COVID-19 and the governmentality of emergency food in the City of Turin

Purpose – This paper shows the accounting, accountability and calculative practices associated with emergency food allocations by the City of Turin through a program to feed the vulnerable during COVID-19. Design/methodology/approach – This is a single case study framed by Foucault ’ s governmentality concept. The data was collected through interviews with key institutional actors and triangulated against decrees, circulars, ordinances and other publicly available documents. Findings – The accountingtoolsofgovernmentalityare alwaysincomplete. Sometimes uniquesituationsand crises help us to revise and improve the tools we have. Other times, they demand entirely new tools. Research limitations/implications – Accounting needs both things to count and a context to count them. Inthecaseoffoodassistance,whatiscountedispeople.InTurin ’ scase,manypeoplehadneverbeencounted – either because there was no need or because they were unaccounted for by choice. Now, the government was accountable for the welfare of both. Thus, new classification systems emerged, as did organisational and accounting solutions. Originality/value – Although the accounting-for-disasters literature is diverse, studies too often favour the macro social, economic and political issues surrounding crises, neglecting the micro issues associated with governmentality and calculative practices. through which authorities seek to embody and give effect to governmental ambitions.

work in tandem with social institutions. Second, we contribute to the governmentality literature with a case on the intersections between accounting and emergency food assistance to city populations during a pandemic. Thus, our case study contributes to an understanding of how governance and accounting practices during a crisis might help save people's lives.

Methods
Before COVID-19, one of the authors was already working on a research project into how charities assist the poor. This research meant we had already undertaken interviews and collected some documents about food assistance, NGOs and Turin's distribution processes. Before COVID-19, national, regional and local government laws prescribed the rules for distributing food activities and roles. This prior research enabled us to identify critical institutional and individual actors to inform our research immediately after Italy declared a State of emergency in January 2020. We analysed over 40 documents, including national, regional and local government decrees, debates, circulars and ordinancesplus more than 20 h of video footage of Turin City Council meetings and local committees. Figure 1 illustrates the period of analysis beginning 1 January 2020, through to the lockdown from 9 March to 18 May, and ending on 31 July. Table 1 provides a list detailing the semi-structured interviews conducted before COVID and from April to June 2020. The interviewees we highlight in Table 1 provide data for identifying our meta themes that identify the apparatus and the technologies adopted for allocating food assistance activities. During the interviews, one author made written notes. In some instances, we conducted follow-up interviews with questions and written responses via telephone and email. From this corpus, we identified new meta-themes explicitly relating to food distribution in the pandemic. We also triangulated common elements (convergence) and discovered complementary elements (extension of the previous knowledge) ( Grafton et al., 2011). Foucault (1991, p. 189) states that "for half a century after the great cholera epidemic of 1832, Europeans were obsessed by fears of epidemic disease". In a subsequent paper, how each province "had to take care of public health, diseases, epidemics and accidents . . ., and it had to manage a kind of insurance for people to be protected against all such accidents" (1994, p. 411). Later still, he explains how, historically, the challenge of governing "population was posed in relation . . . to major human catastrophes" (2007, p. 67). Prior accounting studies on governmentality have already provided insights on healthcare (Sargiacomo, 2015) and evacuee-housing following an earthquake (Sargiacomo and Walker, 2021). However, researchers have yet to explore the governance of populations "after a great disaster, be it an epidemic . . . or food shortage" (Foucault, 2007, p. 61). Rose and Miller (1992, p. 175) argue that academia has a responsibility alongside the fourth and fifth estates to frame political issues and make visible how governments are exercising their power:

Governing populations during epidemics
The problematics of government should also be analysed in terms of their governmental technologies, the complex of mundane programmes, calculations, techniques, apparatuses, Drawing on the case provided by "the town, scarcity and the epidemic" (Foucault, 2007, p. 63), and starting from the assumption that "government is the right disposition of things arranged so as to lead to a convenient end" (Foucault, 1991, p. 93), we analyse how a city government used accounting technologies, calculations and lists to allocate and distribute food to the poor and a broader group of the vulnerable during the COVID-19 pandemic.
