Research and analysis

DCMS Creative Foundations Fund: evaluation framework

Published 17 July 2026

Report authors:

Dr Ricky Lawton, Stephen McSwiney

Abigail Lyons, Jack Philips, Ipsos

Professor Josep Grau-Bove

Dr Richard Higham, UCL

George Barrett

1. Introduction

1.1. Policy context

Public funding for the arts, culture and heritage has tightened sharply over the past decade, with local authority budgets for culture falling by an estimated 50% in real terms since 2009 to 2010.^{[footnote 1]} Many buildings are Victorian or mid-20th century, with few major rounds of capital renewal in the last thirty years. The result is a backlog of unfunded repairs, from structural works to obsolete heating, ventilation and electrical systems that are costly to run, often impossible to repair as parts become obsolete, and expensive to replace.

The COVID-19 pandemic further weakened the sector, accelerating long-term financial pressures as organisations in many instances began to draw down cash reserves. It also changed audience behaviour, including reduced willingness to attend in-person events, shifting preferences toward digital events and greater caution about crowded space.^{[footnote 2]} This placed increased pressure on art organisation finances until pre-pandemic visitor trends returned late 2024. High energy costs since 2019 have compounded the financial pressures on arts organisations. This has had a direct impact on cultural output. The number of plays and musicals staged by the UK’s main subsidised theatres last year down by almost a third compared with a decade ago.

Alongside these financial and operational pressures, cultural organisations are also operating in a more competitive leisure landscape. Advances in at-home entertainment technology, the growth of streaming services, and changing patterns of digital consumption have increased the range of alternatives available to households. This transition from physical to digital engagement has delivered positive results for accessibility and inclusion. Despite this, physical spaces remain vital for upholding the virtual landscape whilst delivering a greater range of indirect financial and health-related benefits.

The above factors threaten the safe and reliable operation of organisations and risk shrinking public access to the arts and culture. Ultimately this undermines the sector’s ability to deliver performances and sustain employment and community engagement over time. A 2024 Society of London Theatre and UK Theatre survey found that one in five venues will require an additional £5 million investment over the next decade just to remain operational. Without substantial investment in the next five years, nearly 40% of venues could close and a further 40% may become too unsafe to continue to operate.

1.2. Rationale for investment

Arts organisations in-scope of CFF produce significant benefits, such as cultural participation, community cohesion and educational value. These types of benefits are often not reflected in ticket income or commercial revenue (i.e. non-market benefits). As such, these non-market benefits often fail to produce a financial return for the organisations – a key motivation of the development of the DCMS Culture and Heritage Capital Framework.

Cultural organisations generate positive externalities of production. Investment in culture, heritage, and visitor economy initiatives can benefit local economies in places across the country, including agglomeration benefits, spillover effects to other sectors: Interactions with supply chains. Cultural organisations also generate positive externalities of consumption, as societal benefits of consumption exceed the relevant private benefits.

Many organisations also find it challenging to borrow commercially to fund maintenance activity. Funding repairs are often unlikely to generate additional revenue which, coupled with financial constraints, means that organisations struggle to demonstrate a clear ability to repay. This means that the market fails to provide capital investment into the arts sector at the socially optimum level.

Furthermore, attracting philanthropic donations for back-of-house activity is difficult as donors largely prefer funding front-of-house projects which are on public display. This is further amplified by the depletion of financial resources available from philanthropic and community supporters following the COVID-19 pandemic.

The organisations in scope of CFF also produce benefit to non-users (i.e. non-use benefits), meaning people gain utility from the theatre within ever visiting. These benefits could include existence value (knowing the organisation(s) exist), altruistic value (knowing others can benefit from the organisations) and bequest value (knowing future generations can use the organisations). This element of the ‘Total Economic Value’ is not captured within the organisations balance sheet.

This leaves organisations unable to generate the surplus required to reinvest in their estate or borrow at the scale needed to maintain ageing or specialised buildings. As a result, essential maintenance is repeatedly deferred, and the cultural estate continues to deteriorate in the absence of targeted public intervention. For example, evidence from the CFF Business Case suggests that nearly 40% of theatre venues surveyed risk closure without significant capital investment

Continued deterioration also affects the quality and reliability of the cultural estate. Many theatres and cultural buildings were constructed through the use of public funds approximately three decades ago, with limited maintenance since their inception. This has contributed to buildings simultaneously reaching the end of their intended life cycles, creating a concentrated wave of time-critical capital needs. When essential maintenance is deferred, issues can compound and become significantly more expensive to resolve, with delays often accelerating structural decline. This limits the ability of organisations to maintain consistent production volumes and standards or host diverse or ambitious artistic work. As the quality and reliability of venues decline, audiences are left with fewer safe and high-quality opportunities to engage with cultural activity. This further exacerbates the cycle of deterioration.

The Creative Foundations Fund (CFF) therefore addresses a clear market failure: essential capital works that secure the reliability, safety and accessibility of cultural buildings often do not generate sufficient cash returns to be financed commercially. This is despite delivering significant non‑market public value (participation, inclusion, wellbeing, community cohesion, place identity) recognised in the Culture and Heritage Capital (CHC) framework. Therefore, under a business-as-usual scenario, the long-term and socially valuable benefits generated by cultural organisations would continue to diminish as the estate deteriorates and market failures persist, leading to issues of underconsumption. Public investment is therefore needed to sustain the benefits that are not fully reflected in private market returns.

The Department for Culture Media and Sport said in its response to the Independent Review of the Arts Council that it would publish additional Culture Priority Places, which will ensure that citizens in areas with historically low levels of arts participation and funding will see a major change in their ability to access and participate in culture’. Therefore in addition to the criteria ‘Meeting the brief,’ ‘Management of the activity,’ and ‘Financial viability,’ applications are also assessed whether it is in a DCMS Culture Priority Place. This is to ensure that funding supports areas of the country that need investment the most. This means activities, outputs and outcomes need to be evaluated against this in mind.

1.3. Aims and objectives

By stabilising the condition of the cultural estate, the Creative Foundations Fund aims to ensure that cultural organisations can continue to generate wider economic spillover effects. These include increased local spending, job creation and stronger creative-sector supply chains. Protecting these cultural institutions also safeguards key public-good outcomes such as improved wellbeing, social cohesion and a strengthened sense of community identity. The objectives of CFF, as set out in the Business Case include:

1. Reduce frequency and severity of asset failure/degradation and resulting losses in cultural output and associated benefits. The Theory of Change set out in later chapters specifies a focus on reducing the risk of building failure and avoiding loss of cultural output.

Measures:

a.  the numbers of performances and number of hours an organisation is open to the public

b.  the number of people attending the cultural organisation

c.  the number of seats unusable or impacted

d.  M2 of property unusable following renewal.

2. Decreasing the financial impact of abandoned or cancelled performances, exhibitions or events from equipment or infrastructure failure, and supporting cost-effective regulatory compliance.

Measures:

a.  reduction in revenue lost over reactive repairs

b.  expenditure on works to improve environmental sustainability

c.  cost of reactive and emergency repairs over an agreed time horizon

d.  cost of hiring emergency and replacement equipment

e.  cost of hiring/renting alternative space

f.  cost of additional staffing

g.  expenditure on planned repairs

h.  reduced costs from investment in environmentally sustainable works

3. Increasing audience engagement, access, diversity and reported satisfaction with cultural organisations through investment in capital assets.

Measures:

a.  regional participation in cultural organisations and activities

b.  number of beneficiaries reporting increased levels of satisfaction

c.  expenditure on planned repairs, asset renewal and accessibility measures

d.  variation in audience demographics

CFF targets creative and cultural organisations in England which have delivered arts activity within the past 12 months. Eligible settings include:

• theatres (producing and presenting), concert halls, music venues, dance house
• multi‑use arts centres and community cultural venues.
• non‑collecting galleries/visual arts centres and exhibition spaces
• production parks/backstage facilities, artist workspaces and writers/literature centres
• museums and libraries are not eligible to apply for the fund

Funding is focused on the renewal, retrofit, and betterment of existing cultural assets. This includes essential fabric repairs, plant and systems replacement, accessibility improvements, and energy-efficiency measures that reduce operating costs and support net-zero pathways. New builds, change-of-use projects, and for-profit entities without an asset lock are out of scope. The CFF focuses purely on performing arts venues, creative production venues and community cultural centres. The fund comprises:

• Strand 1 (Small Capital Awards): £100,000 to £1,000,000 per project (allocation up to £25m), including equipment‑only projects.

• Strand 2 (Large Capital Awards): £1,000,001 to £10,000,000 per project (allocation up to £60m), for building or combined building‑and‑equipment projects (no standalone equipment).

• Total capital: - This first round of funding will distribute up to £85 million in capital grants to organisations across the country^{[footnote 3]}. Minimum match funding: 5% (Strand 1) and up to 15% (Strand 2) from non‑ACE sources (trusts, philanthropy, local authorities, sponsorship, in‑kind, reserves).

1.4. Structure of the Report

The remainder of this report is structured as follows:

Section 2: Theory of Change

This section provides an overview of the ToC for the Fund, and the assumptions, risks and contextual factors influencing the Fund.

Section 3: Key evaluation questions

The section sets out the key evaluation questions that a quantitative impact and economic evaluation should seek to address.

Section 4: Outcome measurement

This section provides an overview of the key outcomes of interest that will underpin an impact evaluation and the data requirements to enable an assessment of these outcomes. This includes consideration of the use of primary data collection methods to fill data gaps.

Section 5: Process evaluation

This section sets out considerations for a process evaluation, exploring how this method can support a 4E’s assessment and evaluate outcomes beyond the scope of a counterfactual impact evaluation.

Section 6: Counterfactual Impact evaluation

This section explores quasi-experimental approaches to establish the causal effect of funding, relative to a scenario where funding was not awarded to organisations.

Section 7: Heritage science counterfactual

This section explores the use of heritage science counterfactual analysis to understand what would have happened to venues in absence of the Fund.

Section 8: Theory based impact evaluation

This section explores how a theory-based method could be utilised to understand the impact of the Fund if standard counterfactual approaches (like comparing funded vs non‑funded organisations) are not possible.

Section 9: Economic evaluation

Details the required elements for an economic evaluation approach to undertaking a 4Es assessment and a Cost-Benefit Analysis.

Section 10: Recommended approach

This section sets out the recommended approach for a counterfactual impact assessment, process evaluation and economic evaluation, presenting the outcomes in-scope for a counterfactual impact evaluation and the data sources required.

2. Theory of change

This section articulates the expected causal processes by which the fund is anticipated to have led to their intended outputs, outcomes and impacts.

2.1. Inputs

The following inputs are required to design, administer and deliver the CFF.

Financial resources

As outlined in section 1, delivery of the CFF is supported through capital investment from central government (DCMS), complemented by partnership contributions from other sources:

• Central government investment: £85 million in capital grants. The Department for Culture, Media and Sport (DCMS) will provide £75 million capital funding, and the Arts Council England (ACE) are providing £10 million capital funding. ACE will distribute investment to funded projects across the lifespan of the project (spanning the 2026/27 to 2028/29 financial years).

• The CFF is comprised of two grant strands:
  • Strand 1 – Small Capital Awards: £100,000 to £1,000,000 per project (total allocation up to £25m).
  • Strand 2 – Large Capital Awards: £1,000,001 to £10,000,000 per project (total allocation up to £60m).
• Partnership funding: Matched funding is required from non-ACE sources, including charitable trusts, philanthropic giving, local authority contributions, corporate sponsorship, in-kind support, or applicant reserves, to leverage additional investment and test financial resilience. The level of matched funding is dependent on the strand applied for:
  • Strand 1: A minimum of 5% of total project costs must come from matched funding.
  • Strand 2: A minimum of 15% of total project costs must come from matched funding^{[footnote 4]}.

Central Government and arms lengths bodies’ time and resources

• Department for Culture, Media and Sport (DCMS): Policy sponsor and funding authority, setting the strategic objectives (economic sustainability, safeguarding cultural output, widening access) and ensuring regional/ sectoral balance with an ongoing oversight role.

• Arts Council England (ACE): Lead delivery body and Accounting Officer, responsible for fund design, application systems, assessment and moderation of Expressions of Interest and full bids, award decision-making, grant agreements, monitoring and compliance.

  • All funded projects must apply to ACE Investment Principles^{[footnote 5]}, requirements for funded artists and organisations to embed excellence, equity, sustainability and resilience into their planning, delivery and evaluation.. For smaller organisations, applying these principles may create additional resource pressures, as limited staff capacity and specialist expertise can constrain their ability to plan, evidence and report.
  • Application assessment: Applications are assessed through a two-stage process: an EoI stage where applicants outline their proposed project, its associated costs and the intended benefits of the investment. Shortlisted organisations are then invited to submit a Full Application providing detailed information about the project, including need, outcomes, beneficiaries, inclusion and diversity data, and a full capital budget. ACE operates the digital submission platform, provides pre-application support through regional briefings, webinars, and FAQs. The EoI round opened in September 2025, with full applications invited by October 2025 and decisions confirmed in March 2026.
  • Grant agreement and monitoring: Following award decisions, ACE issues legally binding grant agreements setting out conditions for match funding confirmation and delivery milestones. Grant payments are released in arrears on verified expenditure, supported by audit checks. ACE oversees performance reporting, risk monitoring, and change management during delivery.
  • Fund launch and guidance development: ACE develops and publishes detailed applicant guidance following DCMS approval of the fund design. This includes ‘How to Apply’ materials, and technical documentation requirements (RIBA Stage 3 drawings, cost plans, environmental and accessibility standards). Guidance is co-designed with DCMS policy, commercial, and analytical teams to align with strategic objectives.

Applicant time and resources

• Professional time: Applicants are required to invest significant professional capacity in developing their proposals. This includes commissioning professionally prepared drawings, specifications, cost plans and risk allowances (typically to RIBA Work Stage 3 or equivalent) to evidence that projects are technically feasible and financially robust. The process is staged, requiring an Expression of Interest (EOI) followed by a Full Application. Applicants are expected to refine cost estimates, risk registers and delivery plans between EOI and full application. RIBA Work Stage 3 should, in principle, provide sufficient design and costing maturity to support delivery on time and within budget. Additionally appropriate contingency allowances should be included to manage unforeseen conditions or unexpected cost pressures. Applicants must demonstrate clear governance arrangements, realistic delivery timescales and a plan for maintaining operations during any closure periods.

• Specialist advice: Projects are expected to engage qualified technical consultants, architects, surveyors, engineers and environmental specialists. These specialists ensure design integrity, compliance with building and heritage standards, and contribution to sustainability and net zero objectives. For heritage or listed buildings, applicants often require additional conservation and structural expertise to navigate planning constraints and unforeseen site conditions. Costs associated with engaging with specialist inputs, which can often be costly, are not covered by CFF funding.

Activities

Following a successful funding announcement, cultural organisations are expected to undertake a range of maintenance and renewal activities. These can be grouped into four broad categories:

• Improving their buildings.

• Improving environmental performance and building systems.

• Enhancing access and user experience; and strengthening technical and production capacity.

It is understood that prioritisation is based on urgency of risk, listed/ heritage status, and safety/ compliance, with linkage to outcomes. As noted, there are differences in what each strand can fund:

• Strand 1 supports equipment-only projects as well as building upgrades.

• Strand 2 is limited to building or combined building-and-equipment projects, with no standalone equipment funding.

Building systems and environmental performance

• Replacement and upgrade of essential building services: This includes heating, ventilation and air conditioning (HVAC), lighting (LED), power, IT cabling, lifts, fire detection, security and alarm systems, building control systems.

• Environmental sustainability measures: Including insulation, renewable energy installations, low-carbon heating, improved thermal performance, and decarbonisation measures.

Building fabric and structural renewal

• Priority asset repair and renewal: For example, roofs, walls, masonry, windows, and external envelope upgrades.

• Flood resilience and climate adaptation: For example, floodproofing, drainage upgrades.

• Provision of permanent inspection and maintenance access: Including safe access points and platforms to enable planned maintenance.

• Development of maintenance plans: A condition of funding is that organisations develop maintenance plans to help organisations understand their short- and long-term maintenance needs.

Technical and production capacity

• Purchase of technical and production equipment: Including the renewal of specialist stage, audio-visual and digital equipment, and replacement of vehicles essential for sustainable touring and production logistics. This funded activity is only available for Strand 1 applicants.

• Renewal and repair of touring equipment and outdoor arts facilities: For example, the repair of touring tents, portable sound desks, lighting rigs, and upgrades to cable infrastructure, equipment for outdoor arts and sustainable touring.

Access, inclusion and user experience

• Accessibility and inclusion improvements: Including ramps, accessible toilets, Changing Places facilities, signage, level landscaping, inclusive circulation spaces.

• Upgrades to visitor experience and workforce facilities: For example, upgrades to front-of-house and back-of-house areas, seating, artist and workforce comfort and safety.

Outputs

These activities would be expected to lead to the following outputs:

Building systems and environmental performance

Upgrades to building systems are expected to deliver immediate operational outputs in terms of infrastructure involving:

• Upgraded building services: For example, HVAC, electrical, lighting, fire detection, security and alarm systems, IT cabling, lifts, building control technologies.

• Energy systems and fabric upgrades: Including LED lighting, insulation, low-carbon heating, renewable energy systems, and climate adaptation works.

• Weather resilience and flood protection measures: For example, drainage upgrades, floodproofing, etc. installed.

Building fabric and structural renewal

Projects are expected to use their funding to undertake essential asset repair and renewal of their buildings. The expected outputs include:

• Repaired cultural buildings or stabilised structures: including roofs, walls, windows, masonry and external envelope works.

• Permanent access and inspection infrastructure: For example, walk-on wire tension grids, catwalks, inspection-friendly rigging, roof/façade access.

• Updated or created building maintenance plans and asset management schedules introduced by funded organisations.

Improved technical and production capacity

Renewal of specialist stage, audio-visual and digital equipment, alongside investment in sustainable touring infrastructure:

• Renewed specialist stage, audio-visual and digital equipment: For example, lighting, sound, projection, digital media systems.

• Low- or zero-emission vehicles: For sustainable production and touring logistics.

Accessibility, inclusion and user experience

Investments to enhance accessibility and visitor experience are expected to deliver more inclusive, comfortable and functional cultural spaces with:

• Enhanced accessibility features: Such as ramps, changing places and accessible toilets, level paths, power-assisted doors, inclusive wayfinding, signage, etc.

• Improved visitor and workforce facilities: Such as seating, circulation areas, artist and backstage spaces, front-of-house comfort and safety.

• Increased usable capacity: For example, more seats brought back into operation, previously closed spaces reopened, additional hours of building use available.

• Improved digital and physical user information: For example, updated signage, wayfinding and visitor support resources.

• Measurement and assessment of these outputs will also need to be set against Cultural Priority Places.

2.4. Outcomes

The outcomes expected to arise from the CFF are set out below. Central to the evaluation framework is the concept of additionality – that is, understanding what would have happened in the absence of CFF funding. In this context, additionality may relate to whether projects would have taken place at all or on the same scale and timescale without support, and whether undertaking maintenance works earlier may prevent significant future costs or asset loss.

The outcomes and benefits of CFF are expected to manifest over different time horizons. As capital works are completed and upgraded infrastructure comes into operation, a range of positive changes are expected to emerge across cultural organisations, audiences, and communities. These outcomes capture both the immediate benefits of physical improvement and the medium-term organisational, environmental, and social shifts that such investment enables. As with outputs the measurement and assessment of these outcomes will also need to be set against Cultural Priority Places. The outcomes, short- and/or long-term, are hypothesised to manifest through seven domains:

Economic

In the short term, the maintenance activity funded by CFF is anticipated to generate wider economic impacts/benefits:

• Demand for specialist workforce: In many instances, capital works undertaken on buildings may require specialist contractors. Similarly, the equipment required to stage productions or exhibitions may require specialist technicians to advise, install and operate. Therefore, CFF funding will produce impacts within the wider culture heritage and arts eco-system.

Over the longer-term, the continued existence of the cultural organisations will likely lead to:

• Economic and employment growth: Well maintained and fully functioning infrastructure underpins both direct and indirect economic benefits. In the long term, CFF will help safeguard and create jobs across the creative and cultural workforce from artists and technicians to front-of-house and operational staff. Upgraded venues will generate ongoing demand for suppliers, contractors, and creative technology providers, strengthening the sector’s supply chain. Increased and more reliable audience attendance may also generate induced economic activity through its impacts on tourism and local visitor spending.
• Local economic growth: At the local level, sustaining audience attendance will support surrounding businesses such as cafés, retail, hospitality and accommodation, creating spillover benefits for local economies. Improvements to cultural infrastructure may also contribute to wider place-based regeneration efforts by enhancing the attractiveness of town and city centres, supporting footfall, and acting as anchors for mixed-use development. There may even be agglomeration benefits through the creation of local cultural and creative clusters which facilitate knowledge and skill exchange, competition, economies of scale and innovation. This outcome is particularly important given that in May 2026, DCMS published a list of Priority places,^{[footnote 6]} where DCMS will focus more activity and investment to ensure that funding is concentrated in the communities where it can make the biggest difference.

Asset level

Over the short-term, the outputs produced by CFF will facilitate:

• Reduced risk of building failure: The CFF funding is expected to reduce the immediate risk of core building failure.

• Reduced risk of equipment failure: The CFF funding is expected to replace or repair outdated or faulty equipment, reducing the risk of system or equipment failure and enhancing operational performance.

Over the longer-term:

• Increased longevity of cultural buildings: Addressing critical building defects and the development of maintenance plans increases the likelihood that the historically and culturally significant buildings which house organisations will retain their cultural usage and continue to exist into the future.

Organisational

Over the short-term, the CFF funding is expected to lead to:

• Reduction in the maintenance backlog: The funded activities directly reduce the maintenance backlog for the organisation, both in terms of:
  • Value of the outstanding backlog.
  • Number of items on the backlog.
• Reduced financial pressure: Investments in building maintenance and essential equipment should reduce short-term financial pressures on organisations through:
  • Mitigating risks of closure or lost revenue.
  • Reducing the value of outstanding (essential) maintenance work.

• This includes:

  • The stabilisation of maintenance budgets through lower planned-maintenance requirements.
  • Reduced reliance on reactive repairs.
  • Decreased energy costs arising from more efficient building systems and equipment.
  • The potential for reduced insurance premiums where upgraded infrastructure lowers assessed risk.

Over the long-term, reduced financial pressure is expected to lead to:

• Improved organisational resilience and survival: Early intervention prevents costly asset failures and help organisations plan with greater confidence. Addressing outstanding maintenance needs may help to attract investment in other parts of the organisation. Over time this enhanced financial stability may enable organisations to adapt strategically to changing economic conditions.
• Reduction in reactive repairs: Where critical maintenance needs have been addressed and with the development of a maintenance plan, cultural organisations may be able to adopt a schedule of proactive repairs. This is as opposed to often more costly reactive repairs.

Cultural

In the short-term, CFF may have the following effects on the cultural sector:

• Avoided loss of cultural output: Priority repairs and equipment replacements should address critical points of failure which would potentially lead to avoided cancellations of performances, exhibitions or events due to building or equipment failure. The avoided reduction or loss in activity from art organisations are therefore anticipated to help preserve the existing levels of cultural output in the short-term.

• Sector resilience: CFF is anticipated to help stabilise cultural infrastructure by preventing the loss of productive assets essential to revenue generating activity and cultural output. By strengthening the financial resilience of arts organisations, the sector should become more resilient. A more financially sustainable and resilient sector will be more capable of maintaining cultural output, retaining skilled staff, and contributing consistently to regional and national economic and social value.

Over the longer-term, the following outcomes may be expected to manifest:

• Preservation of the arts and cultural sector: Through increasing the financial stability of cultural organisations, CFF may help to preserve the arts and cultural sector.

• Skill and knowledge preservation: As organisations are able to maintain their workforce, the skills and knowledge of the workforce will be sustained.

• Quality of output increases: More up-to-date equipment may enable organisations to increase the quality of shows, performances, exhibitions or events

• Increase the UKs reputation for creative excellence: By maintaining and growing the cultural output of the country, and preserving the stock of culture and heritage, the UK will remain internationally renowned for its creative excellence. The high-quality content produced within the arts organisations, which attracts audiences from around the world furthering the UKs soft power will be sustained

Environmental

In the short-term, the building and system upgrades are anticipated to produce the following outcomes:

• Environmental sustainability: In the short term, upgrades such as energy-efficient systems, improved insulation, and the use of sustainable materials may reduce energy consumption and emissions, leading to immediate and long-term cost savings for organisations. There will also be societal benefits through associated reductions in negative externalities.

