Crisis policy enactment: primary school leaders’ responses to the Covid-19 pandemic in England

ABSTRACT This paper explores the enactment of government policy during the Covid pandemic in primary schools in England. Based on interviews with school leaders and teachers across the period 2020–21 (n = 66), drawn from two major studies of primary schools’ priorities during the crisis, we argue that school leaders’ responses can be understood as a distinct form of policy enactment particular to an unprecedented crisis. Policy arrived in schools differently, and was enacted differently. Our findings suggest that enacting policy during the Covid crisis was a process dominated by the need to act at speed, informed by the prioritisation of children’s basic needs and based on a knowledge of the local circumstances. Thus dimensions of context which affect policy enactment were altered during the crisis, with the material circumstances of the school and the values of the headteacher becoming highly significant. The approach we term crisis policy enactment is response to policy which is focused first on coping but is also agentic, demonstrating a commitment to children’s welfare and a belief in the power of schools to make a difference.


Introduction
This paper explores how senior leaders and teachers from schools in England enacted government policies during the COVID-19 pandemic in the spring and summer of 2020 and following the second wave in the spring and summer of 2021. It aims to explore in depth the distinctive nature of school leaders' and teachers' engagement with policy during the crisis, and the extent to which the theoretical framework offered by policy enactment (Ball, Maguire, and Braun 2012) can be applied and is useful in understanding how schools responded.
The closure of primary schools in England to all but the children of essential workers and vulnerable children in March 2020 was an unprecedented event which school leaders were unprepared for (Harris and Jones 2020). As a range of research has indicated, the need for plans to be quickly put in place to provide teaching for some children in school while also providing online or home learning resources in age-appropriate ways presented a huge logistical challenge (Harris and Jones 2020;Moss et al. 2020; Thomson, Greany, and Martindale 2021). These problems had to be solved by a workforce who were in many cases demotivated by years of policy reform (Harris and Jones 2020), and fearful of the impact of the virus on their health and their families. These teachers and staff also often had to manage school closures for their own children and other sudden family and care demands. As the 'lockdown' period continued, school leaders were faced with numerous decisions relating to curriculum provision, eligibility for school attendance, and staffing allocation, all of which had potentially significant consequences, including possible impacts on children's and staff's health and wellbeing. 1 At the same time, the government, who had already faced criticism for enforcing a lockdown too late (Manthorpe 2020) issued a series of sets of guidance and emergency legislation relating to education.
Circumstances changed again as schools were allowed to reopen for some pupils in the summer of 2020, under new restrictions relating to social distancing, cleaning systems and numbers of pupils, and then again in the autumn when they fully reopened. A second lockdown beginning in January 2021 produced further new rules and regulations, necessitating more new plans for schools, before schools reopened again in March 2021. Throughout this period school leaders were -as previously -responsible for enacting government policy in their schools, relating to education, welfare, safeguarding and provision of food, among others. However, our findings suggest that both the delivery of policies and how they were translated into real-life school environments were different from 'normal' times, which leads us the question of how we can conceptualise the distinctiveness of enactment during a crisis.
In this paper we present a range of data from two projects exploring schools' priorities during the pandemic, and consider teachers' and school leaders' interpretation and enactment of these policies. Schools in England are highly relevant in this discussion of enactment given that they are heavily researched in terms of policy pre-Covid, because of the neoliberal accountability-based system that operates, and thus we have some understanding of previous policy enactment (see for example, Pratt 2016; . They also provide an example of a school system which is subject to control by central government in London (rather than federal), and where school autonomy has been emphasised in recent years (Boyask 2018). There is also great diversity within the system, with some schools operating within local authorities, and others independently as Academies, or in groups of schools known as Multi-Academy Trusts (MAT) (Keddie and Mills 2019). Primary schools in England also vary hugely in size, and religious or community affiliation, but they are all subject to the same systems of statutory testing that dominated pre-Covid.
The paper begins with a discussion of the wider literature on how schools cope with crises such as school closures, before explaining the theoretical underpinnings in enactment literature and then the details of the research studies. This broader research evidence helps to frame our discussion of what is specific and distinct about enacting policy during a major crisis, though we do not assume the transferability of this conceptualisation from one crisis to another. We then discuss the ways in which enactment was distinctive during the pandemic, beginning with the way in which policy was received, and then examining the contextual dimensions of how schools 'did' policy (Ball, Maguire, and Braun 2012) during this period.