Lists of people sit at the heart of this caselists of the elderly, lists of the disabled, the poor and food insecure, those in strict isolation and quarantine, old lists, new lists, and people who should be on lists but are not. Homogenous social lists are a standard part of charities' calculative practices that help the poor and needy (Walker, 2004;Servalli, 2013). They are also a standard tool for supporting governmentality programs, for instance, to assist evacuees in the aftermath of natural disasters (Sargiacomo and Walker, 2021, p. 15). Bowker and Star (1999, p. 138) highlight that "when lists are used to coordinate important work that is distributed widely over time and space, a corresponding complex organisational structure and infrastructure evolves". Infrastructure may evolve during "the regulations for times of epidemic" when we observe "measures taken in plague towns, and the quarantines enforced . . .all constituted forms of authoritarian medicalisation" (Foucault, 1994, p. 92).
4. Governance, accounting and calculative practices for distributing food to the vulnerable 4.1 The pre-COVID state in Turin Turin is in north-eastern Italy. It is the fourth most populated Italian city with less than 1m citizens and home to the fourth-largest concentration of immigrants (15%) (Centro Einaudi, 2019). Turin has been described as "two cities in one" because 30% of its population has considerable financial means, while the other 70% fall near or below the poverty line (Centro Einaudi, 2019). Of its population, almost 25% are over 65, and 40% have an annual income of less than V15,000, which, by law, is the average family income threshold needed to apply for essential social service relief (see Table 2).
Among these statistics used to count the vulnerable before COVID, none pertain to food distribution. Private initiatives led by NGOs largely fulfilled this function, sometimes in collaboration with the city, sometimes not.
Notably, since 2015, Turin was among the first Italian cities to sign the Milan Food Pact [2]. The pact stipulates that Mayors are responsible for providing "cash and food transfers, and other forms of social protection systems (food banks, community food kitchens, emergency food pantries)" to vulnerable people.
Consequently, the City of Turin amended one of its statutes to reflect this responsibility [3], i.e.: To promote the implementation of the right to adequate food, as the right to have regular, permanent and free access to quality food, sufficient, nutritious, healthy, culturally and religiously acceptable, which guarantees mental and physical satisfaction, for the individual and the population, necessary to lead a worthy life.
In 2018, Turin established a Social Inclusion Plan involving 110 NGOs and associations to serve as a repository of co-designed projects and a collector of any public or private financial funds to be spent on local welfare. Initiatives to feed the poor, such as soup kitchens managed by local NGOs (13 in total in 2020).
As such, operational control of all food deliveries pre-COVID remained with the NGOs, and the City did not design its lists to track those affected by food shortages. Therefore, when the pandemic hit, new lists had to be drawn up of people who needed feeding. Some identified new COVID-19 and the emergency food "vulnerable populations", such as COVID-positive individuals subject to home recovery or precautionary quarantine and people who had lost their jobs. Others were lists of invisible people who needed to become visible, e.g. illegal immigrants and undocumented workers.

Distributing emergency food assistance to a growing vulnerable population
On 3 February 2020, the Minister of the Interior released the first ordinance to tackle the COVID-19 pandemic (OCCPD no. 630). The ordinance defined responsibilities for various government levels and established an expense chapter with the Civil Protection Department (CPD) to face COVID-19 and prevent it from spreading. A few weeks later, each region, including Piedmont, was required to create a unique COVID-19 accounting centre (OCCPD n. 639, 25 February 2020). Three days later, Turin set up a Local Government Operations Centre to help its 900,000-strong population face an imminent nationwide lockdown (which ended on 9 March 2020). The Centre managed communications, the media, and the flood of calls coming in on an emergency hotline. Food management during the emergency was placed "under the direct control of the City of Turin" (Regional Council Member, Interview). This control meant collaborating with a central Food Bank as a buyer of processed food and a collector of fresh food surplus from ,000-15,000 (13%) (4) V15,000-26,000 (33%) (5) V26,000-55,000 (22%) (6) V55,000-75,000 (3%) (7) V75,000-120,000 (2%) (8) >V120,000 (1%) OCCPD no. 658 Mar 20 Note(s): *e.g. unemployed, drug addicts Table 2.
