The impact of austerity measures on local government funding for culture in England

ABSTRACT In the 2010s, austerity measures had a severe impact on the level of spending on cultural activity. In England, the potential for this was especially great due to the reduction of resources available for local government, the largest funder in this field. Whilst prior work has attempted to establish overall reductions in spending, little attention has been paid to variations in cultural spending across the country in this period of austerity, or which types of cultural services have been cut and why. Here, we compare local authority spending in 2009/10 with 2018/19 to provide further detail on the varied picture that has emerged in this period. Whilst there is indeed a significant reduction in the level of spending on culture in this period, we identify significant variation in local experiences of austerity across England. We also show which types of cultural services are more vulnerable to cuts and explore a range of contextual factors as potential explanations of these patterns.


Introduction
Over the last decade, many arts and cultural organisations have faced increasingly difficult financial circumstances (Mansfield, 2014;Harvey, 2016;Rimmer, 2020;Terwey, 2017). In England local authorities have been subject to significant cuts in this period of "austerity". Yet discussion of the specificities of these circumstances has had a tendency to be alarmist, speculative, predictive, or light on evidence (e.g. Naidoo, 2015;Starkey, 2015). As such, ten years after the onset of austerity policies there has been little analysis exploring the cuts across local authorities. Average figures for expenditure cuts are commonly quoted which give the impression that austerity has hit all areas equally, producing equivalent reductions in spending on cultural provision. As well as masking geographic specificity, much analysis considers only broad categories of "cultural" spending, which may be treated as representing "the arts", despite covering a much wider remit. Commentators have also been quick to explain spending patterns as the consequence of the purposive action of individuals seeking to "protect" or disregard cultural services (Harvey, 2016;Mansfield, 2014). Whilst our methods do not allow us to discount this as a potential explanation, we identify a number of additional factors with a bearing on spending decisions. Studies pertaining to the way decision-making and behaviour are influenced by an interplay between structure and agency within specific institutional settings are welldocumented in political science (see Lowndes, 2002). Here, March and Olsen's insight that "political action is inadequately described in terms of rationality and choice" (1983, p. 738) has informed a large volume of studies including recent scholarship on the influence of institutional factors on the response of local government to austerity (Ferry, Coombs, and Eckersley, 2017). In offering an analysis of a decade of local authority expenditure data informed by such scholarship, our aim is to contribute to a richer and more contextualised understanding of austerity as a political strategy with varied manifestations. This is part of a broader interest in the role of austerity measures in reshaping cultural services.
In so doing, we demonstrate significant geographical variability in local authority cultural spending, and between different types of expenditure. Curiously, research into the nature of spending cuts to cultural services has been limited, with the issue of austerity being explored more through questions of political discourse and individual organisations' responses and less in terms of the transforming shape of cultural funding itself. While we acknowledge that scholarship has not ignored austerity, as a recent literature review illustrates (Rex, 2020a), we suggest that it is equally important to supplement a focus on austerity discourses with a contextualised exploration of the nature and extent of the cuts. The picture that emerges in this article is of austerity as a variable concept rather than a singular empirical reality. We show that austerity has taken different forms from one local authority to the next. We point towards a number of institutional factors as potential explanations as to why certain approaches are advanced and not others. In doing so, our aim is exploratory rather than explanatory. Further empirical work is needed to identify the specific factors in play within local settings. Nevertheless, existing scholarship on how local government works allows us to offer an informed commentary on the patterns in the data on income and expenditure identified by our analysis.
To illustrate the variegated geographies of austerity, we present a categorisation system for local authority spending to show that, whilst certain key patterns certainly prevail, "austerity" has not been experienced uniformly, nor in ways that necessarily fit a prevailing narrative of blanket spending cuts. We find that patterns of cuts have emerged over the past decade as certain types of expenditure appear more vulnerable to cuts than others. We show that, although particularly vulnerable as a non-statutory form of spending, arts and cultural spending has been preserved in some areas of even severe deprivation facing particularly constrained financial circumstances. Relatedly, although the general picture that emerges is, unsurprisingly, one of spending cuts (albeit to different degrees), this sits alongside evidence of increased income generation from arts and culture, which raises important issues about the role of cultural institutions and their beneficiaries going forward.
As we develop an argument about the considerable variation in expenditure following a decade of austerity, it is firstly important to acknowledge that provision in these areas has never been uniform. Since the 1972 Local Government Act gave local authorities the power to provide cultural services as they saw fit, there has been significant policy change across the areas covered by the "Culture and Heritage" subcategory (see Table 1), each of which has its own political and legislative history. The salient point for our analysis, however, is that there has always been variety in the nature and level of activity. This is because provision is discretionary (or permissive), not mandatory, and is therefore contingent upon many factors. For example, although the 1845 Museums Act is often cited as a watershed moment in the history of local authority museums for the permission it gave councils to raise taxes to establish museums, the amount of money councils could raise was small which meant they were reliant on the donations or bequests of wealthy elites, a dynamic which partly explains the patchy nature of service provision today (Hill, 2016, p. 43).
