Collective sensemaking within institutions: Control of the COVID-19 epidemic in Vietnam.

How people make initial and collective sense under crises remains unanswered. This paper addresses this question using the control of COVID-19 in Vietnam as a case study. Our results suggest that sensemaking under crises is influenced by an institutional propensity for prevention that has developed gradually over time. Local governments play a vital role in fostering collective sensemaking which enables concerted actions in epidemic control. However, biases are inherent in sensemaking, including a delay in access to vaccine and a violation of privacy. For policy makers, this study suggests that developing specific prevention policies and programs, building large-scale coordination capacity, and promoting local initiatives are necessary for coping with epidemics. For theory development, the study explores how institutions condition sensemaking and specifies several mechanisms in which local authorities could facilitate collective sensemaking in crises.

INTRODUCTION

COVID‐19 epidemic is an unexpected event of unprecedented magnitude, duration, and reach, requiring governments to make proper sense and actively influence the emergence of collective sensemaking in the nations (Angeli & Montefusco, 2020; Christianson & Barton, 2020). The literature provides ample evidence that in crisis situations, such as COVID‐19, people conceive the problems and react to them through sensemaking processes rather than comprehensive problem analysis (Lu & Xue, 2016; Persson, 2013; Weible et al., 2012). Sensemaking facilitates swift decisions under conditions of high uncertainty and ambiguity, but it can also create blind spots by preventing actors from bracketing contradictory cues until it is much too late (Maitlis & Sonenshein, 2010; Weick, 1995). Individuals may construct different senses for the same crisis, calling for an effort to foster collective sensemaking and concerted actions. However, the literature has been not clear on how collective sensemaking could be facilitated at the national scale and how local governments contribute to this process.

Scholars have proposed that an integration of institutional and sensemaking perspectives will shed light on these questions (Weber & Glynn, 2006). According to this approach, sensemaking is embedded in broader social, historical and institutional contexts (Maitlis & Christianson, 2014; Weber & Glynn, 2006). In fact, institutional theory (DiMaggio & Powell, 1983; Meyer & Rowan, 1977; Scott, 1995) suggests that perception of and measures to cope with a crisis are constrained by existing institutions, which are rooted in the regulative, normative and cognitive systems. According to this view, institutional contexts explain the differences in substances and processes of individual and collective sensemaking.

This research explored how collective sensemaking emerges within institutions to cope with crises and how local governments contribute to the process. We believed the control of COVID‐19 in Vietnam from January to September 2020 is an interesting case study for several reasons. Firstly, the country's experience of Severe Acute Respiratory Syndrome (SARS) 2003 biased the government toward aggressive measures and a swift control of the virus. This turned out to be unsustainable since COVID‐19, compared to SARS 2003, is more universal in scope, lasts for relatively long time, and yet, remains mysterious as different variants have been evolved. Government at all levels had to constantly make sense of the disease and adjust their measures. Secondly, in the country, the desire for strong control from the central government has been constantly contested by the need for local initiatives (Nguyen et al., 2013). While local authorities may not have law‐making power regarding nation‐wide epidemics, their experience of and attitude toward the epidemic greatly influence how they interpret and implement the policies set by the central government. In the Covid‐19 case, local governments play significant roles in managing people's attitude toward the epidemic through influencing the emergence of collective sensemaking. This offers an excellent opportunity to examine the roles of local governments in fostering collective sensemaking under uncertainty.

Our study contributes to the literature in several ways. First, we respond to Weber and Glynn (2006) call for an empirical study on how people's sensemaking and associated biases are conditioned by surrounding institutions. Our paper shows that, under crisis, institutions influence what cues people noticed, how they defined themselves, and what they are committed to and acted upon. Second, we identify various mechanisms local governments used to foster collective sensemaking under crisis. We also develop a concept of institutional propensity for prevention (IPP) and discuss how IPP influences local governments in making sense of the crisis. To our knowledge, empirical studies that integrate sensemaking and institutional perspectives to analyse the role of local governments in fostering collective sensemaking are non‐existent.

The rest of the paper is organized as follows. First, we briefly discuss sensemaking perspective in crises and the roles of institutions in actors' sensemaking processes. Then, we present our Methodologies and provide a description of the institutional context. Next, in the Results, we describe the unfolding of COVID‐19 control in Vietnam and discuss the substance of sensemaking, the role of local governments in the creation of collective sensemaking. We also develop a theoretical model describing the influence of institutional factors on these processes, based on the case of COVID‐19 in Vietnam. Finally, the Discussion provides policy and theoretical implications.

THEORETICAL BACKGROUND

Sensemaking in crises

A crisis is a low probability/high impact event that threatens the viability of an organisation, community or society and is characterised by ambiguity in its cause, effects and means of resolution (Maitlis & Christianson, 2014; Weick, 1988). The conditions of a crisis give powerful occasions for sensemaking (Maitlis & Christianson, 2014) as an inadequate sensemaking process would likely make the crisis appear out of control (Weick, 2005). While sensemaking can be defined in different ways (Maitlis & Christianson, 2014; Starbuck & Milliken, 1988; Weick, 1995), it essentially involves three main aspects: perceptions, interpretations and actions (Wright et al., 2020). Weick (1995) specified seven properties of sensemaking, including (1) grounded in identity construction; (2) retrospect; (3) enactive of sensible environment; (4) social; (5) ongoing; (6) focussing on extracted cues; and (7) driven by plausibility. In this paper, we follow Maitlis & Christianson, (2014, pp. 67), who define sensemaking as:

…a process, prompted by violated expectations, that involves attending to and bracketing cues in the environment, creating intersubjective meaning through cycles of interpretation and action, and thereby enacting a more ordered environment from which further cues can be drawn.(Maitlis & Christianson,

Sensemaking can be a useful lens to theorise how certain resolutions emerge and develop in crises. People ‘make sense’ of situations not by carefully analysing complete sets of data as these data are in no way complete; instead, they do so by enlarging the cues that they notice, at the risk of ignoring other important signals. People cannot rely on existing frameworks of cause and effect to comprehend crises as they do not fit these ambiguous situations. Instead, people try to understand the situations they encounter by redefining who they are (their identities) and taking actions. By doing so, people contribute to the environment that they experience (enactment). People justify and rationalise past actions based on the results (retrospect), and—as such—plausibility is more important than accuracy. While sensemaking facilitates swift decisions, it may also create blind spots (Weick, 2005).