Lists of Turin's inhabitants pre-COVID-19 AAAJ 34,6 the local agri-food centres. It also involved working with the Bank of the Charity Works (Banco delle Opere di Carit a) to ensure that staff could complete emergency deliveries while still complying with social distancing guidelines. By March, only 5 of the city's 13 soup kitchens remained open due to social distancing rules and the demographics of the workers and patrons who frequent food shelters. As the Vice-Mayor of Turin explains, "Most of the volunteers working with our local associations and NGOs are retired, the most in danger during the pandemic, and this has affected the possibility of providing the same quality level of the food services" (Interview, 18 June 2020). As more retail food stores closed and the quantity of surplus food collected declined in both quality and quantity, those relying on the soup kitchens began to protest. In response, the Food Bank began to order fresh food from the local markets instead.

The Torino Solidale network
Toward the end of March, a local government directive formed Torino Solidale, a network of 12 emergency food and essential supply hubs throughout the city for the needy and households without local support. The City of Turin managed the network alongside thirdsector organisations chosen from those already participating in the Social Inclusion Plan. Turin's NGO network was already using some hubs to feed the poor before the pandemic, so Torino Solidale expanded from the existing infrastructure. Foucault (1988, p. 142) would describe this as an "upgrade to government technology" that required a "new mapping and the closer surveillance of the urban space". Figure 2 illustrates a map of the community centres in the most deprived areas and the logistics centres used to distribute the food to the 12 hubs.
During the state of emergency, feeding the poor turned into feeding the new poor. Thus, a larger, more diverse group of people labelled "the vulnerable", amalgamated from various purpose-built lists of citizens, each requiring their own urgent, yet tailored-made, government solution emerged. For example, one list comprised the elderly, the infected, and the quarantined, who were all subject to "authoritarian medicalisation" (Foucault, 1994, p. 92) in that they had financial resources but were not allowed to go shopping. Thus, to those on this list, CPD offered a paid food delivery service.
A second list was of those facing extreme financial difficulty, whether entrenched or for the first time. For these new poor, the city-funded a web portal where people could apply for food e-vouchers and, later, for food baskets. People could also register for assistance through a mobile app, by phone or in person. The most impoverished people were usually registered both by local government officers and a network of volunteers to ensure they received food assistance. A Torino Solidale representative interviewed each applicant before being included on the list. During the interviews, narrative accounts of their situation were a critical factor in approving the applications than other quantitative criteria.

The fund for food solidarity
At the same time, Turin was organising its emergency food distribution plans, and the federal government was organising theirs (OCCPD no. 658). On 29 March, just five days after Torino Solidale commenced operations, the Italian government announced V400m in support for the Fund for Food Solidarity to be proportionally distributed between the municipalities using financial and non-financial criteria. The CPD performed the calculations and funding allocations in keeping with the relevant bookkeeping and accounting guidelines in tandem with ANCI (the Italian association of municipalities). These calculations provided the inscriptions about how the cities would receive and spend public money to organise food assistance.
Specifically, 80% of the money was allocated based on population, with the remaining 20% based on the difference between the average resident's income and the national average.

