Accountability in times of exception: an exploratory study of account-giving practices during the early stages of the COVID-19 pandemic in Italy

Purpose – The purpose of this paper is to explore the role of accountability in times of exception. The Italian government ’ saccount-givingpracticesarecriticallyanalysedwithrespecttothedistinctmodesinwhichduties of accountability are discharged for the exceptional measures taken during the early stages of the COVID-19 pandemic outbreak in early 2020. primarilyondataobtainedthroughpubliclyavailabledocumentsandcoverstheperiodbetweenJanuary1andAugust7,2020. Findings – The paper reveals that the Italian government employed various accountability styles (rebuttal, dismissal, reactive, proactive and coactive). Each style influenced both how the government justified its conduct and how it sought to form distinctive relationships with social actors. Originality/value – The paper uses the notion of “ styles of accountability ” to empirically illustrate how an unprecedentedpublicgovernancechallengecanrevealbroaderaccountabilitytrends.Thepapercontributestoaccountabilityresearchbyelucidatinghowgovernmentstackleambiguityanduncertaintyintheirsystemsofpublicaccountabilityinextraordinarytimes.


Introduction
The COVID-19 pandemic has resulted in a global health crisis that affects all aspects of society (World Health Organization, 2020a, b). The speed of the pandemic's global spread has led many national governments to declare a state of emergency both to allow them to Accountability in times of exception thoroughly understand the situation and to protect the people. As the disease continued to spread, this state of emergency was characterized by the adoption of various measures, including the limitation of individual freedom, the diversion of goods and services, the closure of public and private facilities and the use of mass-surveillance technologies (Ahrens and Ferry, 2020;Andrew et al., 2020;De Villiers et al., 2020;Nemec and Spa cek, 2020). While they differ in terms of their implementation, these measures are exceptional within most governments' legal frameworks (Nay, 2020).
The spread of the COVID-19 pandemic has required urgent reactions from national governments that, at times, have circumvented the usual deliberative processes. However, while accountability studies have become more sophisticated, little attention has been paid to how accountability is shaped by (and also shapes) unparallelled times such as these. The paper focuses on the Italian government's account-giving practices during the early stages of the COVID-19 pandemic (manifested by the communication between government officials and the wider public) and empirically analyses how the pandemic's evolution was related to accountability processes in response to abrupt, conflicting and extraordinary public governance challenges. In times of crisis, such as the COVID-19 outbreak, public governance arrangements play a critical role in governments' ability to respond immediately to contain the damages but also to recover and rebuild. For instance, governments are required to formalize and implement policy options and governance models that promote resilience investments such as health centres, test capacity and track and trace systems. Risk governance and crisis management models, however, must rely on digital infrastructures and live data repositories for evidence that can shed light on how different areas, sectors or individuals behave and implement the necessary regulations. At the same time, regulations adopted through fast-track procedures must comply with extant systems of law and must be subject to careful postimplementation reviews. These challenges illustrate the importance of defining principles that are broader than efficiency or cost-effectiveness to include governments' ability to uphold core values such as transparency and equity. Understanding how duties of accountability come to be discharged in such exceptional times, therefore, has important implications for understanding the potential and limitations of the public governance arrangements that are set out. The COVID-19 pandemic has highlighted how these processes of rationalizing conduct have become more sensitive to the emerging demands of health, safety and security. Consequently, novel processes of accountability aimed at aligning organizational practice with wider societal discourses can be seen to emerge (Ahrens, 1996).
This paper aims to increase knowledge and understanding of accountability in times of exception. Drawing on the notion of styles of accountability, this paper classifies and analyses the ways in which governments' emergent reactions to the COVID-19 challenge can reveal wider accountability trends. This paper contributes to the accountability literature by showing that during times of exception (Agamben, 2005), the emerging "styles of accountability" do not depend so much on cultural determinism or the exigency of aligning "rhetoric and practice with wider public discourses" (Ahrens, 1996, p. 140) as on the need to enhance "shared responsibility" (Sciulli, 2018). The existing literature has shown how governments have mobilized political capital to alter the role of accountability in times of crisis (Demirag et al., 2020). This paper builds on this research by critically analysing the mechanisms through which the Italian government sought to combine public governance and public accountability issues as a means of urging shared responsibility. The notion of styles plays an important role in enhancing our understanding of how shared responsibility is promoted during various phases of the COVID-19 pandemic. For example, people living in areas in which the virus is actively circulating should know the restrictions applied to the populations when the number of active cases is proportionally increasing (i.e. mandatory testing or self-quarantine for travellers; tighter mobility regulations and reduced social interaction). They should also be familiar with the criteria that governments use to determine whether such measures are appropriate. However, while restrictive measures are developed by experts from health authorities in conjunction with government representatives, trade unions and business associations, it is ultimately the population that makes the final decision as to whether or not to follow the recommended protocols. That is, people build their knowledge bases regarding the pandemic's evolution through the justification of operational proposals before deciding to take on (or not) a shared responsibility. The analysis reveals five styles of accountability (rebuttal, dismissal, reactive, proactive and coactive), with each style underlining the different ways in which accounting is used within the context of extraordinary public governance challenges.
The remainder of the paper is organized as follows. The next section will briefly introduce Agamben's "state of exception," which has been used to frame the paper's conceptual underpinnings. It will also review key accountability scholarship to identify research gaps. Section 3 explains the research design, while Section 4 presents the analysis. The final section discusses the findings and draws conclusions.

