Live 24-hours News', the live studio of nation's largest tabloid newspaper, Verdens Gang. This running news feed includes a Twitter feed dominated by statements from the police and emergency services.
One interviewee draws attention to how vital this interplay between social media and news media is in crisis communication. The operations head at OPD believes that if they had used Twitter when the bomb went off in Oslo, they could have informed people much more swiftly about the decisions that they had taken. However, he emphasises that it is through news media that communication with the public is effectuated: We achieve most impact from our use of Twitter by reaching journalists who communicate our messages in online media.
Even though we now have around 300,000 followers on our Twitter accounts, I believe that this dissemination through online media and radio is the most important for us in our crisis communication.
His views correspond to findings in recent research on media channel choices and preferences in the Swedish public during acute crises.
Although social media have an impact in the public's crisis communication, research by Odén, Djerf-Pierre, Ghersetti, and Johansson (2016) clearly point out how traditional news media, including their online editions, still have the key role in crises. It is through daily news media, radio and television that most people are alerted and informed (Odén et al., 2016, p. 234).

Interactivity and Dialogue in Terror Crises?
We need to ask whether Twitter use for the operational police unit is simply for disseminating information. Could it not be the case that it also receives information through Twitter that is vital for the emergency operations, for example, from actors close to or at a crisis scene? The OPD leader of the operational central office admits that if they had incorporated social media use in the crisis communication routines before 22 July, they could possibly have followed the situation on Utøya more closely and may have understood how difficult the situation was at an earlier stage.
However, he does not believe that social media would have changed the planning and execution of their operation. As an example, Twitter could not have helped them in determining how many perpetrators were on the island in the acute situation, because a huge amount of contradictory information about this circulated in the Twitter sphere. According to him, emergency phones provided the police with vital information that gave them a perception of the situation.
Should their role during terror crises include engaging in direct dialogues with victims on social media by responding to their messages? In a study of how Utøya survivors reflect on their experiences with social media during the attack (Frey, 2018), the interviewees express a need for communicators to develop their social media use during crises to include two-way communication, so that victims and crisis authorities can have real-time contact.
On their Twitter accounts, the OPD stress that they do not answer mes- The communications director at POD points out that whereas they are, in general, increasingly engaging in dialogues with users on social media, this is not the case in acute phases of crises:

95
Social Media in Management of the Terror Crisis in Norway A major concern after 22 July was if we could have engaged in a dialogue with the young victims at Utøya on social media, in a way that could have saved lives. Could we have tweeted: 'start swimming'? Because we saw that they wondered if they should hide or not. But we did not know how many perpetrators were there, or if they were on their way or where they were. Above all: the perpetrator or perpetrators could have followed us on social media. Giving advice under such circumstances could have worsened the situation. We would have had to take that into consideration. You need to be very careful.
Consequently, instead of entering into dialogues in acute crisis situations, they choose to draw out vital information by monitoring social media and provide the information and advice that is needed according to their perception and understanding of the particular crisis event.

Monitoring Challenges
Monitoring during terror crises and other crises, however, poses challenges. According to the POD communications director, monitoring of social media requires not only access to and use of applicable digital monitoring tools. Purposeful monitoring in crisis management also largely depends on training the staff to secure sufficient senior competence.
Primarily, the challenge is to draw out the essence of what social media actors are concerned about, and respond accordingly: 'This is what is happening now' and 'this is what we need to handle regarding crisis communication': Although we continuously replace and improve our tools, their use is still person-dependent. Social media communication officers lack knowledge of social media. In a crisis, I may be in need of the two that are competent in social media dynamics, but have only those who are semi-competent in place. In such situations, digital tools will not be of any help in drawing out essences from social media messages. Therefore, we still need to enhance and systematise our monitoring competence.
The OPD interviewees find that their established social media monitoring routines, which include daily monitoring of Facebook and Twitter, will not be adequate in severe crisis situations even though considerably more officers will be involved then. This is due to the enormous amount of crisis-related messages that will be posted, many of which addressed to them: 'There is an overflow of information which the staff cannot efficiently sort out and analyse', says the OPD communications advisor.
Digital tools may, however, be a support in filtering, sorting, validating and spotting patterns in the social media postings. In turn, the communicators at OPD can post public messages providing advice. POD, who are situated at a greater administrative distance from where crises unfold, will, on their part, not see it as their task to post crisis-relevant advice directly to users. Rather, their tweets will signpost relevant websites and thus direct general users and journalists to places where they can find advice and verified information.