Situating current spending patterns in their organisational and institutional contexts, albeit in an exploratory way, is an important concern of this article due to the relative neglect of the influence of these contexts on decision-making processes and outcomes (cf. Gray, 2002;Rex, 2020b). This is in spite of a long tradition of such approaches in political science where explanations based on individual choice have been met with approaches seeking to explore the interaction between individuals and institutions (Lowndes, 2002). Because our method is quantitative, our paper presents findings pertaining to patterns identified from the data on income and expenditure. However, as part of our objective to contextualise these findings, we also suggest a number of ways in which the patterns contained in this data might be understood, focusing on institutional factors as potential explanations. Further empirical work would still be necessary to evidence these claims. Indeed, this paper forms part of a larger project including interview-based research. As such, we are not trying here to make conclusive claims in respect to these observations. Rather, our observations with regard to the potential factors that explain spending decisions highlight that these are constrained and enabled by structural factors beyond individual decision-makers. Within the institutional perspective guiding our observations, spending decisions are not a straightforward matter of individuals acting according to their preferences as institutions "give order to social relations, reduce flexibility and variability in behaviour, and restrict the possibilities of a one-sided pursuit of self-interest or drives" Olsen, 2006, p. 7, citing Weber, 1978). We hope to highlight the specificity of the impact of austerity in local contexts, and stimulate further thinking on the circumstances driving these spending patterns.

Local authority funding in an age of austerity
Since the austerity programme pursued by the UK government began in 2010 local authorities have faced particularly severe funding reductions. Analysis by the National Audit Office found an average 49.1% real-terms reduction in grants from central to local government over the period 2010-2018. This is forecast to increase to 56.3% by 2019-20 (Morse, 2018, p. 15), completing a decade of year-on-year decreases (Brien, 2020, p. 6). The real spending power of local authorities fell by close to 30% over the period from 2010 to 2018 (Morse, 2018, p. 7). Both the revenue support grant, the general funding provided from central to local government, and grant funding for services such as housing have been cut, while demand for councils' services has increased. These cuts and demand pressures have been compounded by ongoing reforms to local government finance with the expressed intention to make local authorities financially self-sufficient (HM Treasury, 2015, p. 4). The ideological content of these policies has been discussed extensively in the literature, where reducing public debt through austerity is understood as part of a conservative political project to reshape the state and welfare system (Blyth, 2013). With these politics in mind, it is unsurprising that local government has been subject to substantial cuts. Despite the rhetoric of "localism" promoted alongside budget reductions, the UK remains a highly centralised state, with devolution resulting in English local authorities having less autonomy over spending than their Scottish, Welsh and Northern Irish equivalents.
This institutional framework creates a particularly challenging dynamic for local governments in England. They retain responsibility for delivering services yet have limited control over their budgets. As recent analysis by Gray and Barford (2018) highlights, cuts have had an uneven impact because some councils rely far more on central government funding than others. However, where these authors find a "notable relationship between grant dependence and the cuts in local service spending" (ibid, p. 553) our findings suggest the relationship between grant dependence and spending on "Culture and Heritage" is less clear-cut.
Some local authorities are better placed to mitigate central government cuts by generating revenue from sources such as council tax, business rates and fees (e.g. for parking) (Amin-Smith et al., 2019). These factors have led scholars to raise concerns about central government ambitions for local authorities to become "self-sufficient" via income generation (Eckersley, Murphy, and Ferry, 2019). These variations in financial circumstance mean that cuts to the amount received by a local authority from the central government grant do not necessarily translate into equivalent cuts in spending. As in our analysis, a more appropriate measure of financial circumstance is the amount councils are spending on services annually rather than grant funding reductions.
The coronavirus pandemic has highlighted the limitations of a financial model which is reliant on earned income and therefore exposes councils (and by extension, service users) to high levels of risk. Currently, council leaders are warning of further spending cuts as coronavirus leads to a drop in income, with recent commentaries indicating the acute impact on non-statutory services such as those examined here. Commenting on its £197.6m overspend, the chief executive of Leeds City Council spoke of cutting spending on non-statutory cultural services completely (Drury, 2020). Elsewhere, charitable trust Vivacity, which has been running cultural services in Peterborough since 2010, ceased to operate due to the impact of coronavirus, as a financial model heavily reliant on income from ticket sales and admissions fees proved unsustainable (Edwards, 2020). Given that it is potentially expedient for political figures to present their circumstances in alarmist terms as part of an effort to obtain additional financial support, such reports should not necessarily be taken at face value; nevertheless, it is likely that the coming years will see further reductions to arts and cultural spending, if not necessarily total withdrawal, coupled with the intensification of income-generation strategies as local authorities prioritise the limited funds at their disposal towards mandatory services or towards investing in commercial ventures with the hope of financial returns (Christophers, 2019).
Existing research has provided valuable insights into where cuts are most likely to fall and why. Both Streeck (2014) and Gray and Barford (2018) show that non-statutory services are more vulnerable than mandatory ones, describing an average cut of 35% to "Cultural and Related Services" spending over the period from 2010 to 2017 (ibid, p. 556). That cultural provision has been severely affected is partly because local councils are under no legal obligation to provide these services. However, even when there are legal obligations, such as is the case with libraries, this has protected services only to a limited extent, given the ambiguous wording of the relevant legal frameworks (Anstice, 2015;Hitchen, 2021).