States of exception and styles of accountability
States of exception have been framed as a government's immediate response to extreme and imminent threat. The concept has recently been studied by Agamben (2005) and earlier by Schmitt (1922), and it is used in this paper as a heuristic to frame the context in which accountgiving practices occurred. This state is characterized by an extraordinary condition, wherein the conventional order is suspended. One of the key tenets of the state of exception concerns the indistinction between the private lives of citizens and the public sphere. This governmental process results in the transformation of the individuals' identity from citizens of a state to subjects over which the state has complete authority (Agamben, 2005;Foucault, 2008).
States of exception entail exceptional measures resulting from a state of crisis that includes a political dimension and may be described as "a point of imbalance between public law and political facts" (Saint-Bonnet, 2001, p. 28 as cited in Agamben, 2005. These exceptional measures can range from suspension of the constitution and assumption of full powers by government to the loosening of the distinction between different powers (legislative, executive, etc.). Fundamental to explaining the existence of a state of exception is the theory of necessity, which implies an exception (dispensatio) with respect to the current legal order aimed at ensuring the "well-being of men" (Agamben, 2005, p. 25). Agamben (2005) characterized a state of exception as "an empty space" (p. 86), implying several risks that should be managed and addressed. This empty space must be filled with unconventional and unused practices or with known practices that assume new meanings and forms. Among these practices, accountability plays a key role, as the suspension of the usual rule of law can create uncertainties regarding the attribution of responsibilities to specific actors regarding key issues, such as human rights protection (Welch, 2007) regarding healthcare for frail and vulnerable people.
The state of exception concept has rarely been invoked in accounting and accountability studies. From an institutional perspective, systems of accountability (Roberts and Scapens, 1985) become not just subjective formulations but rather socially constructed phenomena that provide actors with a common framework for the shared understanding of social reality, and consequently they can be used by governmental actors to construct a shared understanding of social reality.
Analyses of accountability have long been developed in accounting research, yielding considerable insights into the competing types of accountability logics and rationales (Baker, 2014;Dillard and Vinnari, 2019;Everett and Friesen, 2010;Grossi et al., 2019;Rana and Hoque, 2020;Sargiacomo, 2015;Sargiacomo et al., 2014;Sargiacomo and Walker, 2020;Sciulli, 2018;Walker, 2016). Indeed, demands for increased institutional accountability have been Accountability in times of exception regularly framed to assume that it is always and unambiguously desirable (Munro and Mouritsen, 1996). In this realm, the importance of accountability topics has often been considered in terms of the volume of disclosure related to the disclosed issues (Beattie and Thomson, 2007;Unerman, 2000). A connection between the state of exception and accountability has been established in risk management studies, whereby accountability has been considered capably of reducing or amplifying risks when an empty legal space exists (Huber and Scheytt, 2013). The state of exception poses challenges and requires a reconsideration of the risk management of uncertainties at the institutional level (Tan and Enderwick, 2006) and the reconsideration of accountability as a critical issue.
During states of exception, accountability may be connected with the specificities of leadership and culture or with the adoption of tailored plans to face the emergency (Wilson, 2020). In effect, the spread and impact of SARS-CoV-2, the virus that causes COVID-19, is here to stay (Kissler et al., 2020) and the idea of exception in relation to the pandemic may require that accountability be articulated in different ways, depending on the spread of the virus within a specific context to address protean needs. One stream of this body of the literature portrays accountability as a subjectively constructed notion that changes with context (Garfinkel, 1894;Pesci et al., 2020) and consequently can translate into different ways of being shaped. This literature studies the many forms that accountability may take, whose dimensions of meaning depend on the social institutional conditions in which relations of accountability are enacted and experienced (Sinclair, 1995).
The changing nature of accountability has been well documented in the area of management accounting. Ahrens (1996), for example, introduced the notion of the "style of accountability," defining it as "a heuristic device to explicate some of the ways in which [. . .] notions of "good management", to which organisational members hold themselves and each other accountable, can be implicated in the shaping of very different roles for management accountants and their practice" (Ahrens, 1996, p. 140). This construct has helped to theorize the specific "ways in which accounting was implicated in processes of accountability" (Ahrens, 1996, p. 170).
This paper uses the concept notion of "styles of accountability" to empirically illustrate the ways in which an exceptional public governance challenge can reveal wider accountability trends. In this empirical context, each style is embedded in a distinct mode in which the Italian government has discharged its duties of accountability for the exceptional measures taken as well as in the individual actions and individual responsibility required for the measures' success. In particular, the main features of possible styles of accountability related to these modes can be inferred by referring to the literature that, in contexts where humanitarian issues are involved, tries to point out how accountability has been and/or should be shaped. Sargiacomo (2015) used Agamben's (2005) notion of states of exception to examine how calculative practices and classification systems helped guide emergency responses to the 2009 earthquake in Italy's Abruzzo region. The study highlighted that, during the state of exception, "provisional exceptional measures are transformed into "a technique of government"" (Sargiacomo, 2015, p. 70;Agamben, 2005, pp. 2-3). The evolution of calculative practices and the exceptional measures that translate in government techniques may be regarded as key features in relation to the development of different styles of accountability during a state of exception.