Social impacts

• Community cohesion: The role of cultural venues as trusted, inclusive civic hubs, strengthening bonding and enhancing social capital will be sustained and potentially enhanced. This may be evidenced by sustained cross‑community participation and co‑creation, more active volunteer networks, more durable partnerships with local groups and services, and reduced social isolation.
  • This can also include improvements in pride of place, with visible renewal of cultural landmarks and reliable programming reinforcing local identity and belonging, improving perceptions of town centres/high streets.

• Health and wellbeing: Participation in creative activity within safe, comfortable and accessible spaces potentially leads to sustained improvements in subjective wellbeing and reduced loneliness, increased physical activity through dance/movement offers, and stronger links with social prescribing pathways.

• Education outcomes: Formal and informal learning spaces housed or developed within funded organisations will be sustained and potentially enhanced, supporting human capital accumulation, cultural capability, and strengthen progression into creative and technical pathways

Access and inclusion

In the short term, the following outcomes related to access and inclusion are anticipated:

• Preservation and enhancement of access and participation: In the short term, accessibility improvements such as step-free entry, enhanced signage, and inclusive facilities will enable access to audiences regardless of race, sex, gender or disability who may otherwise have been excluded due to deteriorating or inaccessible infrastructure. In addition, ACE aim to ensure inclusivity as a core criterion guiding funding decisions, enhancing inclusion and reach (with positive distributional effects for priority groups and places).

• Improved user experience: Upgrades to the organisations’ building fabric and equipment may enhance audience engagement and satisfaction. Improved physical environments may make venues more welcoming and inclusive, increasing the likelihood that existing users return, and attracting new audiences.

• User experience (production/programme): Audience perceptions of the artistic/technical quality, relevance and emotional impact of what is presented can be enhanced.

Over the longer-term, this is expected to lead to:

• Growth in cultural access and participation: Improvements are expected to strengthen the relationship between communities and their local venues. As organisations operate more reliably, they can programme more consistently, build audience trust, and attract audiences of all demographics. In the longer term, these changes may lead to behavioural shifts, embedding patterns of regular attendance, community engagement, and inclusive design within organisational practice. The result should be a more diverse audience base that supports both the cultural and financial sustainability of England’s creative sector.

Growth in access and participation also supports the emergence of social impacts, by driving community cohesion through attachment, belonging, and pride in place by shaping and representing identity, traditions, memory, connectedness, meaning, and social bonding, that can create intergenerational connections.

Complexity of effects

Cultural output is anticipated to arise from:

• The existence of the building.

• Organisations use of the building.

• Any collections within the buildings.

• Host organisations’ cultural output.

• The characteristics of the organisation, building and output.

Within the production of cultural output, there are several layers of effects which underpin the causal mechanism for CFF. The below narrative is presented in Figure 2.1 below:

• First order effects: The CFF funding will enable and facilitate maintenance activities which are hypothesised not to happen, or not to happen to the same extent, in the absence of CFF.

• Second order effects: The maintenance and remediation of the building and contents of the building (i.e. equipment to facilitate productions) will support the i) existence and ii) financial resilience of beneficiary organisations.
  • Where cultural organisations are housed in culturally or historically significant buildings, the existence and proper maintenance of the building fabric may lead to non-market benefits (e.g. enhanced existence values)

• Third order effects: Cultural organisations remaining housed within a functioning building fabric enables the continued production of (higher quality) cultural output.

• Feedback loops: The maintained or increased production of cultural output, may increase the financial security of cultural organisations, better enabling organisations to remain within and ensure adequate maintenance of the buildings they are housed in, and the contents of the building

Figure 2.1: Complexity of the effects of CFF

Figure 2.1 is a diagram showing the complexity of the effects of the Creative Foundations Fund. The diagram begins with the CFF funding and continues in order of effects, up to third order. Each effect produces sub-effects that return output into the next order of effect, and returns output to the previous order.

2.6. Assumptions

Undertaking CFF funded activities may require a specialist workforce. This could include building specialists, heritage construction specialists, retrofitting specialists, listed building specialists, theatre equipment specialist, etc. There is an assumption that there is sufficient capacity of specialist contractors to absorb the surge in demand. Sector wide demand peaks may lead to inflationary pressure and threaten the ability to deliver on time, within budget and within scope. Alternatively, the funding may enable increased investment in training by suppliers, potentially alleviating future supply side constraints.

• It is assumed that organisations have the time, capacity and skills required to develop, deliver and manager large-value capital projects.

• It is assumed that there are sufficient levels of matched funding (e.g. from local authorities or other sources) to meet match funding requirements.

• It is assumed that sufficient time is available within the application and delivery phases to undertake thorough procurement exercises to ensure:
  • High quality work
  • Fair prices.

• There is also an assumption that organisations have sufficient capacity and cash reserves to consult with often costly specialist advice. If organisations are unable to consult with specialists, this may be prohibitive to submitting an application and obtaining CFF funding, potentially excluding what may otherwise be high priority projects.

• It is assumed that there will be a positive audience response to the enhanced cultural offer, with demand sufficient to sustain the long-term viability of improved facilities. This assumes that upgraded environments and improved user experience translate into stable or increased attendance and do not coincide with broader downturns in cultural participation.

• It is assumed that existing shortcoming and outstanding maintenance issues within organisations have significant impacts in excluding groups with vulnerabilities or needs.

• It is assumed that the design of the application and assessment process enables the fund to reach those organisations most in need of capital intervention and where the greatest benefits would be delivered. This includes the assumption that eligibility criteria, evidence requirements and scoring approaches reliably identify organisations facing the greatest maintenance pressures or risks of asset failure.

2.7. External factors

Several external factors are likely to potentially threaten the effectiveness of the CFF:

• Inflation: Material price inflation, labour market tightness, interest rates, insurance premiums and energy prices are all likely to threaten the ability of projects to deliver within scope and budget. Inflationary pressure is also a risk to the financial stability of art organisations in the long run. In the worst-case scenario where organisations fail to survive, the benefits of the support may be lost.

• Fund‑raising/philanthropy climate: Limitations in donor confidence and financial capacity, corporate/foundation priorities, tax incentives, and competition for funds. These can determine how much match funding, philanthropy and follow‑on investment organisations can raise independent of the programme.

• Economic and fiscal conditions (local and national): The financial health of the local authority and its ability to deliver matched funding, either at the minimum level or above the minimum level, will have likely constrained the scope of the projects for which the organisations have applied. Local authority budgets and funding allocations are also likely to be an important factor in art organisations budgets in future and the ongoing survival of the organisations.

• Audience and revenue context: Cost-of-living pressures, changes in audience behaviour, tourism trends (both domestic and international) will affect the realisation of expected benefits such as increases in audience numbers and revenue.

2.8. Potential unintended consequences

• Temporary closure/disruption: It is likely that delivery of the CFF supported projects will lead to closures, programme cancellations or disrupted community use in the short term. This will likely have negative, temporary, impacts on the number of visitors and organisation revenue and may in some cases lead to – perhaps short-lived - scarring effects on audience behaviour if temporary disruption weakens established attendance patterns.

• Higher operating costs: New systems may require additional maintenance, new software licences and training. This may increase some elements of operating expenditure.

• Short term construction externalities: Where building works are undertaken, there is likely to be elevated levels of noise, dust, traffic and street disruption which may affect local residents and businesses.

• Displacement of cultural activity: Venues experience short-term audience loss or programming clashes due to rescheduling or cancellations.

• Reputational risk: Organisations unsuccessful in obtaining funding may experience reputational damage. Likewise, there may be reputational damage to civic buildings and libraries which were not eligible or not awarded for funding.

• Competition effects: Strengthening the position of funded organisations may adversely affect the volumes of visitors and the financial viability of non-CFF funded competitor organisations, including those within the commercial sector.

Figure 2.2: Logic model

Figure 2.2 is a logic model for the Creative Foundations Fund. It begins with the inputs, and follows through to the activities, and the outputs. Next to it is the Outcomes, which is split across two columns. Below is a list of assumptions and external factors that could affect the implementation of the Creative Foundations Fund

3. Key evaluation questions

This section sets out the key evaluation questions (KEQs) an evaluation of the Creative Foundations Fund should seek to address.

3.1. Process Evaluation

Competition setup

• KEQ1: Were notifications of the funding opportunity to prospective applicants clear and did they provide sufficient time to develop quality applications?

  • KEQ1a: Did applicants find the arrangements for clarifying questions suitable? Which resources were the most/ least helpful?
  • KEQ1b: What were the volumes of ineligible applicants and did these align with expectations? How did the volumes compare to other DCMS or ACE programmes?
  • KEQ1c: How effectively was the sector informed about the fund’s introduction through communication efforts?

• KEQ2: To what extent did CFF reach organisations with the highest priority needs in terms of assets failure/ deferred maintenance?

  • KEQ2a: To what extent did CFF funding reach organisations offering the greatest benefit from the available funding?
  • KEQ2b: Were application volumes in line with expectations?
  • KEQ2c: How far did the exclusion of for-profit organisations limit reach of CFF and its wider impact?

• KEQ3: Did Strand 1 vs Strand 2 rules (e.g. equipment-only eligibility, higher match funding requirement for Strand 2) function as intended?

  • KEQ3a: Was the required match funding level proportionate?
  • KEQ3b: Were match funding thresholds appropriate to need and market conditions, and proportionate for smaller/heritage organisations and priority places?
  • KEQ3c: What sources of match funding were most used (LA, philanthropy, trusts, reserves, in-kind)?
  • KEQ3d: To what extent did requirements lead to exclusion of potential priority projects or impact the volume and characteristics of applications?
  • KEQ4: Did programme timelines give funded applicants sufficient time for projects to undertake capital work to desired specification/ standards?

Application process

• KEQ5: Was the two-stage process (EOI progressing into full application) clear, proportionate and timely for different applicant types?

  • KEQ5a: What support, if any, was needed or provided to smaller organisations?
  • KEQ5b: Was the application process and readiness requirements (e.g., RIBA Stage 3, costed plans, engagement with specialists) proportionate to project scale and risk?
  • KEQ5c: Was the application process disproportionately burdensome on smaller organisations (e.g. cost of reaching RIBA Stage 3, cost of engaging with specialists, knowledge and skills of staff, staff capacity).
  • KEQ5d: Did the volume of information required in the application form discourage organisations from applying?

Project Selection

• KEQ6: Did ACE have sufficient resource to undertake assessment process within the prescribed timeframes?

• KEQ7: Did assessment criteria and scoring reliably identify need, additionality, readiness, risk and potential benefits?

  • KEQ7a: Were the funds allocated as intended and to the organisations most in need of the funding?
  • KEQ7b: Were the potential benefits of funding maximised?
  • KEQ7c: How was funding allocated to different activities? Were there any priority activities?
  • KEQ8: Were moderation, decision-making and portfolio-balancing processes transparent, consistent and free of conflicts of interest?
  • KEQ8a: Did selection decisions achieve regional/sectoral balance and reach priority groups/places without excluding potentially crucial projects?
  • KEQ8b: Did assessment decisions give sufficient consideration to the adequacy of contingency and optimism bias provisions in applicants’ cost estimates?
  • KEQ9: To what extent did selection choices align with fund objectives (safeguarding cultural output, organisational sustainability, widening access, environmental performance)?

Contracting and monitoring

• KEQ10: Did ACE have sufficient resource to efficiently manage the contracting process and deliver funding on time?

• KEQ11: Were funding arrangements (e.g. payments in arrears) manageable for organisations with limited reserves?
  • KEQ11a: Did this impact delivery of the planned works?
• KEQ12: How effective were governance, procurement and risk management within funded organisations? What support did ACE provide during delivery?
  • KEQ12a: Was there sufficient specialist contractor capacity (heritage, retrofit, theatre tech), and what were the consequences if it was constrained?

• KEQ13: What short-term disruptions (closures, cancellations, displacement, construction externalities) occurred and how were they mitigated?

• KEQ14: How effective were monitoring arrangements in managing risk, scope, time and budget?
  • KEQ14a: Were the monitoring arrangements proportionate to the funding received?
  • KEQ14b: Were monitoring arrangements reviewed and refined over time? Did this improve delivery over time?
  • KEQ14c: Did monitoring enable the identification of emerging problems, allowing effective corrective actions to be put in place?

Project delivery

• KEQ15: Were outputs delivered as planned?
  • KEQ15a: Were outputs delivered within expected costs and expected timelines?
  • KEQ15b: What factors may have contributed to or impeded delivery as planned?
  • KEQ15c: How did project sequence works to minimise disruption? Which mitigations were the most effective?
  • KEQ15d: Which procurement strategies best balanced price, quality, and specialist capacity?
  • KEQ15e: How were maintenance plans operationalised post-handover?
• KEQ16: What were the key challenges encountered during project delivery, how were they overcome, and what are the overarching lessons learnt for future capital funding programmes?
  • KEQ16a: How did macroeconomic factors (e.g. inflation, supply chain constraints) impact project scope, budget and timelines?
  • KEQ16b: What mitigation strategies were most effective in overcoming these challenges?
  • KEQ16c: What lessons can future prospective applicants learn in terms of delivering large-scale capital funding projects?

3.2. Impact evaluation

Asset condition and reliability

• KEQ17: To what extent did CFF reduce the risks of building, systems and equipment failure, and the future risk of closure?

• KEQ18: Was CFF successful in reducing the extent of maintenance backlogs?

• KEQ19: To what extent did CFF reduce future maintenance and operational costs for funded organisations?

• KEQ20: Were realistic maintenance plans implemented and embedded into asset management?

• KEQ21: Did CFF allow previously closed spaces to be re‑opened?

Organisational resilience and financial sustainability

• KEQ22: Did CFF reduce the short-term financial pressures on organisations created through outstanding maintenance needs?

• KEQ23: Over time, did CFF improve organisations’ cashflow, reserves, and survival prospects?
  • KEQ23a: What effect did CFF have on the long-term sustainability of organisations?

• KEQ24: Did reactive repairs decrease over time?

• KEQ25: Did CFF help unlock further investment or improve borrowing terms for organisations?

Cultural output and sector resilience

• KEQ26: How far did CFF avoid losses of cultural output (e.g. cancelled performances/exhibitions) and sustain programming levels in the short term? How might this change in the future?

• KEQ27: Were production quality, diversity and ambition enhanced with upgraded spaces/tech and reduced hire/emergency replacement?

• KEQ28: Did touring/production capabilities and resilience of the wider cultural ecosystem improve (e.g. technical capacity, skills retention)? Is it likely that these capabilities will continue to increase or improve in the future?

• KEQ29: Did projects contribute to preserving and/ or enhance the sector’s capabilities and the UK’s reputation for creative excellence?

Access, inclusion and user experience

• KEQ30: Have accessibility improvements (e.g. step‑free access, changing places, wayfinding) enhanced user experience and /or increased participation among underserved groups? Is it likely that this may improve in future?

• KEQ31: Did user satisfaction, trust and repeat attendances improve? Is it likely that this may improve in future?

• KEQ32: Which audience groups benefited most/least? Were there reductions in barriers for people with protected characteristics and in priority places?

Environmental performance and net-zero

• KEQ33: What effect did CFF have on energy consumption, emissions and environmental ratings (EPC/ DEC)? What might be the effects on energy consumption in future?

Social and wellbeing

• KEQ34: Did organisations strengthen their civic role, community cohesion and volunteering networks? How might this change in future?

• KEQ35: What changes occurred in subjective wellbeing, loneliness and pride in place among staff, volunteers, users or communities? How might this change in future?

Economic and employment

• KEQ36: What short‑term employment and contracting impacts occurred during works (including specialist contractor demand)?

• KEQ37: Were sectoral skills maintained or enhanced? Could this change in future?

Local economic growth and place-based spillovers^{[footnote 7]}

• KEQ38: Did improvements contribute to local footfall, visitor spend and town/city centre vitality? Could this change in future?

• KEQ39: Were there signs of creative clustering, knowledge exchange or wider regeneration narratives linked to funded venues? Could this change in future?

3.3. Economic evaluation

At the organisation level, income/profit and broader financial metrics are influenced by multiple concurrent shocks and supports. Given the small number of funded venues, it is not recommended to explore econometric attribution of organisation‑level financial outcomes to CFF. Counterfactual methods should be reserved for operational outcomes (reliability, cancellations, hours open, energy). Organisation‑level finance and resilience impacts should instead be evidenced through theory‑based methods (Contribution Analysis, Process Tracing) triangulated with HS reliability findings.

Economy

• KEQ40: To what extent did CFF deliver its’ objectives using the minimum level of public support?
  • KEQ40a: What was the total cost of CFF?
  • KEQ40b: Did the process for allocating funding support value for money?
  • KEQ40c: Was the level of funding sufficient?
  • KEQ40d: How did CFF compare to similar programmes in terms of costs and outputs?

Efficiency

• KEQ41: To what extent were outputs arising from funding delivered efficiently?
  • KEQ41a: Were outputs achieved as planned, within expected costs and expected timelines?
  • KEQ41b: Were risks and issues effectively identified, mitigated and managed?

Effectiveness

• KEQ42: Did the CFF represent value for money based on its impacts so far?

• KEQ43: To what extent is it likely the CFF will represent value for money based on its likely future impacts?

• KEQ44: How effective has CFF been in delivering its objectives so far? How might this change in future?

• KEQ45: What were the levels of additionality of CFF?

  • KEQ45a: To what extent did the impact of CFF represent a net gain, opposed to displacing activity from other areas?
  • KEQ45b: Is there any evidence of displacement, either nationally or locally?

Equity

• KEQ46: Were the effects of CFF distributed differently among users and non-users, organisations and staff?
• KEQ47: How were the outputs and outcomes of CFF distributed geographically?
  • KEQ47a: Where there any areas which benefitted disproportionately more than others?
  • KEQ47b: How do the value for money conclusions change after accounting for distributional effects?

4. Outcome measurement

This section sets out proposed outcome indicators which can be used to measure changes in outcomes over time. The section also includes an appraisal of data sources which would enable outcomes measurement.

4.1. Potential outcome indicators

The below table maps the outcomes identified in the ToC with metrics which could provide the basis for objective assessment. The recommendation presented in Section 11 produces a shortlist of outcome indicators which a future evaluation may seek to prioritise. It also provides information on timings for the evaluation.

Table 4.1: Linking outcome with outcome indicators

Domain	Outcome	Metric(s)	Considerations/ challenges
Economic	Demand for specialist workforce	Contractor FTEs during works; Number and value of contracts let; Apprenticeships/traineeships started and completed; % local labour within X miles; % contracts to SMEs/micro firms	Likely to be a high level of displacement as the workforce may be employed on other jobs. Need to define specialist taxonomy (e.g., SOC/SIC/CSCS/heritage trade lists); Short-term project roles vs sustained jobs; Local vs national leakage; Wage inflation: Demand for specialist workforce may push wages up if limited resource leading to higher project costs
Economic	Economic and employment growth in organisations receiving funding	Number of FTE, employee wages, organisation revenue, and profit, estimated productivity (e.g. revenue per FTE per year)	Non-profit business models complicate profit/productivity; Additionality (safeguarded vs new jobs) hard to evidence; Timing effects (closure dips during works); Adjustments needed for inflation/ seasonality
Economic	Local economic growth	Local area GVA, local area employment, local area revenue, town/ city centre footfall, visitor spend, culture heritage or creative firms within given distance, culture heritage density, hotel occupancy; business rates yield. Note that local area metrics should be considered in reference to the DCMS Culture Priority places list published in May 2026.	This should be used as supplementary analysis as funding unlikely to influence wider economic metrics. Macro shocks (energy, cost-of-living), although can focus on relative performance. Inconsistent footfall/firm data;
Asset level	Reduced risk of building failure	Number of unplanned closures, number of cancelled performances, number of critical items on maintenance backlog, Insurance premium data, Number of seats unusable or impacted	Establish comparable baselines and seasonality; Need to standardise backlog criticality ratings;
Asset level	Reduced risk of equipment failure	Number of unplanned closures, number of cancelled performances, number of critical items on maintenance backlog	Distinguish equipment vs fabric failures; Comparability may vary significantly across equipment classes due to differences in expected operational life, maintenance requirements and replacement cycles. Failure due to events or strain may be more relevant than degradation Spares/obsolescence complicate comparability
Asset level	Increased longevity of cultural buildings	Estimated lifetime/ lifetime extension	Life‑cycle assumptions uncertain; Integrate HS metrics (subject to availability) Estimated lifetime extension should be treated as a modelled indicator rather than a precise prediction.
Organisational	Reduced financial pressure	Value of maintenance backlog, overheads, avoided revenue loss from cancellation/ closures, cost of reactive repairs	Backlog valuations vary: need to normalise to a price base; Challenge of separating one-off capex from recurring OPEX savings (energy, insurance, maintenance); Accounting/reporting lags
Organisational	Improved organisational resilience and survival	Value of reserves, operating margin, liquidity ratio, survival probability	Define/unify measures (unrestricted vs restricted reserves); Confounding external funding sources (LA funding, energy, philanthropy) confound; Track standard resilience indicators (reserves, liquidity, margin).
Organisational	Reduction in reactive repairs	Number of reactive repairs, percentage of planned maintenance sustained over period of time	Consistent definitions for reactive vs planned; Temporary spike likely during/after works; Depends if asset registers updated and sustained
Cultural	Avoided loss of cultural output	Number of organisation closures (temporary and permanent), number of cancelled performances, number of tickets sold/ audience numbers, m2 of unusable space, lost revenue from cancellations	Account for substitution/rescheduling/digital delivery; Compare like‑for‑like seasons/artforms; Prioritise ticketed/recorded activity; estimate informal participation via samples/case studies where feasible
Cultural	Sector resilience	Number of financially secure organisations, investment and donation volumes	Agree threshold for ‘financially secure’; Donations/sponsorship cycles may be volatile over evaluation period
Cultural	Preservation of art and culture sector	Venues/ organisations avoided closing, retention of staff (FTE), sector GVA growth	Sector GVA is a broad, high-level metric affected by many things beyond CFF; Evidence via selected case studies and management records;
Cultural	Skills and knowledge preservation	Number of FTEs, number of volunteers	Headcount may not equate to skills: Use available HR metrics (retention/turnover, CPD hours); collect detailed skills data only in sample case studies
Cultural	Increase in UKs reputation for creative excellence	Export volumes (£), awards/ nominations, sentiment analysis, international tour volumes/ revenue	Long lags and noise in the data; Robust sentiment methods are still nascent: Optional light‑touch media review; advanced sentiment only if feasible
Environmental	Environmental sustainability	Total energy usage (kWh), energy produced through renewable sources (kWh), environmental standards rating (EPC, DEC)	Where data permit, normalise for weather (HDD/CDD) and hours open; otherwise report kWh and CO2e with caveats; Take into account rebound effects (comfort setpoints/hours)
Social impacts	Community cohesion	Number of active partnerships, local volunteering levels/ volunteer numbers at organisation, audience diversity	Attribution vs wider initiatives; GDPR/data governance
Social and wellbeing	Health and wellbeing	Self reported quality of life, e.g. EQ5D, volume of social prescribing, SWEMWBS, ONS personal wellbeing questions	Can potentially be interpreted as both final outcomes or measurement of intermediate outcomes.
Social and wellbeing	Pride in place	Advocacy scores, perceptions of town/ city/ high street, place attachment scale; local media sentiment	Area‑level confounding factors may be hard to quantify (regeneration, transport, policing); Robust sentiment methods are still nascent
Social and wellbeing	Education services	Number of new spaces created/ refurbished, number of learners engaged, learning hours delivered	Inputs (spaces, hours) do not equate to outcomes
Access and inclusion	Improved user experience	Number of visitors, visitor satisfaction, dwell time, number of hours open, number of performances/ events/ exhibitions, accessibility satisfaction levels	Disruption during works requires construction‑adjusted or pre-construction baselines. Dwell time and visits may not reflect user experience (once a visitor has entered the building, they may stay for the duration even if this does not reflect their satisfaction (sunk cost fallacy).
Access and inclusion	Preservation of access and participation	Number of accessible performances, ticket sold to under-served groups, events maintained during capital works, concessionary/free/companion tickets issued	Define accessible and underserved consistently; Weak historic EDI baselines
Access and inclusion	Growth in cultural access and participation	Visitor numbers – both repeat and unique visits, audience diversity vs local demographics, tickets sold to schools/ youth groups, attendance within priority places	Benchmark against national/local trends; Difficult to monitor displacement from non‑funded venues; Physical capacity ceilings may cap growth

4.2. Appraisal of data sources

4.2.1. Monitoring information

Expression of Interest

Organisations were required to submit an expression of interest (EOIs). The EOIs collected high-level information on region, organisation type and status, whether a charity, funding amount and a description of the proposed works. The EOIs alone will be unable to inform an evaluation, however, will provide a useful sample frame for future primary data collection.