Studies of sensemaking in crises were conducted in various contexts, including a fire in Mann Gulch (Weick, 1993), a gas leak in Bhopal (Weick, 1988), a disaster in entertainment events (Vendelo & Rerup, 2009), mining disasters (Wicks, 2001), local crime (Persson, 2013), or emergency management systems (Lu & Xue, 2016), to name a few. These studies converge in several themes. First, studies have shown that individuals' optimistic evaluations of crisis situations are often a fatal mistake. When people view a crisis as ‘ordinary’ or ‘under control’, they blind themselves from growing evidence to the contradictory. In particular, if these evaluations are made public, people are even more likely to feel bound to them (Maitlis & Sonenshein, 2010), creating dangerous blind spots. Second, previous studies acknowledge the complexities of collective sensemaking during a crisis. Since people have different backgrounds and experience, they are likely to draw on different sources of knowledge to develop different understandings of the same situation. Yet, in some situations, people had not developed sufficient trust with each other when the crisis occurred (e.g., Weick, 1993). In other situations, they may be scattered across locations with poor communication skills and equipment (e.g., Vendelo & Rerup, 2009). These inhibited a social process that facilitates disparate individuals and teams to jointly involve in scanning, sharing, and interpreting information on a particular issue.

Sensemaking within institutions

Scholars have increasingly recognised the role of institutions in shaping and facilitating sensemaking processes (Maitlis & Christianson, 2014; Weber & Glynn, 2006; Weick, 2005). As Taylor and Van Every (1999, p. 275) put it, ‘making sense…is not an accomplishment in vacuum, it is not just context‐free networking’. The dominant view has been that institutions provide contexts for sensemaking, shaping what we expect or take for granted (Maitlis & Christianson, 2014; Weber & Glynn, 2006). According to this view, institutions impose cognitive constraints on the actors who do the sensemaking and the ‘content’ of sensemaking often mirrors the ‘content’ of institutions (Weber & Glynn, 2006).

Weber and Glynn (2006) suggest a framework that integrates institutions in the substance of sensemaking. In this framework, institutions enter the sensemaking process by influencing the construction of identities, frames and actions. Institutions influence an actor's processes of redefining their identity by constraining the types of identities that are acceptable. Similarly, the noticed cues need to be ‘bracketed’ or framed to generate some ‘sense’ of what is going on. Institutions put constraints on acceptable frames. For example, while a frame of political or multi‐party competition is well accepted in western democratic societies, it is nether available nor politically acceptable in a one‐party led society like Vietnam. Thus, institutions supply the substance for an actor to answer both questions of ‘Who am I?’ and ‘What is going on?’ during a turbulent time. These institutionalised identities and frames determine expectations about how actors should act in the situations, thereby giving rise to institutionalised actions. Given the same crisis, actors from different institutions would likely make different sense and commit to different courses of action.

In Weber and Glynn (2006) model, institutions do not only constraint but also facilitate sensemaking through priming, editing, and triggering processes. In priming, institutions serve as filters that lead people to extract cues, which in turn activate identities, frames and role expectation in particular situations. Institutions prime sensemaking by showing available measures that can readily be used to construct a course of actions, suggesting a starting point in meaning‐making, and providing rules in connecting cues with identities and frames. The priming emphasizes the role of situational context as a primary determinant of action formation. Institutions edit an actor's sensemaking in providing social evaluations of the actor's identity performance. Once others' expectation is not met or broken, the actor is under pressure to modify their meaning‐making and performance. Social interactions are essential in this process (Weick, 1995). Finally, institutions trigger sensemaking by directing attention to certain issues and demanding reconciliation of contradictions and ambiguity that are inherent in the institutions. In brief, institutions facilitate sensemaking in filtering cues, directing attention to certain foci, governing the interpretations of the cues, and editing the meaning‐making process. Given the same crisis, actors from different institutions would likely make different senses and commit to different courses of actions.

While the sensemaking perspective originated from the organisational theory tradition (Maitlis & Christianson, 2014; Starbuck & Milliken, 1988; Weick, 1995), it has been evident in some public policy research. Examples include studies of policy influence (Weible et al., 2012), local crime prevention (Persson, 2013), rural planning (Hulme, 1989), discourse on ‘globalization’ (Fiss & Hirsch, 2005), emergency management systems (Lu & Xue, 2016), and evolution of reform policies (Nguyen et al., 2018).

Studies of sensemaking in crises, however, suffer from several shortcomings. First, discussions on how sensemaking emerged within institutions remain largely theoretical (Weber & Glynn, 2006; Weick, 1995). We lack evidence to demonstrate how institutions condition people's sensemaking under crises. Second, most studies have been on crises with small scales, that is, involved a team or groups of several dozens of people, that happened in a relatively short time, that is, from less than an hour to several days. How collective sensemaking emerges in large‐scale crises, such as COVID‐19 epidemic, has not been studied. We are also left with a puzzle of how stakeholders, especially local governments, coordinate collective attention to the issues of the crises.

METHODOLOGIES

Research context

The political system in Vietnam is characterised by a one‐party socialist republic state led by the Communist Party of Vietnam (CPV or the Party). State power appears at four levels, including central, provincial, district and commune/ward. Besides the Party, state and government organisations, there are socio‐political, socio‐professional and mass organisations, which are grouped under the Vietnam Fatherland Front. Every 5 years, the CPV organises a national congress for the setting of strategies for Vietnam's socio‐economic development and the election of new leaders. By the end of 2020, congresses at the various levels of commune/ward, district and provincial were required to take place in preparation for the National Congress scheduled in early 2021. The COVID‐19 outbreak of happened during preparation for this congress.