Daily Routines and Challenges for Future Crises
All our police sector interviewees emphasise that a precondition for good social media monitoring and advice in crises is establishing daily routines for social media communication. Two months after the Utøya tragedy, every employee in the operations central office was expected to tweet a least one message when on duty. This has now developed to 20 tweets posted on average as well as increased daily interaction with users and monitoring of social media postings. They believe that the communicative competence in social media that they acquire through their daily use has prepared them for improved application of social media in crises. Their increased social media usage has also created expectations among their rising number of followers: People expect that we deliver on social media. If a crisis occurs now, we cannot be silent in social media. We see it as a positive challenge that people require more of us after we have started using social media frequently. (Communication advisor, OPD) The crisis communicators in the police sector are also aware that demanding challenges and deliberations will arise in coming crises. One pitfall is that the pressure to communicate swiftly in emergencies may lead to wrong and misleading postings, with potentially unfortunate 97 Social Media in Management of the Terror Crisis in Norway consequences. Another risk is to use an inappropriate tone or voice. The OPD has on occasion been criticised for using humor in their tweets.
POD's communication director is of the opinion that one should be very careful with using irony and humour in tweets posted from the operations command centre: 'Humour and irony may come out wrongly and offend people. We need to maintain our authoritative voice, while at the same time using a language that speaks with "head, heart and hands"' she says.
The OPD head of operations nevertheless wishes to avoid too rigid rules for his staff's daily usage of social media, as this may 'inhibit their production of Twitter messages'.
A potential threat to good crisis management is fake social media accounts, something OPD has experienced, but managed to remove. The operational leader points out that it is probably not a great challenge to hack their social media accounts. Should that happen in a terror situation, it would create a chaotic situation, and the only thing they can do to prevent this is to maintain their well-established routines of continuously changing passwords.
Finally, a major risk for good crisis management and communication would be if key actors, sectors and agencies involved do not coordinate sufficiently and publish univocal advice. Although administrative responsibility is assigned to different sectors in different crises À and the police sector has the leadership in management of terror crises À social media may pose challenges for the different actors involved.
Due to the swiftness of communication in social media, we are afraid that other actors, for example Oslo municipality, could go out with a different message and say: Do not evacuate Oslo! Whereas the police say: Evacuate the city! Then we will really have a crisis. So, to be able to coordinate between the different actors is all-important in severe crises.

HEALTH SECTOR
The national official evaluation made by the 22 July Commission (NOU, 2012), a report from the Swedish Directorate of Health (Socialstyrelsen, 2012) and an evaluation made by the Norwegian Directorate of Health (NDH, 2012) all concluded that the health sector did a good job in the acute phase of the bomb attack in Oslo and the shootings on Utøya.
According to NDH, the health service 'responded very well to the greatest challenge it has ever faced' (NDH, 2012, p. 12). NHD also stated that voluntary rescuers made a great and valuable contribution during the acute phase. Professionals rushed to their work places even without being on call, that is, available to be called to work outside normal working hours.

Hospital Staff Experiences with Twitter: Swift, and Potentially Misleading
However, as was the case with the police sector, social media were not integrated in the health authorities' crisis communication strategies at that time. In their evaluation, NHD found that public health officials overall were absent in social media on the afternoon and evening of 22 July.

NDH (2012, p. 107) concluded that:
Effective communication via social media should be given greater priority in crisis PR. In the event of major crises, health service actors should be more proactive in monitoring, engaging in dialogue and providing information via social media.
The exception from the general social media absence was Oslo University Hospital (OUH). The hospital tweeted several times about press briefings, which telephone relatives could call, and the need for blood donors (NDH, 2012, p. 105). According to the communication officer who was serving OUH's official Twitter account on 22 July, the communication staff experienced that Twitter was faster than the hospital's official lines of communication, the emergency medical system (AMK): We understood through Twitter that there was shooting on Utøya, before we were notified through our own system. It was unreasonable to assume that such a large amount of messages were fictional or based on errors.