Full application

As part of the full application process, organisations were required to submit baseline data, covering the following outcome indicators:

• Income in last financial year
• Audience numbers
• Number of hours open
• Energy consumption (kWh
• Number of performances/exhibitions/events
• Lost activity due to asset failure
• Financial impact of asset failure

It should be noted that this data is only available for organisations who progressed to the full application stage (n=117). Gathering similar data for potential counterfactual organisations will likely require primary data collection.

Ongoing ACE data returns

A condition of funding is that organisations are required to submit quarterly-, annual-, and end of- project data returns. As above, data is only collected for the funded organisations. Whilst funding allocations were not announced during the development of this Evaluation Framework, discussions with ACE suggest that approximately 40 applicants from Strand 1 and 20 applicants from Strand 2 will be funded. The ongoing data returns include:

• Quarterly data returns: These returns collect monitoring information tracking the project spend against expectations and progress updates. However, as they are not designed to capture the full range of outcome indicators, additional data collection methods will be needed to objectively measure changes in outcomes.
• Annual data returns: These returns will collect relevant outcome indicator data on:
  • Audience numbers
  • Number of hours open
  • Number of performances/exhibitions/events - lost activity due to asset failure
  • Financial impact of asset failure
  • Energy consumption

A critical consideration regarding the value of this data is the duration for which post-project data will be provided.

• End of project data collection: As with the quarterly returns, end of project data has a greater emphasis on evidencing project delivery against expectations (monitoring data). Whilst providing useful information for the process evaluation, the data collected at the end of project are unlikely to provide relevant metrics to understand changes in outcomes over time since they do not focus on post-project outcome indicators

Costed condition reports

Costed condition reports provide structured technical evidence regarding the condition, risk profile and maintenance requirements of buildings and associated infrastructure. Typically prepared by qualified surveyors or conservation professionals, these reports identify defects, estimate repair and replacement costs, and prioritise recommended interventions over short-, medium- and long-term maintenance cycles.

Within the CFF evaluation framework, these reports could provide an important source of baseline and follow-up evidence regarding changes in asset condition, maintenance backlog, and risks of operational failure. Where repeated over time using a consistent methodology, costed condition reports may also support assessment of whether CFF investment contributes to slowing deterioration, reducing critical maintenance liabilities, or extending the functional life of cultural assets.

4.2.2. Secondary data sources

A review of secondary data sources was undertaken to explore the extent to which publicly available information could provide timeseries (pre- and post-intervention) data on outcomes of interest. Where possible, publicly available data should be used in a full evaluation of CFF to minimise the burden placed on arts organisations.

The below sets out a non-exhaustive review of potential datasets, which have been identified through discussions with experts from DCMS, the wider stakeholder group, and a desktop review.

ACE National Portfolio Organisation data

ACE National Portfolio Organisations (NPO) are a group of 990 arts and cultural organisations that get regular funding from ACE. It was possible to identify approximate matches between the name of applicants from CFF in the NPO data. The results of this matching exercise indicated:

• 238 of the 722 CFF EOIs could be linked to the ACE 2023-2026 NPO data.
  • 90 organisations applied for Strand 1 and 29 organisations applied for Strand 2.
  • Whilst funding allocations had not been announced at the time of writing, conversations with ACE suggested that approximately half of the organisations who submitted full applications for each Strand will receive funding. Assuming proportions of EOI to full applications, it is estimated that there will be c.28 organisations which submit a full application for Strand 1, c.14 of which will be funded through Strand 1. Likewise for Strand 2, it is estimated that there will be c.14 organisations which submit a full application, c.7 of which are expected to receive funding.

Because many CFF applicants are also ACE‑funded NPOs and are simultaneously affected by other funding and market factors, changes observed in NPO survey metrics cannot be causally attributed to CFF alone. It will therefore be necessary to use NPO data primarily for benchmarking/context and, where used in counterfactual methods, only within carefully specified synthetic controls with strong pre‑trend fit and triangulation from primary/heritage science evidence.

NPOs are required to submit data through an annual survey.^{[footnote 8]} The annual survey includes questions spanning several topics:

• Workforce
• Finance
• Activity
• Audiences
• Learning and participation
• Touring and international
• Sector support and bridges

The annual survey collects data that is directly relevant to a future evaluation of CFF, however it should be noted that employee and volunteer figures are rounded to the nearest 5, which may limit use of NPO data in a full evaluation:

• Number of paid staff
• Number of volunteers
• Earned income
• Sources of public funding
• Expenditure (including overheads and collection conservation costs)
• Balance sheet (assets and liabilities)
• Audience inclusivity
• Number of exhibitions
• Loaning and digitisation of collections
• Audience numbers, including ticked events

ACE Illuminate platform

ACE also produce an audience data platform, which includes data on attendances and events for its funded organisations since 2023. This data is aggregated by region, and discipline, and may provide contextual evidence to demonstrate audience trends for the wider creative and cultural sector.

Charity Commission data

The Charity Commission for England and Wales is the regulator of registered charities. The Charity Commission provides data extracts, covering organisational and financial data of the registered charities. Charities are required to submit annual returns, which are publicly available and available for bulk download.^{[footnote 9]} Relevant data collected through the Charity Commission includes:

• Charity name and address
• Charity number
• Date of registration/date of removal/date of dissolution
• Income
• Government grants
• Expenditure
• Number of volunteers

Charities with income greater than £500,000 are required to provide additional information which would be relevant to an evaluation of CFF:

• Donation income
• Other sources of income
• Cash reserves
• Total asset value
• Value of liabilities
• Number of employees

Scoping work identified that 69% (483 out of 702) of the applicants were registered charities. Of these organisations, a proportion could be linked to Charity Commission data.

Business Structure Database

The Business Structure Database (BSD) provides an annual snapshot of the Inter-Departmental Business Register, covering all firms registered for PAYE and VAT (representing approximately 98% of economic activity in the UK).^{[footnote 10]} The BSD provides annual data, dating back to 1997, and can be accessed at the firm level (alongside the corresponding Output Area code) through the ONS SRS. The BSD provides the following metrics that may be relevant to a future evaluation of CFF:

• Sector level changes in employment and revenue.
• Standard Industry Classification codes.
• Employment (site level).
• Turnover (enterprise level).

There are several known limitations of the BSD, namely:

• Data lags: The BSD is constructed using several data sources (PAYE and VAT returns, the Annual Business Survey or Business Register of Employment Survey returns). This means that the BSD is updated at different points of time depending on when the information arrives from other sources. This means that some of the records can be up to two years out of date. A future evaluation of CFF should consider the significance of this data lag within the timelines of the evaluation.
• Turnover data: Turnover data is only captured at the enterprise level. This means that where firms have multiple sites, it must be assumed that there are equal levels of productivity across all sites. This is not expected to pose a significant threat to robustness for an evaluation of CFF as many organisations operate one site.
• Productivity: It is only possible to derive proxy measures of productivity (i.e. turnover per worker per year).

Meter point gas and electricity data

Meter point data is available through the ONS SRS.^{[footnote 11]} The data contains electricity and gas consumption in kWh and can in principle be linked to organisations using unique property reference numbers. Given publication lags (latest 2021 at time of writing), ONS meter‑point data should be used for contextual benchmarking only. Primary energy analysis will instead use venue‑held monthly meter or smart‑meter exports for electricity and gas.

Business insights and impacts on the UK economy

The business insights and impacts survey by the ONS focuses on various aspects of business performance, including financial health, workforce data, pricing, trade, and business resilience.^{[footnote 12]} It surveys approximately 10,000 businesses, with data collection waves occurring every 2 weeks. The most recent data sample included 431 businesses specifically from the ‘Arts, Entertainment, and Recreation’ sector. The survey captures broader business operation trends and may provide contextual insights and trends on these key areas.

• Cash reserves, debt, insolvency
• Capital expenditure
• Trade
• Carbon emissions
• Supply chains
• Turnover
• Employment

Purple Seven

Purple Seven is a commercial data source for the arts and cultural sector.^{[footnote 13]} Whilst it is not possible to access the data for free, the platform looks to include insights around audiences, revenue and visits.  Central/portfolio analytics licences are typically priced at £25,000 to £75,000 per year for aggregated dashboards/benchmarking across a cohort alongside potential export fees of £10,000 to £30,000 per year.

Social media and other public reviews

It may be possible to conduct sentiment analysis to understand feelings and attitudes towards the sector or specific organisations, based on customer feedback, social media monitoring and other sources. This may provide contextual information around attitudes towards the sector.

4.2.3. Primary data collection options

Purpose and approach

To fill gaps in the evidence base for this evaluation, primary data collection will be required to complement existing monitoring returns and secondary data sources. The data collection should be aligned with existing data (e.g. monitoring returns or secondary data) to ensure consistent outcome measurement across the funded organisations over time.

It is recommended to apply Computer-Assisted Telephone Interviewing (CATI) as the primary collection mode, with secure data follow-up requests for items organisations already hold as management information. This approach has previously proved successful in collecting data from similar organisations.

Sampling frame and respondents

The sampling frame should include all funded CFF awardees for Strand 1 and 2. It is not recommended to gather primary data collection among a comparison group. Through this method, the analysis in Section 6 identified that difference-in-differences designs are unlikely to yield sufficient statistical power.

Additional primary data collection against a comparison group is likely to be expensive, and challenging to achieve a modest response rate.

Waves and timing

The timing of data collection would be expected to occur during the following time periods:

• Baseline (retrospective): Data would be collected from the last fully completed financial year before works began (or the closest comparable pre-works period).
• Follow‑ups: Data would be collected for the first year post-completion of the works (the 2030 financial year), and over a longer-horizon four to five years post completion.

Outcome measurement

To ensure consistent and robust data collection, the survey design should align questions with published CFF fund guidance and definitions. The survey should be designed to contain a core set of standard questions and optional question blocks (modules).  This would minimise burden and ensure consistency to allow for statistical comparison of funded and unfunded organisations. Where feasible under GDPR, data should be collected to enable data linkage to reduce duplicate reporting.

Table 4.2 below contains a comprehensive list of outcomes that could be collected through primary data collection, to supplement data that may be partially covered by other sources. Whilst a QED using primary data is not considered feasible (e.g. a DiD) as noted in section 6, there may still be value in undertaking primary data collection to cover data gaps (descriptively), but also to inform the process and economic evaluation.

Table 4.2: Outcomes for primary data collection

Domain	Outcome
Project identifiers, scope and timeline	Site/venue identifiers; strand applied; award decision; consent to link to programme monitoring/public filings and to upload MI.
Project scope and timeline	Works categories (fabric/systems/equipment/access/energy), start/expected completion dates, closures (dates, full/partial), match sources (type/value).
Asset condition and reliability	Maintenance backlog: number of items, value (£, price base), criticality rating, planned preventative maintenance plan in place. Number of reactive repairs and cost (and proportion of assets requiring repair). Unplanned closures (count/hours). Cancellations: counts with primary reason. For example; power distribution; lighting/rigging/stage machinery; lifts/access equipment; IT/ticketing outage; fire/safety system fault; cast/staffing; demand.
Finance and resilience	Turnover by source (earned/public/philanthropy/other); total expenditure (overheads, collections/programme care, repairs, staff); surplus/deficit. Unrestricted/restricted reserves; insurance premium change. Donations and investment volume and value.
Workforce and volunteering	Employees total FTE; freelancers/contractor days (FTE‑equivalent); volunteers (headcount; hours); Employee wages, partnerships.
Programme, attendance and capacity	Events scheduled vs delivered; programme mix by broad type. Attendance total; capacity used (%) where known and tickets sold. Attendance from under-represented groups(%). m2 (and proportion) of temporarily unusable space. Number of learners and learning hours.
Cancellations and access	Cancellations (as above) and estimated lost revenue (£) where recorded. Accessible provision: accessible performances by type, concessionary/free/companion tickets (counts). Accessibility upgrades now operational (step‑free, lifts, Changing Places; tick‑box).
Environmental	Electricity/gas kWh monthly for last 12 months; metering set‑up (whole‑building/sub‑meter/shared estate); EPC/DEC (rating/date). Energy generated through on‑site renewables (kWh, if relevant). Normalisers: hours open/week; fixed seat/standing capacity; which areas are served by upgraded systems (whole building/main auditorium/FOH/BOH).
User experience	Customer satisfaction scores, dwell times, awards or nominations. Noting this is likely to provide descriptive information for each museum opposed to being considered a robust quantitative measure. The lack of consistency in how organisations collate this information overtime, as well as differences between organisations, means this data is unlikely to be robust enough to support statistical analysis.

Heritage science data

The Heritage science method has been designed to minimise data collection requirements from funded organisations. The data requirements largely rely on data that applicants would have already gathered through the EOI, full application, and costed condition surveys.

• Data collection instrument: It is anticipated that funded applicants will be given a ‘heritage science form’, likely Excel based, to gather the required information. Whist the application is not anticipated be challenging to complete, it is recommended that a heritage science specialist holds time to assist applicants where difficulties arise. This approach has proven effective in previous evaluations.
• The primary data collection items are likely to include:
  • Asset type: Applicants will need to define the specific assets which will be affected by the funding. Applicants will need to select fields from a pre-defined set of categories – which follow RICS categories of building elements (structure, roof, façade, internal finishing, external areas, service installations, or ‘other’ for novel equipment).
  • Baseline condition: For each asset affected by funding, applicants will be required to provide an assessment of current condition.
  • Expected lifetime: Applicants will be required to estimate the number of years of serviceable lifetime left for the funded assets in its current state (i.e. without receiving funding).
  • Rate of degradation: Applicants must provide insights into how quickly their assets are degrading. Applicants can choose from continuous rate of degradation, degradation that increases as assets get older, or rapid late-stage deterioration.
  • Intervention scenario parameters: Applicants will be required to provide data on the intervention start year, the duration of the works, the condition uplift (i.e., the condition of the asset following the completion of the works), and the expected lifetime of the asset following repair.
  • Real life operating factors: Data will need to be gathered on usage intensity (categorised as low, typical or high) and maintenance activity (categorised as reactive, routine or planned).
  • Cultural service thresholds: Applicants will be required to make professional judgements on the physical condition scores (i.e. the mechanism to determine the baseline condition and condition uplift) required to continue to deliver cultural services.

4.3. Linking outcome indicators and data sources

The below matrix maps outcome indicators with data sources. Primary data collection is not included in the table below. Green shading means that the data source provides comprehensive coverage for both successful and unsuccessful organisations. Amber shading means that the data source only provides partial coverage, where the limitations in coverage are noted within the respective cell, and white shading means that the metric is not covered by the data source; this data would therefore need to be collected on a primary basis, if judged important. Area level data is also not included in the below table.

The below matrix maps outcome indicators with data sources. Primary data collection is not included in the table below. Green shading means that the data source provides comprehensive coverage for both successful and unsuccessful organisations. Amber shading means that the data source only provides partial coverage, where the limitations in coverage are noted within the respective cell, and white shading means that the metric is not covered by the data source; this data would therefore need to be collected on a primary basis, if judged important. Area level data is also not included in the below table.

Table 4.3. Data source coverage matrix

	EOI/ Application data	Monitoring data	ACE NPO data	Charity Commission data	Business Structure Database	Meter point gas and electricity data	Purple Seven
Specialist contractor FTE	-	-	-	-	-	-	-
Capital work contracts awarded by firm size and geographic distribution	-	-	-	-	-	-	-
Number of employees	-	-	Rounded values and partial coverage	Partial coverage	Comprehensive coverage	-	-
Number of volunteers	-	-	Rounded values and partial coverage	Partial coverage	-	-	-
Employee wages	-	-	-	-	-	-	-
Organisation revenue	Funded orgs only	Funded orgs only	Partial coverage	Partial coverage	Available at org level not site level	-	-
Organisation profit	-	-	-	-	-	-	-
Local area GVA	-	-	-	-	-	-	-
Local area employment	-	-	-	-	-	-	-
Footfall	-	-	-	-	-	-	-
Visitor spend	-	-	-	-	-	-	-
Hotel occupancy rates	-	-	-	-	-	-	-
Local business rates	-	-	-	-	-	-	-
CI firm density	-	-	-	-	-	-	-
Number of unplanned closures	Funded orgs only	Funded orgs only	-	-	-	-	-
Number of cancelled performances	Funded orgs only	Funded orgs only	-	-	-	-	-
Number of critical items on the backlog	-	-	-	-	-	-	-
Insurance premiums	-	-	-	-	-	-	-
Number of usable seats	-	-	-	-	-	-	-
Estimated lifetime/ lifetime extension	-	-	-	-	-	-	-
Value of maintenance backlog	-	-	-	-	-	-	-
Overheads	-	-	Partial coverage	Partial coverage	-	-	-
Revenue loss from cancellations	Funded orgs only	Funded orgs only	-	-	-	-	-
Reactive repair cost	-	-	-	-	-	-	-
Value of reserves	-	-	-	Partial coverage	-	-	-
Operating margin	-	-	-	-	-	-	-
Liquidity ratio	-	-	-	-	-	-	-
Survival probability	-	-	-	-	-	-	-
Number of reactive repairs	-	-	-	-	-	-	-
Proportion of maintenance which is reactive	-	-	-	-	-	-	-
Number of organisation closures	-	-	-	Partial coverage	Comprehensive coverage	-	-
Number of tickets sold	-	-	Partial coverage	-	-	-	Partial coverage
Audience/ visitor numbers	Funded orgs only	Funded orgs only	Partial coverage	-	-	-	Partial coverage
m2 of useable floor space	-	-	-	-	-	-	-
Investment value	-	-	-	Partial coverage	-	-	-
Donation value	-	-	-	Partial coverage	-	-	-
Export value	-	-	-	-	-	-	-
Awards/ nominations	-	-	-	-	-	-	-
Number of international tours	-	-	-	-	-	-	-
Electricity consumption (kWh)	Funded orgs only	Funded orgs only	-	-	-	Partial coverage	-
Gas consumption (kWh)	Funded orgs only	Funded orgs only	-	-	-	Partial coverage	-
Energy generated through renewable sources (kWh)	-	-	-	-	-	-	-
Environmental rating	-	-	-	-	-	-	-
Number of active partnerships	-	-	-	-	-	-	-
Number of volunteers	-	-	-	-	-	-	-
Attendance from underrepresented groups	-	-	-	-	-	-	-
Self-reported quality of life (e.g. EQ5D)	-	-	-	-	-	-	-
Self-reported life satisfaction	-	-	-	-	-	-	-
Number of learners engaged	-	-	-	-	-	-	-
Number of new learning spaces created	-	-	-	-	-	-	-
Hours of learning delivered	-	-	-	-	-	-	-
Opening hours	Funded orgs only	Funded orgs only	-	-	-	-	-
Number of performances/ events	Funded orgs only	Funded orgs only	Partial coverage	-	-	-	-
Accessibility satisfaction levels	-	-	-	-	-	-	-

Note: ACE NPO data has only partial coverage, and this means that attribution would be caveated, since ACE revenue support and concurrent shocks confound any impacts detected.

4.4. Recommended outcome indicators

The previous section presented a suite of potential outcome indicators, which collectively would provide a complete assessment of the effectiveness of CFF. However, experience on previous evaluations of a similar nature has determined that it would not be proportionate, nor cost-effective, to adopt the full suite of outcome indicators. The table below outlines a set of recommended outcome indicators that could be prioritised for an evaluation of CFF.

Table 4.4: Proposed outcome indicators

Metric	Rationale	Secondary data source
Maintenance backlog	-	-
Value (£)	Identifies whether the maintenance backlog changed in value over time.	No secondary data source
Number of items	This adds depth to the above metric.	No secondary data source
Heritage science metrics	-	-
Risk reduction	Quantifies the extent to which CFF funding was successful in mitigating likelihood of asset failure	Requires heritage science modelling
Asset lifetime extension	Quantifies how the CFF funding preserved the organisation assets	Requires heritage science modelling
Energy usage	-	-
Electricity usage (kWh)	Shows how the funded works impacted electricity consumption	Applications for funded organisations
Gas usage (kWh)	Shows how the funded works impacted gas consumption	Applications for funded organisations
Energy generated from renewable resources (kWh)	Shows how the funded works impacted the adoption of renewable energy sources	Applications for funded organisations
Reactive vs preventative maintenance	-	-
Percentage of annual maintenance budget on planned vs unplanned repairs	Provides insights into how maintenance activity is undertaken within organisations and can show behaviour change over time.	No secondary data source
Annual spend on maintenance (£)	Decreasing trends in maintenance spend over time may reveal that the funding was successful in enabling a more proactive approach.	No secondary data source
Revenue	-	-
Total revenue (£)	A measure of commercial activity for the organisation, allowing insights into how the funding affected the commercials.	BSD, partial coverage in NPO and Charity Commission data
Revenue (£) disaggregated by activity	Can identify which specific attributes of commercial activity were most affected (positively or negatively) from funded capital works.	Partial coverage in NPO
Donations (£)	Understand how the funded affects the donations that an organisation receives	Partial coverage in NPO and Charity Commission data
Lost activity due to asset failure	Provides a direct measure of avoided loss following completion of the works	Applications for funded organisations
Number of closures due to asset failure	Provides a direct measure of avoided loss following completion of the works	Applications for funded organisations
Staffing	-	-
Number of paid staff	A proxy measure for organisational health, access and resilience.	BSD, partial coverage in NPO and Charity Commission data
Number of volunteers	A proxy measure for organisational health, access and resilience, as well as an indicator of public attitude towards the organisation.	Partial coverage in NPO and Charity Commission data
Visitor engagement	-	-
Number of visitors	Indicator of public demand for culture and heritage – and how this changed due to funding.	Applications for funded organisations
Days open to public	Indicator of public supply of culture and heritage – and how this changed due to funding.	Partial coverage in NPO

5. Process evaluation

The process evaluation will assess the implementation and delivery of CFF, examining the effectiveness of the key activities. A key focus of the process evaluation will be exploring challenges encountered by projects, and lessons learned in terms of successfully delivering the capital projects. This section sets out the overarching framework for the process evaluation.

5.1. Process overview

Section 1 of this report describes the overall process for fund delivery. The below figure also presents an overview of the CFF process.

Figure 5.1: High level process map

Figure 5.1 is a map of the evaluation process. The process begins with communications and Engagement, followed by the EOI stage, the Full Application Stage, the Portfolio Balancing, the beginning of the Funded works, and then splits into Ongoing monitoring and payments in arrears.

DCMS and ACE are the two key organisations involved in the design, delivery and oversight of CFF, and their specific roles are outlines below:

• DCMS: Have overall responsibility for CFF, ensuring that the delivery aligns with the Business Case, which set out how CFF will support the cultural sector. DCMS are not involved in programme assessment or decision-making on individual grants. DCMS will provide £75 million of capital funding (88% of total capital funding).

• ACE: Are delivering CFF, on behalf of DCMS. ACE are responsible for running the application processes, assessing Expressions of Interest (EoIs) and full applications, convening and chairing the decision panel meetings, drawing up grant agreements, monitoring and paying grants, including counter fraud activities. ACE will provide £10 million of capital funding (12% of total capital funding).

5.2. Data requirements

It is anticipated that the process evaluation will draw on two key sources of data, i) interviews with a range of relevant stakeholders, and ii) monitoring data held by DCMS and ACE.

5.2.1. Interviews

Depth interviews will be conducted with key stakeholders who engage with CFF. This is anticipated to include:

• Central government/ ACE policy and analysts: To understand the rationale for intervention (including why the chosen policy option was considered advantageous against alternatives) and perspectives of how CFF has been delivered.

• Assessor interviews: To understand the views from assessors on adequacy and relevance of documents submitted for the purposes of the application, the apparent need and urgency of the work the applicants applied for and the quality of the applications.

• Successful applicants: To understand the end-to-end process of CFF from the perspective of successful applicants, including experiences in delivering projects and lessons learnt for potential future cohorts/other capital maintenance programmes.

• Unsuccessful applicants: To understand the views of the application process from the perspective of unsuccessful applicants, including whether any constructive feedback was provided and how failure to obtain funding impacted their organisations.