Understanding how schools responded to crises before and during Covid
There is an existing international literature on the importance of understanding crises in the context of education reform, which focuses on economic crises, natural disasters and terrorism events (Slater 2015); for example, Buras' work in New Orleans identifies the neoliberal reforms following Hurricane Katrina and their racialised effects on schooling (Buras, Randels, and Ya Salaam 2010;Buras 2014). While this work focuses on the organisation of schooling and national policies, there is also more relevant research on how schools respond to unexpected closures, based on natural disasters such as earthquakes (Mutch 2016), snowstorms (Stuart et al. 2013) and again Hurricane Katrina (Howat et al. 2012). This work, as summarised in Harmey and Moss' (2021) review of this field, demonstrates the importance of schools in providing a return to normalcy, the need for a curriculum which is responsive to needs, and the significance of schools in supporting the mental health of both students and the wider school community. Particularly relevant here is the point from this review that school leaders' local knowledge was essential in response to a crisis. For example, Fletcher and Nicholas (2016) found that principals, who knew the community well, had an important role in supporting students' learning during and after natural disasters, and that they played a pivotal role in ensuring social cohesion. Mutch's (2015Mutch's ( , 2016 work following the Christchurch earthquake and Stuart et al.'s (2013) work on school closures due to the H1N1 flu pandemic and snowstorms found similarly that local knowledge is vital in recovery. School leaders have a better understanding of local needs which puts them in a more advantageous position than national governments to plan for recovery (O'Connor and Takahashi 2014).
The growing field of research on the impact of COVID-19 on schooling internationally focuses largely on the disruption to children's formal learning and issues relating to the shift to remote or online learning (see for example, Yates et al. 2021 on New Zealand;Yan et al. 2021 on the Guangdong Province of China). A review of existing research in the UK (conducted by some of the same research team), mainly drawing on survey data, found however that there were diverse impacts on student mental health and safeguarding as well as on learning (Moss et al. 2021a). While there has been a general awareness of the multiple impacts of school closures (itself an inaccurate term given that schools stayed open for many children) in the media and in research, the evidence on how schools responded and their priorities during the crisis is still emerging, and evidence of impact remains uncertain (Moss et al. 2021a).
International research on school leaders thus far suggests a high level of stress and anxiety (Arnold, Rahimi, and Riley 2021;Thomson, Greany, and Martindale 2021;Reid 2022). School principals in New Jersey, US, were affected by the 'ongoing and evolving nature' of the crisis, and coped by suppressing these feelings in front of staff and students (Reid 2022, 63). In Australia, school leaders felt they had less clarity over the nature of their roles, and their work was less predictable, though interestingly they also reported lower levels of work-life conflict (Arnold, Rahimi, and Riley 2021). There are some contradictions within this research, for example between this finding on work-life balance and media reports of teachers struggling to make distinctions between home and work while teaching from their homes; there will likely be many more contributions to this body of work as researchers begin to publish their studies.
In the UK, the study of school leaders by Thomson, Greany, and Martindale (2021) has found significant issues of work-related stress, including a feeling of 'surviving' (p298). Using survey and interview data, Thomson et al. describe 'challenges unimaginable prior to the pandemic' (2021, p296) for school leaders. Work-related stress and change fatigue have had negative impacts on leaders' wellbeing and their ability to solve problems, and government advice has not been timely or straightforward, leading to a potentially long-term distrust of government. They point to the threats of legal action made against schools wishing to close in late December 2020 as evidence of this declining relationship and the problem of a shift from local autonomy to national decision-making -or 'over-centralisation' (Greany et al. 2022, 92). These findings echo the speculative comments made by Harris and Jones (2020) about leadership in disruptive times, specifically that in these circumstances 'the pressure is relentless, the options are limited and the sleepless nights are frequent' (p244). They predicted that headteachers were the 'pinch point' in the system, subject to pressures from above and below in a situation with 'no precedents and no guides to leading schools during a pandemic' (p244); Thomson, Greany, and Martindale (2021) and Greany et al.'s (2022) work suggests their prediction was accurate.
A major cause of stress for headteachers has been identified in several studies as arising from government guidance (Moss et al. 2020;Fotheringham et al. 2022;Thomson, Greany, and Martindale 2021). School leaders at both primary and secondary level received on average three policy updates per day in the period immediately following school closures, and there were ten days where five or more policy documents were published (Fotheringham et al. 2022). In interviews focused on this topic headteachers described struggling to keep up and being confused about what was statutory and what was merely guidance (Fotheringham et al. 2022). Similarly, the vast majority of Thomson et al's 1491 survey respondents in 2020 felt the Department for Education had 'not provided timely and straightforward advice' (p298).
As we see in later sections, the way in which guidance was delivered to schools is hugely important in understanding what school leaders chose to do. In the follow up phase of Greany and Thompson's research in 2021, interviews with 42 senior leaders suggested that: leaders' individual agency had been hampered by the 'top down', centralised nature of decision-making by government, in particular where this was also last minute and/or poorly matched to the needs of particular schools and contexts (Greany et al. 2022, 30). This point on context chimes with our findings throughout the pandemic that the disjuncture between local circumstances and generic national policy was a source of tension (Moss et al, 2021b).