Theoretical enquiries have also challenged some positions set out in the existing accountability literature. These challenges are represented by the "burden that accountability may place on the accountable self who is expected to provide a convincing account even in situations where this is extremely difficult or even impossible" (Messner, 2009, p. 919). The provision of convincing accounts can acquire particular urgency during times of exception that originate from the need to protect human beings. Nevertheless, established and sound accountability practices for facing such situations are lacking; moreover, accountability practiceseven those relating to humanitarian fieldshave been criticized for their focus on accounting only for positive performances and for their lack of focus on true compassion. Thus, an examination of the apparent lack of accountability in such contexts can help in understanding which elements accountability should encompass in states of exception relating to humanitarian issues. In particular, the literature (Everett and Friesen, 2010) suggests that a proper accountability style when dealing with humanitarian needs should focus on sentiments such as "compassion". The word "compassion" derives its original meaning the Latin cum pateo, which suggests identification with the pain of others, which is particularly crucial when humanitarian needs are at stake.
Similarly, some authors have noted that accountability during times of exception may not be effective owing to the excessive emphasis on technical devices and calculations that are unable to assess the impact on responsibility. Baker (2014), for example, studied breakdowns in accountability during and after Hurricane Katrina and concluded that governments "relied to an excessive extent on a calculative accountability. . . instead of a calculative accountability they should have relied more on the "potential of accountability to enhance levels of responsibility for the other"" (Baker, 2014, p. 621;McKernan, 2012, p. 259). Consequently, accountability styles in time of exception should also include the development of a linkage between calculative accountability (or accounting practices) and their effective use in enhancing responsibility.
Other papers have investigated accounting in the context of disasters that determine a state of exception, emphasizing the role of individuals who participate in the accountability process (Sargiacomo and Walker, 2020) and introducing the notion of "shared responsibility" (Sciulli, 2018), wherein each actor "has a responsibility for their well-being" (Sciulli, 2018, p. 42). Sciulli's (2018) work focuses on the fact that "shared responsibility" must be communicated to the community to allow actors to assume an active role in public risk governance (Sciulli, 2018, p. 41). The individual participation in the accountability process that culminates in a shared responsibility could be considered a key desirable feature of the styles of accountability that may develop during a state of exception associated with humanitarian issues.
In sum, these studies indicated that accountability could constitute one technique for enhancing citizens' responsibility in facing humanitarian problems in times of exception. Certainly, the difficulty in translating accountability into effective responsibility at the individual level suggests that in order to build styles of accountability that can be considered effective in time of exception, it is fundamental that an approach reaches the individual sphere to enhance a shared responsibility.
Notably, the COVID-19 pandemic has not had the usual temporal characteristics of disasters (which tend to occur within a limited period and have a clear beginning and end) and consequently, in a similar context, it is argued that different styles of accountability are developed and evolve over time. The following section sets out the research design.

Research design
This paper is based on a single exploratory case study that examines the Italian government's account-giving practices during the early stages of the COVID-19 pandemic. This approach was selected for the flexibility it allowed in acquiring new insights into the emerging phenomenon under investigation (Yin, 2004). Italy represents a suitable context for this analysis because it was the first European country to be significantly impacted by SARS-CoV-2 and thus had to cope with this unprecedented situation before other Western nations (Boccia et al., 2020;Molinari, 2020).

Accountability in times of exception
The analysis addressed by this research covers the period from January 1, 2020 to August 7, 2020. The starting date was chosen because it represents the development of global awareness that China was experiencing a new and dangerous health crisis. The end date coincides with the date on which the Italian government issued a decree outlining economic support measures to assist the country's recovery and resumption of activities. It also represents the conclusion of the first cycle of the state of exception (Agamben, 2005) caused by the spread of COVID-19. In this first cycle, it is possible to observe a progressive development of norms and calculative practices in response to the spread of the pandemic, the institutionalization of a shared responsibility among all citizens and the first results of the application of norms and the assumption of shared responsibility, thanks to which the gravity of the health emergency fades, leaving more room for economic recovery measures.
This case study analyses publicly available government data, including reports, press conferences, websites, legal acts and speeches. The classification of the data focused primarily on the material obtained through analysis of the press conferences held by the Prime Minister and the Italian Civil Protection Department (hereafter CPD), which were subsequently complemented by, and integrated with, other material. The CPD supports the Italian Prime Minister and the Italian government in coordinating all national resources to protect the population in the event of a serious emergency. Within this study timeframe, the CPD arranged 68 press conferences, while the Prime Minister arranged 26 press conferences, thus developing the Italian government's accountability to the population during the period under analysis (see Appendixes 1 and 2).
The researchers transcribed all press conferences given by the CPD and Prime Minister using publicly available online subtitle extractor software and downloaded video recordings. The transcriptions of the 94 press conferences served as input for the coding process, resulting in five overarching phases consolidation, each representing a separate analytical dimension of enquiry and, consequently, a different style of accountability. The content of the conferences was classified according to the development of specific issues identified by the literature: the development of norms and calculative techniques (Sargiacomo, 2015); the volume of topics related to informing the population on the spread of the virus (Beattie and Thomson, 2007;Unerman, 2000;Baker, 2014); the use of rhetorical language addressed to foster compassion (Everett and Friesen, 2010;Baker, 2014) and the exigency of sharing responsibility (Sargiacomo and Walker, 2020;Sciulli, 2018). The study's inductive nature permits the detection of differences in the development of the aforementioned issues and leads to the definition of five styles of accountability: rebuttal, dismissal, reactive, proactive and coactive.