• LA stakeholders: To gather views from external stakeholders, particularly around match funding requirements and the availability of match funding. These stakeholders may also provide an understanding of how the overarching aims of CFF align with local authority priorities.

• Eligible non-applicants: To understand the rationale for not applying for CFF, such as insufficient need, challenges in completing the application (e.g. capacity, resources or skills), or challenges in obtaining required levels of matched funding.

5.2.2. Monitoring information

The monitoring activities undertaken by the DCMS and ACE will directly inform the process evaluation. It is anticipated that this will include key programme metrics. These include application volumes, volumes of clarification questions, application data and ongoing data collection and progress reports from funded projects. The monitoring activities will provide qualitative and quantitative measures for the process evaluation. These can be used to assess the implementation in terms of activities undertaken, outputs produced, process costs and performance against targets where specified.

5.3. Process evaluation framework

The below table maps each of the key evaluation questions, and sub-questions, with potential assessment criteria, to the relevant evidence sources.

Table 5.1: Process evaluation evidence mapping

Evaluation questions	Assessment criteria	Data sources: MI Data	Data sources: Programme team interviews (DCMS and ACE)	Data sources: Assessor interviews	Data sources: Successful applicant interviews	Data sources: Unsuccessful applicant interviews	Data sources: LA stakeholders	Data sources: Eligible non-applicants
KEQ1: Were notifications of the funding opportunity to prospective applicants clear and did they provide sufficient time to develop quality applications?	Guidance clearly explained eligibility, strands, timelines, documentation and criteria. Most prospects understood purpose, aims and eligibility. Time window was seen as adequate to prepare good bids and secure match funding.	-	Y	-	Y	Y	-	Y
KEQ1a: Did applicants find the arrangements for clarification questions suitable? Which resources were the most/least helpful?	Most applicants knew about Q&A channels and how to use them. Channels were accessible to different organisations. Questions were answered promptly and helpfully. FAQs/guidance was updated to reduce recurring confusion.	-	-	Y	Y	Y	-	-
KEQ1b: What were the volumes of ineligible applicants and did these align with expectations? How did the volumes compare to other DCMS or ACE programmes?	Ineligible share within expected range. Main ineligibility reasons were minor/borderline, not widespread misunderstanding. No big clusters of ineligible applicants from particular organisations or areas. Ineligible volumes were manageable within assessment capacity. Ineligible volumes are comparable to other programmes.	-	Y	Y	-	-	-	-
KEQ1c: How effectively was the sector informed about the fund’s introduction through communication efforts?	Most applicants knew about the introduction of the fund and how to apply, through DCMS communications.	-	-	-	Y	Y	-	Y
KEQ2: To what extent did CFF reach organisations with the highest priority needs in terms of assets failure/deferred maintenance?	Significant share of applications/awards from organisations with clear asset failure or high risk/deferred maintenance. Awards concentrated on projects meeting “priority need” definitions. Stakeholders generally viewed recipients as those with the most pressing capital issues.	Y	Y	Y	-	-	-	-
KEQ2a: To what extent did CFF funding reach organisations offering the greatest benefit from the available funding?	Most funded projects showed strong expected cultural, social and/or economic benefits. Impact/benefit clearly used and applied in assessment decisions. Portfolio balanced urgent capital need with wider benefits. Stakeholders saw the overall portfolio as good value and strategically aligned.	Y	Y	Y	-	-	-	-
KEQ2b: Were application volumes in line with expectations?	Total applications broadly matched forecasts. Application coverage across strands, regions, org types and priority places. Limited evidence of major unmet demand or weak demand.	Y	Y	Y	-	-	-	-
KEQ2c: How far did the exclusion of for-profit organisations limit reach of CFF and its wider impact?	Stakeholders reported few important missed opportunities due to for‑profit exclusion and provide clear justification for exclusion. No strong pattern of systemic gaps or lost impact from the exclusion.	Y	Y	-	-	-	-	Y
KEQ3: Did Strand 1 vs Strand 2 rules (e.g. equipment-only eligibility, higher match funding requirement for Strand 2) function as intended?	Applicants generally understood strand differences and chose correctly. Applicants understood which stand they needed to apply to with no need to a clarification question.	-	Y	-	Y	Y	-	Y
KEQ3a: Was the required match funding level proportionate?	Match funding as proportion of project value. LAs viewed match funding requirement as proportionate. Applicants didn’t report any challenges in securing value of match funding/raise concern around whether this was prohibitive in submitting and application.	Y	-	-	Y	Y	Y	Y
KEQ3b: Were match funding thresholds appropriate to need and market conditions, and proportionate for smaller/heritage organisations and priority places?	Smaller organisations/organisation in priority places did not report that this was prohibitive to applying. Ratio of match funding to project value for smaller organisations or projects in priority places.	Y	-	-	Y	Y	-	Y
KEQ3c: What sources of match funding were most used (LA, philanthropy, trusts, reserves, in-kind)?	Clear MI on type and value of match funding used (e.g. LA, philanthropy, trusts, commercial, reserves, in‑kind). A small number of match sources account for most match value, with patterns that broadly reflect expectations and market conditions. Distinct patterns are visible by organisation size, type and geography (e.g. smaller orgs using more reserves/in‑kind; some areas relying more on LAs). Stakeholders can explain why particular sources were more/less used (e.g. local authority capacity, strength of local philanthropic base). Limited evidence that heavy reliance on any one source created systemic risks (e.g. over‑stretching reserves).	Y	Y	-	-	-	-	-
KEQ3d: To what extent did requirements lead to exclusion of potential priority projects or impact the volume and characteristics of applications?	Evidence that requirements were/ were not a key reason for not applying, withdrawing, or scaling back proposals. Representation of intended priority groups among actual applicants compared with the expected pool of potential applicants.	Y	Y	Y	-	Y	-	Y
KEQ4: Did programme timelines give funded applicants sufficient time for projects to undertake capital work to desired specification/standards?	Most projects completed within prescribed timelines. Minimal delays and minimal changes to scope. Changes were due to unanticipated factors. Quality standards were not routinely compromised to meet deadlines.	Y	Y	-	Y	-	-	-

Application process

Evaluation questions	Assessment criteria	Data sources: MI Data	Data sources: Programme team interviews (DCMS and ACE)	Data sources: Assessor interviews	Data sources: Successful applicant interviews	Data sources: Unsuccessful applicant interviews	Data sources: LA stakeholders	Data sources: Eligible non-applicants
KEQ5: Was the two-stage process (EOI progressing into full application) clear, proportionate and timely for different applicant types?	Guidance clearly explained EOI vs full application purposes, criteria, and progression rules. Applicants generally understood what was expected at each stage and why some EOIs do/do not progress. Timings between EOI and full stage were seen as reasonable to develop bids (especially for smaller/less‑resourced organisations). Limited evidence that good‑quality projects dropped out mainly due to unclear, burdensome or poorly timed staging. Number of applicants advised to submit full application which did not. Number of applicants not advised to submit full application which did.	Y	Y	Y	Y	Y	-	Y
KEQ5a: What support, if any, was needed or provided to smaller organisations?	Common typology of FAQs from known smaller organisations. Clear evidence of support offers (guidance, webinars, Q&A sessions, templates). Smaller organisations that accessed support reported it as timely, relevant and helpful in progressing through stages. Limited evidence that lack of support alone caused capable smaller orgs to drop out or not apply.	-	Y	-	Y	Y	-	-
KEQ5b: Was the application process and readiness requirements (e.g., RIBA Stage 3, costed plans, engagement with specialists) proportionate to project scale and risk?	Upfront cost as proportion of funding award Number of projects which faced delivery challenges. Number of projects which failed due to insufficient readiness.	Y	-	Y	-	-	-	-
KEQ5c: Was the application process disproportionately burdensome on smaller organisations (e.g. cost of reaching RIBA Stage 3, cost of engaging with specialists, knowledge and skills of staff, staff capacity).	Small organisations perceived resource and time costs as reasonable relative to award size. Variation in time and resource costs as a proportion of grant size, by organisation size.	-	-	-	Y	Y	-	Y
KEQ5d: Did the volume of information required in the application form discourage organisations from applying?	Few eligible non‑applicants cited form length/complexity as the main deterrent. Most applicants saw information requested as broadly necessary/proportionate. Collected information was used and added value to assessment.	-	-	Y	Y	Y	-	Y

Project selection

Evaluation questions	Assessment criteria	Data sources: MI Data	Data sources: Programme team interviews (DCMS and ACE)	Data sources: Assessor interviews	Data sources: Successful applicant interviews	Data sources: Unsuccessful applicant interviews	Data sources: LA stakeholders	Data sources: Eligible non-applicants
KEQ6: Did ACE have sufficient resource to undertake assessment process within the prescribed timeframes?	Actual assessment workload and staffing/capacity versus expectations. Application assessments completed within prescribed timescales. Number of applications not assessed in prescribed timescales. Limited evidence of quality issues (e.g. rushed assessments, generic feedback/ comments) due to resource constraints. Few significant process changes (e.g. emergency triage, drastic simplification) driven solely by under resourcing.	-	Y	Y	-	-	-	-
KEQ7: Did assessment criteria and scoring reliably identify urgency, additionality, readiness, risk and potential benefits?	Criteria and scoring framework clearly mapped to core constructs (urgency, additionality, readiness, risk, benefits). Assessors reported criteria were understandable and usable in practice. Scoring showed reasonable consistency both within and between assessors. Limited evidence of high-quality/high need cases systematically “falling through the gaps” due to criteria design. Limited evidence of low need cases being funded.	-	Y	Y	-	-	-	-
KEQ7a: Were the funds allocated as intended and to the organisations most in need of the funding?	Proportionate distribution of awards across different dimensions (e.g. organisation type, priority places, geographic spread). Funded organisations generally showed higher levels of need and urgency than unsuccessful applicants.	Y	Y	Y	Y	Y	-	-
KEQ7b: Were the potential benefits of funding maximised?	Portfolio included a variety of projects with high expected cultural, social, economic and/or environmental benefits. Evidence that assessment and decision-making actively considered maximising aggregate benefits within constraints. This could include, for example, evidence of scoring matrices or Multi criteria analysis that consider budget, geography, need, deprivation. Early indications (or projections) suggest funded projects are on course to deliver expected benefits.	Y	-	-	Y	-	-	-
KEQ7c: How was funding allocated to different activities? Were there any priority activities?	Clear breakdown of awards by activity type (e.g. remedial maintenance, upgrades, accessibility, environmental performance, equipment, new build). Limited evidence that important intended activity types were systematically under‑funded without clear rationale.	Y	-	-	-	-	-	-
KEQ8: Were moderation, decision-making and portfolio-balancing processes transparent, consistent and free of conflicts of interest?	Clear, documented moderation and decision-making processes, including portfolio balancing steps. Robust conflict of interest (COI) processes in place with no evidence of breaches. Unsuccessful applicants understood why they were not awarded funding. Decisions perceived as transparent.	-	Y	Y	-	Y	-	-
KEQ8a: Did selection decisions achieve regional/sectoral balance and reach priority groups/places without excluding crucial projects?	Portfolio showed a reasonable spread across regions, sectors and priority groups/places, which reflected stated aims. No strong evidence that achieving balance significantly reduced support for very high‑need or high‑impact projects overall. Stakeholders broadly agreed that regional/sectoral balance was appropriate and did not undermine the sustainability of the cultural ecosystem.	Y	Y	Y	-	-	-	-
KEQ8b: Did assessment decisions consider the adequacy of contingency and optimism bias/ provisions in applicants cost estimates?	Guidance and criteria explicitly referenced treatment of contingency and optimism bias. Assessors reported that they routinely reviewed costings, contingencies and delivery assumptions, challenging unrealistic budgets. Evidence in assessment records of adjustments or concerns raised about optimism bias and insufficient contingency. Limited instances of funded projects which faced severe cost overruns or viability issues attributable to under‑estimated costs that should have been apparent at assessment.	Y	-	Y	-	-	-	-
KEQ9: To what extent did selection choices align with fund objectives (safeguarding cultural output, organisational sustainability, widening access, environmental performance)?	Clear mapping from fund objectives to assessment criteria and decision‑making tools. Portfolio demonstrably contained projects contributing to core objectives (safeguarding output, sustainability, access, environment). Stakeholders broadly perceived that funded projects collectively advance the fund’s key objectives.	Y	Y	-	-	-	-	-

Contracting and monitoring

Evaluation questions	Assessment criteria	Data sources: MI Data	Data sources: Programme team interviews (DCMS and ACE)	Data sources: Assessor interviews	Data sources: Successful applicant interviews	Data sources: Unsuccessful applicant interviews	Data sources: LA stakeholders	Data sources: Eligible non-applicants
KEQ10: Did ACE have sufficient resource to efficiently manage the contracting process and deliver funding on time?	Contracting timelines met planned schedules. Internal capacity was adequate to process contracts and complete due diligence. No evidence of issues which would have been picked up during due diligence. Few reports from organisations of delays or confusion in contracting.	Y	Y	-	Y	-	-	-
KEQ11: Were funding arrangements (e.g. payments in arrears) manageable for organisations with limited reserves?	Most organisations, including those with limited reserves, could manage cashflow without severe strain (e.g. emergency borrowing, staff non‑payment). Limited evidence that arrears payments were a primary deterrent to applying or accepting funding.	Y	Y	-	Y	-	-	Y
KEQ11a: Did this impact delivery of the planned works?	Funding arrangements did not cause significant delays, de‑scoping or quality compromises due to payment timing/cashflow constraints. Organisations with weaker reserves were not systematically more likely to under‑deliver due to payment structures.	Y	Y	-	Y	-	-	-
KEQ12: How effective were governance, procurement and risk management within funded organisations? What support did ACE provide during delivery?	Funded organisations had appropriate governance structures and procurement processes for project scale and type. Key delivery risks were identified, owned and actively managed at project level. ACE provided clear guidance and oversight on governance/procurement expectations and risk management. Where issues emerged, ACE support (advice, conditions, variations) was timely and constructive. Appropriateness and effectiveness of the governance process could be assessed through development of criteria. For example, High: Clear risk registers and evidence of meetings/ oversight, Low: No evidence of risk registers and meetings/oversight.	-	Y	-	Y	-	-	-
KEQ12a: Was there sufficient specialist contractor capacity (heritage, retrofit, theatre tech), and what were the consequences if it was constrained?	Project could secure appropriate specialist contractors within required timeframes and budgets. Extent to which capacity constraints led to cost increases, scope changes or quality changes. No strong evidence that lack of specialist capacity systematically Mprevented high priority works.	-	Y	-	Y	-	-	-
KEQ13: What short-term disruptions (closures, cancellations, displacement, construction externalities) occurred and how were they mitigated?	Projects identified and planned for likely short-term disruptions to audiences, communities and neighbours. Recorded disruptions (closures, reduced programming, noise, access issues) were time limited and broadly proportionate to works. Bounce-backs were evident post-completion. Mitigation measures (alternative venues, scheduling, communications, support to affected groups) were used.	Y	Y	-	Y	-	-	-
KEQ14: How effective were monitoring arrangements in managing risk, scope, time and budget?	Monitoring requirements (reports, milestones, financial returns) are clear and consistently applied. Monitoring returns were timely and of sufficient quality to track risk, scope, time and budget. Programme team proactively used monitoring data to identify issues and intervene where needed, and provide effective corrective actions. Limited instances of unforeseen challenges without prior monitoring signals.	Y	Y	-	Y	-	-	-
KEQ14a: Were monitoring arrangements reviewed and refined over time? Did this improve delivery over time?	Monitoring burden (frequency, depth, evidence requested) scales was perceived as reasonable given the size of the grant awarded. Limited evidence of projects being significantly burdened or threats to delivering on time due to monitoring demands.	-	-	-	Y	-	-	-
KEQ14b: What feedback loops were in place? Did these improve delivery over time?	Clear processes for feeding monitoring/learning back to projects (e.g. progress meetings, written feedback, guidance updates). Evidence that recurring issues were identified and addressed through updated guidance, FAQs, or process changes. Projects report that feedback from ACE helped them improve delivery, risk management or reporting. Over time, common problems (e.g. specific compliance issues) reduce in frequency or severity.	Y	Y	-	-	-	-	-
KEQ14c: Did monitoring enable the identification of emerging problems? Did monitoring allow corrective actions to be put in place?	Monitoring data routinely flagged emerging risks (delay, overspend, scope creep, governance issues) before they become critical. ACE took timely action in response (e.g. support, re‑profiling, conditions). Projects generally perceived monitoring/ACE engagement as helpful in resolving issues, not just compliance‑driven. Limited examples where serious project failure could reasonably have been anticipated but was missed by monitoring.	Y	Y	-	Y	-	-	-
KEQ15: Were outputs delivered as planned?	Most planned outputs delivered. Consistency between planned output indicators and actual deliverables.	Y	-	-	Y	-	-	-
KEQ15a: Were outputs delivered within expected costs and expected timelines?	Most outputs delivered within budget, and on schedule.	Y	-	-	Y	-	-	-
KEQ15b: What factors may have contributed to or impeded delivery as planned?	Internal drivers and barriers to delivery. External drivers and barriers to delivery.	Y	-	-	Y	-	-	-
KEQ15c: How did projects sequence works to minimise disruption? Which mitigations were most effective?	Detailed project plans, including contingency and mitigation. Evidence of engaging with specialist project managers, or evidence in in-house expertise.	Y	-	-	Y	-	-	-
KEQ15d: Which procurement strategies best balanced price, quality, and specialist capacity?	Evidence of detailed procurement process, receiving multiple quotes from multiple organisations. Organisations can describe the process or present their frameworks for weighing up price and quality.	Y	-	-	Y	-	-	-
KEQ15e: How were maintenance plans operationalised post-handover?	Projects can identify how and where maintenance plans have been implemented. Projects can present costs of implementation. Over the longer term, reactive repair costs fall.	Y	-	-	Y	-	-	-
KEQ16: What were the key challenges encountered during project delivery, how were they overcome, and what are the overarching lessons learnt for future capital funding programmes?	Broad identification of the types of challenges (internal vs. external) faced across the portfolio. General consensus among stakeholders on the overall success of project delivery despite these challenges. Synthesis of overarching lessons learnt documented at project close.	Y	Y	-	Y	-	-	-
KEQ16a: How did macroeconomic factors (e.g. inflation, supply chain constraints) impact project scope, budget and timelines?	Evidence of budget variances, cost overruns, or timeline delays directly linked to inflation, material price increases, or labour market tightness. Extent to which project scopes were reduced (e.g. value engineering) or altered due to cost increases. Number of projects requiring formal change requests or extensions due to macroeconomic factors.	Y	Y	-	Y	-	-	-
KEQ16b: What mitigation strategies were most effective in overcoming these challenges?	Clear documentation of actions taken by projects (e.g. phased delivery, re-scoping, drawing on unrestricted reserves, or securing additional match funding). Assessment of which strategies successfully kept projects on track versus those that were less effective. Evidence of whether local authorities or other partners had to step in to bridge inflation-related funding gaps.	Y	Y	-	Y	-	-	-
KEQ16c: What lessons can future prospective applicants learn in terms of delivering large-scale capital funding projects?	Clear documentation of peer-to-peer advice and best practices for managing large-scale capital projects (e.g. establishing robust governance boards, internal resourcing, and the value of hiring dedicated project managers). Insights into effective procurement strategies, managing specialist/heritage contractors, and sequencing works to minimise venue disruption and loss of cultural output. Recommendations regarding realistic budgeting, contingency planning, and timeline expectations from the applicant perspective.	Y	-	-	Y	-	-	-

5.4. Evaluation timing

CFF projects are expected to commence from 1 April 2026 onwards. Projects are expected to finish within 36 months of commencement. ACE guidance identifies that the first payment will be scheduled for three-months post agreement, and that interim payments will be made in quarterly instalments. It is important to ensure that monitoring and contracting aspects can be comprehensively assessed within the process evaluation. Therefore, it is recommended that the process evaluation is conducted approximately 12 following the start of the capital work (i.e. April 2027).

There may also be value in undertaking a second-stage process evaluation, revisiting key process evaluation questions and developing insights. These can be fed into project completion, final payments/ contract end, and provide a mechanism for understanding short-term impacts. This can be aligned with the first impact evaluation in April 2030.

6. Counterfactual impact evaluation

This section provides recommendations on constructing a credible counterfactual and undertaking econometric analysis to understand the impact of CFF. It should be noted that at the time of writing, the funding projects and Cultural Priority Places (CPPS) were not announced. The final approach will also need to take account of CPPs. The below recommendations were made based on the review of EOIs and organisations which submitted full applications.

6.1.Definition of a counterfactual

A counterfactual impact evaluation would seek to establish the causal effect of the programme, relative to a scenario where funding was not awarded to arts organisations. To proxy the counterfactual scenario, a group of organisations (comparison group) that did not receive funding should be identified. This group of organisations can be considered equivalent to the organisations which received funding at the point at which the allocation of funding was decided.

6.2.Counterfactual considerations

There are two distinct selection processes that affect counterfactual construction. First, self‑selection into the applicant pool (applicants vs non‑applicants): organisations that choose to apply are likely to differ on unobservables (e.g. leadership capacity, risk management, board effectiveness, fundraising networks) from non‑applicants; this creates selection bias if non‑applicants are used as comparators. Second, award selection within the applicant pool (successful vs unsuccessful): the assessment process is designed to prioritise projects with higher need/urgency/readiness and strategic fit; this creates systematic differences between funded and unfunded applicants. The first pathway mainly threatens designs using non‑applicants; the second mainly threatens designs using unsuccessful applicants. We distinguish and mitigate these separately below.

The design of the impact evaluation, and identification of a counterfactual, will need to consider the following key issues:

• Applicant self‑selection (applicants vs non‑applicants): Organisations opt in at EOI/full‑application.^{[footnote 14]} Applicants may differ on unobservable managerial and governance capabilities, risk appetite, and growth trajectories relative to non‑applicants, making non‑applicants a weak counterfactual without very strong covariate control (often infeasible) or instruments (unlikely here).
• Award selection (successful vs unsuccessful applicants): Among applicants, assessment and moderation favour higher‑priority, higher‑readiness, and higher‑benefit projects. Funded organisations may therefore have greater need/urgency (e.g. asset criticality), stronger delivery capacity (e.g. RIBA 3, governance), or different financial positions than unsuccessful organisations. The EOIs and other organisation characteristics (possibly collected through the full application and primary data collection) could be used to select a group of organisations which could be considered comparable.
• Additionality of outputs: The counterfactual needs to consider the outputs, as well as outcomes of the programme. This provides a test for the extent to which the capital works would have happened in the absence of CFF, including whether there were differences in scale and quality of the work.
• Comparability of Strand 1 and 2: As set out in Section 2, there are two parallel funding streams: Strand 1 for grants between £100,000 and £1,000,000 and Strand 2 for grants over £1,000,000 and up to £10,000,000. Whilst the overarching aims and objectives of the funding are the same, it would be anticipated that the size, scale and potential impact of projects will differ between the strands. The organisations applying for both Strand 1 and 2 may also systematically differ.
• Sample size: Sample sizes may pose a challenge to implementing a quasi-experimental design (QED). In particular, samples sizes are likely to be prohibitive for sub-group analysis. The table below presents the volumes of applicants by EOI and full application stage, disaggregated by Strand 1 and 2.

Strand	Expression of Interest volumes	Full application volumes	Funded
Strand One	589	90	58
Strand Two	133	29	16
Total	722	119	74

• Prioritisation by need/urgency/readiness: Because awards target the highest priority projects, unsuccessful applicants may be systematically less urgent’ or less ready at baseline. Without mitigation, comparisons will understate effects (downward bias). This can potentially be addressed by:
  • Stratifying within strand/venue type/region/grant band/priority place.
  • Matching/weighting on pre‑treatment need/urgency/readiness proxies (e.g. costed condition surveys, backlog value/criticality, recent asset‑related cancellations/closures, RIBA stage, liquidity/reserves, insurance/risk notices).

However, feasibility may be challenging if sample sizes are low.