There is some international literature which emphasises the shifting priorities of teachers and school leaders during the crisis across both time and place. Survey data from the United States suggests that planning for future closures, academic achievement gaps, and student physical and mental health have become higher priorities during the crisis (Trinidad 2021). International research has similarly found variation among schools' responses -for example, among schools in Northern Italy where Fedele, Iacuzzi, and Garlatti (2021) found that 'one size does not fit all' because schools had very different local responses. Increasingly, there is an awareness of the variation and evolution of practices through the different phases of the pandemic, suggesting awareness of the temporal as well as geographical/local variations. In recent analysis, Thomson (2022) has described the school system in England as having gone through three phases, with different demands on school leaders: Phase 1 (March to August 2020) which involved the first lockdown and partial reopening of schools, where the main issues were welfare, food and home learning; Phase 2 (September 2020 -August 2021) when schools fully reopened with attendant challenges, pressure to 'catch up' and then further school closures; and Phase 3 (September 2021 to spring 2022) where all restrictions have gradually been removed. Furthermore, the ongoing nature of discussions about restrictions and infection rates and the still unknown wider impacts of lockdowns mean that this, as we write in 2022, is still an ongoing crisis.
Certainly, there is much still to understand and learn about how school leaders experienced and coped with the pandemic. While there increasing volumes of academic research on schools and Covid, what is often missing from this range of literature is a detailed analysis of how policy was enacted in real schools, during the crisis itself and in the phases of recovery, drawn from in-depth studies of the schools as a whole. It is this area which we focus on here.

Understanding policy enactment
The circumstances of the Covid pandemic changed the nature of education policy, as schools came under emergency legislation. Some of what the government decided was fixed in law, but much was issued as guidance. Here we use the term 'policy' to describe both of these, taking the lead from our participants who described being engaged in interpreting both statutory and advisory information.
Despite this unique situation, we argue that we can use theoretical perspectives from pre-pandemic policy sociology to analyse policy enactment in a crisis; in fact, these ideas alert us to the distinctive features of enacting policy in such circumstances. We draw here on the work of Ball, Maguire and Braun on 'how schools do policy' in the secondary sector (Ball, Maguire, and Braun 2012). This scholarship provides a number of key insights into the ways in which policy is understood, interpreted, acted upon and resisted in real world situations. Most relevant here is the point that context matters in how schools respond to policy: policies are enacted in material conditions, with varying resources, in relation to particular 'problems'. They are set against and alongside existing commitments, values and forms of experience. In other words, a framework for policy enactment will need to consider a set of objective conditions in relation to a set of subjective 'interpretational' dynamics and thus acknowledge that the material, structural and relational are part of policy analysis in order to make better sense of policy enactments at the institutional level. (Braun et al. 2011, 588) Braun et al. (2011) go on to detail the facets of context which affect policy enactment, in four dimensions: Situated contexts (such as locale, school histories, intakes and settings). Professional contexts (such as values, teacher commitments and experiences, and 'policy management' in schools).
Material contexts (e.g. staffing, budget, buildings, technology and infrastructure). External contexts (e.g. degree and quality of local authority support, pressures and expectations from broader policy context, such as Ofsted ratings, league table positions, legal requirements and responsibilities). (Braun et al. 2011, 588) These dimensions are presented as a 'heuristic device to encourage investigation and questioning' (p587) based on the critique that many policy analyses fail to take schoolspecific contexts seriously enough, rather than as a 'comprehensive model'. The authors encourage a broad view of context through these dimensions, including social factors alongside physical and historical aspects of schools, and recognise how these can overlap and interconnect.
This schema has been used in subsequent analyses of policy enactment (Singh, Heimans, and Glasswell 2014;Keddie 2013;Braun and Maguire, 2019) to frame detailed discussions what schools do with policy, and to emphasise the multiplicity of responses from schools to reforms. It is useful, we argue, in exploring what was different about how school leaders interpreted policy during the crisis period, as it provides a framework to think about how different elements of context became more significant. As Braun and colleagues note, context is 'dynamic and shifting, both within and outside schools' (2011 p595); but Covid, we would argue, brought sudden and more dramatic shifts than ever before.
We note that policy sociology is in a process of responding to the new Covid era, with new lines of analysis arising from developments such as the growth of new policy networks (Peruzzo, Ball, and Grimaldi 2022), to detailed studies of how children framed by policy as 'low attaining' coped with home learning (Buchanan, Hargreaves, and Quick 2022). These studies demonstrate how established concepts -like policy enactment -can be used to further our understanding of the pandemic era. Broader sociological perspectives are at the same time being used to conceptualise what Covid has demonstrated about the operation of the neoliberal state (Lupton 2022), notably that 'Neoliberal political systems could no longer rely on the free market to solve the broad and deep problems associated with the pandemic' (p43) and to develop how we understand the emerging multiple 'Covid societies' as the pandemic recedes. This wider emerging field of policy sociology during and after Covid informs our approach here, which is to consider the particularities of the crisis as significant in themselves, as well as shedding light on broader issues in education.

The research studies
The two research projects discussed here were motivated by a desire to find out what was happening in schools as the crisis unfolded, based on the principle that this information is most useful in planning for the future. The first project -A duty of care and a duty to teach: educational priorities in response to the COVID-19 crisis (hereafter DCDT) -was inspired by news footage of a primary teacher with a huge backpack delivering free school meals to his local community, suggesting a prioritisation of fundamental needs over everything else. Our approach prioritised gaining information from as many teachers as possible via nationwide surveys (see Moss et al. 2020), alongside getting in-depth information from interviews with teachers. As we reported in June 2020, schools were prioritising welfare and care over academic learning (Author 1 and Author 32020 Moss et al. 2020).