According to the World Bank's World Development Indicators, Vietnam's health expenditure reached US$357 per capita in current international Purchasing Power Parity in 2016, equivalent to 6.4% of that in high‐income countries, and about half of those in East Asia and the Pacific (excluding high‐income countries). According to the World Health Organization, by 2015 Vietnam had 25.6 hospital beds and 12.2 health workers per 10,000 population. These coverage ratios appear similar to those of the neighbouring country of Malaysia, but limited resources might affect the quality of healthcare delivery.

Data collection

We studied how people at various occupations and localities perceived, interpreted, and responded to COVID‐19 cues in early days (January to September 2020). During that time, COVID‐19's categories and concepts were not set but were built up and constituted in and through interactions. People's perception and interpretation of COVID‐19 varied by both time and space. Therefore, we considered a qualitative approach to be the most appropriate for this study as it utilizes an interpretive, naturalistic approach to the subjects and phenomena (Denzin & Lincoln, 1994). Qualitative approach allowed us to approach respondents from different positions, to reveal their assumptions underlying their perception of COVID‐19 with specific reference to a local context. Document reviews, interviews, and observations were our main sources of data.

Document review

We obtained extensive documentary data from regulations, reports and newspapers. Regulations on COVID‐19 control were extracted from various archival sources within the central Government Office, Ministry of Health (MoH), and Offices of People's Committees and Departments of Health in three different cities and provinces, with various degree of exposure to the epidemic, including Hanoi, Da Nang and Son La. These regulations revealed government's beliefs, intentions, and commitments in relation to COVID‐19 control. In total, 173 central and local regulatory documents, including instructions and directives, were reviewed.

Regulation review was supplemented by expert analysis and government leaders' statements. We obtained four analyses, two from medical experts and two from economists. The medical expert analyses were written and posted in late February 2020, expressing different viewpoints about COVID‐19. Economic analyses came from well‐known organisations in Vietnam, such as the National Economics University and the Vietnam Chamber of Commerce and Industry. The reports revealed divergent views on the future of COVID‐19 and the pressures it would put on the economy. We also collected government leaders' statements from relevant meetings and interviews that were made public by national television (VTV) and mainstream newspapers.

Interviews

We interviewed members and leaders of COVID‐19 Prevention Committees at both central and local levels, staff members serving in quarantine centres, infected and quarantined citizens, as well as business managers at the time of the epidemic (March to May, 2020, and July to September, 2020). We conducted online semi‐structured interviews with these informants, using Facebook Messenger or Skype applications. In total, we conducted 31 interviews which lasted between 30 and 90 min. With government officials, we asked them to reflect on how their perception of COVID‐19 had formed and changed, the key events they experienced and were responsible for, and the commitments and actions of their agencies. We also asked specific questions on why and how the government came up with these commitments and actions. Other questions were tailored to each informant's specific role. Notes were typed or hand‐written nearly verbatim and transcribed into the computer within 24 h. With citizens and businesses, we asked how they learnt about the COVID‐19, their specific experiences—such as being quarantined or having to close businesses—and their views on government measures. We obtained permission to audiotape interviews with citizens and businesses.

Observations

Our observations and diary notes from February 1 to August 20 were also included in the data. Two of our research assistants were quarantined from March 20 to April 4 after they returned from the United States. One author took a summer holiday in Da Nang from July 20–25 and was subsequently quarantined and tested in Hanoi. These members noted the arrangements and quality of services, the attitude and cooperation of quarantined people, and daily information provided. We also spent time observing people's behaviour at supermarkets and pharmaceutical stores on key dates when important events were announced.

Data from multiple sources were then triangulated to reduce inconsistencies. An iterative approach was used to cross‐check the descriptions and views given by different participants. Data sources and uses are summarised in Table 1.

TABLE 1

Description of data

Data types	Description	Uses in analysis
Secondary data‐ Government regulations, instructions, and reports ‐ Local (Hanoi, Da Nang, Son La provinces) instructions ‐ Expert analyses	173 regulations, instructions from central and local (Hanoi, Da Nang, Son La) government 04 expert analyses	Formal viewpoints of the government on the epidemic over time
Primary data Interviews
11 interviews with government officials, lasting from 30 to 60 min.‐ NCPC: 3 ‐ Ministry level: 2 ‐ LCPCs: 6	125 pages of text typed directly from the online interviews or transcribed from tape recorders	‐ Participants' perception of the virus and of the epidemic. ‐ Participants' thoughts of possible scenarios in the near future. ‐ Participants' accounts of government prevention actions, and their own participation.
05 interviews with healthcare and medical experts
10 interviews with quarantined citizens
05 interviews with enterprises
Observation and diary notes	Notes on how NCPC statements changed (mostly on national TV and newspapers). Notes on how people reacted to the epidemic	How meanings were constructed and changed. How actions were modified
Notes on evolvement of the epidemic and government instructions.
Observation of people's reactions

Abbreviations: LCPC, Local COVID‐19 Prevention Committees; NCPC, National COVID‐19 Prevention Committee.

Data analysis

We followed abductive approach to analyse our data (Dubois & Gadde, 2002; Timmermans & Tavory, 2012). Abductive analysis is a continuous movement between an empirical world and theoretical world. By pushing the data against existing theories, we identified changed circumstances, additional dimensions, or misguided preconceptions.

We organised data according to the three phases described in a previous section: ‘Aggressive measures, swift control’ (January 23–April 16), ‘Complacence’ (April 17–July 24) and ‘Strong measures, long‐term control’ (July 25–September 30). In each phase, we coded data according to sensemaking concepts, including ‘extracted cues’, ‘frame’, ‘identity construction’, ‘commitment’ and ‘action’. While these sensemaking properties varied greatly across locations, ultimate actions were rather concerted, suggesting some collective sensemaking mechanisms were in place. We then explored how collective sensemaking was made. Three mechanisms emerged from the data, including propaganda and participation; coordination and knowledge sharing; and sense‐bridging. The interactions between central and local governments were termed as ‘sense‐giving’ and ‘sense‐updating’. We developed working definitions of these terms and three authors independently coded data based on these concepts. We discussed any discrepancies in the coding until consensus was reached.

Table 2 summarises the coding scheme.