• Use of non-applicants: There is currently no comprehensive administrative frame of non‑applicant cultural organisations, and self‑selection into applying is material. Whilst in principle it is possible to undertake significant volumes of desk research to construct a sample, this would likely be prohibitively expensive. The sample would also likely be systematically different to the funded organisations  (and other applicants) which may undermine any potentials comparisons. As such, the use of non-applicants is not recommended.
• Future funding rounds: At the time of writing, funding was not committed to future funding rounds of CFF. Future funding rounds likely affect a counterfactual design through:
  • Unsuccessful applicants re-apply: Unsuccessful applicants from round 1 may reapply for round 2. This reduces the effective size of the comparison group.
  • Accounting for applicant and project maturity across funding rounds: It is important to carefully consider the composition of future funding rounds to ensure robust pooling of data for the DiD analysis. Specifically, distinguishing between:
  • Repeat organisations (Same project): Organisations that applied in Round 1, were unsuccessful (serving as our initial control group), but successfully secure funding for the same maintenance issue in a later round. This requires a staggered-adoption DiD model as they transition from control to treatment.
  • Repeat organisations (New project): Organisations that applied previously but are now applying for a newly emerged maintenance need.
    • New organisations: Organisations that did not apply in earlier rounds. There will be a need to test if these new entrants have systematically different baseline characteristics or urgency profiles compared to early recipients.
    • Methodological implication: Distinguishing between these groups ensures the control group is not contaminated and enables the analysis to accurately pool data across rounds while controlling for varying levels of project urgency.
  • Pipeline counterfactual design: Additional funding rounds may enable comparisons between early and late recipients (i.e. pipeline designs), creating additional scope to implement a QED.

6.3. Recommended counterfactual design

Following the above considerations, there is in principle a feasible counterfactual design which would facilitate a quasi-experimental analysis of CFF:

• Primary comparator: unsuccessful full‑application applicants. This attenuates applicant self‑selection and allows close matching on pre‑treatment need/urgency/readiness and organisational characteristics. Secondary comparators (EOIs) should be used only after a documented screening protocol confirms proximity on need/urgency/readiness. non‑applicants are out of scope for primary counterfactual estimation due to self‑selection biases and practical challenges in building a sample.
  • Pooling EOIs and full applications (only where screened): EOI‑only applicants could be included only if a pre‑agreed screening protocol (using costed condition surveys/backlog criticality/recent asset‑related disruptions/RIBA stage) indicates sufficient similarity to funded projects. Analysis can control for remaining observable differences using regression/matching/weighting and stratify analyses by strand/venue type/region/grant band/priority place (assuming a sufficient sample size).
  • Pooling Strand 1 and 2 organisations: To estimate dose‑response models (e.g. continuous treatment using project cost or ACE grant) with strand fixed effects and interaction terms, enabling pooling while allowing for different slopes by strand. Sensitivity analyses will re‑estimate effects within strands.
  • Staggered (pipeline) comparisons: If multiple rounds exist, it may be possible to compare early vs later awardees within the applicant pool, which further reduces bias from applicant self‑selection and exploits common timing shocks.

Appraisal of econometric methods

Following the identification of an appropriate counterfactual design, this section explores the feasibility of applying a quasi-experimental design to understand the causal impact of CFF funding. A summary of the feasibility of implementing the quasi-experimental approaches is presented below.

Table 6.1: Summary of feasibility of implementing quasi-experimental approaches

Quasi-experimental design	Feasibility assessment
Difference-in-Differences	Analysis unlikely to have sufficient statistical power.
Synthetic Control	In principle a feasible approach to understand additionality of CFF on key outcomes
Regression Discontinuity Design	No clear scoring cut off to exploit.

6.4.2. Difference-in-Differences

A Difference-in-difference (DiD) approach compares how outcomes change over time between the treatment group and the control group to estimate the causal impact of CFF. The treatment group would be taken as the funded organisations, and the control group would be taken as all unsuccessful (and therefore unfunded) organisations.

A power analysis was undertaken (results in Annex A) to estimate the effect sizes required to identify statistically significant results. Outcome measures used for the purposes of the power analysis included annual visitor numbers, number of full-time equivalents, commercial income, philanthropic investment and unfunded but necessary repairs.

In most instances, for the sample sizes available, outcome metrics would likely have to nearly double before statistical significance is achieved at the 95% confidence level.^{[footnote 15]}

It is therefore not considered feasible to implement a DiD design for the purposes of evaluating CFF.

6.4.3. Synthetic controls

The synthetic control approach is based on the idea that when the number of treated units are small, a combination of unaffected units provides a more meaningful comparison than any single unaffected unit alone. Synthetic controls can be built from multi‑year secondary panel data (e.g., ACE NPO annual survey, Charity Commission annual returns). It is anticipated that primary survey waves cannot support synthetic control construction. Notably, NPO workforce figures are rounded and financial granularity is limited. It may therefore be necessary to position synthetic controls as case‑study level causal estimates on a subset of venues with strong pre‑trend fit and adequate coverage. This would complement (not replace) the broader theory‑based assessment.

The synthetic creates a ‘clone’ of a funded venue which is meant to represent the venue in its unfunded, counterfactual, state. This clone, or rather synthetic control, is a weighted average of all other venues which did not receive CFF funding. The aim is for the synthetic control to mirror the characteristics of the treated venue before the treatment was received, as closely as possible. The weighting process is a key part of creating a synthetic control. The process for determining the weights involves a mathematical optimisation problem. This seeks to minimise the ‘distance’ between the treatment unit and the control units by applying a weight to each unit. The weights are such that they are non-negative and sum to one. This reduces the need for extrapolation, and instead produces a counterfactual informed by interpolation of control units.

Synthetic controls have extensive panel data requirements covering at a minimum outcome metrics, and ideally variables which predict the outcome level. This limits the usefulness of primary data collection for underpinning synthetic controls. Additionally, gathering longitudinal data can be prohibitively expensive as well as disproportionate on funded and unfunded venues. As such, the implementation of a synthetic control will largely depend on data availability within secondary data sets.

Two potential datasets are likely to provide sufficient coverage and completeness of data:

• ACE NPO data.
• Charity Commission returns.

Not every applicant for CFF can be linked to the ACE or Charity Commission data. However, it may be possible to undertake a synthetic control analysis for each of the funded organisations which returned a match. This is done by using unfunded organisations which returned a match as the control units. Whilst this approach would not include all organisations, it would provide a case study approach to understanding the causal effects of CFF on key outcomes. Annex B provides more detail on an initial scoping exercise on the synthetic control approach.

Considerations for implementing a synthetic control approach

Whilst the synthetic control approach in principle represents a feasible approach to implementing a quasi-experimental design, there are some considerations which must be understood:

• Scope of inference: Findings should be presented as case‑study estimates. Averaged effects should be treated as indicate opposed to a true ‘average treatment effect’.

• Generalisations of findings to wider CFF: Consideration should be taken as to how findings are applied to the wider CFF. This could include a simple average effect size, or results could be weighted by funding amount.

6.4.4. Regression discontinuity designs

The funding was not allocated on the basis of a score with a clearly defined threshold. As such, it is not possible to make comparisons between those that were ‘just awarded funding’ and those that ‘just missed out’.

6.5. Evaluation timing

Funded projects are anticipated to be completed by March 2029. However, it is common for large-scale capital projects to overrun, so it would be prudent to allow for this within the evaluation timing.

There may be benefits in undertaking both a short- and longer-term impact evaluation:

• Short-term evaluation: To understand the immediate post-completion benefits, a short-term impact evaluation could be undertaken approximately two years after works are scheduled to finish (i.e. April 2031). This should enable, in the majority of cases, a full year of post-completion data.
  • DCMS should allow sufficient time within the evaluation to design necessary primary data collection elements, finalise the design, and ensure the strands align. Commissioning a full evaluation in April 2030 will provide sufficient time for a quantitative assessment to be full operationalised by April 2031.
• Long term: It is not known with certainty how long it will take for benefits to realise, and the persistence of benefits. As such, there will be value in undertaking a longer-term evaluation four to five years post-completion. This will enable an exploration of how benefits evolve over time and should provide sufficient time for the new ‘steady state’ to emerge.
  • It is recommended that from April 2033 a longer-term assessment count be conducted. Where, as above, DCMS allow 1 year for impact evaluation planning and set up.

7. Heritage science counterfactual

This chapter presents the heritage science counterfactual developed for the Creative Foundations Fund (CFF). The counterfactual provides a structured method for estimating how funding buildings, systems and equipment, affects the delivery of associated creative and cultural activities.

The rationale for this approach follows directly from the objectives of the Creative Foundations Fund. Applicants must demonstrate that proposed investment is business-critical to delivering creative activity, reduces the risk of asset failure, and improves the long-term sustainability of the organisation.

The evaluation must therefore assess the difference between the funded scenario and a plausible counterfactual in which these works did not take place. In practice, this requires estimating how buildings, technical systems and production equipment would deteriorate over time without intervention. Furthermore, it requires estimating how this deterioration would affect the ability of organisations to deliver cultural activity. The model therefore focuses on how CFF interventions preserve or extend the capacity of venues to continue generating cultural, social and economic value over time.

The model compares two trajectories:

• No-intervention scenario, in which assets continue to deteriorate.
• Intervention scenario, where repair, renewal or replacement funded through CFF improves asset condition and extends its functional life.

By modelling these trajectories, the framework estimates the difference in asset lifetime and the additional years of cultural activity enabled by the intervention. This allows the evaluation to assess avoided deterioration, reduced risk of closure, and the period over which cultural venues can continue delivering creative, social and economic value.

The model also supports the wider evaluation framework by identifying when outcomes are likely to emerge. This links capital investment to the continuity of cultural services, and provides a structured counterfactual consistent with the Magenta Book.^{[footnote 16]}

The approach is also designed to remain proportionate, drawing primarily on information already available in application materials, condition surveys and technical documentation.^{[footnote 17]} The model can be populated using a small number of structured inputs, largely drawn from application materials and technical documentation. In practice, this would involve asking applicants to provide the following information:

• What asset of system was the focus of the intervention?

• What was the condition of the asset prior to funding?

• What was the expected timeline until the asset because unusable or hazardous (in the absence of intervention)?

• How long are the planned works expected to take?

• What is the expected condition of the asset following completion of the works?

• What is the expected lifetime of the asset after intervention (based on professional judgement, warranty information and building plans)?

• What level of maintenance is planned following the works?

• What is the expected level of use of the asset following intervention (e.g. frequency of use, type of activity supported)?

This use of readily available information from applications aligns with Magenta Book guidance that evaluation evidence generation should maximise learning while minimising burden on participating organisations.

7.1. Background and rationale

This problem is well suited to a heritage science approach. In heritage science literature, a counterfactual describes the condition of an asset if conservation or maintenance work had not taken place.

The MEND/PBIF evaluation used this approach to create a collections demography model^{[footnote 18]}. This model estimates how deterioration risks affect the lifetime of museum collections in the absence of maintenance and conservation activity.

The same logic can be adapted for CFF, while recognising an important difference between the cultural assets being funded. Cultural organisations funded by CFF - principally performing arts venues, creative production facilities and community cultural centres - where the primary asset of interest for modelling is the enabling infrastructure (building fabric, MEP plant and controls, lifts, life‑safety systems, rigging/stage/AV) and the cultural service being protected is the reliable delivery of live programmes and public access. Whereas in the MEND/PBIF framework the cultural service being protected is the collection itself, which is also the subject of the deterioration modelling.

In the case of CFF, the assets being supported - buildings, technical systems and production equipment - do not themselves constitute the cultural output. Instead, they function as enabling infrastructure which allows creative and cultural activity to take place.

Deterioration of these venues and/or elements of the building fabric therefore affects cultural services indirectly. Failure of structural elements, environmental systems or production equipment may prevent performances from taking place, restrict building access, or ultimately lead to venue closure.

The counterfactual model therefore focuses on the relationship between asset condition and cultural activity. By modelling how deterioration affects the functional viability of buildings and equipment, the framework estimates how long venues would remain capable of delivering cultural services without intervention, and how CFF investment changes this trajectory. The change in cultural services can be used to measure the impact of the CFF intervention in an approach suggested in the DCMS paper, ‘Culture and Heritage Capital: using economic valuation methodologies and heritage science to measure the welfare impact of ongoing conservation, protection, repair and maintenance of culture and heritage assets’.^{[footnote 19]}

7.2. Defining the cultural activities

Before modelling how degradation affects cultural activities, it is necessary to consider how the model will assess which cultural activities are enabled by the assets (venues and elements of the building fabric) supported through the CFF intervention. There are a multitude of cultural values that are potentially associated with different venues. For example, they could include aesthetic, architectural, historical, educational, social, artistic, commemorative, symbolic, spiritual, ecological and environmental values ^{[footnote 20]} to different groups at different times in their lives.  

There could also be intangible heritage values that may be associated with venues and influenced by CFF interventions. These include oral traditions, performing arts, social practices, rituals, festive events, knowledge and craftsmanship.^{[footnote 21]}

The wide range and history of the multiple potential cultural values and the history of using values in decision making concerning the arts culture and heritage, is explored in the DCMS Scoping Culture and Heritage Capital Report^{[footnote 22]}, part of the DCMS Culture and Heritage Capital (CHC) Programme.^{[footnote 23]}

The counterfactual model therefore draws on the information from CHC and the emerging ACH Taxonomies Project^{[footnote 24]} to classify and interpret the cultural outcomes of CFF interventions.

In the context of the CHC framework shown in Figure 7.1, CFF interventions can be understood as interventions or enabling activities, that maintain the physical assets or stocks, required for cultural service benefits or flows to occur despite the pressures of deterioration and operational use. The increase in these service flows resulting from CFF investment in cultural venues is used by the counterfactual model to assess the effectiveness of the intervention.

Figure 7.1: The Culture and Heritage Capital Framework (CHC)

Figure 7.1 is a diagram of the culture and heritage capital framework. Pressures and interventions affect culture and heritage asset and benefits, which goes on to affect asset value.

Therefore, the CHC categories of cultural benefits and service flows shown in Table 7.1 are used to characterise and measure creative and cultural activities impacted by the CFF intervention. Allowing for an assessment of how intervention in cultural capital stocks generates value through changes to service benefits and flows. This analysis also includes capital services as these are also a key service generated from an asset. Therefore they are included in the evaluation of the impact of the CFF intervention.

Table 7.1: Culture and heritage services

Service	Description	Examples
Aesthetic services	Provide individuals with sensory, emotional and intellectual stimulation as a result of the asset’s appearance - design, beauty, and architectural character and artistic endeavour.	Aesthetic enrichment, architectural character, attractiveness, beauty, captivation, congruence, design, distinctiveness, escapism, pleasure, reflection, spirituality.
Authenticity services	Considers authentic value of culture and heritage assets, derived through their meaning, existence and historical significance, and reflect a unique and irreproducible character, meaning identity, symbolism and significance.	Atmosphere, distinctiveness, experiences, historical significance, inheritance, meaning, motivation, symbolism, uniqueness.
Communal services	Reflect the qualities of the relationships between humans and culture and heritage that foster a sense of identity, belonging and pride.	Attachment, civic engagement, communal meaning, community identity, networks, peace, political dialogue, reflection, sanctuary, social bonding, social connectedness and contact.
Inspirational and creative services	Services that inspire and motivate individuals, providing satisfaction and information through performing, celebrating and promoting culture and heritage.	Aspiration, captivation, creativity, curation, design, escapism, expression, hope, imagination, innovation, intellectual exploration, intellectual stimulation, motivation, pleasure, reflection, resonance, spiritual uplift.
Identity services	Recognise, protect, and promote diverse cultural expressions, traditions and histories that contribute to a community’s identity and sense of belonging.	Connectedness, cultural interpretation services, curiosity, empathy, empowerment, inquisitiveness, introspection, purpose, reflection, spiritual meaning, traditions, understanding of diverse cultures.
Knowledge (educational) services	Formal and informal learning that involves the acquisition, creation and dissemination of knowledge and learning that supports personal development, creativity and innovation.	Access, comprehension, education, flourishing, historic understanding, knowledge sharing, memory, research and development, skills development, teaching.
Health services	Services that improve overall health, quality of life and mental wellbeing of individuals, providing supportive and preventative interventions.	Physical and mental health services, wellbeing practices.
Environmental services	Services provided through the interactions of culture and heritage with nature (including natural resources and natural processes and functions underpinning them).	Diversity in species, habitat, reductions in flood risk and soil erosion, removal of urban air pollution by trees, settings for recreation, water supplies and regulation.
Capital Services	Flow of productive services from an asset.	For example, service flows from a building including floorspace accommodating recreation activities or employment space.

7.3. How the model works

This section provides an overview of how heritage science modelling can be used to underpin an evaluation of CFF. A full technical discussion can be found in Annex C.

The aim of the CFF is to reduce the degradation of an existing asset stock or introduce new asset stocks. This will impact on the flow of cultural and heritage services in different ways. Before and after the intervention, the venue will continue to deteriorate over time, as the ageing of buildings and equipment is an inevitable process.

Once assets degrade beyond a certain point, the asset reaches a condition where it can no longer fulfil the functions and requirements for which it was originally intended. This is referred to as functional obsolescence.

Heritage science modelling can be used to understand how the intervention changes the expected lifetime of the asset and, in turn, shift the point of functional obsolescence. In practice this requires estimating the time period over which a venue would remain functional under two scenarios:

1. No-intervention scenario in which deterioration continues.

2. Intervention scenario in which asset condition improves following CFF investment.

The difference between these trajectories represents the additional years of functional asset life enabled by the funded activities. These additional years of asset functionality can then be translated into the additional years during which cultural activity (and flow of services) can continue to take place within the venue. This can be used to calculate the lifetime multiplier of the CFF intervention.

This provides a practical mechanism for linking capital investment in infrastructure to the continuity of cultural services. Additionally, it supports the evaluation of avoided deterioration and reduced risk of venue closure.

7.3.1. Step-by-step approach

In practise, the following steps are taken to implement heritage science modelling. More technical detail can be found in Annex C:

1.  Define the asset: CFF supports a wide range of activities. A typology of the works, and the elements of the buildings which are affected can be created by drawing on the RICS categories of building elements for planned preventative maintenance. The categories include structure, roof(s), facades, internal finishings, external areas, service installations and an ‘other’ category to capture novel interventions or equipment. Funded organisations would then be asked (via primary data collection) to identify the categories of the works being undertaken.

2.  Define baseline condition: To quantify the improvements made by CFF, the baseline condition must be established. This can be achieved by asking projects to ‘score’ each asset on a continuous 0 to 100 scale. The continuous scale is based on the RICS condition rating guidance, where condition is assumed to decline towards zero overtime. This provides the starting point of assigning the lifetime and degradation rate of the asset.

The scale has four bands, where the highest band represents a good condition, and the lowest band represents a hazardous condition. A full description can be found in Table 8.3 in Annex C. Applicants will be able to use information contained in the EOI, full applications and costed condition surveys to inform their decision making.

3. Assign material lifetimes and degradation behaviour: Once the baseline condition has been established, assets can be assigned an expected service life and degradation profile. Applicants are required to estimate the approximate time until the asset becomes hazardous in the absence of intervention. Given the nature of the application process, this information is typically available from building condition surveys submitted as part of the funding request. Applicants are also asked to indicate the expected behaviour of deterioration, selecting whether the deterioration profile is linear, accelerating or rapid.

This information can be inputted into a deterioration function to estimate the condition at a given point into the future.

4. Model the intervention scenario: It is assumed that the intervention does not permanently stop deterioration. Instead, it is assumed to improve the starting condition before the asset continues to deteriorate. To estimate this ‘step change’, the same approach taken in Step 3 above can be implemented, varying the model inputs to represent the post-funding conditions. This requires knowing the intervention start year, duration of works, condition uplift and asset lifetime.

a. Include real-life operating factors: In practice, the service life of an asset is influenced not only by the inherent properties of the material or component, but also by the conditions under which it is used and maintained. It is possible to incorporate measures of usage intensity and maintenance regime into the condition functions. Different levels of either usage or maintenance hold different uplift factors to the lifetime. For example, lower levels of usage are associated with higher uplift factors; and higher levels of maintenance are associated with higher uplift factors. The uplift factors are borrowed from BS ISO guidance.

b. Include uncertainty: Building deterioration is inherently uncertain. Future degradation rates depend on environmental conditions, usage intensity, maintenance regimes and unforeseen failures. As a result, long-term deterioration cannot be predicted with absolute precision. To reflect this uncertainty, multiple trajectories can be produced, which represent a central, pessimistic and optimistic scenario, as opposed to a single deterministic estimate.

5. Link physical condition to cultural activity: The distinctive feature of the CFF counterfactual model is that it does not assess asset condition in isolation. Instead, it links the physical condition of cultural infrastructure directly to the continuity of the cultural activities that those assets enable.

In the context of the CFF, the primary function of the asset is to CHC service delivery. To represent this relationship within the model, each cultural service identified within the CHC framework is associated with a minimum condition threshold, defined by the applicants who will understand the needs of the cultural services in relation to the asset best. This threshold represents the level of condition required for the asset to continue supporting a particular cultural activity. As asset condition falls below the threshold, it is assumed that the service can no longer be reliably delivered.

Using this premise, the number of years during which each cultural service remains viable under both the counterfactual and intervention scenario can be calculated. The impacts of CFF can be expressed as the difference in these years and as a lifetime multiplier.

7.3.2. Model outputs

The heritage science modelling approach will generate three key metrics that can be used to understand the impacts of CFF and inform the value-for-money assessment of CFF (see Section 9):

• Additional years of life gained for each asset/stock

• Additional years of resulting cultural and heritage services produced by the asset/stock

• Lifetime multiplier for cultural and heritage services produced by the asset/stock

These outputs will help measure the extent to which deterioration is prevented and the risk of venue closure is reduced. In turn, they will enable the assessment of indirect impacts arising from this avoided deterioration - such as health, social, and educational outcomes (as outlined in the theory of change) - that are linked to the continued provision of CHC services.

Conclusion

This chapter has presented a heritage science counterfactual model for the Creative Foundations Fund. The HS approach is used here because CHC treats cultural assets as stocks that generate service flows over time. Within this CHC framework, maintenance and renewal are conceptualised as enabling activities that protect those flows by slowing deterioration and avoiding functional obsolescence. Heritage science provides a proportionate, evidence‑based way to model the counterfactual trajectory of asset condition and, crucially, to translate that into the continuity of cultural service flows. This is consistent with DCMS’s CHC guidance on using heritage science to measure welfare impacts of conservation/repair.^{[footnote 25]} Additionally, it creates quantitative inputs that can be linked to CHC valuation in the economic evaluation. The model begins by identifying the relevant assets, establishing baseline condition and assigning expected service lives and degradation profiles. It then defines the condition thresholds at which deterioration begins to compromise cultural activity, models how intervention alters asset condition trajectories, and incorporates operational factors affecting asset lifetime.

Finally, the model translates physical changes in asset condition into additional years of cultural service viability and a lifetime multiplier by linking condition trajectories to cultural service thresholds.

The approach provides a credible and proportionate method for modelling avoided deterioration and reduced risk of asset failure and directly addresses the challenge of estimating what would have happened in the absence of CFF investment. Taken together, this means the heritage science counterfactual can play a central role in the next phase of the evaluation.

8. Theory-based impact evaluation

This section sets out theory-based evaluation approaches which could be used to understand the impact of CFF.

8.1. Overview of theory-based methods

Theory-based approaches provide an alternative, or complementary, approach to impact evaluation – particularly where counterfactual methods are found to be infeasible. Theory-based methods typically don’t generate quantitative causal estimates of impacts in most cases. However, they do seek to ‘go behind the black box’ of the intervention to consider the logic of how and why it affects specific outputs and outcomes.

Theory-based methods typically start at the Theory of Change, applying different sources of evidence to test whether activities, outputs and outcomes have evolved in the way the intervention logic intended or was anticipated. This can build understanding of how and why an intervention works, rather than just if it works. This can add significant value and insights for future policy development purposes.

8.2. Relevant theory-based methods to evaluate CFF

The approaches suggested below are considered to be the most relevant and commonly used practical possibilities for the impact evaluation which are outlined within HM Treasury Magenta Book. These are best considered more as ‘families’ of approach which have been developed and applied in different ways – with different strengths and weaknesses - rather than as single techniques.

Examples of detailed criteria which might be utilised in the CFF evaluation are set out below. However, further development work on the chosen approaches will be required in light of the information collected so far from applicants, the profile of projects which are funded. Additionally the scale, shape and timing of the eventual evaluation will also require further development work - and thus what further information it will be feasible to collect going forward.

It is anticipated that a theory-based impact evaluation will largely be underpinned by qualitative interviews. MI and secondary data will be used in addition to contextualise and support findings where necessary and relevant.