COVID-19 and the emergency food
Subsequently, "the mayors of the largest cities, [. . .] started working on the subsequent elaboration of an implementation guideline to help them manage distributing funds" (Regional Council Member, Interview). However, as there are no official counts of homeless people for any city, these people were not on the central government's lists and, therefore, not factored into this temporary relief program calculations. Turin received around V4.6m of the total V400m budget and added another V663,000 of its funding to support the program's activities in the local area. The funds were apportioned as follows: V4.29m on food vouchers; V120,000 on prepared meals for home delivery, the  homeless and immigrants; and V880,000 to Torino Solidale to buy and distribute food baskets. Deciding how to apportion the money was a complicated process, as was how to distribute the vouchers, meals and baskets equitably. For example, to qualify for food vouchers, applicants had to declare that they were suffering financial distress caused by the pandemic; and no-one in the family was benefitting from another public subsidy. Only one family member could request the food vouchers and the dollar amount depended on the number of people in family (1-2 5 V300; 3-4 5 V400; ≥5 5 V500). Table 3 outlines the income and expenditure of these funds from January through July. As part of the approval process, officials cross-checked applicants against social services lists to ensure no-one received double support. But Turin decided to "not apply any criteria linked to income during the emergency, and all the controls would be performed ex-post, to guarantee an expedited action for the vulnerable people" (Turin Vice-Mayor, Interview, 18 June 2020). Also, the web portal listed 150 accredited local shops where recipients could redeem their vouchers. A cloud platform tracked where the vouchers had been redeemed, including the distance between the voucher holder's residence and the shop. This system reflects a "buyer-supplier" governmental technology that connects assisted populations to suitable local suppliers. Other items tracked were daily redemptions, family composition and the type of entitlement. The tracking provided some accountability over the funds. For example, the city constructed maps reporting voucher use concentrations across Turin's districts. As one City Councillor stated: The portal opened in the week from 6 April to 10 April. Once the news was released, we received more than 1,000 calls per hour. On 6 April, we distributed nearly V1m, V1.38m the second day, V1.2m the third day and nearly V0.6m the fourth and fifth day, and we went rapidly over budget during the weekend. We distributed vouchers to 12,752 families in four days [. . .] Table 3. Emergency food relief funds income and expenditure 31 January to 31 July 2020 COVID-19 and the emergency food concentration of shops, but an equal distribution across the territory: 15% in small shops, 60% in local supermarkets and only 25% in larger supermarkets.
By 9 April, Turin had spent the budget allocated to food vouchers, and there were still 7,928 unfulfilled applications. Hence, the list of assumed vulnerable people was much longer than initially thought. On the same day, the city announced that these unlucky applicants could instead apply for a food basket through the web platform. As a result of this plan B strategy by the city and NGOs, 4,000 more families received home-delivered food baskets (Minutes of the City Council of 27 April). Plan B emerged on day one of the voucher program when the overwhelming numbers of vulnerable people became apparent. That afternoon, the Food Bank, DamamarNGO, and the Bank of Charity Works signed an MoU to boost the collection and distribution of free and surplus food as food baskets. However, the type of food basket and its quality soon became an issue. In part, this was because the kinds of people requesting food during the pandemic did not necessarily fit the standard mould of people the charities typically helped. Traditionally, food assistance meant providing a high calorie, high protein meal. However: A family with a baby needs other types of food because usually, a food pack has 1 kg of rice, 400 gr of beans, and one can of tuna and tomato sauce that can last for one week only. (City Council Minutes,27 April) 4.5 Accountability and reporting practices Meeting the conditions stipulated by the emergency food directive, the Italian government asked cities to open a new accounting ledger entitled the Urgent food solidarity measure. Accordingly, the cities used the same account to collect other public and private donations to assist with the emergency. By so doing, utterly novel government technologies were implemented, including "techniques of notation, computation and calculation" (Rose and Miller, 1992, p. 183). On 6 April, the City of Turin opened its first "food solidarity" account.
We find evidence of specific accounting for emergency relief by the CPD within their reporting infrastructure. For instance, on 23 May, they asked all municipalities, including Turin, to complete and forward a new multi-table report in Excel to accounting for the funds and how they spent them.
One of the compulsory reports, "Home assistance", required an accounting of the expenses incurred for the food's cost and delivered for the vulnerable (period 31.01.2020-31.05.2020). Another, "Hotel assistance", asked for expenses sustained for the allocation of meals.
Another example is Turin's compliance with L.G.D. No. 986, when the city reported a variation of the financial forecast and budget for 2020/2022 to include other donations received from other public and private institutions (Table 3). Turin accounted for the money using two additional sub-accounts. One was for purchasing food by transferring funds to Damamar NGO. Torino Solidale and the management network used the second account for transactions made in providing emergency food relief.