In addition, a quantitative content analysis supported the definition of the five styles of accountability. This analysis was performed using the downloaded press conference videos to capture duration, topics discussed, speakers and the number and length of questions allowed. The number and length of the press conferences was a key object of investigation in this study as a proxy for the "volume" of disclosure, which is considered in the literature as directly linked to the importance attributed to a certain topic (Beattie and Thomson, 2007;Unerman, 2000). Analysis of the videos yielded output on how many minutes speakers spent on each discussed topic. From the Prime Minister's conferences, six main topic categories emerged (health, safety, economics, national policy, education and foreign policy), while, from the CPD's conferences, six categories emerged (health, safety, economics, national policy, education and foreign policy). This approach allowed the researchers to obtain a first impression of the importance of communicating COVID-related information. In this regard, particular emphasis should be placed on the repetition of forms of communication and concepts that are considered able to "impress" the information's receiver (Pesci et al., 2015).
The second source of data was the public online databases through which the Ministry of Health and the CPD updated data on the evolution of the pandemic. This database contains data on the spreading of the infection, such as the epidemiological curve and data on the virological testing undertaken. These conferences and databases were considered the main accountability tools adopted to inform the Italian population regarding the state of exception arising from the COVID-19 pandemic (Agamben, 2005). Other documents that were analysed included two reports from international health bodies (European Centre for Disease Prevention and Control (ECDPC), 2020; WHO, 2020a) and one report from the Italian National Social Security Institute [Istituto Nazionale della Previdenza Sociale (INPS), 2020].
The last source of data was the legal acts enacted by the Italian government and Parliament during the period analysed. A total of 30 regulatory sources were downloaded and summarised (see Appendixes 1 and 2).
The following section presents the analysis of the five phases in the context of the styles of accountability heuristics (Ahrens, 1996).

Empirical analysis and discussion
This section examines how the CPD discussed the circumstances surrounding the evolution of the COVID-19 pandemic by revealing five styles of accountabilityrebuttal, dismissal, reactive, proactive and coactive.
In particular, the different styles of accountability show a progressive development along some issues: (1) Techniques of governance (the CPD's role and technicians' role) and accounting (development of dashboard showing the numbers of the spread of the pandemic) that, as suggested by Sargiacomo (2015), are tools for governing the state of exception; (2) The volume and topics of accountability devoted to informing the population on norms and reasons for them (Beattie and Thomson, 2007;Unerman, 2000;Baker, 2014); (3) The willingness to establish an emotional link with the population (Everett and Friesen, 2010;Baker, 2014); (4) The development of a concept of shared responsibility between government and citizens (Sargiacomo and Walker, 2020;Sciulli, 2018).
The fact that a style of accountability embeds points 3 and 4 represents an evolution of that style in contrast to those that do not embed such points because, as previously explained, the failure of accountability systems in the humanitarian context has been attributed to the absence of such conditions (Everett and Friesen, 2010;Baker, 2014;Sciulli, 2018;Sargiacomo and Walker, 2020). Table 1 summarises several key findings that help interpret the evolution of accountability in five phases.

Phase 1 -Rebuttal style of accountability
During this phase, documents from international and national bodies, such as the WHO (2020a, b), the ECDPC (2020) andshortly thereafterthe INPS (2020), were made available. In January 2020, the WHO alerted the world to the potential rapid global spread of the COVID-19 infection (Jamieson, 2020;WHO, 2020a), urging all countries to respond actively (The National Post, 2020). The ECDPC also published a report on January 26, stating that "the potential impact of 2019-nCoV outbreaks is high [as] further global spread is likely" (ECDPC, 2020, p. 2), and that "the impact of the late detection of an imported case in an EU/EEA country without the application of appropriate infection prevention and control measures would be high, [. . .] the risk of secondary transmission in the community setting is estimated to be very high" (ECDPC, 2020, p. 2 increasingly receiving global attention, the Italian government remained relatively inert and silent. The only conference convened by the Prime Minister during this phase was organized on January 31 as the inevitable conclusion to the accountability silence of this phase. This style of accountability is best described by the term "rebuttal." In the "rebuttal" style of accountability, political institutions resist justifying the possible implementation of strict measures, such as those limiting freedom of movement or mandating the closure of public and private facilities. In addition, the lack of accountability, demonstrated by the distinct lack of any organized information made available to the population, indicates that the government did not perceive a concrete societal need for information during this phase (Walker, 2016).
This state of inertia was abruptly interrupted on January 31 (the end of the first phase) by the hospitalization of two Chinese tourists in Rome due to an initially suspected (and subsequently confirmed) COVID-19 infection: "Two Chinese tourists are now hospitalized at the Spallanzani Hospital in Rome [. . .] the Coronavirus has reached Italy" (Il Messaggero, 2020). The test results shook the rebuttal phase into an initial less static approach, which resulted in the Italian Accountability in times of exception government ordering the cancellation of all direct flights coming from China (Sole24Ore, 2020). The decision was criticized because of the impossibility of checking people who were coming to Italy from China using non-direct flights. In this regard, Sargiacomo (2015) suggested that accountability action during the first phases of an emergency can assume trivial forms. Late in this phase, the government released the first act specifically designed as a regulatory response to the emergency. The legal form chosen by the government was represented by a "government decree", rather than a Parliamentary Act to facilitate a more rapid response to the exceptional situation. From that moment onwards, government decrees have become the standard system used in governing the state of exception; indeed, exceptional measures have been effectively used as "a technique of government" (Agamben, 2005, pp. 2-3;Sargiacomo, 2015). The unique and brief (4.16 min) press conference organized by the Italian government at the end of this phase marked the initiation of phase 2 with a reassuring message: "we have one of the best health services in the world [. . .] these are the conditions for managing this event in a very positive and transparent way" (Prime Minister's quick press conference, January 31, 2020). This rhetoric of accountability in this initial phase recalls the need to make the narrative conform to the expectations of the receivers, or to align "the rhetoric and practice with wider public discourses" (Ahrens, 1996, p. 168).