8.2.1.Contribution Analysis

Contribution Analysis (CA) is one of the most commonly applied, systematic, theory-based approaches to economic programme evaluation. It focuses on testing potential ‘contribution claims’ that a programme has helped to bring about the particular changes which it was designed to achieve. A key strength of contribution analysis is its focus on identifying and testing alternative explanations for the outcomes observed. This can help increase certainty a programme has worked as intended, where long term outcomes might be influenced by numerous contextual factors. In the case of CFF, for example, longer term financial resilience might also be influenced by market and consumer trends. CA can help develop explanations about how these factors and the programme interact to lead to desired outcomes.

CA is generally based upon a six-stage process:

1. Setting out the attribution problem to be addressed. In this case, the objectives specified in the business case argue for a focus upon whether the programme has:

• Reduced the loss of output of the arts sector as a result of asset failure /degradation
• Improved the economic sustainability of the sector
• Delivered value for money by decreasing the financial impact of abandoned or cancelled performances, exhibitions or events as a result of equipment or infrastructure failures
• Helped to grow audience engagement, access and satisfaction.
• Second order, or longer-term impacts such as effects on local economic growth would be a further potential aspect for investigation.

2.  Develop the Theory of Change/Logic Model (Impact Framework) and relevant contribution hypotheses. Building upon the ToC developed as part of this Evaluation Framework, as appropriate.

3. Populate the model utilising existing data and evidence.^{[footnote 26]} Likely drawn from a combination of:

• Survey and case study evidence on the evolution of the overall financial performance of the beneficiary organisations subsequent to the receipt of grant.

• Evidence gathered on:

  • a. Equipment and building condition/improvements funded by the CFF grant.
  • b. The consequential reported savings in relation to aspects such as maintenance and operating costs from the funded improvements.
  • c. Changes in the extent of maintenance backlogs.
  • d. Observed reductions in the incidence of abandoned and cancelled performances as a result of asset related problems.

• Evidence on the evolution in audience/attendee numbers (including disabled or diverse attendees) and feedback on audience satisfaction with the improvements which have been funded.

4. Assemble and assess the performance story. Involving a critical assessment of the evidence that the CFF funding and the improvements it has enabled have been at least significant causal drivers of the observed outcomes. This should also involve identifying and considering the roles of confounding and other contextual factors in influencing the observed changes.

5.  Seek additional evidence. Involving wider stakeholder consultations, to clarify emerging issues, help to strengthen understanding of causal links and resolve any conflicting evidence. Areas for investigation might include:

• Gathering data on how survival rates of beneficiary venues evolve compared with the experience of potentially comparable venues elsewhere in the sector.
• Assembling and analysing evidence on the evolution of the relative performance of the assisted venues in maintaining/increasing their audiences/revenues.
• Gathering and analysing evidence on audience experience/perceptions of improvements, for example from Google reviews.
• Consideration of how far the enhancements which have been enabled have contributed to local regeneration objectives, for example through discussions with local property professionals, analysis of property transactions, local retail spend data, footfall data (e.g. from GPS or mobile phone data).

6. Revise the performance story. As necessary, with further iterations of other aspects of the evidence gathering process.

Contribution Analysis represents a potentially useful tool for the impact assessment. However, it is important to stress that it cannot generally alone ‘prove’ whether an intervention is responsible for an observed outcome or for a particular proportion of the outcome. Rather it is ‘designed to reduce uncertainty about the contribution the intervention is making to the observed results through an increased understanding of why the observed results have occurred and the roles played by the intervention and other internal and external factors’.^{[footnote 27]} The strength of the inferences which can be drawn from the approach can potentially be enhanced through combining it with other methods such as Process Tracing and Bayesian Updating methods (discussed below). For this reason, for capital maintenance, HS reliability outputs (failure probabilities; expected avoided cancellations/closures; life‑extension) will serve as the central factual basis for additionality, complemented by Process Tracing of counterfactual actions absent CFF (e.g., partial closures, short‑term hire, further deferral).

8.2.2. Process tracing

Process tracing is the systematic examination of diagnostic evidence selected and analysed considering research questions and hypotheses posed by the investigator. In practical terms it involves the development of a systematic framework of tests of the strength of the evidence in favour of a particular causal link or contribution claim. The variety of tests of strength can be combined to provide an overall qualitative judgement on the strength of a contribution claim.

The following example – based upon the formulation of tests by Collier^{[footnote 28]} - illustrates a potential framework of tests to assess the strength of the evidence that the CFF has led to at least a commensurate increase in the relevant investment by the beneficiary organisations (i.e., that the programme has substantially resulted in additionality). Establishing that this is the case is important for the credibility of the overall Logic Model. However, even if the claims were found to be false in many / some cases, the programme could produce some of the same effects by supporting the development of the beneficiary organisations in other ways. For example, through releasing funding to strengthen the organisations’ creative work. In most practical applications, the Hoop and Smoking Gun tests are the main focus of Process tracing approaches.

Table 8.1: A Process Tracing approach to assessing programme additionality

Test	Potential additionality tests	Implications
Straw in the wind	Beneficiaries had developed but been unable to take forward previous investment plans	a) Passing: Affirms relevance of hypothesis but does not confirm it b) Failing: Does not eliminate hypothesis but slightly weakens it c) Implications for rival hypotheses: Passing: Slightly weakens them Failing: Slightly strengthens them
Hoop	Timing of commitment to the investments involved consistent with claims that availability of CFF funding crucial to taking them forward (‘Consistent Chronology’) Beneficiaries affirm substantial importance of CFF funding to the investment decisions involved Financial data re individual beneficiaries’ profitability, liquidity, reserves, available lines of credit and potential to attract alternative sources of funding confirm they mostly likely faced major issues in self-funding the investments involved	a) Passing: Affirms relevance of hypothesis but does not wholly confirm it b) Failing: Potentially eliminates hypothesis c) Implications for rival hypotheses: Passing: Somewhat weakens them Failing: Somewhat strengthens them
Smoking gun	Weak Version: Evidence of unsuccessful efforts to secure alternative sources of funding for investments involved Strong version: Efforts to secure alternative sources of funding appear to have exhausted all realistic possibilities	a) Passing: Tends to confirm hypothesis b) Failing: Does not eliminate hypothesis but tends to weaken it c) Implications for rival hypotheses: Passing: Potentially weakens them Failing: Potentially strengthens them
Double decisive	Independent sources all tend to confirm significance of funding to beneficiaries’ capability to make the investments involved (‘Convergent Triangulation’) No contradictory evidence emerges pointing to frequent deadweight funding of investments which would have been made anyway	a) Passing: Confirms hypothesis but does not eliminate possibility other explanatory factors involved b) Failing: Does not necessarily eliminate hypothesis but substantially weakens it c) Implications for rival hypotheses: Passing: Potentially weakens them Failing: Potentially strengthens them

Adapted from Collier, 2011

8.2.3. Qualitative Comparative Analysis – A potential complementary tool

QCA is a hybrid theory-based methodology which combines findings from case study research with a Boolean Algebra and Set Theory based analytical approach. It uses these to identify combinations of intervention characteristics and contextual ‘conditions’ which are associated with particular outcomes, typically the success or failure of a policy intervention. It was designed to analyse situations where standard econometric methods cannot be utilised because key ‘conditions’ cannot be defined in terms of continuous variables but the number of cases is too large for the analyst to handle using less formal methods.

The approach could potentially be utilised within the CFF evaluation. It could analyse what combinations of project characteristics, planning and delivery and contextual conditions appear to have been most critical to achieving successful outcomes. Additionally, what combinations have tended to produce more problematic results. It could potentially be utilised in relation to an analysis for just the largest projects or for a larger sample, dependent on the number of case studies / the scale of the fieldwork which is to be undertaken.

The issues to be considered are how far its inclusion in the evaluation would be likely to ‘add value’ and whether it would be proportionate. In these contexts, it is noted that it would not help with the task of assessing the impact of the programme. However, it would have potential value in highlighting what has worked, or not worked, well which may be useful in designing future programmes in this type of policy space. The possible doubts are around:

• How satisfactorily /unambiguously it will be possible to define success in relation to what are likely to be a diverse set of projects and outcomes.

• The fact that there is already a strong understanding from other evaluations of the characteristics of the delivery process which predispose projects to success or are likely to prove problematic.

8.3. Evaluation timings

The timings of the theory-based evaluation should be such that they align with the counterfactual impact evaluation, to provide a complementary narrative and rationale behind quantitative measures:

• Short term: From April 2030, Evaluators will be required to undertake a detailed theory-based scoping phase, including scoping interviews, Theory of Change (ToC) workshops and documentation review. These activities inform the design of research tools, the development of the ToC, and a detailed process map (or maps) for process tracing. At this stage evaluators would also scope the conditions (e.g. delivery and contextual factors) needed to support later Qualitative Comparative Analysis (QCA).

• The first round of interviews and data collection could begin in early 2021, where by this point, evaluators will be able to:
  • Refine and validate the process map.
  • Test and strengthen contribution claims.
  • Capture early and short‑term outcomes as projects complete.
  • Identify and assess plausible alternative explanations for emerging impacts.
  • Review the contextual and delivery conditions influencing outcome realisation.
• Longer-term: A second round of interviews and data collection would then be undertaken across 2032 and 2033. This phase would focus on gathering evidence on longer‑term, down‑the‑chain outcomes. This phase would assess the role of alternative explanations in shaping these outcomes, and assemble detailed delivery evidence to support robust process‑tracing. QCA can only be undertaken once this second round is complete and outcome data are available across the portfolio.

9. Economic Evaluation

This section focuses on how an overarching economic evaluation of DCMS’ CFF could be undertaken. This includes identifying what questions an economic evaluation should seek to address, the degree to which a 4E’s assessment of value for money is likely to prove possible using scheme level evaluation evidence, and the extent to which an overarching Cost-Benefit Analysis is likely to be possible.

9.1. 4E’s Value for Money Assessment

The main assessment of whether the CFF delivered value for money will be guided by the National Audit Office’s (NAO) 4E’s Framework.^{[footnote 29]} The 4E’s framework largely draws on monitoring and qualitative data (collected through the process and theory-based evaluation), meaning that VfM assessments can be undertaken with or without an accompanying quasi-experimental design. A full 4Es framework for CFF is set out in section 9.1.3.

A 4E’s assessment would assess the CFF against four dimensions:

• Economy: The degree to which the cost of resources is minimised whilst having regard to quality and objectives. This would likely be informed by the outputs of the process evaluation.

• Efficiency: The extent to which outputs arising from funding (e.g. services) were delivered efficiently (i.e., at minimum cost, using minimum resources and without delay). This would likely be informed by the outputs of the process evaluation.

• Effectiveness: The extent to which outputs arising from the funding leads to their intended outcomes and impacts. This will draw on analysis of monitoring information, the process and impact evaluation.

• Equity: The extent to which outcomes are distributed geographically, and reach all intended organisations or people. This would likely be informed by the outputs of the process evaluation, as well as some aspects of the impact evaluation (e.g. audience composition).

9.1.1. Key sources of information

A 4E’s assessment would be expected to draw on a range of different data sources and strands of the evaluation:

• Programme level data: Data collected by ACE or DCMS throughout the delivery of capital works. This is expected to include:
  • Volumes of EOIs and full applications.
  • Outcome of assessments, including applicant scoring/ funding decision notes.
  • Costed condition surveys to understand the extent to which funded projects can be considered a priority.
• Funding data: Data covering the funding arrangements to enable the funded works. It is anticipated that ACE will be able to provide data covering:
  • The value of the grant to each organisation.
  • The value of matched funding, including the source of the matched funding, for each organisation.
  • The value of funding, if any, contributed by organisations
  • Other sources of public funding, which may or may not be used to fund the activities undertaken as part of CFF.
• Outcome data: This will combine several key sources of information including baseline data, ACE data returns, primary data (if collected as part of an evaluation) and will be synthesised and analysed within the three impact evaluation strands:
  • Heritage science modelling: Recognising that econometric proof of additionality for capital maintenance is inherently limited, this approach (as outlined in section 7) will serve as the primary quantitative method for establishing the counterfactual. It will utilise reliability modelling to estimate failure probabilities, asset life-extension, and expected avoided failures or closures in the absence of CFF. This directly links physical asset conditions to the continuity of cultural services. The approach therefore provides a way to quantify social and educational outcomes included in the cost benefit analysis.
  • Quasi-experimental design: This approach (as outlined in section 6) would seek to establish the causal effect of funding, relative to a scenario where funding was not awarded to organisations.
  • Theory-based impact evaluation: This approach (as outlined in section 8) will be used in the event that other quantitative counterfactual approaches are not possible to evaluate CFF. In this case, the evaluation will instead rely heavily on theory‑based methods built around the programme’s Theory of Change. The approach would use multiple evidence sources to test whether activities, outputs and outcomes actually happened as expected, the additionality, and why these occurred.

• Process information: Qualitative data and quantitative monitoring data, drawn from the process evaluation (including interviews with stakeholders, beneficiaries and unsuccessful applicants) will provide evidence of the process that have been used. Additionally, how and why these supported the achievement of outputs or outcomes at a low cost.

• Contextual information: To properly contextualise the evaluation, it will be essential to account for broader external factors and market conditions. This includes monitoring macroeconomic and price pressures that could impact project delivery, such as fluctuations in construction and material costs (tracked via the BCIS Tender Price Index and Materials Cost Indices), the UK Construction PMI, and wholesale energy prices. We will also consider the wider health of the sector using DCMS Economic Estimates for GVA and employment, alongside general labour market capacity. On the consumer side, we will look at audience demand and ticketing trends using data from the Audience Agency, UK Theatre and SOLT, and tourism metrics from VisitBritain and VisitEngland, all contextualised by GfK consumer confidence scores. Finally, to ensure our sensitivity analysis is robust, we will factor in potential exogenous shocks that could disrupt venue operations or project delivery. Examples include extreme weather events, transport or sectoral strikes, or broader national emergencies like public health crises or security threats.

This will also include consideration of place-based effects. For example, if a venue is the main economic driver in an area or the only cultural venue, this will likely have a disproportionately larger impact on the supply chain and the social impacts of the venue closing compared to areas higher cultural density.

9.1.2. Place-based impacts

As part of the 4Es assessment, evaluators should examine place‑based impacts on both the supply and demand sides. On the supply side, this includes downstream impacts on the construction sector and specialist equipment technicians. On the demand side, it includes impacts on visitor spending and tourism in local areas. The aim is to capture the wider economic effects of CFF funding beyond the direct benefits to funded organisations, particularly the indirect benefits to related sectors and local economies.

When assessing place‑based impacts, evaluators should prioritise areas where CFF‑funded organisations are considered a significant driver of local economic activity. For example, instances where a theatre or arts organisation are the only cultural offering in the local area. In doing so, they could draw on the DCMS‑published list of Priority Places, where DCMS intends to focus more activity and investment to ensure that funding is concentrated in communities where it can have the greatest impact.

The analysis could consider, and input into both the ‘Effectiveness’ and ‘Equity’ Es:

• Survival in the absence of CFF: Exploring the counterfactual of ‘no CFF’ to understand how the absence of CFF funding would have affected demand for specialist construction services, visitor spending and associated demand‑side impacts in the local area. This includes assessing whether there are temporary decreases in visitor spending during works or increases as a result of the works and considering potential substitution effects (i.e. whether other work was forgone in order to deliver CFF‑funded contracts).

• Geographic distribution: Understanding where the impacts are located, providing insights into how the capital grant funding supports businesses and local economic growth across the country.

• Supply chain effects: Exploring the value of other contractors or suppliers to fulfil the capital works contract. This would provide insights into how the capital works provide revenue for different firms across the economy. Where possible, these firms will be linked to a SIC code.

There are, however, a number of challenges associated with this approach. In practice, it is likely to require primary data collection, combined with ONS Input–Output tables or assumptions, to estimate local‑level impacts. More robust, granular assessment would involve collecting additional data from the surrounding area to identify multiplier effects and better understand the wider economic impacts of CFF funding.

A further consideration is how the results are presented. This should be agreed with DCMS, including whether findings are reported separately from the programme cost–benefit analysis (CBA) in the wider 4E’s assessment, or included in the CBA. To incorporate into a programme CBA, a sampling approach may be required to estimate an ‘average place effect’, which can be used to evidence place-effects. This can be presented alongside the ‘core CBA’ which would take a net national accounting approach.

9.1.3. 4E’s Framework

The table below links the key evaluation questions to:

• The relevant ‘4Es’ (economy, efficiency, effectiveness and equity).

• The data sources that would inform the value for money assessment.

• The proposed qualitative and quantitative methods for assessing value for money.

Table 9.1: Overview of value for money assessment

Evaluation question	Evidence for assessment	Quantitative assessment of CFF VfM	Qualitative assessment of CFF VfM
Economy	-	-	-
To what extent did CFF deliver its objectives using the minimum level of public support? Sub questions:  What was the total cost of CFF? Did the process for allocating funding support value for money? Was the level of funding sufficient? How did CFF compare to similar programmes in terms of costs and outputs?	Total CFF funding (including grant funding, staff costs and other public costs) provided through monitoring data. Funding costs for comparable interventions (if available). Qualitative interviews	Using desk research, CFF funding and costs could be compared with the funding allocated to similar schemes with comparable objectives. This would allow an assessment of whether CFF funding is lower than, similar to, or higher than that of comparable schemes.	Qualitative interviews from the process evaluation will provide information on the extent to which funding was considered sufficient to achieve aims. Interviews could also provide evidence on whether stakeholders viewed the process for allocating funding as supporting value for money.
Efficiency	-	-	-
To what extent were outputs arising from funding delivered efficiently Sub questions: Were outputs achieved within expected costs? Were outputs achieved as planned and within expected timelines? Were risks effectively identified, mitigated and managed?	Total CFF funding (including grant funding, and other public costs) provided through funding data Admin costs Staff time Outputs data Number of applications  Qualitative interviews	This could be assessed through various calculations to understand how efficiently resources were translated into delivery and outputs. For example: A calculation of overall resource costs (including admin costs or time) per application processed. A calculation of overall costs (including grant funding, resource costs or time) per output delivered e.g. repair or renewal (as outlined in section 2.3.3). Noting the challenge of comparing costs for the heterogenous range of activities and outputs funded by CFF.	Qualitative interviews from the process evaluation could provide evidence on the extent to which the delivery process was efficient and timely (as outlined in table 5.1). This could include an assessment of: The extent to which projects were completed within timeframes The extent to which there were delays and changes to scope The extent to which quality standards were compromised to meet deadlines. How were issues and risks identified and resolved?
Effectiveness	-	-	-
How effective was CFF in delivering its objectives so far? How might this change in future?	Number, and scale of change in outcomes (compared to counterfactual) provided through primary data collection, monitoring data and secondary data sources.  Outputs from heritage science modelling Qualitative interviews	This will be revealed through impact evaluation findings. Because econometric proof of additionality is inherently limited, this will primarily involve assessing changes in outcomes relative to a counterfactual using heritage science reliability modelling and structured theory-based methods (Process Tracing), supplemented by Quasi-experimental design where feasible. These outcomes can be mapped broadly to the overarching objectives.	Qualitative interviews from the process evaluation could provide evidence on extent to which: The portfolio included a strong mix of projects with high expected cultural, social, economic and/or environmental benefits that deliver against the core objectives (safeguarding output, sustainability, access, environment). Stakeholders broadly perceived that funded projects collectively advanced the fund’s key objectives. Early indications (or projections) suggest funded projects are on course to deliver expected benefits.
What were the levels of additionality of CFF? Sub questions: To what extent did the impact of CFF represent a net gain, opposed to displacing activity from other areas? Is there any evidence of displacement, either nationally or locally?	Number, and scale of change in outcomes, both in aggregate and  by location (compared to counterfactual) provided through primary data collection, monitoring data and secondary data sources. As set out in section 9.1.2. this should include measurement of place based impacts. What proportion of funded activity and outcomes was additional (timing, scale, scope), and what was the minimum necessary public contribution to secure them? Project‑level counterfactual evidence: application ‘what would happen without CFF’ statements, condition surveys/backlog criticality, HS reliability modelling (expected failure/closure/cancellation without works).	This will be revealed through impact evaluation findings as above, and supplementary place based analysis. Compare awarded grant to evidenced funding gap	Plausibility of counterfactual: triangulate applicant claims with HS reliability, assessor notes, and unsuccessful funding attempts
Did the CFF programme represent value for money? Either at present, or has the potential to represent value for money in the future? Did the (present or future) benefits justify the costs?	Number, and scale of change in outcomes (compared to counterfactual) provided through primary data collection, monitoring data and secondary data sources. Monetary unit values (Benefit Transfer from secondary data sources or primary data collection). Issues of transferability (high transfer error) and the challenge of dealing with heterogeneity of projects may mean such analysis can only be indicative.	This will be revealed through the cost-benefit analysis (as outlined in section 9.2). The cost benefit analysis will be informed by findings from the impact evaluation. Overall findings could be compared to other benchmarks to determine whether VFM is high, medium or low.	Triangulate the CBA result with qualitative evidence on additionality, plausibility of counterfactual, and persistence of benefits to judge VfM as high/medium/low.
Equity	-	-	-
Were the effects of CFF distributed differently among users and non-users, organisations and staff?	Number, and scale of change in outcomes, by beneficiary – provided through primary data collection, monitoring data and secondary data sources.	The cost–benefit analysis could be disaggregated by beneficiary group, allowing comparison of the benefits received by different types of beneficiaries against the associated costs. Distributional weighting may also be applied to account for differing levels of marginal utility (as per the Green Book 2026).	Qualitative interviews from the process evaluation could provide evidence on extent to which: Grants were distributed among levels of need and urgency.   The portfolio showed a reasonable spread across regions, sectors and priority groups/places, which reflected stated aims
How were the outputs and outcomes of CFF distributed geographically?: Where there any areas which benefitted disproportionately more than others? How do the value for money conclusions change after accounting for distributional effects?	Number, and scale of change in outcomes, by location – provided through primary data collection, monitoring data and secondary data sources. This should include place based impacts.	Evidence of changes in outcomes by location should be provided through the impact evaluation. Distributional weighting may also be applied to account for differing levels of marginal utility (as per the Green Book 2026).	n/a

9.2. Cost benefit analysis

As part of the 4Es value for money assessment, it is also worth considering the potential for undertaking a full cost-benefit analysis of the DCMS CFF. A CBA assesses all economic and social costs and benefits attributable to the intervention to determine whether, on balance, the benefits exceed the costs. This approach should align with the principles set out in HM Treasury’s Green Book (2026).

9.3. Overview of the approach

An overall CBA calculation requires the calculation of the estimated benefits, as well as the costs of the CFF. To understand the changes in outcomes and therefore calculate benefits, inputs are needed from the heritage science counterfactual and quantitative impact evaluation:

• Heritage science (HS) reliability modelling: Recognising that econometric proof of additionality for capital maintenance is inherently limited, this modelling will form the main quantitative basis for establishing the counterfactual. This enables the quantification and monetisation of long‑term benefits of additional years of life for each asset/ stock. This will include impacts such as health and wellbeing, social and community cohesion outcomes outlined in the theory of change.

• Counterfactual impact evaluation: This approach seeks to identify the causal impact of the funding on outcomes, by comparing funded organisations to a counterfactual scenario in which they did not receive funding. If this approach allows for estimation of the additionality for specified outcomes, it can be applied to relevant unit values and incorporated into the cost–benefit analysis.

9.4. Benefits

The Green Book recommends valuing all material benefits (including avoided costs) to UK society in a social cost–benefit analysis (CBA). This covers impacts on users, producers, the Exchequer, third parties, and the environment. The core benefits that should be included in a CBA include:

Productivity

In the short-term, the funded works are expected to preserve the productivity of the funded organisations compared to the counterfactual. For example, building degradation or failure may result in productivity decline or in extreme circumstances full closure. In this instance, the avoided productivity loss is taken as the economic benefit.

Over the longer-term, productivity gains may occur if upgraded infrastructure enables higher-quality cultural output, or if the funding stimulates knowledge exchange and innovation within local creative clusters. The additional productivity over and above the counterfactual is taken as the economic benefit.

Typically, productivity is measured as economic output, or gross value added, (GVA) per worker.^{[footnote 30]} In the absence of firm level GVA estimates and data on intermediary costs, proxy measures such as turnover per worker will be used to estimate productivity changes. This requires data on how organisation levels of revenue and labour supply change over time. This data will be collected through secondary data sources and supplemented by primary data collection (see Section 4). The analysis should however be mindful that in the context of non-profit organisations, turnover can be significantly influenced by factors beyond cultural output (e.g. grants or donations). Turnover alone also does not price in the existence of non-market benefits and so does not reflect the true economic and social output of a museums. For this reason, evaluators should consider this when conducting this analysis.