TABLE 2

Coding guidelines

	Evidence in the data
Sensemaking
Extracted cues	Types of information and signals to be noted and used to interpret and/or construct meanings or issues of the epidemic (which types of information/signals got noticed, and which ones got ignored)
	Evidence of comprehensive, systemic analyses to support prevention actions
Frame	How virus/epidemic were presented in the documents, statement, propaganda.
	‐ Languages, terminologies
	‐ Relations to existing tasks
	‐ Emotion related to the framing
Commitment	Stated goals set by the government
	The match between stated goals and actions
Identity construction	Redefinition, negotiation, or debate on related parties':
	‐ Missions, mandates, functions, roles
	‐ Beliefs and values
	‐ Distinctive characteristics or entities
	Correspondence between epidemic meaning making with changes in related parties' identities
Actions (Enactment)	Prevention measures in practice:
	‐ Planning
	‐ Prevention actions: Communication, propaganda, monitoring
	‐ Reactions to infected cases:+ Timing and seriousness of the responses to any infected case + Quarantine and block‐down actions + Testing of high‐risk areas
	‐ Acceptance of high direct costs
	‐ Acceptance of breaches of some basic rights
	‐ Biases, extremeness, and irregularities in actions
Construction of collective sensemaking
Coordination	Mechanisms for cross‐functional coordination
	Mechanisms for cross‐level coordination
Propaganda	Dominant messages in the propaganda
	Creativities in propaganda
	Debates and oppositions to the propaganda
Participation	Participation from citizens and businesses:
	‐ Contribution of resources
	‐ Provide information on neighbourhood issues related to COVID‐19
	‐ Contribute to social programs
Sense‐bridging	The act of making sense of how different entities notice cues, interpret, and give meanings to a situation and then find ways to create shared meanings among these entities
Sense‐giving	The act that one party (mostly central government/NCPC) influences others in understanding the situations and envisioning actions
Sense‐updating	The act that one party elicits inputs from others to adjust their interpretation of COVID‐19 and modify their commitment, identities, and actions
Institutional propensity for prevention (IPP)	‐ Readiness to engage with potential problems (e.g., time to issue scenario plans, resources mobilised)
	‐ Tendency to engage in aggressive control
	‐ The presence of three factors (that is, experiences in similar epidemic, local healthcare system,

Abbreviation: NCPC, National COVID‐19 Prevention Committee.

RESULTS

Prevention of COVID‐19 in Vietnam and roles of local authorities

The evolution of COVID‐19 in Vietnam up to September 30th, 2020 could be categorised into three broad phases. The first phase began with the first two COVID‐19 cases reported on January 23rd. On January 30th the National COVID‐19 Prevention Committee (NCPC) was established. The NCPC was chaired by a Deputy Prime Minister and members were representatives from the MoH and other related ministries. Local COVID‐19 Prevention Committees (LCPCs) were established at all local administrative levels. Initially, most of the infected cases were international arrivals. Later, infected cases from communities with no clear traces triggered concerns amongst the public. Nationwide social distancing was implemented from March 16th to April 1st. The Central Government persistently delivered messages of ‘putting human life as the foremost priority’, and ‘fighting the epidemic like fighting invaders’.

The second phase from April 17 to July 24 marked 99 consecutive days of no transmission within communities. The government considered re‐opening the economy, with top priority given to domestic tourism. Line ministries and local government were required to revise their plans, focussing on economic recovery.

The third phase began with four new cases officially reported on July 25th. All of these were patients or patients' caregivers from three hospitals in the inner city of Da Nang, a major tourist attraction in the central region. The subsequent cases were people travelling through and tourists returning from Da Nang. In this phase, group testing was implemented for the first time in the country, speeding up the tracing of COVID‐19. The overall strategy changed to location‐specific quarantine upon virus eruption instead of nationwide social distancing as previously taken place in the first phase. However, the overall guiding principle; namely, state‐led effort which aimed at a complete quarantine of the virus, did not change during 2020. More detailed descriptions of the three phases are provided in the Appendix.

While the overall principles and policies in controlling Covid‐19 were set by the central government, local authorities had notable discretions in the implementation of these measures. Vietnam's COVID‐19 control strategy adhered to the ‘Four On‐Site’ principle, including: (1) ‘on‐site’ leadership; (2) ‘on‐site’ human resources; (3) ‘on‐site’ facilities and equipment; and (4) ‘on‐site’ basic supplies and logistics. Local leaders took full responsibility for handling COVID‐19 issues in their areas. The Government's directives numbered 15, 16, and 19 allowed provincial governments to specify measures that fit with the specific local contexts. For examples, provincial government made decisions on which locations to be quarantined, the lengths of quarantine, conditions for travels to and from the provinces, and what were ‘essential goods/services’ (hang hoa thiet yeu) that were allowed to trade during quarantine. This approach promoted strong local initiatives and facilitated coordination among provinces. However, variation in local implementation of the policies was noteworthy. In addition, coordination between provinces appeared to be problematic, particularly in interprovincial travel and transportation.

Sensemaking and updating regarding COVID‐19 by the Prevention Committees

We focus on the senior party‐state officials and NCPC as the main actor in sensemaking of the COVID‐19. The LCPCs and other stakeholders played more passive roles as they were greatly influenced by NCPC in understanding the situations and envisioning actions (i.e., sense‐giving). LCPC were also periodically elicited inputs for NCPC to adjust their interpretation of COVID‐19 and modify their commitment, identities, and actions (i.e., sense‐updating). The main roles of local governments, therefore, were to make sense of how central government (NPCP) perceived, interpreted, and responded to COVID‐19 and then foster collective sensemaking and actions in their localities. Table 3 summarizes the three phases of COVID‐19 and associated sensemaking properties.