The second project -Learning through disruption: rebuilding primary education using local knowledge (hereafter LTD) -aimed to collect in-depth data on how schools were responding to the crisis as children returned after a second lockdown. Here the schools were the main focus rather than individuals, and this shaped our research design. Similarly, we found a prioritisation of welfare as well as huge variation in schools' 'Covid stories' in 2021 (Moss et al. 2021b). This paper draws on interview data collected through both projects.

DCDT project
This project was funded under the UKRI scheme to promote rapid Covid research, through the Economic and Social Research Council, and involved collection of data during the first phase of the crisis in March-August 2020 (ES/V00414X/1). The project had several strands of data collection: a rapid evidence assessment (REA), two national surveys, and interviews (see Moss et al. 2020). This paper uses only the interview data but is informed by the other data sources and particularly the conclusions of the REA, already discussed (Harmey and Moss 2021). The aim of the interviews was to gather detailed data about how the crisis was affecting schools and their priorities. In total we conducted 17 interviews, with teachers, headteachers and representatives from a Multi-Academy Trust and Local Authority, across different regions of England (summarised in Table 1 below); we focus here mainly on the data from school leaders.
These interviews were part of an iterative process of data collection, informed by and informing large-scale surveys, the REA on learning disruption and documentary data collection focused on the media representation of educational issues. As such, though they are presented alone here, they cannot be separated from the larger datasets produced. Here we use mainly data arising from the senior leaders, with some relevant data from other teachers also included. The interviews were conducted by members of the research team using a standard schedule that we adapted to the different roles of participants and to the unfolding events of the summer term of 2020 (including announcements about which children would return to school). The project adhered to the ethical standards of the British Education Research Association, and was approved by the UCL Institute of Education Ethics Committee.

LTD project
The Learning Through Disruption project built upon the findings of the DCDT project and further work undertaken by the team on schools' use of teaching assistants during the crisis (Moss et al. 2021c), and was again funded by UKRI/ESRC (ES/W002086/1). Data collection commenced in May and completed in August 2021. The design for this project involved seven case study schools (labelled S1-7) in different areas and of different types, with the sample based on the principle of maximum variation. Each case study involved interviews with the headteacher, school staff, parents and representatives from middle-tier organisations linked to the school, such as the Local Authority or Multi-Academy Trust ('brokers'). In total we interviewed 49 participants, but here we focus on the 10 headteachers (three of the seven schools had joint headteachers) and three deputy and assistant headteachers (summarised in Table 2). While the number of school leaders here is relatively small, these interviews are nestled within larger case studies, so their data are understood in the broader context of data provided through interviews with teachers, parents and brokers. Note that the regions are not revealed to ensure anonymity.
Data from both projects were analysed using a variety of systems including thematic coding in NVivo and hand-coding, with findings discussed throughout the research team. Here we present the results of the analysis focused on senior leaders' responses to the crisis, decision making processes, government guidance and their networks of support and guidance. While both projects were limited by the Covid conditions, time available to the researchers, and the need to produce findings at speed, the volume of data collected through both studies provides many rich insights into the lived realities of the period for teachers and families. Enacting policy in a crisis: a distinctive approach In the following sections, we discuss the distinctive ways in which policy was received and enacted during the pandemic, based on our interview data. The differences between pre-Covid enactment and enactment during the crisis (in all its phases) are summarised here:

Policy delivery
The first point of distinction between before and during the crisis period was the sheer volume of policies sent to school leaders. As noted by Fotheringham et al. (2022), 74 documents were published for primary and secondary schools in 90 days following the first school closures. Headteachers were sent both information relating to statutory requirements and guidance documents far more frequently, via email, and often late at night or at the weekend.