It is anticipated that the counterfactual impact evaluation will provide direct estimates of changes in turnover (revenue) and employee numbers which can be attributed to CFF. Where statistically significant changes are identified, these metrics will feed into the estimate of productivity, as described above.

Employment and local economic benefits

CFF may likely influence employment through the below channels:

• Jobs safeguarded in the short term. By keeping venues operational and preventing loss of cultural output, CFF could safeguard direct employment in cultural organisations.

• Job creation in the long term. By supporting organisations directly to remain operational this could improve prospects for job creation in cultural organisations in the longer term.

• CFF funding may directly support local specialist construction jobs and jobs in local businesses associated with visitor spending.

At the national level, the inclusion of employment effects has two key factors which limit its inclusion in the estimated benefits:

• Displacement: As set out in the Green Book (2026), it should generally be assumed that interventions do not increase national labour supply. Unless robust evidence can be produced to the contrary, it is assumed that any job created is filled by a worker leaving another job elsewhere.

• Double counting: Economic benefits are typically already captured through productivity changes (such as increased Gross Value Added or wage premiums). Monetising employment changes on top of this would ‘double count’ the economic value of that labour.

The CBA will be undertaken at the programme level and therefore should follow the above guidance.

At the place-based (or local) level, provisions are made within the Green Book to account for employment effects. Place-based impacts should therefore be explored with DCMS, and considered as to where and when they should be included as part of a CBA or wider 4Es assessment. For example, demand-side effects can matter locally if a theatre organisation is the only one in the area. If this theatre were to close, the area could lose demand, supply chain activity and associated jobs since another nearby theatre may not absorb that activity.

Organisational level benefits

The reduction in the maintenance backlog can be considered through two mechanisms:

• Monetary value of the backlog: This would measure the estimated monetary change in the backlog of projects for each organisation. However, since the projects will be directly funded by CFF, this is considered a transfer and would not be included in the CBA.

• Avoided future costs: This will need to be gathered through primary data collection, where implementing a QED is not viable. Instead, impacts will be estimated through a before and after analysis. As above, this would need to incorporate assumptions on the levels of additionality to avoid overstating the benefits.

Estimating the avoided future maintenance costs will need to consider the initial monetary value of the backlog to avoid double counting and overstating the cost savings.

Cultural and social wellbeing benefits

In addition to the economic benefits, there may also be positive social welfare benefits (e.g. community cohesion, health, wellbeing and pride in place) arising from the programme. Social welfare benefits are often non-market in nature, in that they are not expressed in current market prices. However, they may be observable in people’s behaviour, preferences, or other measures of wellbeing, using methods provided in the HM Treasury Green Book (2026). Some of the non-market or intangible benefits include:

• Users: In the absence of a robust assessment in the change in visitor numbers, use value is anticipated to arise from an extension to the lifetime of the organisation. The calculation for estimating the avoided welfare loss is set out below:
  • Avoided welfare loss = WTP for cultural service provided by organisation x (additional years of services gained from CFF funding/year of service in counterfactual) x annual number of visitors

The relevant use values are set out below. It is recommended that the final evaluator undertakes additional checks to ensure the correct value for benefit transfer is used. The authors did not know the funded projects nor had access to the full applications at the time of writing. The final evaluator will also need to provide a judgement to whether the use or non-use value is the most relevant. For example, if the life-time extension far exceed the lifetime of the average person then the benefits are likely to be predominantly non-use value. In this instance, the non-use value could be applied to users, and non-users omitted from the analysis.

• Theatres: Lawton et al. (2021) estimates a WTP value for a one-off payment of £11.08 (in 2020/21 prices) for maintaining a theatre. This value is recommended for benefit transfer.
• Galleries: Lawton et al. (2021) estimates a WTP value for a one-off payment of £6.22 (in 2020/21 prices) for the expansion of a gallery. However, the authors do not recommend this value for benefit transfer.

However, no benefit transfer values exist for performing arts venues, concert/opera halls, live music venues, multi‑use arts centres/community cultural hubs etc. In the absence of comparable cultural sites in the CHC database, it is recommended to design a bespoke Stated Preference survey to generate evidence on galleries and performing arts venues. It is also likely to be worthwhile undertaking primary Stated Preference survey research to understand the use values for theatres. Whilst values exist for benefit transfer, it is unknown in terms of how robust they are for application in this specific context.

• Non-users: Likewise, the funding may also help to preserve the social welfare benefits of the organisation into the future for those who don’t necessarily use the organisation (i.e. non-use benefits). Non-use benefits can include the value of knowing the organisations exist (existence value), knowing other can use the organisation services (altruistic value) and knowing future generations can use the organisations services (bequest value). There is also likely to be value for the person knowing that they have the choice to use the organisation into the future, even if they are not using it in the present (option value).

The same studies above provide relevant benefit transfer values for non-users, which can be used in the same ‘Avoided welfare loss’ calculation above. However, it should be noted that these values are not bespoke to CFF.

• Theatres: Lawton et al. (2021) estimates a WTP value for a one-off payment of £4.32 (in 2020/21 prices) for maintaining a theatre. This value is recommended for benefit transfer for non-users.

• Galleries: Lawton et al. (2021) estimates a WTP value for a one-off payment of £3.20 (in 2020/21 prices) for the existence of a gallery. It should be noted that this is not specific to maintenance or extending the lifetime of the building. Benefit transfer of this value should note this a core limitation.

A core challenge in the inclusion of non-use values is that of aggregation. DCMS research on the application of non-use value^{[footnote 31]} recommends a cut-off to capture the non-use population. In the absence of quantitative data, the theory-based evaluation will gather qualitative insights from case study organisations into how far the ‘reach’ of the organisations extend.

The benefit transfer values are not bespoke or tailored to CFF, and therefore there are significant unknown in terms of their robustness. Thus, there is likely to be good value in undertaking primary Stated Preference survey research on the general population to identify a set of benefit transfer values. This should include scope to compare non‑user values for local and non‑local populations at varying distances from the site.^{[footnote 32]}

Prior to undertaking primary data collection, it is recommended to explore the extent to which there is heterogeneity within the CFF portfolio by undertaking a review of funded projects. This would help assess the risks of generalising non‑use values estimated for a small number of case studies to the wider, diverse programme. If substantial heterogeneity is identified, primary stated preference data collection should only be considered where specific, high‑profile projects justify bespoke valuation and where sufficient budget is available, rather than as a programme‑wide requirement.

• Educational value: Funded venues provide a space for formal and informal learning services. Created or upgraded accessible spaces and equipment could lead to the accumulation of human capital (skills, experience, qualifications), cultural capability, and strengthen progression into creative and technical pathways. Work in progress for DCMS has collated new evidence on the impact and value of cultural educational activities which could be applied to this evaluation based on the type of activity, number of learners engaging, and frequency of their engagement, learning hours delivered.

As outlined in Section 9.3, the unit values used to estimate non‑market impacts can be applied directly to the heritage science outputs. In practice, this means applying unit values for cultural services to the additional years of service delivered by cultural organisations as a result of CFF, to generate an overall estimate of non‑market value.

Environmental

Upgrades to more energy‑efficient systems, improved insulation and the use of sustainable materials may reduce energy consumption and emissions. Reduced energy use could generate two main benefits.

• Upgrades could deliver positive financial gains for organisations through lower energy bills, where lower costs will reflect reductions in resource costs.

• There may also be reduced negative externalities through reductions in emissions. These can be valued by converting observed changes in emissions and multiplying by the UK Green Book approved carbon values.

However, the counterfactual impact evaluation is unlikely to be able to attribute changes in energy consumption to CFF, so it is recommended that these benefits are not included in a CBA.

9.5. Costs

In line with HMT Green Book guidance, it will be necessary to focus on the net costs associated with the delivery of the fund. This will be derived from the following, as outlined in the inputs included in section 2.3.1:

• Central government and partnership funding costs provided by the public sector and other bodies. In line with the Green Book, CBA should estimate the net resource costs of delivering CFF (i.e. the real resources used, irrespective of funder), relative to a without‑CFF counterfactual.

• Administrative costs. This includes the resources used in setting up and delivering the fund, incurred by DCMS as well as Arts Council England and specialist advisers. These advisers will spend time and resource designing the fund and application process and carrying out monitoring and compliance activities. In addition, the time and resource required from beneficiaries to apply for the fund should be included.

9.6. Adjustments

In a typical cost benefit analysis, there are general principles that are recommended to adjust for (as outlined in the Green Book):

• Deadweight – This considers the extent to which the fund has reduced the loss of cultural output caused by asset failure. It will be addressed primarily through the impact evaluation methods set out in Sections 6 and 7. If, for any outcome metrics, this reduction cannot be measured directly, deadweight adjustments based on findings from previous evaluations may be applied instead if there is a reasonable expectation of comparability.

• Displacement – Displacement considers how much of the value created is as a result of shifting spending or demand from other sites or areas. As noted, demand side economic impacts (e.g. changes in revenue and output) are largely seen as displaced from elsewhere in the economy and should therefore not be included in a social cost benefit analysis. Furthermore, some of the wider benefits of CFF may also involve elements of displacement, though this may be something that can only be addressed narratively in the evaluation.
  • When considering through a place-based lens – i.e. the place-based considerations within the 4Es – displacement assumptions can be relaxed. This allows for a more nuanced assessment of how the funding affects local areas around the organisations.
• Leakage – Leakage considers how much of the benefits are flowing outside of the UK. It is anticipated that this may not be relevant given that the organisations operating in the cultural assets are likely to be UK based.

9.6.1. Evaluation timing

The economics evaluation should follow the timings of the counterfactual and theory-based impact evaluation, drawing on evidence of both to provide informed judgements about the value for money:

• Short-term: To leverage the findings of the other evaluation strands, an initial 4E’s assessment should be undertaken in April 2031. It is important to commission the economic evaluation alongside the other evaluation activity. This is to ensure the design of the economics is embedded in all evaluation approaches and the required information obtained. The design of the economic evaluation can be finalised from April 2030 alongside other elements of the evaluation.
  • At this point, it may be possible to undertake a CBA, depending on the statistical significance of the findings from the impact and heritage science evaluation.
• Longer-term: The final value for money assessment should be undertaken from April 2032 to leverage the wider evaluation findings. The 4E’s assessment can be updated to account for longer-term findings which may have materialised.
  • It is anticipated that a CBA will be viable from this point in time, subject to statistically significant findings within other evaluation strands.

10. Lessons learnt from previous evaluations

Ipsos, supported by UCL, delivered the 2022 to 2026 dual evaluation of the DCMS Museum Estate Development Fund (MEND) and the Public Bodies Infrastructure Fund (PBIF). MEND and PBIF, like CFF, provided funding to museums to fund outstanding maintenance backlogs. The evaluation provided several relevant recommendations which have direct relevance for CFF:

1. Consideration should be given to longer-term evaluation frameworks that can better capture impacts as they materialise over time. There is no definitive way to quantify the most appropriate time to undertake further evaluation activity. However, five-years post-funding should provide a sufficient period of time for outcomes to materialise (or at least begin to materialise).

2. Where possible funding decisions should consider the use of a quantitative scoring systems. These would be more amenable to establishing a counterfactual and enhance the ability to attribute outcomes to the interventions.

3. Improving data collection processes could enable more comprehensive impact measurement in future evaluations. For example, a condition of the grant funding could be annual reporting requirements on key metrics relevant to the ToC and Business Case.

4. The evaluation would have benefitted from higher response rates. This would have enabled:

a. A larger sample size.

b. Provided a more comprehensive account of the changes to museums following the funding.

One way in which engagement with external evaluators could be increased could be to make engagement a contractual obligation upon receiving grant funding.

11. Recommended evaluation approach

This section sets out the conclusions and recommended approach to delivering an evaluation of CFF. It should be noted that at the time of writing, the funded projects had not been announced. This study was therefore developed through analysis of the EOIs.

11.1. Mixed method evaluation

An evaluation of CFF should utilise qualitative and quantitative evaluation evidence to explore the efficiency and effectiveness of the delivery process, the impact of the capital funding and the associated value for money.

11.1.1. Process evaluation

The process evaluation would need to cover elements such as fund design, the application process, application assessment, contracting and monitoring, and final payments/ project close. The process evaluation would utilise qualitative depth interviews with central government, successful and unsuccessful applicants, assessors, LA stakeholders, and eligible non-applicants. Ongoing monitoring information would provide the basis for a quantitative assessment of some process elements.

11.1.2. Theory-based impact evaluation

The primary theory-based approach would be Contribution Analysis – which would test the contribution claims for each link in the logic model of how CFF brought about change. However, Contribution Analysis alone is unlikely to provide sufficient evidence of impact. Furthermore it is recommended that Contribution Analysis is combined with other methods such as Process Tracing and Qualitative Comparative Analysis.

Process Tracing seeks to develop tests for the strength of the evidence in favour of contribution claims. Through the use of several tests (straw in the wind, hoop, smoking gun, double decisive) an overall qualitative judgement on the strength of the contribution claim can be made.

Finally, Qualitative Comparative Analysis could be used to analyse what combinations of project characteristics, planning and delivery and contextual conditions appear to have been most critical to achieving successful outcomes and what combinations have tended to produce more problematic results. It could potentially be utilised in relation to an analysis for just the largest projects or for a larger sample, dependent on the number of case studies / the scale of the fieldwork which is to be undertaken.

11.1.3. Counterfactual impact evaluation

The recommended approach to delivering a counterfactual impact evaluation is to utilise a synthetic control methodology. The analysis would be undertaken on the funded organisations which can be linked into the ACE NPO data and/or Charity Commission data. Control organisations would be unsuccessful applicants which could also be linked into the relevant data sources. Control variables could be sources from either of the data sets, EOIs, or primary data collection.

The analysis would predominantly be focused on financial metrics (income, expenditure, assets and liabilities, government grants, donations, cash reserves), but also could cover employee numbers, number of volunteers and audience numbers.

The principal risk of this approach is that the analysis is only undertaken on organisations which can be linked, opposed to a full programme wide assessment. Therefore, there is a risk that the findings are not generalisable to the entire cohort of funded organisations.

11.1.4. Heritage science counterfactual

Heritage science can be used to provide a practical basis for estimating when programme outcomes are likely to:

• Become measurable.

• Support the assessment of additionality.

• Create a bridge between capital investment, cultural service continuity and subsequent economic evaluation.

Heritage science can be used to understand how funded activities result in physical change, for a range of maintenance works including:

• Roofs.

• Facades.

• Internal finishes.

• External areas.

• Service installations

• Other novel interventions such as equipment needed to conduct cultural activities.

11.1.5. Economic evaluation

The economic evaluation should be framed within a NAO 4E’s evaluation, to provide the basis for synthetising the qualitative and quantitative data gathered through the mixed methods approach.

It may also be feasible for a CBA to be undertaken. The CBA would compare benefits such as productivity uplifts and extension of organisation lifetime (and associated use and non-use value) against programme cost.

11.2. Outcome measurement

11.2.1. Key outcome indicators

The recommended key outcome indicators to provide objective assessment of CFF includes:

• Value of maintenance backlog (£)
• Number of items on maintenance backlog
• Asset lifetime extension
• Electricity usage (kWh)
• Gas usage (kWh)
• Energy generated from renewable sources (kWh)
• Total revenue (£)
• Donations (£)
• Lost activity due to asset failure (£)
• Number of closures due to asset failure
• % of annual maintenance on planned vs unplanned works
• Annual spend on maintenance (£)
• Number of paid staff
• Number of volunteers
• Number of visitors
• Number of days open to the public

11.2.2. Data sources

The data sources required to inform the evaluation include:

• Monitoring information: The EOIs, full applications, costed condition surveys and annual data returns are anticipated to provide baseline and outcome data on many of the above metrics.

• Primary data collection: Primary data collection will be required to inform the Heritage science analysis. Whilst a QED using primary data is not considered feasible (e.g. a DiD), there may still be value in undertaking primary data collection to cover data gaps (descriptively), but also to inform the process and economic evaluation.

• Secondary data sets: The ACE National Portfolio Organisations data and Charity Commission data is anticipated to provide sufficient coverage of funded and unsuccessful organisations to undertake a synthetic control case study analysis.

• Qualitative interviews: The process and theory-based evaluation will be underpinned by a series of depth-interviews with applicants and stakeholders.

11.3. Timeframes

The funded projects are required to be complete by March 2029. However, it is common for large scale capital projects to overrun, so it would be prudent to allow for this within the evaluation timing.

The recommended evaluation approach includes:

• Short term process evaluation in April 2027: Process evaluation, commencing approximately 12 to 18 months following the start of the capital work to provide insights into the design, application and delivery process. This will also enable an initial assessment of whether projects are on track according to their project plans.

• Mixed-methods evaluation in April 2030: Commissioning a full mixed methods evaluation to explore the short-term impacts of CFF, and understand the project close elements of the process evaluation:
  • April 2030: From this point, the longer-term process evaluation and the theory-based evaluation could be completed. It would also be prudent to build in sufficient time to plan and operationalise the quantitative elements of the evaluation, including, the counterfactual impact evaluation, heritage science impact evaluation, value for money assessment and any primary data collection which may be required.
  • April 2031: From this point, there will be a full year of post completion data. This will allow for quantitative assessments undertaken within the counterfactual impact evaluation, heritage science impact evaluation and value for money assessment.
• Longer-term mixed-methods evaluation April 2032: A longer term approximately 4-5 year after projects finish should enable a new ‘steady state’ to have established. From April 2032, a longer-term theory-based evaluation could be undertaken, and the planning and operationalisation of a longer-term counterfactual impact evaluation, heritage science impact evaluation and value for money assessment.

From April 2033, the longer-term counterfactual impact evaluation, heritage science impact evaluation and value for money assessment could be conducted.

11.4. Risks

There are risks associated with the above recommendations, many of which are associated with the long timeframes over which the evaluation activity takes place:

• Engagement: Many of the qualitative interviews and primary data collection waves will take place following the completion of the works. As such, organisations may not be full engaged with the evaluation, and ACE and DCMS poses less levers to encourage participation than if funding was still being administered.

• More rounds of funding: Whilst additional rounds of funding would increase the scope to implement quasi-experimental methods (e.g. larger sample sizes, pipeline/ staggered DiD designs), it may change the recommendations presented in Chapter 6. DCMS should therefore re-assess the feasibility of implementing DiD designs should additional funding rounds happen.

• Additional sector support: In the future, there may be other sector support packages, providing funding for similar outstanding maintenance issues. Future work should therefore be mindful of conflating CFF with other support measures.

• ACE NPO portfolio updates: ACE NPO organisations change every few years. As such, there is likely to be sample attrition across the different data years, reducing the usable sample for statistical analysis. However, this is not anticipated to undermine the synthetic control approach.

Annex A: Statistical power calculations

Difference-in-difference statistical power calculations

Whilst in principle there are viable comparator units to the c.60 funded organisations, the statistical power of the DiD analysis is likely to be constrained by the sample size. To provide insights into the minimum detectable effect size, a series of power calculations were undertaken on key variables. These included annual visitor numbers, number of full-time equivalents, commercial income, philanthropic investment and unfunded but necessary repairs, to understand the extent to which indicators would need to increase to be detected.

The DCMS sector benchmarking dataset (sample) of organisations was used to provide distributional information on the key outcome variables (e.g. standard deviation). The following assumptions were also made:

• Sample size: Sample sizes are dependent on the response rate to primary data collection. For simplicity, it is assumed that all 60 funded organisations respond to the survey. The response rate of unsuccessful organisations is modelled over 100% to 10%, decreasing in 10% increments. Separate power calculations are undertaken assuming:
  • Strand 1 and 2 can be pooled.
  • Assuming Strand 1 and 2 cannot be pooled.

• Power: The statistical power is the probability that the test will correctly detect an effect, if it truly exists. Standard practice is to assume 80% power for the purpose of power calculations.

• Number of time periods: It is assumed that there will be a single pre- and post-period data point. This approach minimises the burden placed on organisations.

• Alpha: Set at 0.05, this assumes effects are detectable at the 95% confidence level.

Standard deviation: Calculated using the DCMS sector benchmarking dataset (sample).

A key consideration is the extent to which Strand 1 and 2 can be pooled. For the purposes of these power calculations, separate analyses are undertaken assuming

• Pooling of Strand 1 and 2.
• Strand 1 and 2 cannot be pooled.

The large increase required in outcome indicators below is largely driven by the significant variability in outcome measures high variability means high relative changes in outcome variable are needed to detect statistically significant changes.

The power calculations found the following results:

• Minimum detectable effect size when pooling Strand 1 and 2: The plot below shows the change from baseline for effects to become statistically significant at the 95% confidence level. For the given response rates (informing the sample size for the available analysis), key outcome indicators need to double before becoming statistically significant at the 95% confidence level, even with 100% response rates among control organisations.

Response rates around 20 to 30%, representing more realistic response rates for unsuccessful organisations, yield broadly similar results in terms of minimum detectable effect size.

Figure A.1 Change from baseline for effects to become statistically significant at the 95% confidence level, pooling Strand 1 and 2 funded organisations

Figure A.1 is a line graph. It shows the percentage change from baseline for effects to become statistically significant for strand 2 funded organisations, by the assumed response rate. This is across visitor numbers, full time equivalents, commercial income, philanthropic investment and unfunded but necessary repairs. Across all categories, the higher the assumed response rate, the lower the change from baseline required for effects to become statistically significant.

• Minimum detectable effect size for Strand 1: Similar results to the above are found when the treatment sample consists of only organisations who applied to Strand 1. With the highest response rates, outcome indicators need to increase between 2 and 2.5 times for impacts to be detected at the 95% confidence level. When looking at 20 to 30% response rates, outcome indicators need to increase by 2 to 3 times to be detected at the 95% confidence level.

Figure A.2 Change from baseline for effects to become statistically significant at the 95% confidence level, Strand 1 funded organisations

Figure A.2 is a line graph. It shows the percentage change from baseline for effects to become statistically significant for strand 1 funded organisations, by the assumed response rate. This is across visitor numbers, full time equivalents, commercial income, philanthropic investment and unfunded but necessary repairs. Across all categories, the higher the assumed response rate, the lower the change from baseline required for effects to become statistically significant.

• Minimum detectable effect size for Strand 2: Even larger effect sizes are required to return statistically significant results at the 95% confidence level when considering only Strand 2 funded organisations. At a more realistic response rates of 20 to 30% outcome indicators need to increase by approximately 2.5 to 3.5 times for statistically significant effects to be detected.

Figure A.3 Change from baseline for effects to become statistically significant at the 95% confidence level, Strand 2 funded organisations

Figure A.3 is a line graph. It shows the percentage change from baseline for effects to become statistically significant for strand 2 funded organisations, by the assumed response rate. This is across visitor numbers, full time equivalents, commercial income, philanthropic investment and unfunded but necessary repairs. Across all categories, the higher the assumed response rate, the lower the change from baseline required for effects to become statistically significant.

Annex B: Synthetic control scoping

The feasibility of undertaking a synthetic control approach using ACE NPO and Charity commission datasets has been explored in turn below. For both datasets, treatment status was randomly allocated to five venues (treatment allocation varied by dataset). It is beyond the scope of this Evaluation Framework to identify the exact econometric specification for the synthetic control. Instead this Study seeks to understand whether a synthetic control can viably be estimated and used to provide causal insights.

As such, synthetic controls were produced using past outcome measures to inform the weights as the starting point. Where the synthetic control produces a ‘close enough’ match to the pre-treatment outcome indicators, it is determined that the synthetic control can provide a viable route to credible impact evaluation. The econometric specification of the synthetic control can be ‘tightened’ through the use of additional venues controls (both time variant and invariant) and area level controls (e.g. local economic conditions or trends in attitudes and participation in heritage and culture).

To assess the goodness-of-fit of the synthetic control, three checks are undertaken:

1. Does the synthetic control visually match the trends of the treated unit? In the fictitious pre-treatment period, the synthetic control is expected to closely resemble the trend and level of the chosen treated unit for each iteration of the analysis. Divergence is anticipated in the fictitious post-treatment period as these periods are not included in the weighting. However, trends closely following the synthetic control provide confidence that weights derived in earlier time periods are still representative in later time periods.

2. Is the reported effect size statistically insignificant? It is expected that there is deviation from the treated unit and synthetic control in the hypothetical post-treatment period. However, the magnitude of this deviation would be expected to be within the bounds of normal outcome variability and therefore not exceed the threshold for statistical significance.