TABLE 3

Sensemaking properties at different phases

	Phase 1: Aggressive measures, swift control	Phase 2: Complacence	Phase 3: Strong measures, long‐term control
Cues and frames	Overall theme: The virus is deadly and needs to be fully controlled ‐ The virus was named as ‘new SARS virus’ ‐ The virus was seen and communicated by officials as ‘unknown’, ‘uncertain’, and ‘deadly dangerous’ ‐ COVID‐19 control was framed as ‘fighting the epidemic as fighting invaders’, and as ‘political responsibility’	Overall theme: The virus has been controlled and we need to recover the economy ‐ The virus was ‘successfully’ controlled ‐ Recovering the economy, especially small and medium‐sized businesses ‐ Domestic tourism is safe and encouraged ‐ Some symptoms of COVID‐19 were ignored	Overall theme: We have to live safely with the virus for much longer time ‐ Fighting against the virus is a long‐term battle and is everybody's job ‐ The virus should/could be controlled while minimizing economic loss
Commitment	Overall theme: Fully controlled the virus to protect citizens and show superiority of the system ‐ Zero death from COVID‐19 ‐ Keep infected cases under 1000 by Apr 1st, 2020 ‐ Sacrifice economic development to control COVID‐19 ‐ The state takes care of all expenses related to COVID‐19 control	Overall theme: Recover the economy and maintain a positive growth rate ‐ Launch a recover subsidy to business ‐ Promote domestic tourism	Overall theme: Achieve dual objective of controlling the virus and maintain a positive economic growth rate ‐ Dual‐objective: COVID‐19 prevention and economic development ‐ Keeping ‘zero death’ is not feasible ‐ People entering Vietnam from abroad have to pay for testing and quarantine fees
Identities	Overall theme: People and public organizations are ‘soldiers’, fighting against COVID‐19 ‐ State takes care of people's lives ‐ Public organizations, military, polices, local governments provide services to COVID‐19 control ‐ Citizens monitor and report compliances COVID‐19 regulations	Overall theme: Back to old normal ‐ people and public organizations unconsciously swing back to the old mandates and routines ‐ Popular perception: ‘we are the winner’ in fighting COVID‐19 ‐ People and organizations have missions to recover the loss and foster development	Overall theme: Living in the ‘new normal’ Individuals and organizations have to revise their missions and routines to balance between normal mandates with COVID‐19 prevention
Actions	Overall theme: Strong control approach at all levels ‐ All public entities were required to conduct scenario planning ‐ Aggressive preventive measures were required, for example, facemasks and distancing ‐ Centralized and subsidized quarantines the infected and F1 contacts. Home quarantines for F2 contacts	Overall theme: Complacent in COVID‐19 control at local and organizational levels ‐ Local governments promoted economic recovery ‐ Some hospitals loosened the diagnostic and/or treatment procedure ‐ High‐risk activities were loosely controlled, for example, illegal immigrations, gathering in bars, karaoke	Overall theme: Prevention of COVID‐19 is part of everyone's normal activities ‐ Government agencies made plans to reflect the dual objective of economic development and COVID‐19 control ‐ Promotion of the ‘five Ks’ prevention measures, that is, ‘khẩu trang’ (facemask), ‘khoảng cach’ (distance), ‘khai báo’ (reporting), ‘khử khuẩn’ (pasteurize bacterium), ‘không tụ tập’ (no gathering)

Abbreviation: SARS, Severe Acute Respiratory Syndrome.

Phase 1—Aggressive measures, swift control

Extracted cues and frames

The information of COVID‐19 was filtered according to a clear pattern. Specifically, the government attention was directed to cues that signalled potential dangers of the virus. COVID‐19 was initially framed as a ‘new variant’ (chủng mới), ‘dangerous’ (nguy hiểm), and ‘potentially deadly virus’ (virus có khả năng gây chết người). In addition, the weakness in medical facilities was well recognised. These cues pointed to a strong belief that ‘when Vietnam has more than 100 infected cases per day, the system will break down quite fast’ (an in expert of the NCPC). This strong, pessimistic construction of the situation pointed NCPC and government officials at all levels to growing evidence of the deadly nature of COVID‐19.

Similarly, fighting against COVID‐19 was framed as ‘fighting the epidemic as fighting invaders’ (chống dịch như chống giặc), and as a ‘political responsibility’ (nhiệm vụ chính trị) of all public organisations and officials. As the 13rd Party Congress was approaching, this framing meant a mistake or evidence of ignorance in controlling the disease was considered a potential career‐ending for political leaders.

Commitment and actions

The government showed an unequivocal commitment to control the virus. The Prime Minister’s directive No.13/CT‐TTg (March 13, 2020) clearly stated that ‘the first priority is to protect citizens”health’. Vietnamese leaders committed to keep zero death, under 1000 infected cases by April 1st (Deputy Prime Minister Vu Duc Dam on VTV, March 27th, 2020). One implication was ‘to prioritize people’s safety over economic development’. For local leaders, this also meant that overreaction was better than inaction. As a local official put it, ‘in this COVID‐19 fight, you don’t want to take any risk’.

In this phase, actions to control COVID‐19 were quick, drastic and unequivocal, albeit costly. ‘The most important indicator of success is speed’, a member of NCPC believed. Government at all levels and public organisations were required to make scenario plans to cope with COVID‐19. Based on experience of SARS in 2003, a series of actions were introduced, including requiring citizens to use facemasks, preparing quarantine centres, and fully locking down areas that had infected cases. In early February 2020 these measures were highly controversial and had not been recommended by the WHO. Many by‐effects of these measures, such as excessive use of medical facemasks or psychological effects of lockdowns, have not been sufficiently evaluated.

Identities

During this phase, stakeholders' identities were redefined in coping with COVID‐19. First, the State shifted back to its traditional socialist role of ‘taking care of people’s lives'. As this paternalistic role clings to the memories of the country’s socialism, the shift was quick and unprotested.

Second, public organisations also redefined their identities to take unconventional roles in handling the epidemic. Medical organisations and staff became advisors, communicators, and monitors to help citizens maintain calm and follow preventive measures. Military forces became motel runners and service providers where they arranged lodging, food, and all necessary supplies for quarantined people. Police became tracers of potential virus‐carriers, tracking down people who had been in contact with infected patients.