Any guidance was released late at night, they didn't ever show what the guidance changes were. So you had to reread all the guidance where you get told you had to open the school and they give you the guidance 24 hours before (Head B S1, LTD) The problem of simply sending out information without identifying where changes had been made to guidance compounded the sudden delivery of new information to school leaders, adding to their workloads. The most significant policy decision made by government, to close schools, was delivered in a particularly different manner to most education policy, due to the evolving nature of the crisis. The public were made aware at the same time as school leaders, via an announcement on national television from the Prime Minister, Boris Johnson. I remember we were sat and we had Boris on the big screen, and he said that we were closing and it was kind of an instant 'What do we do now? How do we go about this?' So I remember that. And then we've just worked together really to figure out how we were going to do things. (Deputy Head S4 LTD) Given that most education policy is long anticipated, as it is trailed in political manifestos and speeches, and in the press and through consultation, the sudden and dramatic nature of this policy of closing schools to all but children of key workers was shocking. As one headteacher commented, 'It was a nightmare. We were told to close down in 48 hours.' (Head, NW, DCDT). The fact that this announcement also required complex arrangements to be made, notably working out who the key worker children were and how on-site provision would be staffed, exacerbated the impact of this sudden announcement. This coheres with findings on school teachers in the first six weeks of lockdown, which have noted a 'powerful shared sense of not knowing what is going on' (Kim andAsbury 2020, 1070) A similar pattern of events happened in late December 2020 and early January 2021, when there was uncertainty over the announcement of a further 'lockdown' and period of school closures. Decisions over the closure of schools in some areas came a few days before the start of the new term in January, but in many areas of England schools returned for one day, and then were told to close to all but vulnerable and key workers' children again. Again, the lack of time to prepare was problematic for school leaders: the beginning of January [2021], when we were all sitting there planning and we did our risk assessment for coming back the next day, and then nine o'clock at night, Boris Johnson said, "Oh, no. You're not going back to school, actually." [. . .] They didn't tell us. We heard first on the news on that evening (Head S5, LTD) As a head teacher, I'm sitting at home waiting for announcements. I heard it first when there's the news conference at 5 o'clock. By 8 o'clock at night the NAHT have then given you the union update. That is just not right. It's not about getting inside information and tip offs, but to only hear, oh you're back at school, or you're not back at school, or you're wearing masks, on the news, is just wrong. That's been very challenging -the amount of information. Everybody's like ducks swimming. (Head S3, LTD) The distinctive nature of finding out government decisions which would impact on schools the next day via the media was unprecedented for school leaders, who were used to receiving information in advance and having time to communicate it to parents in appropriate ways. In particular, the fact that schools were told at the same time as parents caused problems: A lot of people I've spoken to outside of teaching hadn't realised that we found out what was happening at the same time as everybody else. That was really hard. So you had to kind of watch the news to find out what your plan was, whether you might be open or not. And that was tricky. (Deputy Head S4 LTD) Dealing with the decision to close schools again in January was made more complex by the inconsistencies by area, as some regions were under different levels of restrictions due to higher infection rates, and so the date when the decision was made varied by local authority. In London, this meant it was announced that schools in some boroughs were shutting a few days earlier than others. This lack of consistency was also distinctive, in a system used to having central government make policy for the whole of England; unusually, schools could not assume that other schools nearby were enacting the same policies. At times, school leaders felt they could not see the overarching plan for schools during the pandemic; this apparent lack of coherence contrasts with the kinds of policy usually enacted in schools which are set out in manifestos and white papers, usually a plan of reform.
Finally, the lack of consultation was a key factor in how school leaders felt about receiving new policy during this period: often they felt that their views were not listened to, because new rules were simply given to them without an opportunity to comment, as in pre-Covid times.
[Government should be] listening to individuals who are on the frontline, to see what is actually essential for their circumstances (Head S2 LTD) I think just generally to anyone away from education trying to make these decisions for people and children in the education system that . . . It's not a blanket or a one size fits all. Every decision that we make and have to make is according to each family of each child and that is the way that they need to work, and stop thinking that they can pluck these answers out of the air and think that we can make it work because we know our children. (Assistant Head S1 LTD) As this second quote suggests, the use of generic solutions to contextually specific problems was a problem for school leaders, who had in-depth knowledge of their pupils and local communities. Without consultation, they felt that policy was often irrelevant or inappropriate, making it more challenging to enact in practice.
While the difference between how policy is received by schools in normal time and during Covid was a necessity of the situation, these points of distinction are important in understanding how school leaders responded to the high volume of policy change. As we return to in the discussion section, these points demonstrate the disjuncture between normal conditions and the Covid period, the feelings of shock and anxiety, and the distance felt between those in schools and those making decisions.

Policy responses
As suggested in earlier sections, the literature on policy enactment emphasises the importance of context in determining how schools respond to policy, both as professionals and in more practical ways (Ball, Maguire, and Braun 2012). Here we argue that our findings on policy responses during the Covid crisis suggest that context remains important, but that new aspects or facets of context became more significant during the pandemic. These include the increased importance of material and situated contexts, but we begin the overriding significance of school ethos in relation to children's welfare. This can be seen as part of what Ball, Maguire, and Braun (2012) call the professional context of a school, in that they are indications of the values that matter in that institution.

Professional context
The overwhelming priority of all the school leaders in both studies was the welfare of children at all stages of the crisis, particularly in terms of being safe at home and having enough food. New guidance and information were understood through this lens, so that immediate plans on the announcement of school closing focused on providing food and monitoring vulnerable children's wellbeing.