Phase 2 -Dismissal style of accountability
Unlike the "rebuttal" style, the "dismissal" style of accountability reflects an acceptance and internalization of the impact of the SARS-CoV-2, albeit in such a way that it did not drastically affect confidence in the government's ability to protect the population from the outbreak.
The second phase began just after the issuing of the first emergency decree ordering that all direct flights from China be cancelled: the moment from which accountability (also associated with the comment, to the just issued decree, during following institutional press conferences) evolved in a new form. During this phase, three important actions were taken by the government, the first of which was the establishment of the Operating Committee (February 5, 2020). The Operating Committee included several experts covering multiple scientific fields. This action constitutes the first development of governance techniques for the state of exception (Sargiacomo, 2015).
The second action was a brief press conference (1.46 min) at the beginning of the phase (February 6, 2020), with the Prime Minister speaking (in line with the "dismissal" style of accountability) of the Italian government's adoption of "utmost precautions and the principle of maximum precaution" to "reassure citizens and protect their health as much as possible" defining the "Civil Protection Department [CPD] an international flagship of monstrous efficiency" (Prime Minister's Quick Conference, February 6, 2020) which has "not been reported suspicious cases because the mechanism works" (Civil Protection Department Press Conference, February 5, 2020). An initial form of dialogue with the population is implemented even if no exigency of enhancing compassion or sharing responsibility is either mentioned or felt (Everett and Friesen, 2010;Sargiacomo and Walker, 2020;Sciulli, 2018).
The third action comprised two press conferences led by the head of the CPD to report to the population on the emergency's evolution. The first COVID-19-specific CPD press conference was held on February 5, 2020 and the second was held on February 18, 2020. These conferences' topics were mainly related to safety issues (e.g. pandemic-spread control measures, social distancing measures and train station, port and airport controls) and policy issues (e.g. the explanation of new governmental laws adopted to halt the spread of COVID-19 in Italy). This phase was characterized by several restrictions, limited to the control of Italian citizens coming from China or resident in China. For citizens coming from China, temperature screening at airports was mandatory. However, the need to test the Italian resident population for coronavirus at this stage was neither considered nor discussed.
The accountability mechanisms translated into several press conferences (Beattie and Thomson, 2007;Unerman, 2000;Baker, 2014) in which some initialbut limitedsafety and policy measures were presented. It is worth noting that, during these press conferences, no questions were allowed, suggesting that the accountability was mono-directional and that "shared responsibility" (Sciulli, 2018) was neither considered nor provided at this stage. This limited account-giving is not unsurprising at this point. As existing research suggests (Ahrens, 1996;Garfinkel, 1984), accountability rhetoric and practice are usually aligned with wider public discourses, which in this period were mainly related to the Chinese context and to how the pandemic was dealt with in China.
Until February 20, 2020, indeed, it seemed that COVID-19 could affect only people that had been in China because the only two hospitalized people to date were the two Chinese tourists in Rome. Consequently, accountability was mainly centred on control measures regarding people coming from China.
However, on the night of February 20, the first Italian was hospitalized in Codogno (near Milan). Until that point, the limited accountability regarding COVID-19 had been provided in an attempt to avert potential concerns due to the pandemic by controlling the population's informational needs with a "dismissal" approach (Walker, 2016).

Phase 3 -Reactive style of accountability
The hospitalization of patient 1 gave rise to the third phase. During this phase, the COVID-19 outbreak began to spread in some areas of Italy (WHO, 2020b), and the government undertook exceptional measures to contain the infection. From February 22, 2020, all schools were closed, and the government imposed mandatory quarantine for all citizens resident in three cities in Northern Italy and for all those people who tested positive for COVID-19, irrespective of the area in which they lived. The mandatory quarantine was extended to 11 cities in Northern Italy on March 1, 2020. The number of people tested for COVID-19 per day in this period was between 3,000 and 4,000 (Corriere, 2020).
Accountability regarding the development of the state of exception (Agamben, 2005) was provided initially on February 22 by a first press conference held by the Prime Minister (37.31 min) along with the Health Minister and the head of the CPD. Five further Prime Minister's conferences were held (in three conferences, the Prime Minister was alone, while in two conferences, he was joined by the Ministers of Health and Economics and Finance). Two additional press conferences with the Ministers of Justice and Economics and Finance were scheduled. The conference with the Justice Minister clarified the urgency of providing a legal basis for the developing state of exception (Agamben, 2005).