Phase 2—Complacence

The initial results of the first phase were highlighted in the Prime Minister's report to the Assembly meeting on May 19, 2020. While the report warned people ‘not to let our guard down (chủ quan)’, the overall tone was quite positive and optimistic. The Prime Minister reported 324 infected cases, zero deaths, and used such evaluative phrases as ‘we have basically controlled the epidemic’, ‘initial victory’, or ‘important victory’ to refer to the fight against COVID‐19 in the first phase.

At local levels, the attitude increasingly moved towards ‘success in COVID‐19 control’ rather than ‘caution with the virus’. Signals of potential COVID‐19 infected cases were largely lost, in contrast to the first phase. As a case in point, patient number 416 visited a hospital in Da Nang several times during July 8th–20th before being diagnosed with COVID‐19th on 23rd July. The authorities confirmed this case only after four consecutive tests that all showed a ‘positive’ result for the virus. As one medical staff in Da Nang reflected, authorities at that time ‘could not believe that there would be an infected case in the community’.

The frame of ‘success’ in controlling COVID‐19 in the first phase triggered a commitment to make up for economic loss. The Prime Minister emphasised several times (in May 5th and 19th) the need to recover the economy. At the same time, there were signs of loosening control of the virus, evidenced with the re‐opening of bars, karaoke or international commercial flights. The control of illegal immigration was less strict than in the first phase. A statement from the Da Nang leader recognised some complacence in COVID‐19 control1: ‘the epidemic broke out in two major hospitals…This was a general subjectiveness and we had drawn a lesson’.

In this phase, people gradually reverted to the ‘old normal’, that is, old mandates and routines. A popular perception that ‘we are the winner’ in fighting COVID‐19 was evidenced in media reports as well as people's attitudes. Experience of swift success in controlling SARS (2003) reinforced this attitude. Memories of the virus became increasingly blurred as days went by, until July 25, 2020.

Phase 3—Strong measures, long‐term control

The confirmation of patient 416 on July 25th, 2020 started a new phase of COVID‐19 outbreak in Vietnam. The NCPC quickly shifted back to the frame that the virus was ‘dangerous’, ‘complex’, and ‘uncertain’. However, the NCPC recognised that it was not possible to stop this virus as swiftly as with the SARS outbreak in 2003. As the Deputy Prime Minister remarked at the NCPC meeting (July 25, 2020), ‘fighting against COVID‐19 is a long‐term battle’. Media and government propaganda conveyed the message that ‘we have to live safely with the virus’.

Recognition of the persistence of the pandemic triggered a revised approach in coping with COVID‐19. On the one hand, the control of COVID‐19 remained timely and even aggressive in specific locations where an infected patient was diagnosed. The use of group tests and location‐based quarantines helped reduce the costs. On the other hand, the overall approach was revised to ensure long‐term commitments. First, the dual objectives of controlling the virus and maintaining economic growth were enclosed in the plans of line ministries and local governments. Second, the government cut subsidies relating to several expenses. Third, ‘zero deaths’ was no longer seen as a feasible or reasonable commitment in the control of COVID‐19. However, the approach still aimed at eliminating the virus through ‘locating, tracing, and quarantining’ the infected cases. Until the end of 2020, the government had been equivocal in its plan to access vaccine.

In this phase, people had a clearer sense that they need to fulfil their mandates while taking actions to prevent the spread of COVID‐19. According to the NCPC's instruction (on July 28 and from related meetings), hospitals resumed and were repeatedly required to review COVID‐19 prevention protocols. State agencies, businesses and the public were urged to include COVID‐19 prevention measures to their operational routines. The ‘Five Ks’ measure was an example (See Table 3).

The roles of local governments in the creation of collective sensemaking

An important issue in the control of COVID‐19 in Vietnam was the creation of a collective sensemaking that resulted in shared meanings and concerted actions in the first and third phases. In this case, ‘shared meanings’ did not indicate completely overlapping or agreed‐upon understandings, but these were close enough to trigger concerted actions. We discuss three mechanisms and the roles of local governments in fostering collective sensemaking at the national scale in Vietnam.

Propaganda and participation

In phase 1 and 3, getting people to cooperate with the strong control approach was the ultimate objective of the propaganda. Our data has shown several characteristics of the propaganda on COVID‐19 control in Vietnam. First, the messages on COVID‐19 were communicated with high intensity and frequency in all localities. A variety of media was used, including the once‐abandoned ‘communal loudspeakers’. Some people did feel ‘overwhelmed’ or ‘tired’ by the messages. Others recognized that the intensity of the communication made them believe that the danger is real. Second, potential dangers of the epidemic were emphasized. This created a fear of the disease which in turn induced supports for aggressive actions. Third, the propaganda was primarily top‐down and action‐focused. National experts who had experienced with similar epidemics, such as the SARS outbreak in 2003, frequently appeared in the media to share their purviews on the disease. The government's metaphor ‘fighting the epidemic as fighting invaders’ rekindled people's memories of war times and triggered societal instinctive reactions that supported the aggressive control approach in the first phase. LCPCs employed contextualised communication strategies, such as using local languages and respectful figures to get the messages to ethnic communities.

The propaganda programs thus contributed to social participation from citizens and businesses in the first and third phases. First, citizens acted as members of local information network, reporting any ‘suspicious’ cases of infection and monitoring the compliance of quarantined people. Second, citizens actively contributed resources to the COVID‐19 control program through various social outlets. As a case in point, by May 10th citizens had contributed about 2,000 billion Vietnam dong (equivalent to 90 million USD) to the Fatherland Front's fund to support the COVID‐19 control program. Interviews suggested that these contributions further bound people to the government's program.

Critiques, however, pointed to at least two serious biases in this approach. First, privacy right was violated in many cases where details of infected patients were widely published in many local public media outlets. Second, this strong but reactive control approach may delay possible access to vaccines as it focuses too heavily on ‘tracing the virus’.

Coordinated actions and knowledge sharing

The second mechanism of collective sensemaking in the case is enactment, that is, meaning‐making through actions. The government urged people to act as early and quickly as possible.

This is translated to a generally strong coordination within provinces. One district leader put it, ‘we asked people to delay their opposition [to government measures] because there is no time to argue’. The scenario planning practice drove authorities' attention to possible actions. It also facilitated coordinated actions, especially when infected cases were diagnosed. A member of NCPC commended that scenario planning helped people better understand their superior's constructed meanings of the disease.