Our main priority was making sure that our children were safe and that they were okay at home. In terms of that we were making sure that they were going to have access to food, so for our free school meal children, that they weren't going to be missing out (Head S6 LTD) Really, they were families that we just wanted eyes on every week, so there was somebody who had been suffering domestic abuse, and we just really wanted to see her. So, the food . . . was a necessity, but it was also a really good reason to go and visit, to deliver and see them." (Head S2 LTD) Throughout all of this the children's safety and well-being has been paramount [. . .] if they're not well in their minds they can't function, let alone learn. (Head S4 LTD) School leaders employed various strategies as result of this prioritisation of welfare: visiting homes, making regular phone calls, monitoring online engagement, and using systems to flag up staff concerns about particular children. The ability to provide for 'vulnerable' children in the school building was a clear priority, as this leader of a group of schools explained: I think right from the beginning there's been a number of challenges that we felt we really needed to step up to [. . .] We absolutely made sure that we were in a position to open to vulnerable children and the key worker bit as well, right from the start, and we did that all the way through, fulltime places, no messing around. And we stayed open all the way through Easter [holidays] as well, and the bank holidays, and actually, at one point, we were ramping up to be doing weekends and overnight care and stuff if we needed to. (MAT leader, London, DCDT) For this group of schools, as well as the vulnerable children, the needs of a small group of parents working in medical roles became important because 'we thought potentially it was going to completely overwhelm the hospitals, [so] we were then quite anxious about the kind of critical workers', and wanted to provide childcare so that they could work longer shifts. The identification and prioritisation of these 'challenges' also related to providing food, the MAT leader went on to explain, through whatever means were seen as most appropriate: We made food available on a takeaway basis, and then the hard lockdown came and, at that point, it felt like that wasn't appropriate, so we moved to our own voucher scheme before the government got its act together. [. . .] We did that for not just for the free school meal eligible, but also the universal free school meal children.
Aware of local poverty levels increasing, this group of schools acted in advance of government systems of providing vouchers instead of meals, and extended this provision to all children who would normally get a free lunch (which is all children up to age seven), not just those who were eligible via a means-test. Other schools made similar decisions to allow them to them to provide children with food, such as working with local suppliers, charities and food banks. These were then distributed through various means: . . . in terms of poverty, we had a lot of food hampers going out from our school weekly, for our high demand for those. And the school was very good organising that. We kept the canteen running as well and we made packed lunches and food to be delivered out. (Teacher on SLT, Lon, DCDT) Before the government stepped in and did any of their vouchers, we were already providing vouchers and linking up with our local charity, [name] who supplied food parcels to our families (Head A S1 LTD) Thus we see how the moral and ethical choices of headteachers, which were largely consistent across schools, determined the enactment of policies, especially in the immediate aftermath of the decision to close schools. Significantly, this prioritisation of welfare included making decisions about staff and parents as well as children, so that school leaders were required to balance a number of priorities: We were dealing with some things that [. . .] I've never experienced in my career, and that's all added into the whole mix as well as your most important aspect is keeping your staff, your children, and your wider community safe, in a situation which all of us are just like thrown into, as you say. (Head, S2, LTD) As we have seen, these decisions were not only driven by a concern for welfare, but also affected by the specifics of the local context, especially the needs of the local community; we now turn to an exploration of how the situated contexts of schools affected policy enactment.

Situated contexts
Right from the beginning of the crisis, the Situated dimensions of context became particularly significant, with schools making decisions on their approaches based on their local community's needs. For example, this headteacher noted that their parents did not have means of receiving an automated message from school due to low incomes, but were reliant on the school for guidance: We serve an area like I said of high deprivation and high social poverty and distal poverty and our parents to date have relied really heavily on us for information. [. . .] All of a sudden we're placed in a situation where we have to communicate quickly, at speed and we're spending huge amounts of time on the telephone making sure everyone's got a message. We don't have a system where we can send out a communal email or a communal app. Some of them [parents], quite a lot of them will say my phone's got no data and I've got no money to buy a top up. It was a logistical nightmare. (Headteacher, NW, DCDT) In response, instead of sending out one message to all, this school leader and their staff spent time phoning individual families in an attempt to communicate key messages. Here the situated context determines their approach to the guidance, how long it takes to enact, and who has to be involved.
Another example of schools enacting policy in ways which were informed by local circumstances arose from the key worker policy from government, which stated that children of key workers were able to attend in person to allow their parents to work. The interpretation of this guidance varied by school, due to the variation in employment patterns among parents and practical context of staffing. In this example, the high volume of key worker children meant the school had to draw up a priority list of occupations they saw as most important: . . . we do have a lot of key workers. So that was quite tricky so that in that first instance when we were only allowed to have 15 in a bubble [class]. What we had to do was make a priority list of key workers so, for example: the ambulance drivers, the people in the hospitals, those sorts of people got priority over somebody in Costa [coffee shop] (Head B, S6, LTD) Here policy is enacted in ways which are dependent on local context, but also the values of the school leaders tasked with deciding who should be able to attend school. Presumably informed by the political directive in England to 'protect the NHS' (National Health Service), they decided to prioritise emergency workers over those in retail. This was not a unique approach: We spent the next two days [just before the first lockdown] waiting for government guidance about what constituted a key worker, and it never came. So we made our own decision by the Friday morning. (Head S4, LTD).
In the absence of information, these schools made sensible, locally appropriate decisions on how they would enact the key worker policy -often ahead of government guidance where necessary, as also seen in the example of free meals above. Similarly, some deployed some flexibility as time went on to the definition of 'vulnerable' children, which was supposed to be linked to the involvement of social care services: I'll be really honest, we bent the rules a little bit for other families we knew didn't have social care involvement but were really vulnerable [. . .] What we also saw was increasing levels of need. More and more parents coming to us in crisis. (Headteacher, NW, DCDT) Here we see the entwinement of the importance of the professional context (values) and the situated context of local vulnerable families. This headteacher also had the staff and space to allow these children to attend if they felt it necessary, which leads us to consider further the material dimensions of context.