The Prime Minister stressed the justifications for the state of exception, leveraging the issue of responsibility: "we are aware of our responsibilities [. . .] we know the imminence of the health issue and the implications on economics [. . .] it is a hard decision and we asked for an opinion to our technical and scientific committee" (press conference with the Economics and Finance Minister, March 5, 2020). Nonetheless, in this early third phase, the accountability orientation was based on the government's assumption of responsibility that was joined and entrusted to the technicians (Agamben, 2005;Sargiacomo, 2015). During this period, a further nine Prime Minister's conferences took place. He was joined in four of these by other Ministers to respond to key issues related to the consequences of the lockdown (i.e. economics and finance, labour and social policies, education and economic development).
From February 22, CPD daily press conferences took place every evening at 6.00 p.m. The daily broadcasting of the CPD press conference reflects an evident increase in the public emergency accountability effort to inform and communicate with the citizens. In the same way, the volume of accountability rapidly inflated during this phase (Beattie and Thomson, 2007;Unerman, 2000). Regarding the content of the public disclosure, the accountability system Accountability in times of exception expanded from being solely the responsibility of the government, technicians and experts to a more participatory approach in which responsibility was discussed with the aim of enhancing the understanding of the main issues and forging an emotional link with people who had become accustomed to the daily conference broadcast (Everett and Friesen, 2010). The accountability system's progressive switch from mere technical to more emotional content was reflected in two aspects. The first is the regular cadence with which the government and the scientific committee addressed the population. The recurring daily broadcast at the same time seemed to establish an almost familiar "tradition" between the emergency managers and citizens. The second aspect is the rhetoric used by the Prime Minister, who emotionally linked his government's operation with the growing sense of trust in the citizens' behaviour by stating "we have made a choice in terms of health policy, very consistent, very linear; we believe that trust with our citizens is the essence of our political action" (Prime Minister Conference Press, 25 February 2020). Furthermore, the Prime Minister aimed to link the public sentiment to trust in him as an individual: "Right now the disputes have no value for me because I have to stay focused with the maximum concentration of my physical mental strength to pursue the goal of protecting the health of Italian citizens" (Prime Minister Press Conference, 25 February 2020). The issue of the population's safety and governmental responsibility developed during this phase, increasing in salience. During this phase, from February 22 to March 6, 42 press conferences were broadcast (with an average duration of 15.43 min). The first conference of this phase included the head of the CPDwho was generally the main speaker and coordinated all the press conferencesthe Prime Minister, the Minister of Health and the head of the National College of Medicine. Additional institutional representatives participated with the purpose of providing more sophisticated justifications for the government's actions to the population (Bovens, 2007).
The topics covered by the CPD's conferences began to include mainly quantitative health data (coronavirus accounting) regarding the spread of the pandemic and qualitative information on the measurement and policies undertaken by the government to counteract it. The progressive intensification of accountability data reporting reflects Sargiacomo's (2015) observation that accounting systems tend to become more technical when the state of emergency consolidates and is structured. Indeed, during this phase, the accounting numbers became increasingly technical/calculative (Baker, 2014) and were mainly related to health data regarding the spread of the infection (see Appendixes 5 and 6), encompassing the epidemiological curve, the number of people currently affected by the virus, the number of patients in intensive care, the number of deaths attributed to the virus, the number of recovered patients and the number of coronavirus tests performed (c. 3,000 on average per day). About half of each press conference was devoted to commenting on these figures. National policy and safety topics that had been the object of accountability in the previous phase were relegated to a minor part of the conference. Other topics also emerged as economic and foreign policy issues began to be disclosed. During this phase, the account-giving during conferences realised by the head of the CPD and by the plethora of experts chosen by the government was followed by questions from journalists representing the wider discourses (Ahrens, 1996;Garfinkel, 1984) related to public opinion.
Health and safety were the main topics discussed in this phase, supporting the government's justification of their decision to restrict citizens' freedom in some regions (Bovens, 2007;Roberts and Scapens, 1985).
During this phase, in which the quarantine of some national areas was enforced, efforts to provide a more structured (Sargiacomo, 2015) and reactive accountability style were observed. This style was built upon the multiplicity of conferences that took place, increased coronavirus accounting reporting loss of lives, the infection rate, the number of hospitalized people, the number of tests performed, etc., during the conferences, and on the expertise of scientists who had been called to participate in the public declarations of the CPD to develop accountability (Ahrens, 1996;Bovens, 2007). The multiplicity of events, people, and explanations created a repetition effect on the public's memory and emotions that helped to manage public opinion (Pesci et al., 2015) and helped to develop the public's emotional connection with the health emergency (Everett and Friesen, 2010;Baker, 2014).

Phase 4 -Proactive style of accountability
During this phase, all regions of Italy were affected by the spread of the coronavirus. On March 8, 2020, the government adopted a new decree (Dcpm 8/03/20) extending the restrictive measures to the entire Lombardy region and 14 provinces in Central and Northern Italy. The following day, the Prime Minister addressed the nation in several conferences to the nation with the aim of providing evidence of the emergency and justifying (Bovens, 2007;Roberts and Scapens, 1985) the tough decisions regarding the state of exception that had led to the lockdown (Agamben, 2005). On March 11, 2020, the government decreed a nationwide lockdown: all professional and private activities, unless essential, were banned.