Social Media in Management of the Terror Crisis in Norway
Twitter, in other words, played an important role in raising awareness of the crisis among hospital employees.
However, OUH's experiences in 2011 also demonstrated the need for a close monitoring of social media. During the terror crisis, the hospital tweeted to get in touch with blood donors with the blood type 0 Rh (À) (see also Ottosen and Steensen in this volume). This message did not specify that the hospital blood bank only wanted to get in contact with registered blood donors. When similar incomplete messages were spread by news media, this led to long queues of people outside OUH. As stated by their communications advisor: When I was outside at 10:30 PM, the situation by the main entrance was surrealistic. There were queues with cars and pedestrians. Among them were people that had never donated blood before. Nor were all of them sober, but this did not matter because new blood donors with the wrong blood type could not help us.
OUH then posted a new message pinpointing that only people that already were registered as blood donors at the hospital should phone the blood bank. The communication staff also phoned the most important TV news programme, NRK Dagsrevyen, and asked for assistance. 'We also tweeted that we had enough blood donors that evening. We had an enormous spread of news that night', says the communications advisor.
Ringerike Hospital was the hospital closest to Utøya, and the one that treated the largest number of patients. However, this hospital and the hospital trust it belonged to, did not have either a Facebook or a Twitter account. The hospital was almost 'invaded' by people that had heard there was a need for blood donors. With social media, this message could have been coordinated with OUH and corrected (Socialstyrelsen, 2012, p. 100).

'Every Crisis is Unique'
The communication advisor at OUH says that the hospital has not made any major changes, only ongoing adjustments, in their strategies for using social media in crises. In his opinion, 'every crisis is unique'. Therefore, making extensive procedural changes in the aftermath of a single situation could make the procedures less useful when a new and different situation evolves. Instead, he highlights the importance of being active on social media on a daily basis, participating in full-scale exercises and being proactive and able to improvise.
As an example of the last point, Twitter was used to alert staff to precautionary action in 2014 when a gunman was thought to be inside the huge hospital area. A police action was taking place without any information given to the hospital: 'It looked like a warzone. I could see 7À10 police cars, and 10À20 heavily armed policemen', said the communications' officer who phoned a neighbour working as a communication officer in the police. She soon found out what was going on and phoned him back.
We then posted a message on Intranet and a Tweet stating that there was an unexplained situation on OUH, and we recommended patients and employees to stay indoors.

(Communication advisor, OUH)
The episode testified to the importance of swiftly finding effective ways to solving challenging situations. 'It was not by the book. But it worked', the communications' advisor said. As we understand him, this 'gunman episode' also showed that Twitter was the quickest and most flexible channel for warning about a possibly life-threatening situation.

Social Media on the Scene: A First Responder's Experiences
Our first responder interviewee was employed by OUH, but had his workplace on an ambulance helicopter base situated outside Oslo. He was one of the first health responders to arrive on Utøya, even before the area was 'cleared' by the police SWAT unit (Måseide, 2013, p. 87). He was also one of the last health responders to leave the island, together with police officers, after checking that the casualties were dead. Based on two quite different episodes that he experienced in connection with his Utøya work, he has drawn the conclusion that health personnel treating patients should avoid using social media during a crisis and leave communication to the appointed professionals in the organisation.
The first episode took place on the island of Utøya. The situation was unclear for a long time after the perpetrator surrendered: there was a fear of several perpetrators hiding there. The first health responder was repeatedly ordered by the police to lie down on the floor of the island's main building because of the prevailing fear. When he sent a 'neutral tweet' from the island, communicating that he was safe, a journalist soon contacted him by SMS, trying to get sensitive information.
Another experience, this time with tweeting in aftermath of the attack, led him to evaluate critically his own social media use during crises. When he returned from Utøya to the mainland at 2 a.m, he became aware of a helmet with the inscription 'Medical doctor, Oslo'. On the helmet, there were some solidified drops of blood. To the first responder, a photograph he took of the helmet became a symbol he wanted to share on Twitter.
The communication unit at OUH regarded the tweet as inappropriate and got in touch with the first responder's leader, explaining that they wanted to have it deleted. The first responder deleted it but also posted a new one, asking those who might have saved the photo, to delete it too.
This led to a debate on Twitter about his ethical judgement.
When the health responder was interviewed again in 2016, he admitted: 'I cannot act uncritically as an individual on Twitter, as I did on 22 July', referring to how he had used his private account in 2011.
As related by him, the two episodes demonstrate how difficult it may be to anticipate the consequences of using Twitter during an acute crisis.
Health personnel using Twitter in such situations could also create the impression that they have an informational role towards the public and the press, a role that they should not have in his opinion. As we interpret him, health personnel like him, who do not have a communication task on behalf of the organisation, can acquire a dual role if they communicate via social media from a crisis site. On the one hand, they may be seen as individuals, on the other hand, as professionals. Although they want to do something good by tweeting, tweets can be misunderstood by recipients. This is one reason why health personnel should avoid using social media during crises.