However, while the accounting and accountability practices over the use of emergency funds appear to be adequate, there was no accountability or counting of Turin's invisible denizens, i.e. its 1,200 illegal immigrants and undeclared workers (Hajer and Ambrosini, 2020). Nor was there any accounting for those who did not receive assistance by Torino Solidale. However, Torino Solidale implemented several operating controls. First, operators tried to discourage any opportunistic behaviours by talking to the applicants by phone or face-to-face meetings. Second, the shared database allowed all the hub managers to perform a real-time check of any other support issued to a family member. Third, a second real-time cross-check with the Registry Office gave immediate warning of any transgression.
During the first wave of COVID-19, roughly 10,000 families received a monthly food basket. Subsequent interviews were conducted personally by Torino Solidale operators to ensure each family's continued emergency status and, in turn, their entitlement to support. Nearly 1,500 families were declared to be past crisis point (Minutes of the Social Policy Committee of 4 June 2020). The intention was for Torino Solidale to be transformed from an emergency measure to a permanent public policy after the first pandemic wave with monthly support to a vulnerable population (Interview with the Vice Mayor of the City of 18 June 2020).

Discussions and conclusions
In a scenario characterised by "regulations for times of epidemic . . . measures taken in plague towns, and the quarantines enforced . . ." (Foucault, 1994, p. 92), our analysis of the response to the epidemic in Turin fills a gap in governmentality studies on crises concerning the intersections between accounting and towns, food shortages and epidemics (Foucault, 2007, p. 63). Prior studies on crises have mostly offered investigations of the macro social, economic and political issues surrounding bushfires (Taylor et al., 2014), floods (Lai et al., 2014) or earthquakes (Sargiacomo et al., 2014;Sargiacomo, 2015). The unveiling in our case study of micro-accounting techniques and technologies deployed at the local government level permits a better understanding of the micro issues related to specific population's segments (see Perkiss and Moerman, 2020, p. 11).
The list of technologies called into action, in this case, was extensive, confirming that "government is the right disposition of things arranged to lead to a convenient end" (Foucault, 1991, p. 93). Indeed, there is no doubt that the classification of space and society is pivotal to governmentality during a pandemic. The government was required to launch new tools and technologies, such as the Torino Solidale network. This apparatus established welldispersed food hubs by expanding an already-existing infrastructure. The government used its relational capital with local actors and designed a map based on the town cartography after identifying its constituents' needs across the territory. Our study complements and extends prior literature by adding novel pieces of evidence in a case where a metropolis is facing a severe pandemic and a population either cannot get to or cannot afford to buy food despite its abundance.
In our case, what emerged is an example of how the government is "a science of endless lists and classifications" (Bowker and Star, 1999, p. 16). During Turin's handling of the emergency food assistance program, list-making was of paramount importance as a government technology for "coordinating activity distributed in time and space" (Bowker and Star, 1999, p. 138). Before the pandemic, there were already numbers and lists of the poor. The city had to extend its previous knowledge by creating purpose-built lists, such as the elderly not yet assisted, the disabled who have difficulties with shopping, the infected who must not be allowed to leave their house, the parents with babies who lost their job, and so on. One of the most challenging tasks was tracking the people who would prefer to be uncounted, such as illegal immigrants and undocumented workers, in any other circumstance. Concerned by legal matters, the presumption was that many would not apply for emergency food. The City helped these citizens by providing cooked meals at immigrant centres where visitors could remain anonymous. However, this issue is broad enough and intriguing enough to warrant further research. Unfortunately, for the time being, these individuals remain invisible.