Actions of one stakeholder influenced the sensemaking of the others. Aggressive instructions and serious attitude from superior agencies were transferred to local authorities. In turn, aggressive, swift, and concerted actions at the grassroot authorities convinced citizens of the danger of the epidemic. As a F2 citizen put it, ‘seeing such seriousness [in authorities' responses to COVID‐19], I thought the danger must be real.’

Coordinated actions were enabled by an intensive use of cross‐functional and cross‐level coordination within provinces; in fact, the LCPCs comprised leaders across different functions and mass organisations. During the epidemic, online cross‐level meetings were organised quite frequently. For example, in Hanoi, an online cross‐level meeting was held at least every two days, involving Prevention Committees from the City, 30 districts and 584 communes. These meetings served as a forum for updating information, giving instructions and exchanging different viewpoints across levels. Local extreme preventive measures were also discussed and corrected at these meetings.

Cross‐functional coordination was also apparent in the Emergency Task Forces (ETFs) and management of quarantine centres. Each commune or residential building established an ETF that was responsible for responding to infected cases. The ETF comprised experts from different functions, including local medical staff, police, the military, and mass organisations. If there was one infected case, the ETF would inform all citizens in the area, lock down the building or street/village of the infected case, trace all direct (F1) and first‐level indirect contacts (F2) of the infected case, organise the quarantine and frequently test the F1 and F2 cases.

By contrast, coordination between provinces was problematic. Variations in provincial implementation of the measures created difficulties for inter‐provincial transportation and travel. Some provinces went for extreme measures, such as blocking interprovincial roads to limit the travel in and out of the provinces. As a result, supply chains were broken and people's right to travel was breached at times. One good practice that helped foster collective sensemaking is knowledge sharing across localities. Experienced staff members were sent to help newly diagnosed ‘epidemic areas’. When the epidemic broke out in Da Nang, medical staffs from Hanoi and Ho Chi Minh City, who got experience in COVID‐19 control in phase 1, were sent to help. Besides technical supports, the presence of these experienced staffs bolstered citizen's confidence in the COVID‐19 handling of their local authorities.

Sense‐bridging of LCPCs

Our data suggests that LCPCs frequently acted as sense‐bridging between different levels of administrations. Sense‐bridging refers to the act that LCPCs tried to make sense of how others make sense of the situations and then found a way to coordinate actions. For example, in phase 1, a member of a district CPC noticed different views between the provincial CPC and his district chairman and communes. While the central and provincial CPC were very serious about the potential danger of the virus, the chairman and communal authorities in the district tended to be quite relaxed.

… I then talked to the District Party’s secretary about my concern, citing the SARS 2003 as an example of how fast the virus could travel. The Party’s secretary then told the Chairman: ‘you are responsible for this district. I would take Dr. C’s words seriously.’ The Chairman then listened to my advices.(Quoted from a director of district healthcare center)

In another example, a member of a district CPC noticed some extreme propaganda activities in communes as in the following quote:

Some communal authorities asked people to collectively tidy‐up their communal and village roads [as a measure to prevent Covid‐19]. They viewed COVID‐19 control as a political agenda to promote their own image. Such type of activities was dangerous as it may facilitate the transmission of the virus. We had to correct this kind of activities.(Quoted from a member of district CPC)

In the first example, the member of district CPC used political frame to drive his Chairman's attention to the urgency of COVID‐19 issues. In the second example, the local CPC member had to drive people away from politicalising COVID‐19 actions. In both examples, members of LCPC interpreted how others viewed the epidemic before they could find a way to coordinate.

Theoretical model of collective sensemaking under crises

Based on the case study, we developed a concept of IPP that facilitates initial and collective sensemaking in crises. Institutional propensity for prevention refers to the readiness to engage with potential societal problems before they arise or become acute, as opposed to when they are damaging and expensive. It also reflects the tendency to implement state‐driven, aggressive control measures that focus mainly on ending the pandemic, as opposed to more citizen‐centric, multi‐objective control approach.

Our data suggest that a country's disease IPP depends on three key factors, including experiences in coping with epidemics, a nation‐wide disease prevention system, and a political agenda that promotes transparency and cooperation. At the national level, the country has extensive experience in coping with epidemics. A long history of wars has trained people in top‐down, strong control coordination. Recent epidemics, such as the SARS (2003), H1N1 (2009) and H5N1 (2009) further strengthened the country's coordination capacities. Second, the maintenance of a nation‐wide disease prevention system facilitated the strong control approach. This multi‐layer prevention system was rooted at every commune, and the communal medical staff members knew almost every household in their area. These local medical staffs participated actively in the government's COVID‐19 control program. As capacity influences sensemaking (Weick, 1988, 2005), prevention capacity would bias people towards strong control measures. The third factor, political agenda, refers to short‐term political concerns of individual leaders. The year 2020 was critical for government officials at all levels as the Party's Congress was approaching. Successful control of Covid‐19 would be necessary for their career advancement.

At the local level, provinces vary in their epidemic experiences and disease prevention capacities. In addition, provincial leaders have different advancement opportunities. These factors contributed to differences in provincial IPPs, which induced variations in the creation of collective sensemaking and implementation of Covid‐19 control measures.

Figure 1 presents the theoretical model we developed from our findings. At the national level, the IPP enables risk‐oriented sensemaking, that is, noticing cues of potential dangers, engaging in early and drastic preventive measures, creating collective sensemaking and updating meaning in light of changes. In turn, risk‐oriented sensemaking fosters results of crisis management, at least in the short‐run.

FIGURE 1

Process model of collective sensemaking under epidemic

First, the country's IPP influences the cues that get noticed. Vietnam's IPP drove people's attention to cues that signalled potential risks of the disease. These factors were then blurred in the second phase when the experience of coping with a long‐lasting virus was not vivid and the need for economic recovery was considered more important. Second, an IPP influences the substance of sensemaking. In our case, the IPP triggered commitments to strong control at the price of economic development and certain basic human rights. The resulting sensemaking—that is, commitment, identities and expectations—oriented the NCPC towards early and drastic reactive measures but may also delay longer‐term solutions, for example, access to vaccines.