THE NEED FOR COORDINATION
All three emergency services À health, fire and rescue and police À have traditionally operated separately in Norway. Our interviewees in both the health and the police sectors stress that different authorities' social media use needs to be both proactive and coordinated when a crisis evolves. The NHD's communications director underlines the importance of having clear duties assigned to the different services during a certain crisis and the need to build a strong following on social media, with a web page that is operative and used in non-crisis situations providing coordinated information. The first health responder points out how all three emergency services have a shared responsibility as information facilitators.
According to him, this effort should be coordinated to a greater extent than it is today: The police are very good and very active on Twitter, but I do not think they have any good plans for using Twitter in a crisis. I believe the same is true for the fire and rescue services, and it is certainly the case when it comes to the health services.
In his opinion, crisis communicators in institutions such as OUH need to communicate immediately on Twitter that they are present and will follow the situation closely when a crisis evolves. They should share validated information, call for additional personnel when necessary, and communicate with the public, as they did when there was a need for getting in touch with registered blood donors during the 22 July.

SUMMARY AND CONCLUDING DISCUSSION
We posed the question of how key crisis managers/communicators in the police and health sectors understand and evaluate their social media use during the terror crisis in Norway 2011. Furthermore, we asked how they have improved their use and what they see as challenges and possibilities for using social media. The following tables ( Table 2 and Table 3) summarise perspectives followed by a short discussion.
Our police interviewees' reflections on their communication practices during 22 July clearly demonstrate how their unpreparedness regarding social media use in 2011 has alerted them to the urgent need to include social media in their emergency communication. The social media routines they have established testify to how they acknowledge the importance of engaging social media in daily communication activities to strengthen public presence and trust, and thus prepare for effective use of social media during crises. However, although the interviewees stress that social media can contribute to saving lives and resolving crises, they do not  Given how social media provide crisis communicators with the opportunity to bypass traditional gatekeepers such as news media and distribute information and advice directly to the public during crises, it interesting to note how the police interviewees stress how vital the interplay between news media and social media is in crisis communication.
When the operational leader finds they achieve most impact from their Twitter messages by reaching journalists who communicate them in online media, his acknowledgement corresponds with findings from significant crisis communication research. Schultz, Utz, and Göritz (2011) and Utz et al. (2012) find that people are more willing to share messages from traditional than from social media, because they interpret traditional media as more credible.
Contrary to the police interviewees, our health interviewees actually used social media À particularly Twitter, when the crisis unfolded. In their different roles, however, they used Twitter in quite different ways. For the communications advisor at OUH, improvements implemented in current routines largely follow up on the practices that were beginning to establish themselves at the time of the crisis, with routines of daily use and monitoring strengthened.
Although the first responder's Twitter experiences in connection with his emergency work on 22 July taught him that health personnel on the scene should avoid using social media in emergency operations, we believe that information officers may have an important role in using social media to raise the awareness of an evolving crisis. One possible solution could be that communicators operating on the crisis scene 1 can use social media if they strictly follow all relevant ethical and legal rules of confidentiality, and show consideration for the consequences social media postings may have for patients and the public.
Finally, we observe that the health informants in their different roles and functions in crisis management, strongly underline the need for better coordination between different actors and agencies in their own sector, and À equally important À between the different sectors' emergency services. There is a risk that social media's swiftness may lead to postings of contradictory messages from different services during emergencies, with potentially devastating consequences. To avoid escalating a crisis through social media use, it is vital that the different emergency services coordinate and strengthen their plans for using social media in crises. One suggested step towards securing a univocal message is that all key agencies have a web page providing coordinated information to which their social media accounts can direct their users.

NOTE
1. An agency in Norway that has demonstrated that this can done in practice is the Oslo Fire and Rescue Service, whose communications' director has tweeted during emergency operations (source: OUH's communications advisor).