Before the severe lockdown imposed on March 11, two consecutive Prime Minister's press conferences to the nation had been held (21.27 min on March 8 and 18.38 min on March 9). These conferences were broadcast simultaneously by all main national TV channels. This approach to communication was unusual for an Italian Prime Minister and was intended to have maximum effect on the population by offering evidence of the need to extend the state of exception, thus encouraging the citizens' collaboration (Agamben, 2005). The first conference emphasized the necessity of "sacrifice" and "responsibility" for helping society's most vulnerable (Prime Minister's press conference, March 8). These words were intended to provoke an emotional response and to prepare the population for the sacrifices required from all citizens to build a "shared responsibility" (Sciulli, 2018).
During the second conference, the need to "renounce something for the sake of Italy, of our loved people, parents, and grandparents" was also stressed, along with the need to act "immediately" and to be "collaborative to adapt to the more restrictive laws" by underlining that the government had no option but to act in such a manner (Prime Minister's press conference, March 9, 2020). The Prime Minister's statements included words such as "sake" and "love," appealing to the citizens' emotions and using accountability to enhance feelings of compassion towards vulnerable people (Everett and Friesen, 2010;Baker, 2014). This style of speaking, unusual for Italian politics, aimed to appeal strongly to the public sentiment of a population traditionally reputed to be passionate as a means of establishing an emotional connection and invoking the citizens' sentiments (Everett and Friesen, 2010).
In addition to these two initial impactful conferences, the Prime Minister held five further conferences alone and four others with other Ministers (Minister of Economy, Minister of Labour, Minister of Economic Development and Minister of Education). These press conferences placed particular emphasis on economics, which soon became a key accountability issue. Indeed, economics was the main topic discussed by the Prime Minister at regular intervals, mainly during the press conferences held on March 11, 21 and 28 and April 6, suggesting a proactive attitude aimed at addressing and preventing the economic consequences of the lockdown. On March 7, the CPD made available on its website an online dataset dashboard reporting quantitative information about the spread of the pandemic (See Appendix 3).
The daily CPD press conferences continued and, predictably, focused primarily on health issues. The numbers cited during these press conferences were those available via the online database. Special emphasis was given to the effort to increase virus testing, which implied investments in laboratory equipment and in the recruitment of the necessary professional skills. Another important topic was the need to increase the number of intensive care beds in hospitals.
At this stage, more information types were provided (such as the online database), and it became increasingly evident that the relationship between accounting and accountability Accountability in times of exception systems had become closer, implicating accounting in the processes of accountability (Ahrens, 1996;Baker, 2014). The health experts called to participate in these conferences justified the lockdown as a crucial element aimed at saving lives: "measures undertaken by the government are the crucial element now" (head of the National College of Medicine, CPD's daily press conference, March 10), with the result that they were perceived as both necessary and justified.
Other topics connected with safety, economics and national and foreign policies continued to be the object of disclosure during both the CPD's and Prime Minister's conferences (see Appendixes 4-7). However, the Prime Minister's conferences devoted more space to economic issues, while the CPD's conferences remained focused on health information, with increasing space devoted to this topic. Furthermore, the Prime Minister's press conferences during this phase included increasingly rhetorical language aimed at persuading the audience and enhancing their engagement: "Soon, I am absolutely confident that we will reap the benefits of these sacrifices when everything is over, we will all live a new spring together for Italy" (Prime Minister Press Conference, 4 April 2020), "we are learning what being means [. . .] even more now in this period of difficulty, being Italians means to be courageous to be supportive, to be determined and cohesive" (Prime Minister Press Conference, 4 April 2020). Words such as "sacrifice", "courage", "support" and "cohesion" became part of the government's vocabulary aimed at developing a common sense of shared responsibility (Sciulli, 2018).
In summary, accountability became more sophisticated by developing the topics in detail with the help of experts who were able to sustain different discourses (Bovens, 2007;Sargiacomo, 2015). The other reason for this development was the population's informational need, reflected in the questions posed by journalists, which also began to shift towards economic and governmental policy (Garfinkel, 1984). During this phase, the press conference speakers' expertise broadened, and discussions with journalists were always allowed.
The increasing need for justification (Roberts, 1991) of the government's restrictions led to a proactive accountability style that encompassed the government's use of a rhetorical language (Ahrens, 1996) in the conferences, the increasing number of conferences (both by the Prime Minister and the CPD), the channels of communication and accounting regarding the spread of the coronavirus. Taken together, these elements of accountability style suggested an initial shaping of a sense of responsibility that was meant, as much as possible, to be shared between account givers and receivers (Everett and Friesen, 2010;Sciulli, 2018) and in which the distance between accounting and accountability was reduced (Ahrens, 1996).
The reactive phase suggests a progressive "emotionalization" of the link between the emergency managers and the population that formed the basis for building a shared responsibility during the proactive phase. By extension, the language used in the increasing number of conferences became increasingly "rhetoric" during the proactive phase.

Phase 5 -Coactive style of accountability
The Italian government's style of accountability during this phase of the pandemic may be described as "coactive." In this phase, the political institutions discussed the relaxation of some restrictions as long as the population continued to abide by the government's guidelines. The core aspect of the coactive style here is that the accounts given by the government aimed to "inform" and "explain" the conduct of the political players as well as to "assess" and "validate" the population's efforts to respect the guidelines [under the possibility of further and (allegedly) stricter lockdowns].