Scholarship has also examined why local authorities have tended to implement budget cuts rather than resist them. Analysis of responses to austerity in Leicester conducted by Davies, Bua, Cortina-Oriol, and Thompson (2020) makes an important point as to why local authorities adhere to these requirements. As part of their diagnosis of the "austerian realism" which produces this compliance, the authors link the diligent delivery of austerity to the pressures resulting from the spectre of central government intervention should an authority fail to set a balanced budget, which is a legal requirement. Central to their analysis is the way fear of repercussions from central government (ibid, p.61) in the present day combine with the shadow cast by the failure of previous attempts to resist successive central government decisions. This results in what they term "legalism"strict adherence to legal requirements meaning councils tend to "proceed" "diligently with cuts and restructuring manded by national government" with little resistance (ibid, p. 63).
Taken together, these studies develop our understanding of the general circumstances facing local authorities and why austerity has taken the form it has. But they do not examine the impact of austerity on local authority spending on cultural services as it has emerged across England, as we do here. Notably, some of these overall patterns are not mirrored in spend on "Culture and Heritage". Although Gray and Barford's study is one of the first to highlight cuts to "Cultural and Related Services", figures under this heading are somewhat problematic when used as the basis for evidencing a "shift away from discretionary spending … towards mandatory spending" (2018, p. 543) as library spending, a mandatory service, is included here. Our analysis of local authority spending between 2009/10 and 2018/19 allows us to provide a comprehensive picture of the impact of austerity on those discretionary services within the "Culture and Heritage" category, indicating uneven impacts across types of cultural service. Building on the contextual detail contained within these and related studies allows us to suggest factors that may be influencing these patterns at the local level.

Methodsexamining expenditure categories
We draw on the Ministry of Housing, Communities and Local Government's (MHCLG) data on local authority revenue expenditure and income. This data is published on an annual basis and is publicly available. It relies on individual local authorities collating and submitting data on what they actually spent and the income they generated in the previous year. Although the provision of clear guidance by MHCLG does not preclude variation between local authorities in terms of what they include in each category (see MHCLG, 2019b), we base our analysis on the reported figures. Expenditure as a whole is recorded against a number of macro categories. Although these have changed over the decade under scrutiny, by 2018/19, the size of gross expenditure in each reported category across England is given in Table 2.
The picture of spending by local authorities on arts and culture in the period of austerity has often been opaque, as it has been measured in many different forms in this period.
The key distinction between different measures is which categories from those available for the recording of local authority expenditure are deemed appropriate to represent "expenditure on arts/cultural activity". As noted above, sometimes analysis pertaining only to these macro-categories is presented, but often the "Cultural and Related Services" category is broken down further by those seeking to establish what patterns of spend relate specifically to "art and culture". The subcategories of expenditure available for consideration within this macro-category are shown in Table 1.
The "Funding Arts and Culture in a Time of Austerity" study (Harvey, 2016) reports figures on four subcategories: "Arts Development & Support", "Museums & Galleries", "Theatres/Public Entertainment", and "Library Service". As a counterpoint, figures produced for the National Campaign for the Arts (NCA) Arts Index (NCA, 2020) used only three of these subcategories (excluding "Library Service" from the four above), whereas data in Arts Council England's (ACE) online dashboards uses five subcategories of expenditure (the four above, plus "Archives" [ACE, n.d.]). Each of these organisations operates within different contexts with their own respective priorities and responsibilities, influencing decisions made as to which figures are reported. Issues of remit and perspective aside, however, such differing definitions can result in figures for spend on culture which differ drastically, by up to £80 per person. Gross expenditure in these subcategories across England in 2018/19 is shown in Table 3.
It should be noted that our focus here is on revenue expenditure only (i.e. annual operating costs). This covers the majority of expenditure (MHCLG, 2010-2019), but also avoids distortions by one-off capital projects. Our aim is to track the patterns in year-on-year employment and running costs captured in this data. Also, we do not, as is common, consider net expenditure, but rather consider patterns of income and expenditure separately here. 1 As our focus is on the nature of a climate of cuts in spending, it would potentially be misleading to interpret a rise in income as a "cut" in spending, as would be the case when considering net expenditure. As such, we firstly consider outgoings on employees and running costs, then consider income generated by cultural activities. For reference, however, Table 4 shows the relevant levels of income under the same categories as Table 3. The categories of "Theatres & Public Entertainment" and "Sports and Recreation Facilities" show proportionally higher levels of income, with "Library Service" proportionally lower.
This article is necessarily limited to a discussion of patterns of service spending and variations between different types of authority and between different types of provision. We have been able to draw on existing research into the general context local government employees operate within but do not control (e.g. Morse, 2018), which has allowed us to speculate on potential explanations behind these changes. However, our methods do not allow for an in-depth investigation into the specific circumstances which bear on individual local authorities, producing particular patterns of spending and approaches to service provision. Rather, we present analysis of national patterns of change in expenditure, differential geographical patterns of change, and patterns in expenditure and income levels. Our analysis reveals the variation between different parts of the country and different forms of cultural spending that, to date, have been obscured by a tendency to understand austerity through averages.