Material context
As already suggested, different material dimensions of the school context became important during the Covid crisis, as school leaders had to adapt to new ways of working and priorities. Material dimensions are described by Braun et al. (2011) as 'tangible yet random factors' such as buildings, infrastructure and budgets, which, in this situation, became vital considerations. For example, the guidance in the summer term of 2020 that stated that children returning to school had to work in 'bubbles' of up to 15 pupils and maintain social distancing was consistent across schools, but individual schools had to interpret this differently depending on the staff available, their classrooms, numbers of entrances and exits, and many other practical considerations. The scale of this endeavour was emphasised by one headteacher: The schools have got to be completely rejuvenated. You've got to clear it out, you've got to get rid of everything. You've got to work out how all the bubbles are going to interact and all the rest of it. That was phenomenally stressful. (Headteacher, NW, DCDT) How exactly this 'rejuvenation' was to operate, depended on factors which had previously been seen as insignificant. School leaders explained how factors such as levels of ventilation in large classrooms, location of hand-washing facilities, and outside spaces affected how they organised their school day, including which children could attend school and for how long. At one school, for instance, the number of playgrounds allowed them to stagger lunchtimes and do lessons outside, and therefore take in more children: We've got lots of playgrounds, we've got one, two, three, we've got five playgrounds, we did as much as we could outside and staggered the lunch times a little bit . . . (Head S7 LTD).
The availability of two entrance points to the school also ensured 'the children aren't bottlenecked and it's all outdoors and that works really, really well'. These issues of physical space came to the fore, given the airborne nature of the virus, as previously unimportant material factors became significant in how decisions were made. As emerging analyses of the Covid period drawing on new materialist perspectives have suggested, during this time the 'sociomaterialities of air, breath and the ventilation of spaces have become central features of COVID risk discourses' (Lupton 2022, 131).
Further material aspects relating to staffing and funds were also important to policy enactment. While used to managing staff in normal times, during the Covid period some school leaders had to deal with the complexities of teachers' personal situations, which had an influence on who could be in the building: [Because of the number of pupils on site] we had to have all staff who were available in school.
[But] some were clinically vulnerable, some vulnerable, some pregnant, staff who were just so anxious they didn't want to be in school. (Head S4,LTD) Thus the physical and mental health of the staff became a further factor determining how provision was organised.
In other cases, funding was a key factor: School 5 in the LTD project was heavily affected by the loss of earnings from renting out school property, a reduced school roll due to students not arriving from overseas, and the additional cleaning costs, to the extent that they had to reduce their staffing by putting their specialist teachers into the classroom and the headteachers both teaching in class. Their particular budgetary context thus had implications for the financial impact of Covid, and how they were able to respond to the diverse needs of children returning to school; for example, one of the headteachers commented how without these financial effects, 'we'd have these reading and maths intervention teachers doing what we first employed them to do' (Head A S5, LTD).
For others, access to additional funding became important, in order to allow for extra money to be spent on new issues such as food provision: Eventually we found our way into the kind of food share, food bank system and we've been getting free food that I go and collect every week to give out to families from a more centralised thing in [city]. . . . We were paying for it ourselves, at one point for a couple of months we were spending about £1000 a month on food parcels for our 25 most vulnerable families (Head S7 LTD) As well as having the funds to allow for this spending on food parcels, this system was also dependent on the ability of the headteacher to collect food and deliver it to families in her car. As we have argued elsewhere (Moss et al. 2021a) schools experienced fragmented support during the pandemic, with often ad hoc solutions operating in different ways in different places. A further example of this kind of vulnerability came from another school in the LTD project, which was reliant on the goodwill of a single local post office worker to deliver the home learning packs to families every week.
This drawing on local resources -also seen in use of local charities and food bankswas a further demonstration of how distinct aspects of material context became important in how policy was enacted.

Discussion
Lupton and Willis have argued that 'The Covid crisis is a complex and ever-thickening entanglement of people with other living things, place, space, objects, time, discourse and culture' (Lupton and Willis 2021, 4). In our data, this is demonstrated through the complex and shifting array of factors that determined how school leaders responded to policy: from personal values, cultures which emphasise a duty of care, and growing levels of poverty, to physical space, staff health and available funds. The entanglement, to use their phrase, of health policy with education policy, resulted in an entirely different method of interaction between government and schools, forcing both into unfamiliar territory. As one headteacher explained, there was no plan for this occurrence: . . . nobody had sent something through, 'There you go. These are the things you should think about as a school.' (Head B S1, LTD) Another contrasted their approach during the crisis to 'peacetime', likening the pandemic to the status of being on a war footing. What we wish to emphasise here is the distinctiveness of policy enactment in primary schools during the crisis, but the continued relevance of understanding this enactment as determined by aspects of context.