At local levels, variations of IPPs across localities help explain differences in the process of collective sensemaking. Localities with extensive epidemic experiences and strong prevention capacities, such as Hanoi, Quang Ninh, Hai Phong, were more attentive to the danger of the virus and took stronger measurements more aggressively. Some of their measures were viewed as ‘extreme’, such as blocking the inter‐provincial roads or requiring more paperwork for people entering the provinces. By contrast, localities with less epidemic experiences and coordination tradition, such as those in the Mekong Delta region, tended to relax and faced more resistance in enforcing the strong approach. Leaders' political agenda is another factor that influences the creation of collective sensemaking and the implementation of the strong control approach. Local leaders with brighter career advancement would be more motivated to follow central government's sensemaking and epidemic control approach. As a case in point, Hanoi was very aggressive in enforcing the strong control approach in Stage 1 (before April 15th, 2020) when the City's Chairman Nguyen Duc Chung maintained a hope for a future political career. However, after Mr. Chung stepped down for his past wrongdoings, his temporary successors exerted a much lower level of coordination in the implementation of Covid‐19 control in the second Stage.

DISCUSSION

In this paper we have addressed the question of how sensemaking is made, shared and updated during crises and how local governments contribute to the process. We used the control of COVID‐19 in Vietnam as a case study and relied on government documents, real time interviews with stakeholders and our own observations to analyse the creation of initial and collective sensemaking in coping with the epidemic.

Our results suggested that initial and collective sensemaking during crises are influenced by an IPP that has been gradually developed over times. In the case of Vietnam, the control of COVID‐19 was characterised by the government's early attention to the danger posed by the virus, a strong reactive, state‐driven control focus, and an aggressive construction of shared understanding and concerted actions across the whole nation. While achieving relatively good results in the short‐run, this approach also showed biases in sensemaking, including breaches of privacy and a possible late access to vaccination as a long‐term measure.

There were several limitations of this study. First, as our single‐country study did not allow us to compare the sensemaking processes in different institutional settings, our insights on the roles of institutions in sensemaking should be treated as exploratory. Second, while real time interviews allowed us access to the thoughts and feelings of the interviewees at the time of the epidemic, access to high‐position interviewees was not possible. We could only infer the sensemaking of the leadership indirectly from their decisions, statements, and subordinates' accounts. Despites these limitations, our study offers important implications for policy and research.

Theoretical implications

Our study contributes to sensemaking perspective by suggesting that people's sensemaking and associated biases are conditioned by surrounding institutions. Despite the provocative discussion of Weber and Glynn (2006), the role of institutions in sensemaking remains understudied (Maitlis & Christianson, 2014). Our study suggested that an IPP drives people to notice potential risks and quickly commit to strong preventive actions, which appears to be immediately effective. The inherent biases of this sensemaking include high costs, extreme and irregular actions, and uncertainty regarding long‐term sustainability.

Our paper also highlights the roles of local government in fostering collective sensemaking under uncertainty. Our model suggests that local governments vary in their use of sensemaking mechanisms, including social propaganda and participation, coordination and knowledge sharing, and sense‐bridging. This variation could be attributed to a local IPP which consists of local epidemic experiences, prevention capacity, and local leaders' political agenda. Future research could further examine how a province's IPP influences its government's responses to epidemics. For example, it could be hypothesized that the state‐led, aggressive approach would be more likely adopted by provinces that used similar approach in controlling past epidemic and those with strong prevention capacity, compared to those with less experience and capacity. Similarly, this state‐led, aggressive control could be enforced more strongly by provinces whose leaders have longer term(s) in office or greater opportunities for career advancement.

Epidemic update and policy implications

Vietnam's achievement in controlling Covid‐19 in 2020 did not last through 2021. We argue that the context of 2021 caused a big change in the formation of initial and collective sensemaking on Covid‐19. At the national level, political commitment was particularly strong in organizing offline Party's congresses and the Parliament election at the end of May. These crowded meetings went against the country's ‘strong control approach’ spirit. The achievement in 2020 further delayed the country's access to vaccines. By May 2021, Vietnam was among the least vaccinated countries. In addition, past achievement also drove actions to follow practices in 2020, that is, ‘locating, tracing, and quarantining’. This approach was neither effective with the Delta nor with the Omicron variant. At the local level, the creation of collective sensemaking was seriously challenged in 2021. Newly promoted local leaders did not have experience in Covid‐19 handling. They mainly followed past practices which proved to be rather ineffective with the Delta variant. As numbers of infected and death cases mounted, coupled with the exhaustion from following the ‘strong control approach’, public trust in local government eroded. This triggered a change in the country's approach to ‘living safely with the virus’ which is in many ways similar to approaches in the US and European countries.

Our study offers several policy implications. First, fostering collective sensemaking is a critical success factor for coping with crises. Our paper recommends governments at all levels to build horizontal and vertical coordination capacities. An application of technology, such as online video conferencing or real time data reporting software, would facilitate the coordination process. However, people's willingness to exchange information and perspectives is most important in the coordination. This can be nurtured through official training programs and/or frequent practices in normal conditions. Second, developing an IPP is a critical factor determining success in handling crises. The IPP depends largely on the society's sensitivity to risks and the existence of prevention systems. Given that ‘prevention’ often seems to be vague and inefficient (Cairney & St Denny, 2020), it needs to be converted into specific policies and programs such as a nationwide preventive healthcare system. Nurturing societal sensitivity to risks and the capability of large‐scale coordination requires serious learning from past experiences. However, short‐term political agenda may seriously distort the effect of these long‐term investment in IPP.

Local government plays a critical role on creating collective sensemaking during epidemics. Frequent cross‐level and cross‐functional meetings, strong social participation and local initiatives are good mechanisms for sense‐giving and sense‐updating. However, biases and extremeness are inherent in sensemaking and need to be corrected as the crisis unfolds. In that process, local leaders' political agenda could hinder the learning and correction of biases.

Supporting Information S1

Click here for additional data file.