National patterns of change in expenditure
In analysis conducted mid-way through the period covered here (2009-2019), Harvey noted that "the biggest surprise" when considering expenditure data is that cuts were not particularly excessive in the area of culture (2016, p. 9). Overall patterns of expenditure continued their downward trajectory following this point, however, and later analysis suggests that reductions to areas of discretionary spending have been substantial. Harris et al. note that "culture and related services continue to be an area where the vast majority of councils (around 90%) have cut spending by more than average" (2019, p. 36). Likewise, Cooper observes "councils in every region have cut their annual arts and culture budgets by more than a thirdand by more than £15 per person" (2020, p. 13). Before taking a closer look at disaggregated data, this section establishes the national pattern of change in expenditure in relation to both the macro-category of "Cultural and Related services" and also its major subcategories at the national level.
The overall pattern of reduction in each category here is to be expected. Figure 1 also shows that each major subcategory in the "Cultural and Related Services" category has seen a disproportionate, yet differential, reduction in spending in comparison with Total Service Expenditure. In line with other studies of how austerity measures from central government have translated into spending cuts within local authorities (Streeck, 2014;Gray and Barford, 2018) we find a shift away from discretionary spending. However, Figure 1 shows clear differences between discretionary services that become visible only when we consider trends in expenditure between the subcategories.
The shift away from discretionary spending shown in Figure 1 entails a diminished role for local governments as social democratic institutions providing benefits and services as part of a concern with equality of access to public goods. As existing scholarship illustrates, the financial constraints of austerity have forced local authorities to prioritise the limited funds at their disposal towards basic statutory services provided on a selective basis to those deemed "in need", or towards investing in commercial ventures with the hope of financial return (Gray and Barford, 2018;Christophers, 2019). Or, in the case of libraries, included in the "Other" category in Figure 1, revising the meaning of their obligations entirely, paving the way for cuts. These changes have implications for the image of local government and the orientation of people towards its redistributive role as a guarantor of universal benefits and welfare as well as the legitimacy of representative democracy more generally (Lawson, 2008, cited in Gray andBarford, 2018, p. 559). A context where local governments find themselves forced to play a diminished role in sustaining the social and physical infrastructure of placeits libraries, community spaces, museums, parks and theatresrisks the slow erasure of a broad understanding of these spaces as part of a public sphere to which everyone is entitled. These processes also fuel the dismissal of the importance of the local state itself, furthering a political project that seeks to repurpose the local state away from social democratic ends. However, given the legal frameworks within which they operate, it is not surprising that councils focus on delivering balanced budgets instead of services which they are under no obligation to provide or, as has often been the case with libraries, where obligations are ill-defined. Looking further at the overall spending patterns presented in Figure 1, we observe that the major subcategory of "Culture and Heritage", which we concentrate on in the rest of this article, has been the least affected element within the "Cultural and Related Services" macro-category and some "levelling off" in overall decline is perceptible here from around 2016 onwards. We can also see that the "levelling off" of decline in total expenditure from around 2014 is not mirrored in the macro-category, which continues its decline. This levelling off has been interpreted as evidence of an overall trend for local authorities to "protect" cultural provision of a certain type at the expense of other forms. Harvey, for example, suggests that "councils have tried to protect these services where they can" (2016, p. 9). Likewise, Mansfield's research argues that "local authorities continue to value arts and culture" due to belief in their potential to support local economic development (2014).
We would argue that there is another way to look at these trends, however. This relative pattern may in part be a result of the fact that absolute expenditure in the "Culture and Heritage" major subcategory (comprising the five subcategories set out in Table 1) began at a lower level in this period and thus significantly more expenditure was available to cut in other subcategories such as library services, parks and sports and recreation, as Figure 2 shows. To focus on the example of library services, local authorities would have spent a larger proportion of their budgets on fulfilling their statutory obligation to provide a "comprehensive and efficient" service as per the 1964 Act prior to austerity measures. However, with the pressure to cut spending came the redefinition of "comprehensive and efficient", allowing councils to make substantial and ongoing reductions in expenditure on library staff and buildings. With spending on "Culture and Heritage" starting at a lower level, officials may have had difficulty identifying further expenditure cuts without leaving their budgets insufficient to maintain services of any sort. Rather than illustrating a preference to protect services, as other commentators have claimed, we would suggest that these relative reductions are much more likely to be happening due to organisational and institutional factors of the sort outlined here. Figure 3 shows how spending on "Culture and Heritage" is distributed within its constituent subcategories. Looking at the percentage change in spending across these subcategories it is clear that "Arts Development and Support" is the most vulnerable for spending cuts, constituting only 13% of this shrinking grouping in 2018/19, down from 17% in 2009/10. This translates into a reduction in total spending from £182.7m in 2009/10 to £110.4m in 2018/19. Whilst the share of expenditure in other categories has increased, this is nevertheless a larger part of a smaller wholefor instance, the increase from 38% to 43% in "Theatres and Public Entertainment" represents a reduction in spending from £420.4m 2009/10 to £354.5m in 2018/19. Although we return to this line of analysis below, these decisions are taken in the face of fiscal and political constraints. As such, we would suggest that the particular vulnerability of "Arts Development and Support", a category comprising "direct expenditure on the arts, including photography, purchases works of art, and temporary exhibitions", excluding those purchased for museums and galleries, "the costs of artists in residences" and "grants or other contributions to individuals or voluntary organisations for the development and support of the arts" (MHCLG, 2019b, p. 57) is more likely to result from a political preference for low-visibility cuts and/or those that are more straightforward to implement rather than a straightforward disregard for the value of these services.