Expanding the governmentality literature on "social lists", one final amalgamated list of "the vulnerable" emerged for Turin to encompass the myriad human situations alluded to above (Foucault, 1994, p. 92). That list identified people as a heterogenous "distribution of cases in a population circumscribed in time or space" (Foucault, 2007, p. 60). Classifications and lists were thus pivotal during the administration of food assistance. The amounts of COVID-19 and the emergency food money calculated for food-vouchers linked to the number of people in families seeking assistance became "integral to any working infrastructure". They served "the needs of multiple local, national and international information systems" (Bowker and Star, 1999, p. 16). Turin also made extensive use of accounting and "techniques of notation, computation and calculation" (Rose and Miller, 1992, p. 183) as government technologies. At the national level, in the aftermath of Italy's declaration of a state of emergency, the CPD's COVID-19 expense ledger and Piedmont's unique accounting system (OCCPD n. 639) represented the first act of the launch of novel accounting technologies of government and the inscription of calculative practices. The national Fund for Food Solidarity program, established in March, was distributed using a formula at the regional level and executed at a micro-territorial level. These new forms of data collection, inscription, accounting, and calculative practices were pivotal in administering food deliveries to a diverse population. Thus, purpose-built specific guidelines and technical notes were distributed to local governments, showing both the way to get public money and the new ledger' urgent food solidarity measure' and portraying the way to use the destined public money. The City of Turin followed the new guidelines and then undertook a range of food assistance solutions, including the supply of food vouchers, foodbaskets, and cooked meals. The procedures used to account for the food distribution resonates with Rose and Miller (1992, p. 183), who stressed that government technologies might also include "the standardisation of systems for training, and the inculcation of new habits" as found in these new measures.
Additionally, "procedures of examination and assessment, the invention of . . .presentational forms such as tables" Rose and Miller (1992, p. 183) appeared as well as technologies that involved both reporting and control practices. For example, the novel and purpose-built multi-table compulsory reporting required to document the funding used for food assistance to population did not exist before the pandemic. Further, by triangulating data from the City of Turin web-portal, the government exerted control over public money flows, directing them toward food vouchers and baskets. Importantly, opportunities for control were available both top-down and horizontally. For example, any Food Hub could impose real-time controls on people by ascertaining whether they had already received assistance from another hub. Foucault (2007) defines these as space-time controls on territory and population.
In our case, the classification systems emerged, as did organisational and accounting solutions, which "hold a memory of work that has been done (laboratory, organisational, epidemiological, sociological) and so permit the recommendation of a reasonable due process for future work" as affirmed by Bowker and Star (1999, p. 252). These classification systems illuminate new practices from the Italian experience that may inform how responses to pandemics and other crises could be adopted or adapted in different settings (Parker and Northcott, 2016). More importantly, this case shows how governmentality can govern the wealth, health, and happiness of populations with all its good intentions. However, if that population is invisible, there is no accountability for their wealth, health and happiness. Thus, the concept of governmentality needs to expand to consider both the visible and the invisible to ensure accountability to the entire population. Moreover, governmentality should not only apply to government spending, it should also apply to NGOs.
Through this case study, we find that meticulous attention to something as minor as food distribution to vulnerable population confirms that "managing the population does not mean just managing the collective mass of phenomena or managing them simply at the level of their overall results; managing the population means managing it in depth, in all its fine points and details" (emphasis added) (Foucault, 2007, p. 107). These fine points and details are the most important because these fine points are real people.
In these crises, it is the people that matter because if we exclude someone from a list, there can be dire consequences for them. In Turin, many people under pre-pandemic economic conditions chose not to be on lists because of their situations. Many of these are the thousands of illegal immigrants who would also have lost their income and access to food (Hajer and Ambrosini, 2020). Even if they did want to join the list, they are unlikely to have the same access to the Internet or a mobile phone or landline or even apply in person as would others. So, they miss out on being included in new lists and miss out on any potential emergency assistancethen, now and probably into the future.
What we can especially learn from this case is that the accounting tools of governmentality are always incomplete. Sometimes unique situations and crises help us to revise and improve the tools we have. Other times, they demand entirely new tools. We must remember that accounting needs both things to count and a context in which to count them and that some people are unaccounted for by choice. When a new context arises that fundamentally changes those preferences, governments must work harder to fulfil their welfare obligations because accounting for the unaccounted in a pandemic is a matter of life and death.