On April 23, 2020, the Prime Minister, during a long press conference (43.1 min), informed the population of the gradual reopening of business and private activities. The Prime Minister spoke about the necessity of maintaining surveillance of the spread of COVID-19 during this new phase of the state of exception and the need to consider "what each individual can do for the community to allow a faster recovery" (Prime Minister's press conference, April 23). This phase, during which many restrictions were abandoned, necessitated reminders to citizens that greater freedom meant greater responsibility at the individual level (Sciulli, 2018).
During this phase, five further Prime Minister's press conferences were held, for one of which he was joined by the Economics and Finance Minister, the Education Minister and the Economic and Development Minister.
In particular, during the April 26, 2020 press conference, the Prime Minister thanked citizens for their sacrifice, but underlined the necessity of "respect for the recommendations" to ensure that all the success achieved thus far was not undone. The Prime Minister appealed to the "love" for the people and for Italy to stress the need to maintain social distancing and to use protective masks (Prime Minister's conference, April 26, 2020). This period was called by the government the "second phase of the COVID-19 pandemic counteraction plan" and began with the progressive reopening of business activities (set for May 4, 2020).
The Prime Minister called again for "love" and "sacrifice", reinforcing feelings of compassion and the need for shared responsibility (Everett and Friesen, 2010;Sciulli, 2018) in the accountability rhetoric of this last examined phase.
The population's new needs were also addressed by the Prime Minister's conferences. In particular, issues connected with education could be resolved were perceived as urgent by the population, as evidenced by the questions from journalists to the government. During this phase, the CPD's conferences were reduced to twice weekly (from April 20-30, 2020) and eventually ceased in May.
The online dashboard was maintained and constantly developed with updated quantitative data by the CPD. These data were commented on during the CPD's conferences and again, they remained the focus of these conferences, although disclosure regarding the economic situation increased. During the CPD conferences in this phase, time was invariably and increasingly allocated for questions (on average, lasting 35.2 min). The topics discussed during question times in the witnessed an increasing interest in national policy actions, which seemed to have become increasingly urgent in terms of public opinion (Ahrens, 1996;Garfinkel, 1984).
Coronavirus testing capacity was maintained at the highest level throughout this phase. This phase reflected diminished efforts to provide accountability through public events such as press conferences. Finally, during the conferences, a change in the orientation of the disclosure became apparent, with the increasingly salient role of information aimed at addressing the population's pressing economic needs and the need to address future issues related to education and national policy actions to address the information required to mitigate future risks (Tan and Enderwick, 2006).

Conclusions
This paper aimed to increase our knowledge and understanding of accountability in times of exception. To achieve this aim, the paper critically analysed the mechanisms through which the Italian government sought to combine public governance and public accountability issues in an attempt to urge shared responsibility in times of exception. Drawing on the styles of accountability concept, this paper demonstrated how the government responded to the COVID-19 challenge. Through the analysis of publicly available government data complemented by, and integrated with, other material, the paper provides empirical evidence of how the evolution of the pandemic was related to accountability. The analysis revealed five overarching phases, each representing a different style of accountability. By identifying five styles of accountability (rebuttal, dismissal, reactive, proactive and coactive), the findings show that the state of exception was characterized by the different ways in which the government justified their conduct. In particular, during the initial phases of the Accountability in times of exception pandemic, the accountability styles were focused on the government's responsibility, while in the last two phases, accountability focused on citizens' responsibility (Everett and Friesen, 2010;Baker, 2014) by creating a stronger sense of "shared responsibility" (Sciulli, 2018). This paper contributes to the literature by offering the understanding that times of exception lend distinguishing features to accountability; in particular, it shows that the protean (Walker, 2016) nature of accountability in times of exception related to humanitarian needs must develop towards more shared forms that can result from a previously established emotional link with the actors involved. The more the process of communication is effective in creating an emotional ground for entrusting responsibility to citizens, the more successful the result is likely to be. In this regard, the paper builds on the existing literature that evidenced the need for an accountability that was more focused on compassion (Everett and Friesen, 2010;Baker, 2014), but in this paper, the role of the government in enhancing such sentiments is investigated. This paper also advances existing research on shared accountability (Sciulli, 2018) by showing that accountability can translate into an effective tool to help citizens assume responsibility when the individual emotional level is involved in the process. "Shared responsibility" when the need to protect the population surpasses the usual legal boundaries can also exert a powerful impact by protecting the population through drifts in authoritarian forms (Welch, 2007) of governing the state of exception. Finally, the paper also contributes to developing studies on the effects of governments' responses to the COVID-19 pandemic (Grossi et al., 2020), showing that accountability styles may influence the ability to govern by enhancing the responsibility of the whole of society in relation to the necessary promotion and protection of universal health care (De Villiers et al., 2020). This paper has limitations that open up avenues for future research. An important limitation is that the analysis developed in this paper is specific to the Italian setting and concerns a limited time window. In addition, future studies should consider the linkages between accountability styles and factors such as leadership, culture and plans to manage the state of exception in one or more countries (Wilson, 2020 Figure A2. The CPD's press conference topics (amount of disclosure by topic): phases 2-5 Table A2.
Number of CDP press conferences in each phase and average duration JPBAFM Appendix 4 Figure A3 amount of disclosure for each category of topics discussed between the CPD and journalists (phases 2 to 5).  Figure A3. The CPD's press conference topics (amount of disclosure by topic based on journalists' questions):

Corresponding author
phases 2-5 Accountability in times of exception