The continual restructuring occurring within local authorities may also be relevant here. As other studies have identified, a move towards having fewer, larger departments means that those responsible for budgetary decisions are unlikely to have in-depth knowledge of these services and their value (Lawley, 2003, p. 77;Rex, 2020b, p. 190). "Culture and Heritage" in Bradford council, for example, sits in the "Department of Place" which also comprises planning, transport, highways, economic development, customer services, neighbourhoods, sport and waste. The Libraries, Museums and Galleries Manager (responsibility for libraries was added to this role in 2017) is not a chief officer with two layers of management between their role and the chief executive. With the hierarchical nature of decision-making in local government in mind we can see how even if senior management did have the time to consider the value of supporting a range of cultural resources to create a functioning ecology (Gross and Wilson, 2020), their responsibilities to deliver a balanced budget would prove more dominant.
With the national picture established, the next section draws out the significant variation across local authorities in terms of how spending within "Culture and Heritage" has changed over the decade considered here. As our analysis shows, the effects of austerity at local level call into question the frequency with which aggregate figures have been used as a proxy for the impact of austerity on local authority cultural provision.

A decade of cuts? The uneven experience of austerity measures across England
This overall picture of a somewhat disproportionate reduction in spend on "Culture and Heritage" over time hides much variation. Although it is true that most areas have implemented spending cuts on "Culture and Heritage", the level of these ranges from 1% to 94%, while a minority have increased expenditure over time. The classification system we offer here aims to foreground these variations. By drawing out the range of patterns of expenditure on the subcategories within "Culture and Heritage" in the decade between 2009/10 and 2018/2019 we question the way "austerity" has been interpreted to date. 2 Our analysis demonstrates that "austerity", as it has impacted expenditure on cultural provision by local authorities since the implementation of budget cuts by central government towards the end of the 2000s, is not a fixed concept which can be read as synonymous with savage cuts alone.
There are substantial variations between authorities in terms of how spending on "Culture and Heritage" has changed during this period, mirroring the uneven impact of austerity generally. Additional empirical research would be needed to identify how these spending patterns translate into organisational change and its effects on those using these services. Nevertheless, our analysis shows that the state of local cultural provision is not a simple picture of decimated local cultural services but rather a complicated diversity made up of a range of responses to a national government mandate. Our analysis is indicative of the substantial disparities between local authorities as regards their ability to cope when implicated in broader trends not of their choosing.
We propose that these issues can be best explored by sorting local authorities into categories relating to different patterns of expenditure change between the start and end of the decade 2009/10-2018/19. Analysis of expenditure change over time is important for this period because this is where we can start to explore the effects of austerity policies. It also demonstrates the extension of austerity far beyond the dose of "short sharp pain" promised at the beginning (Hitchen and Raynor, 2020). However, it is important to establish that many areas with low cultural spending at the end of this decade have always had low (or no) expenditure per person in certain subcategories which constitute "Culture and Heritage" spending. As such, while austerity measures have undoubtedly prompted the dismantling and recalibration of cultural provision for some, there were already many local authorities that, for reasons that remain underexplored, did not spend in these areas. In line with the aim of our analysis, our maps represent (uneven) expenditure change over time, illustrating differences between places and types of spend, rather than pre-existing regional discrepancies.
The categories in Table 5 provide a simple comparison of expenditure per person in 2009/10 with 2018/19 as follows: (1) No expenditure at all in this decade, or rare "one-off" expenditure 3 (2) Despite repeated prior expenditure, expenditure at zero by the end of this decade 4 (3) Expenditure in 2018/19 over 50% lower than in 2009/10 (4) Expenditure in 2018/19 under 50% lower than in 2009/10 (5) Expenditure in 2018/19 higher than in 2009/10 5 The 152 local authorities considered here fall into these categories as shown in Table 5 for expenditure. Table 6 considers changes in income using equivalent categories.