The interview data discussed here reveal a form of policy enactment in a crisis, which is specific to this period. As we have seen, enacting policy during the Covid crisis was a process dominated by the need to act at speed, based on a knowledge of the local circumstances, and informed by the prioritisation of children's basic needs. Policy was received in novel ways which caused feelings of shock and anxiety, and the lack of effective communication and consultation means that many school leaders felt decisions were being made without an understanding of the real issues. Notably, our findings do not suggest a process of policy interpretation motivated by accountability, Ofsted or schools' reputations, as in normal times (Bradbury, 2019). Nor is it a process of careful thought-through consultative decision-making. Instead, crisis policy enactment is a process of making ethical and moral decisions in response to outside rules and guidance, within the confines of resources and in the light of local priorities. These decisions demonstrate the agentic nature of school leaders -in line with the idea of autonomous schools (Boyask 2018), but the extent to which they felt empowered is less certain. Several participants characterised their approach as surviving or fire-fighting; some said they were constantly 'on the back foot', coping rather than driving things forward. Others, in line with the emerging international research (Arnold, Rahimi, and Riley 2021;Thomson, Greany, and Martindale 2021;Reid 2022), were under huge stress and pressure during this period. In interviews they spoke of crying in their cars, being exhausted, and a 'heartbreaking' situation.
Our findings suggest that this form of crisis policy enactment was affected particularly by the professional, material and situated contexts of schools (Ball, Maguire, and Braun 2012), though in different ways during the crisis, as new aspects of context became significant and others less so. However, we also note that the external context (support from local networks, pressures from broader policy context) may also have shaped responses. Support networks varied hugely due to the fragmented nature of the education system in England, with schools operating in Local Authorities, Multi-Academy Trusts, as stand-alone academies, and within faith organisations. For reasons of space this is the topic of a further paper. The wider policy context of high stakes tests and inspections, in contrast, became far less significant as testing and inspections were suspended, providing a degree of freedom unseen since the 1980s. Thus we have demonstrated how context still mattered, but in a different way, adding to our overall understanding of the responses of the education sector in England to the challenges of the Covid period. By conceptualising crisis policy enactment and its distinctiveness, we have also added further nuance to the idea of policy as something enacted in real spaces by real people, subject to various constraints and with varying degrees of power.

Conclusion
Overall, the findings of this analysis of how primary school leaders responded to the Covid crisis of 2020 and 2021 in England show that schools 'did' policy in a very different way, and that policy itself was very different. The abrupt decisions to close schools and sudden changes in rules and restrictions lead to unique and unprecedent policy enactments in primary schools, which were guided by new aspects of context. These enactments were responsive in nature, adapting to changing circumstances on a near-daily basis at times, but were guided by ethical and moral principles which, in our view, show the school leaders in a very positive light. School leaders were operating under immense pressure, in 'constant crisis management mode' (Fotheringham et al. 2022), in a policy context which was wholly unconducive to effective home-school communication and consultation and measured, deliberative decision-making.
We characterise what school leaders were doing as crisis policy enactment; that is, the enactment of policy quickly based on the immediate priorities of the school community, knowledge of local circumstances and a clear ethical and moral stance. It suggests a response to policy which is both coping (and reactive rather than proactive) but also agentic, demonstrating a commitment to children's welfare and a belief in the power of schools to make a difference. Relatedly, the kinds of enactment explained here also reinforce the unique nature of primary schools as public institutions which are embedded within their local communities.
These findings suggest an important practical lesson for any future crises: that the government's practical systems of providing guidance and regulations were unhelpful and in places, damaging to schools. One lesson from the crisis could be a well worked out system of consulting with schools on urgent matters of policy, and providing information in advance and in a form that allows school leaders to make decisions in their schools' best interests. Producing guidance and regulations which appeared at times to be distant from the realities of life on the ground, and announcing it via the media, caused resentment and frustration among some school leaders, and caused tensions between schools and parents.
While we do not assume applicability to alternative crises given the unique nature of the pandemic challenges face, we note that our findings cohere the existing scholarship on school closures following disasters, which suggested that school leaders' knowledge of their local contexts is vital (Harmey and Moss 2021); this should be reflected in how guidance is issued. Exceptional crisis circumstances require governments play close attention to the experiences of school leaders and teachers in their localities, or risk becoming a further problem when support is needed most.
School leaders came under huge amounts of pressure, tasked with decisions which were high risk as they related to children's and staff welfare, within a system which was completely unprepared for mass school closures. There were no models, research or examples for headteachers to work from in England, meaning that they each had to invent a system of coping based on their own and their staff's capacities and their school communities' needs. In this context, this form of crisis policy enactment can be seen as both a triumph and a disaster. That school leaders were ever in this position was disastrous, indicating a system that was massively unprepared for a crisis, and a government that was not interested in listening to teachers' experiences and opinions. However, their responses -their priorities, effort, sheer hard work and determinationwere a hugely positive sign that when needed, primary schools were able to provide for their communities to an extraordinary degree.