Here we begin to see the considerable variation in how local authority finances have changed since 2009. Figure 4 shows the geographical distribution of the categories of expenditure change using a hexmap where 1 hexagon equates to 1 authority. 6 Further nuance is added to categories 4 and 5 above as per the key in Figure 5. As the overall tendency is to reduce  spending, some authorities have reduced spending overall but have nevertheless risen in their relative ranking against other authorities of the same type, whilst others have fallen. So category 4 is split to show those who have reduced spending but may nevertheless be interpreted as having "protected" spend more than areas of a similar type. For instance, Figure 5. Key to maps. whilst spending per person in Oldham is £0.20 lower in 2018/19 than 2009/10, it has nevertheless risen to be the 10th highest spending Metropolitan District from 25th, despite this reduction. Similarly, whilst spending might be higher in 2018/19 than in 2009/10 in a given authority, this higher figure may nevertheless be lower than the average year-on-year spend over this entire decade, and so Category 5 is also split to capture this nuance. Authorities in the rare "5b" category are named on the maps below. The presence of a range of categories across England highlights the variation in spending patterns between 2009/10 and 2018/19. Aside from the constancy of the Isles of Scilly (with zero annual expenditure on "Culture and Heritage"), the map shows much variation (a full collection of expenditure and income maps for each subcategory is available at http://www.petersoc.com/maps/). There are clear clusters of major reductions in expenditure on "Culture and Heritage" in the North West, the West Midlands and in parts of the South West where we see the red, hatched hexagons of Category 3. However, 14 of the 32 London boroughs also fall into this category. Although the concentration of public funding in London has been a long-standing issue for bodies such as Arts Council England (see Cooper, 2020) this map highlights that London has not been immune to cuts to local service spendingfar from it. For example, in 2018/19 Enfield Council spent £665,000 (£1.99 per person) on "Culture and Heritage", down from over £4m the previous year, and Hammersmith and Fulham Council spent £177,000 (£0.95) per person, a fall of 94% from the £2.7m (over £15 per person) spent in 2009/10 and 2010/ 11. Even in authorities in the capital, therefore, we can find severe cuts.

Expenditure patterns
Looking within category 3 where spend has dropped by over 50%, there are 12 areas where spend has dropped by over 75%, 4 of which are London boroughs. What we find is that these councils have experienced something of a "cliff edge" year, where spending on "Culture and Heritage" drops dramatically, and does not recover. This "cliff edge" comes at different points. York City Council, for example, was spending between £17 and £20 per person from 2009/10 to 2014/15, dropping to £9 by 2015/16, prompting its then chief executive to describe the trust running the council's museums as more of a "cultural business" than a "cultural service" (Barnes, 2015, quoted in Atkinson, 2015. For Wandsworth, the imperative to make savings seemingly came much earlier, with spending falling from £20 per person in 2009/10 to £7 per person in 2010/11, dropping further to £4 per person in 2016/17. These differences are indicative of increasing inequality between places in terms of their ability to mitigate the cuts and sustain cultural services. When we consider the subcategories within the "Culture and Heritage" category, there are some clearer trends. The majority of local authorities in England have made drastic cuts to "Arts Development and Support"; we can see the increased prominence of Category 3 in Figure 6 compared to Figure 4. Despite prior spending, twenty local authorities cut spending entirely during this period (Category 2). While this pattern appears to conflict with apparent shifts within cultural policy towards increasing participation in cultural activity, the main factor at play here is likely the type of spending covered by this category. Various factors make this type of spending more vulnerable to cuts. Budget rules require councils to deliver a balanced budget, meaning functions that are more reliant on staffing and other types of annual expenditure are more likely to be cut. If we work on the basis that reducing spending on collections care, building maintenance and operational costs might mean a museum or gallery has to close, itself an expensive process practically and politically, then we can see why cutting spending on services predominantly comprised of staff costs and grant programmes is an attractive option. The revenue generating potential of buildings ("capital assets" in council parlance) which is greater than functions relying on staff and grants is also likely to be an influential factor. Local authorities may also feel it is politically expedient to cut spending on less visible forms of provision. This is not to say that the impact of cuts to "Arts Development and Support" would not be keenly felt by external organisations that may have been reliant on council funding. However, we can see how a story about the decision to wind up a grants programme for the arts would receive less media scrutiny than the closure of a museum, gallery or theatre, as has been demonstrated by the predominance of the former in recent years (e.g. Kalia, 2019). Finally, councils are unable to use money raised from asset sales (e.g. from the sale of a museum building or piece of the collection) to fund "ordinary" revenue expenditure, such as staff salaries or maintenance costs, adding a further disincentive to closures.
Although not all cuts to services delivered from buildings necessarily result in closure and buildings being sold, the legal and institutional contexts within which local authorities operate do disincentivise this type of cut. Until 2015, proceeds from asset sales could only be used to repay debts or finance new capital expenditure (e.g. new buildings or roads). Although this rule was temporarily suspended in the 2015 Autumn Statement (see MHCLG, 2018), councils were still blocked from using funds from asset sales for "ordinary spending". Further empirical study is needed to draw conclusions about which factors are shaping decision-making in particular local contexts. Nevertheless, taken together, these contextual factors point towards potential explanations as to why some types of expenditure are more vulnerable to cuts than others, as borne out in our findings. The removal of local authority support for artists, creative practitioners and small-scale voluntary organisations is concerning due to the potential impact on the wider cultural ecosystem, yet it may also be possible for spending to be reintroduced should future budgets allow.
Previous research has shown that it is the councils most dependent on central government support that have suffered the greatest cuts in overall service spending and that higher deprivation correlates with bigger cuts in service spending, due to the uniform approach taken to cuts to the grant (Gray and Barford, 2018;Harrison, Hodge, and Phillips, 2019). However this overall pattern does not hold true for cuts in spending specifically within "Culture and Heritage". There is no straightforward relationship between grant dependence and cuts in spending in this area either, meaning that reductions to the grant received from central government do not translate into cuts in spending on "Culture and Heritage" in the way that we might initially expect. The relationship between deprivation and cuts that applies to overall service spending also does not hold for spending on "Culture and Heritage" either. When the Indicies of Multiple Deprivation (IMD) are used as a proxy for deprivation, no overall pattern is discernible, and although increased spend is rare, one is more likely to find increased spending from more deprived than less deprived authorities. 7 The headline that those areas with higher grant dependency or deprivation often see lower cuts in "Culture and Heritage", and vice versa is significant as it indicates that councils facing high levels of deprivation are retaining some sort of culture and heritage provision in spite of facing high demand for social services. There is scope for further work here as alternative measures of deprivation may produce different results.
What these findings do not show, however, is that local governments making lower than anticipated cuts in "Culture and Heritage" are still providing the same level or type of service. Central government cuts have changed the role of local government, increasing the imperative for revenue raising from business rates, council tax, and other commercial endeavours. A local authority in a post-industrial area with limited ability to generate revenue through collecting business rates will therefore be less able to compensate for the impact of central government cuts than its counterpart in an area with a stronger local economy. Such a dynamic points towards the possibility that those councils that are highly dependent on the central government grant and without a strong local tax base to compensate may be inclined to maintain some sort of provision in the "Culture and Heritage" field, particularly as this domain includes museums, galleries and theatres, all of which are services with buildings. Unlike "Arts Development and Support", which is principally comprised of people, projects and grants, the buildings from which the "Museum and Galleries" and "Theatres and Public Entertainment" service is delivered can be reorientated towards the generation of income from room hire, events and suchlike. How this alters the substance of the service itself and who loses out as a consequence of these shifts is a major question for discussions of public culture following an age of austerity where economic responsibility looks set to become more localised.
Looking to the other end of the scale, in category 5 there are the local authorities that have higher spending in the "Culture and Heritage" category at the end of this period than the start. Six authorities in this category have increased their spending by over 100 per cent: Islington, Haringey, Waltham Forest, South Gloucestershire, Blackpool and Nottingham. Statistics on revenue expenditure do not include information about how such spending was resourced, however, so we cannot discern the steps these councils have taken to enable them to swim against the tide of spending cuts. By considering the subcategories above, however, we can say that in Islington, Haringey and South Gloucestershire it is the category of "Heritage" which sees a major rise in spending, whereas in Nottingham it is "Theatres and Public Entertainment", and in Blackpool and Waltham Forest it is both "Museums and Galleries" and "Theatres and Public Entertainment".

Income patterns
Overall, whilst local authority expenditure on "Culture and Heritage" in the decade 2009-2019 dropped by 26%, income dropped by only 2%. Again, this average measure does not capture variations in levels of income generation across local authorities. 61 of 152 local authorities generated a higher level of income from "Culture and Heritage" in 2019 than in 2009. This is reflective of a shift towards income generation as councils look to replace lost grant income through new ways of working. Of these 61, only 19 also had higher rates of expenditure in 2019 than in 2009 (5 of these 19 areas were in London). As discretionary services it is in the gift of local authorities to determine what fees and charges are set across the services covered by the "Culture and Heritage" category. Figure 7 shows the distribution of income earned by local authorities on a similar basis to the key in Figure 5 relating to expenditure; as such all named locations have an annual income from "Culture and Heritage" that was higher in 2019 than in 2009 and also above their average annual income for the period 09-18. This data refers to income earned from admission charges and school visits, ticket sales, commercial sponsorship of events and other income sources.
Our analysis complicates the picture of London and the South East benefitting the most from a shift towards earned income: local authorities seeing comparatively higher levels of income in 2018/19 are spread relatively evenly across the country. Rather than ceasing to provide services, many local authorities have restructured cultural service departments, employing staff with commercial expertise who are tasked with meeting income targets. In 2009/10 Liverpool City Council generated £2.6m (£5.69 per person) from "Culture and Heritage", increasing to £22.9m (£46.23 per person) in 2018/19, a rise which is largely down to increases in income generated from "Theatres and Public Entertainment". This has been supported by a commercial team within the cultural service consisting of ten members of staff. Bath and North East Somerset Council have also increased income from culture but unlike Liverpool their increase from £14.6m (£84.33 per person) in 2009/10 to £23.1m (£120.26 per person) in 2018/19 is concentrated in the "Heritage" and "Museums and Galleries" sub-categories, representing an intensification of an existing approach, largely enabled by the economic opportunity presented by tourism strategies relating to the Roman Baths, admission to which costs £63.00 at the weekend for a family of four. 8 The patterns discernible in the map are important not just for the variation demonstrated. The mixed picture also prompts us to question what these different income generation outcomes are underpinned by, both in terms of operational structures and strategies and the extent to which concerns over accessibility and participation inform the approaches being developed.

Conclusion
Our main aim in this article has been to examine variation in spending and income patterns in "Culture and Heritage" by local authorities across England across the period from 2009/10 and 2018/19. Drawing on quantitative data on local authority expenditure and financing submitted to MHCLG, our analysis allows us to re-frame "austerity" as a national government mandate that has deleterious, but nevertheless different configurations in each place as it is implemented by local authorities operating in different political, economic, social and historical contexts. Crucially, we have shown that cuts to local government spending on culture ranged from 1% to 94%. By mapping the unequal distribution of cuts to cultural spending, our analysis tempers narratives of austerity as a singular empirical reality of "decimation", "decline" and "dismantling" that remains constant from one place to another. We should clarify that this analysis is not meant to dismiss the very real material effects of the previous decade on people and places but to foreground what is in the background in existing studies: the way varied experiences of austerity coexist alongside one another and give shape to an "emergency" in cultural provision which is not fixed, but manifests in different ways. More substantively, we have sought to build on recent studies of the extent and distribution of local authority spending reductions by paying closer attention to variations between local councils and between categories of spending. We argue that this is a crucial first step towards developing a deeper understanding of the shape of local government cultural provision after a decade of austerity, and the foundations from which future challenges such as operating after the Covid-19 pandemic will be faced.
Focusing initially on spending in the "Cultural and Related services" macro-category, our analysis shows a disproportionate reduction in this area in comparison to "Total Service Expenditure". Within this macro-category, however, "Culture and Heritage" was found to be the least affected subcategory, with some "levelling off" in overall decline perceptible from 2015/16 onward. At first reading this is surprising given that the subcategory of "Other" includes libraries, a statutory service, unlike those within "Culture and Heritage", all of which are discretionary or non-mandatory services meaning that, theoretically, a council can choose to withdraw these services completely. To date, that discretionary services within the "Culture and Heritage" category have been less affected by austerity has been interpreted as evidence of local government efforts to "protect" these activities (Harvey, 2016;Mansfield, 2014). Looking more closely at the data and drawing on existing research into the contexts spending decisions within local government are made, we suggested a range of factors as potentially driving these changes. These include budget rules, there being higher levels of spending in other areas at the beginning of the austerity period, meaning savings were more readily identifiable in these areas, and the imperative to maintain services delivered from buildings which can be repurposed towards income-generation. Moreover, previous research has argued that decisions about expenditure on "Culture and Heritage" are unlikely to involve those who work in these services (Lawley, 2003;Rex, 2020b), raising the possibility that being the least affected subcategory is an unintended consequence of decisions taken elsewhere and in relation to other services rather than evidence of the agency and intentionality of advocates for spending on certain forms of culture. As such, although our analysis tallies with existing findings that discretionary services have been the most common target of austerity cuts (Streeck, 2014;Gray and Barford, 2018), we found that this interpretation requires nuance when it comes to the "Culture and Heritage" subcategory as by some measures, services in the "Other" subcategory (libraries, open spaces, tourism) have experienced a much steeper decline when set against "Culture and Heritage". This challenges interpretations which identify discretionary funding as especially vulnerable to cuts whilst statutory spending is protected, particularly as librariesfar too often mislabelled as discretionary serviceshave not clearly been "protected".
This does not mean that discretionary services have not been hit hard by spending cuts. Our analysis shows that certain types of provision are more vulnerable than others and explores, through reference to existing scholarship on local government contexts, a number of factors which potentially explain this. "Arts Development and Support" has been particularly adversely affected over the period examined here. Although this is troubling if we believe that these funds form an important part of local networks of people and organisations contributing to sustaining cultural opportunities, we suggest that this likely stems from it being more straightforward and reversable to make arts officers redundant and/or pause funding programmes than to close museums, galleries or theatres. We also show that several London boroughs have been severely affected by spending cuts to "Culture and Heritage", troubling the assumption that this is a story of "London and the rest", although where higher expenditure can be found this is often, but not exclusively, in the capital. Further, we found no discernible relation between grant dependence and cuts, and to the minimal extent that there is a relationship between deprivation and cuts, when the IMDs are used as a proxy for deprivation, it is a negative one, raising questions about the purposes and assumptions informing spending decisions. Further research is clearly needed to ascertain the extent to which the differences between local authorities that have fared better (or "less worse") than those which are reliant on the central government grant are reflected in the nature of service provision itself. Likewise, we need to examine the impact of a shift towards income-generation on the social composition of visitors and the working practices of council officers who may be adopting strategies to ensure this shift does not disadvantage their audiences or influence attendance.
While the impact on "Arts Development and Support" is a clear pattern, our main aim in this article has been to present a classification system of expenditure and income patterns as evidence of variation across England as to how austerity has affected spending on "Culture and Heritage" and to foreground potential explanatory factors. There is much variation in this data, which underlines the need for more research on how employees have negotiated or navigated this emerging terrain and the complicated diversity of operational contexts that give rise to the variation identified in this paper. As we knew when we began this analysis, this data can only tell us so much. Attention must be paid to how structures and institutions enable and constrain the capacities of local authorities to adapt to a changing funding environmentthis is a question for future empirical work to address.

Notes
Funding