3. Is the scaled L2 parameter close to zero? The scaled L2 parameter is a goodness-of-fit measure^{[footnote 33]}, and provides an indication of how well the synthetic control on average tracks the observed treated unit in the pre-treatment period. A parameter closer to 0 denotes a good match.

The results of this feasibility assessment are presented below.

Feasibility of synthetic Controls with ACE NPO data

The modelling results using the ACE NPO data are presented below. The results suggest that:

• Visual inspection: Visually, the synthetic control in most instances almost perfectly managed to replicate the outcome trajectory of the treated unit using just past outcome measures. This provides confidence that a fuller specification of the optimisation algorithm will enable a credible estimation of the counterfactual of treated venues.

• Significance of reported effect size: In all instances, the reported effect sizes return p-values well above the 0.05 cut-off for statistical significance at the 95% confidence level. In most instances, p-values are relatively close to 1.

• Scaled L2 parameter: In all instances, the scaled L2 score is very low – suggesting on average that the on average, over the pre-treatment period used for this modelling exercise, the synthetic control tracks the treatment unit well relative to its variability.

Figure B.1 Example synthetic control output and summary of key model fit statistics for randomly selected EOI using ACE NPO data

Figure B.1 is a line chart. It shows an example of a synthetic control with key summaries. Across the random sample, the Scaled L2 ranges from 0.0281 to 0.0598, indicating a strong pre-intervention match. Across the random sample, p-values range from 0.571 to 1, indicating no spurious differences outside the funding period. The two categories, synthetic control and randomly treated organisation, are broadly in line with each other.

Feasibility of synthetic controls with Charity Commission return data

The modelling results using the Charity Commission data are presented below. The results suggest that:

• Visual inspection: Visually, the synthetic control in most instances almost perfectly managed to replicate the outcome trajectory of the treated unit using just past outcome measures. This provides confidence that a fuller specification of the optimisation algorithm will enable a credible estimation of the counterfactual of treated venues.

• Significance of reported effect size: In all instances, the reported effect sizes return p-values close to 1, indicating that the observed variability is likely within the bounds of random chance.

• Scaled L2 parameter: In all instances, the scaled L2 score is very low – suggesting on average that the on average, over the pre-treatment period used for this modelling exercise, the synthetic control tracks the treatment unit well relative to its variability.

An example of the synthetic control plot, including a summary of the key model fit statistics is presented below. Plots for the other randomly selected organisations are very similar to the below, and are not presented.

Figure B.2 Example synthetic control output and summary of key model fit statistics for randomly selected EOI using Charity Commission data

Figure B.2 is a line chart. It shows an example of synthetic controls with key summaries. Across the random sample, the Scaled L2 ranges from 0.0231 to 0.0622, indicating a strong pre-intervention match. Across the random sample, p-values range from 0.308 to 1, indicating no spurious differences outside the funding period. The two categories, synthetic control and randomly treated organisation, are broadly in line with each other.

Overall, the above results provide confidence that a synthetic control approach could be used to undertake a credible assessment, on an organisation-by-organisation basis, of the causal effect of CFF funding on key outcomes of interest.

Annex C – Heritage science modelling

The aim of the CFF intervention is to reduce the degradation of an existing asset stock or introduce new asset stocks which will impact on the flow of cultural and heritage services in different ways. Before the intervention and following the intervention, the venue will continue to deteriorate over time, as the ageing of buildings and equipment is an inevitable process typically caused by atmospheric influences, environmental conditions and wear arising from use.^{[footnote 34]}

Once an asset degrades beyond a certain point it reaches ‘functional obsolescence’. Functional obsolescence is the condition in which the building no longer fulfils the functions and use requirements for which it was originally designed and loses its utility.^{[footnote 35]} In the case of the counterfactual model these functions and requirements are the ability of the venue to provide the cultural and heritage services in Table 8.1. This deterioration may affect different cultural services in different ways depending on the function of the asset in relation to the cultural services.

The model is designed so that projects of different types can input information about the asset (venue and/or elements of the building fabric) and estimate how the intervention changes the expected lifetime of the asset and the cultural services it enables.

In practice this requires estimating the time period over which venue would remain functional under two scenarios:

1.  No-intervention scenario in which deterioration continues.

2.  Intervention scenario in which asset condition improves following CFF investment.

The difference between these trajectories represents the additional years of functional asset life enabled by the funded activities.

These additional years of asset functionality can then be translated into the additional years during which cultural activity (and flow of services) can continue to take place within the venue.

This provides a practical mechanism for linking capital investment in infrastructure to the continuity of cultural services. Additionally, it supports the evaluation of avoided deterioration and reduced risk of venue closure.

Defining the asset

A core analytical challenge of the counterfactual is that CFF supports a wide range of maintenance and renewal activities across different types of assets (Chapter 2.3.2).

Funded projects may include:

• Building systems and environmental performance upgrades.

• Building fabric and structural renewal.

• Improvements to technical and production capacity.

• Works to improve access, inclusion and user experience.

Applicants must first identify the asset type in their funding application. Definitions of assets are outlined in DCMS CHC Programme with definition of assets and services expected to evolve following feedback and further discussion.^{[footnote 36]} Rather than spend time categorising different asset types, this counterfactual model allows CFF applicants to define their own assets. It can be suggested that any building or equipment modified or introduced by CFF could be defined as an asset. This is by virtue of them receiving funding due to the criteria and aims of the CFF applications.

At the time of writing, the list of funded organisations had not yet been announced. Illustrative example asset categories shown in Table C.1 are based on the RICs categories of building elements for planned preventive maintenance.^{[footnote 37]} These serve as examples of types of assets that can be inputted in the counterfactual model. There is also a ‘Other’ category provided which novel interventions or equipment could be inputted.

Table C.1. Example categories of stocks affected by CFF intervention after the RICs categories of building elements

Stock	Common Sub-elements
Structure	Substructure; Superstructure (where superstructure is the component constructed above ground level, while the substructure is the component built below the ground level).
Roof(s)	Roof coverings; Parapet walls; Rainwater goods; Roof lights; Chimneys; Flues; Lightning conductors; Access provision; Walkways; Lifelines / safety fixings; Plant room compounds and doors; Tenant installations
Facades	External walls and cladding; Windows and doors; Entrances; Other façade elements
Internal finishes	Ceilings; Internal walls and partitions; Floor structures; Floor finishes; Stairs; Internal joinery (windows, screens, doors, skirtings); Decorations; Sanitaryware; Fixtures and fittings; Passive fire precautions (compartmentation)
External areas	Access routes / entrances; External paving / pathways; Parking; Service / delivery yards; Landscaping; Street furniture; Boundary treatments
Service installations	HVAC supply and distribution; Electrical supply and distribution (power and lighting); Sanitary / water supply, distribution and evacuation; Plumbing; Storage tanks; Fire detection, alarm and active fire installations; Lift installations; Utility connections; Renewable / sustainable energy systems; Automatic sunscreens / power-assisted brise soleil; Specialist services (trigeneration, photovoltaic cells)
Other	Novel interventions such as replacing equipment needed to conduct cultural activities.

Defining baseline condition

Each asset begins from a baseline condition assessment. This is the condition at which the venue or building/system element is when the CFF fund was applied for.

This baseline condition is derived from the condition surveys provided by ACE or DCMS, schedules of repair, risk reports and technical documentation already submitted by applicants within their CFF applications. By using existing documentation, this minimises time expended in additional data collection and aligns with the Magenta Book principle that evaluation should maximise learning while minimising burden on participating organisations.

The model expresses asset condition on a continuous 0 to 100 scale. This scale is anchored in professional survey practice and derived from RICS condition ratings ^{[footnote 38]} with the condition is assumed to decline toward zero in the absence of intervention.

The scale can be interpreted as follows:

• 100 = fully functional asset in very good condition.
• Mid-range values = serviceable but deteriorating assets.
• 0 = asset no longer capable of safely or reliably performing its intended function.

Table C.2. showing the condition scoring framework used in the counterfactual model

Condition band	Score range	Definition
Good	75 to 100	Item currently in good condition and with no outstanding repair or maintenance requirements. Works of a cyclical nature may be applicable, for example periodic cleaning or decoration.
Fair	50 to 75	Item is in serviceable and generally acceptable condition, but may exhibit signs of age-related wear and tear, weathering or superficial damage. Repair or renewal may be required.
Poor	25 to 50	Item is in serviceable and generally acceptable condition, but may exhibit signs of age-related wear and tear, weathering or superficial damage. Repair or renewal may be required.
Hazardous	0 to 25	Item is in a dangerous condition and imminent works are required to rectify the problem, or to comply with health and safety or other statutory regulations.

After defining their assets, applicants must assign a condition score using the information available to them. This provides the starting point of assigning the lifetime and degradation rate of the asset.

Though condition is a possibility many surveyors also use risk as a measurement. This could potentially be used as an alternative and then allow for a greater inclusion of probability and risk into the degradation counterfactual.

Assigning material lifetimes and degradation behaviour

Once baseline condition has been established, the model assigns each asset (venue, building fabric or service, as defined in Table C.1) an expected service life and degradation profile. This draws on standard service life guidance and professional judgement provided within CFF applications. Applicants are required to estimate the approximate time until the asset becomes hazardous in the absence of intervention. Given the nature of the application process, this information is typically available from building condition surveys submitted as part of the funding request. Given that surveyor-derived information is available within applications, this is considered more robust than modelling best, worst and central case scenarios. However, scenario-based approaches could be explored as a complementary method where there is uncertainty in the different expected services life and degradation profile. Applicants are also asked to indicate the expected behaviour of deterioration, selecting whether the deterioration profile is linear, accelerating or rapid. These categories correspond to different values of the shape parameter 𝑘 within a simplified Weibull-type deterioration function.

Weibull models are widely used in reliability engineering and infrastructure asset management to represent ageing processes in engineered systems. In these models, the shape parameter controls how quickly deterioration accelerates through time. Values greater than one represent ageing or wear-out processes, which are typical of building components and infrastructure assets.^{[footnote 39]} Table C.3 shows the 𝑘 values (all greater than 1) that applicants select from to indicate the degradation of the asset.

Using the baseline condition 𝐶0, the estimated service life 𝐿, and the deterioration parameter 𝑘, the model generates a counterfactual condition trajectory using the following degradation function:

Where:

𝐶𝑡 = condition at time
𝐶0 = baseline condition
𝐿 = expected service life (time until hazardous condition)
𝑘 = degradation curve parameter
𝑡 = time since baseline condition.

Table C.3 showing the values of 𝑘 dictating the simplified Weibull curve based on

𝑘 value	Meaning	Curve Behaviour	Interpretation for assets
2	Moderate ageing	Condition declines steadily through the asset life, with deterioration distributed relatively evenly across the service life.	Represents assets where wear accumulates gradually over time and deterioration begins relatively early.
2.5	Accelerating ageing	Condition remains relatively stable initially but deterioration accelerates as the asset approaches the end of its service life.	Represents assets where defects compound over time and maintenance issues begin to escalate in later years.
3	Late-life deterioration	Condition remains relatively stable for much of the service life before deteriorating rapidly near the end of life.	Represents assets where deterioration is initially slow but structural or system failures can escalate quickly once degradation begins.

The Weibull curve is calibrated so that the deterioration trajectory passes through both the observed baseline condition and the estimated year at which the asset becomes hazardous. The shape parameter therefore controls the distribution of deterioration through time rather than the overall lifetime of the asset. Figure C.1 shows the different degradation curves based on the inputs of Table C.3.

Figure C.1 Degradation curves based on the different 𝑘 values.

Figure C.1 is a line chart. It shows the degradation curves associated with assets with moderate aging, accelerating ageing or late-life deterioration. The greater the ageing “k” value, the quicker the condition worsens.

This reflects observed behaviour in building systems where deterioration typically accelerates once defects accumulate. Table C.4 and Figure C.2 provide example of the inputs and the degradation curve of a Stock in this case a roof.

Table C.4 showing examples of model input for an asset

Asset	Current Condition (0 to 100)	Survey Timescale (years to complete failure (0))	Curve Behaviour
Roof example	50	70	Late-life deterioration

Figure C.2 shows an example of the degradation over time without an intervention.

Figure C.2 shows an example of degradation over time without intervention, using a roof as an example.

Modelling the intervention scenario

The previous section described how the counterfactual deterioration path is estimated using a simplified Weibull-type deterioration curve. Step three is to model how a funded intervention changes this trajectory.

The intervention scenario assumes that once works are completed the condition of the asset improves immediately, and deterioration then continues over time. In other words, the intervention does not permanently stop deterioration, but it improves the starting condition and extends the future service life of the asset.

The intervention trajectory is therefore modelled by modifying the parameters of the deterioration curve. Four parameters are required to describe this change:

• Intervention start year.

• Duration of works.

• Condition uplift.

• Reference service life (RSL).

The intervention start year represents the year in which works begin relative to the application year (year 0). For example, if works are expected to begin two years after application, the intervention year is recorded as 2. The duration of works represents the expected time required to complete the intervention. During this period the model gradually increases the asset condition from the counterfactual level to the improved level at the completion of works.

The condition uplift represents the improvement in condition achieved through repair, renewal or replacement. The condition immediately following the intervention is calculated as:

Where:

𝐶𝑡𝑖= condition immediately after intervention

𝐶𝑡𝑖𝐶𝐹= counterfactual condition at the intervention year

𝑈= condition uplift

Once the works are complete, deterioration resumes using the same Weibull-type formulation described earlier with the option of being able to re-enter a new deterioration value. Through this method, if the intervention is expected to change the rate of degradation this is accounted for. However, the curve is recalibrated using the service life of the repaired or replaced component rather than the expected lifetime of the initial survey. This service life is represented by the Reference Service Life (RSL).

After intervention, the deterioration function resets using a new service life:

Where:

𝐶𝑡𝐼𝑁𝑇= condition after intervention

𝐶𝑡𝑖= condition immediately after intervention (replacing baseline condition)

𝑅𝑆𝐿= reference service life of repaired asset replacing (expected service life (time until hazardous condition))

𝑘post= degradation curve parameter after intervention. 𝑡= time since baseline condition. 𝑡𝑖 intervention completion year

RSL can be drawn from published construction and lifecycle guidance. Typical examples include:

• Structural frames and primary structural elements at around 60 years.

• Many façade elements in the 30–60-year range.

• Roof coverings at around 30 years.

• Windows and doors in the 20–30-year range, with and finishes or fixtures at shorter intervals.^{[footnote 40]}

Equivalent assumptions can also be applied to specialist technical equipment, where warranty periods, replacement cycles or manufacturer guidance provide a practical basis for estimating lifetime.

Table C.5 showing model inputs of the degradation and the intervention

Asset	Current Condition (0 to 100)	Survey Timescale (years to Hazardous)	Curve Behaviour	Enable Intervention	Intervention Start Year	Intervention Duration (years)	Uplift (points)	RSL (Reference Service Life, years)
Roof example	50	70	Late-life deterioration	Yes	7	8	100	80

Figure C.3 shows an example of the degradation over time with an intervention.

Figure C.3 shows an example of degradation over time with intervention, using a roof as an example.

Including real-life operating factors

In practice, the service life of an asset is influenced not only by the inherent properties of the material or component, but also by the conditions under which it is used and maintained.

Lifecycle modelling guidance from BS ISO 15686-8 (Service Life Planning)^{[footnote 41]} indicates that asset lifetimes are influenced by a range of factors relating to design, environment and operational use. The standard proposes a factor method in which the estimated service life of an asset is calculated by multiplying the reference service life by a set of adjustment factors representing real operating conditions.

Therefore, the model is improved by including real world operating factors using Effective Service Life (ESL) in the model.:

The BS ISO 15686-8 expresses Effective Service Life as:

𝐸𝑆𝐿 = 𝑅𝑆𝐿 × 𝜙𝐴 × 𝜙𝐵 × 𝜙𝐶 × 𝜙𝐷 × 𝜙𝐸 × 𝜙𝐹 × 𝜙𝐺

Where the factors represent those set out in Table C.6. The BS ISO guidance notes that factor values typically lie close to unity, with values generally falling within the range 0.8 to 1.2, and preferably 0.9 to 1.1, depending on whether conditions reduce or extend service life.

Table C.6 Factors influencing asset lifetimes after BS ISO 15686-8

Factor	Description
A	inherent performance
B	design level
C	work execution level
D	indoor environment
E	outdoor environment
F	usage conditions
G	maintenance level

For the purposes of the CFF model, only a subset of these factors are used. The factors relating to inherent performance, design level and work execution (A to C) are excluded. These parameters relate primarily to engineering design and construction quality and are therefore unlikely to be known reliably by applicants at the application stage. In addition, applicants may themselves be commissioning the design and delivery of works, which could create an incentive to overstate performance.

Similarly, factors relating to indoor and outdoor environmental exposure (D to E) are recognised as potentially significant drivers of deterioration, particularly for historic fabric and structurally sensitive assets. However, collecting these parameters consistently across all applicants without detailed technical surveys may not be proportionate within the scope of the programme. The simplified model therefore does not include these inputs as including these parameters could introduce unnecessary complexity and reduce the usability of the model.

The model focuses on two operational factors that applicants can reasonably assess:

• Usage intensity (F).

• Maintenance regime (G).

Maintenance planning should start at the design stage; at that point, expected use levels are generally knowable.^{[footnote 42]} The model therefore adjusts the reference service life using the simplified formulation:

Where:

ESL = effective service life

RSL = reference service life

𝐹𝑢= usage factor

𝐹𝑚= maintenance factor

Tables C.7 and C.8 translate levels of operational demands^{[footnote 43]} and maintenance regimes ^{[footnote 44]} into a small number of inputs that can reasonably be assessed by applicants and imputed into the model to generate estimated service life of the asset.

The current framework models a single intervention event for simplicity and comparability. However, in practice, cultural assets typically undergo repeated maintenance and renewal cycles over their operational life. Future development of the model could therefore incorporate multiple intervention cycles to better reflect long-term asset management pathways.

Table C.7 Levels of usage used to calculate ESL

Usage category	Description	Value
UL1	Low use intensity Intermittent or low-footfall use; low wear; low servicing demand; limited environmental loading from occupancy and operations.	1.1
UL2	Typical use intensity-Normal occupied use for the building type; standard wear; standard servicing demand; no unusual operational stress.	1
UL3	High use intensity-High footfall, intensive occupation, frequent events, heavy operational use, or conditions likely to accelerate wear, inspection frequency, cleaning demand, or replacement cycles.	0.9

Table C.8 Maintenance regimes used to calculate ESL

Maintenance category	Description	Value
CL1	Reactive / poor maintenance. None or poorly developed maintenance plan; non-original cementitious mortar or cement elements may be present within the building fabric; masonry in poor condition, including detachment; temporary structural propping may be present; limited access for routine checks or maintenance; rainwater goods may not be fully functioning; damp, vegetation build-up and graffiti dealt with mainly in an ad hoc, reactive way.	0.9
CL2	Routine / adequate maintenance. Maintenance plan in place; most or all non-original cementitious mortar or cement elements removed; masonry in fair condition, with limited cracking or detachment; access available for routine checks and maintenance; most rainwater goods functioning; preventive maintenance for damp, vegetation and graffiti undertaken routinely.	1
CL3	Planned / good maintenance.-Detailed maintenance plan in place; non-original cementitious mortar or cement elements removed; masonry in good condition; easy access for routine checks and maintenance; rainwater goods working effectively; preventive maintenance for damp, vegetation and graffiti undertaken routinely and proactively.	1.1

Figure C.4 Figure shows the degradation curve after the introduction of ESL instead of RSL with inputs UL1 and C3

Figure C.4 shows an example of degradation over time with intervention, using a roof as an example.

Including uncertainty

Building deterioration is inherently uncertain. Future degradation rates depend on environmental conditions, usage intensity, maintenance regimes and unforeseen failures. As a result, long-term deterioration cannot be predicted with absolute precision.

To reflect this uncertainty, the model produces three trajectories rather than a single deterministic estimate. These are expressed as P10, P50 and P90 estimates, representing optimistic, central and pessimistic outcomes respectively.

Table C.9 showing the levels uncertainty estimates

Estimate	Meaning
P10	Optimistic deterioration scenario
P50	Central estimate
P90	Pessimistic deterioration scenario

The P50 trajectory represents the expected deterioration path calculated using the Weibull deterioration function described in the previous section. The P10 and P90 trajectories provide an uncertainty band around this central estimate, representing plausible variations in deterioration rates.

This approach avoids presenting the model as more precise than the underlying evidence allows and reflects standard practice in infrastructure modelling and economic appraisal, where uncertainty in long-term forecasts is often expressed using percentile estimates.

In the Excel implementation, the P10 and P90 trajectories are generated by adjusting the scale parameter of the Weibull deterioration function, producing slightly faster and slower deterioration rates around the central estimate while preserving the same underlying degradation model.

The width of the uncertainty band is controlled by a user-defined input labelled Band Spread (±). This parameter represents the proportional variation applied to the service-life parameter used in the deterioration model.

The effective service life used in the P10 and P90 scenarios is therefore calculated as:

Where:

𝐸𝑆𝐿= effective service life used in the central estimate

𝑠= user-defined band spread parameter

For example, if the Band Spread is set to ±0.2, the model generates:

• A P10 scenario where deterioration occurs more slowly (service life extended by 20%).

• A P90 scenario where deterioration occurs more quickly (service life reduced by 20%).

These adjusted service-life parameters are then applied within the same Weibull deterioration function used in the central estimate, producing three deterioration curves that share the same shape parameter but differ in their rate of decline.

This allows the model to show not only the central estimate of years of viable cultural service gained from the intervention, but also the range of plausible outcomes under different deterioration assumptions.

Presenting results in this way provides a practical basis for sensitivity analysis and helps ensure that the evaluation reflects the inherent uncertainty associated with long-term asset deterioration.

Figure C.5 shows the asset degradation after intervention uplift with P10 and P90 values introducing alternative lifetime scenarios.

Figure C.5 shows an example of degradation over time with intervention, using a roof as an example. Three scenarios are modelled, one where deterioration is quicker and one where it is slower, affecting how long it takes for degradation to set in after intervention. With P10, degradation takes more than 100 years, where as with P90 values, degradation takes around 80 years.

Linking physical condition to cultural activity

The distinctive feature of the CFF counterfactual model is that it does not assess asset condition in isolation. Instead, it links the physical condition of cultural infrastructure directly to the continuity of the cultural activities that those assets enable.

This approach reflects the principle of functional obsolescence,^{[footnote 45]} whereby assets progressively lose their ability to fulfil their intended purpose as their physical condition deteriorates.

In the context of the CFF, the primary function of the asset is to CHC service delivery (recall Section 7.2).

To represent this relationship within the model, each cultural service identified within the CHC framework is associated with a minimum condition threshold, defined by the applicants who will understand the needs of the cultural services in relation to the asset best. This threshold represents the level of condition required for the asset to continue supporting a particular cultural activity. Table C.10 and Figure C.6 provide an example of the minimum conditions inputted into the model in relation to the degradation curve.

Table C.10 Example minimum asset conditions required for CHC services to function

Authenticity services	Communal services	Inspirational & creative services	Identity services	Knowledge (educational) services	Health services	Environmental services	Capital services
5	12	20	25	40	30	60	35

Figure C.6 graph showing the shows the asset degradation after intervention uplift with P10 and P90 values introducing alternative lifetime scenarios in relation to the minimum conditions required for CHC services.

Figure C.6 shows an example of degradation over time with intervention, using a roof as an example. Three scenarios are modelled, one where deterioration is quicker and one where it is slower, affecting how long it takes for degradation to set in after intervention. With P10, degradation takes more than 100 years, where as with P90 values, degradation takes around 80 years. Dashed lines cross the graph showing the minimum asset conditions required for CHC services to function, as discussed in Table C.10.

The cultural viability of a service at any point in time is therefore defined as:

Where:

𝐶𝑡= asset condition at time 𝑡 

𝑇𝑠= threshold condition required to deliver cultural service 𝑠

If the condition of the asset remains above the threshold, the service is assumed to remain viable. When condition falls below the threshold, the model assumes that the service can no longer be reliably delivered.

Using this framework, the model calculates the number of years during which each cultural service remains viable under both the counterfactual scenario (without intervention) and the intervention scenario.

The impact of the CFF intervention is then expressed as the difference between these two trajectories:

𝑌𝑒𝑎𝑟𝑠𝐼𝑁𝑇= years during which the cultural service remains viable with intervention 

𝑌𝑒𝑎𝑟𝑠𝐶𝐹= years during which the cultural service remains viable without intervention

This provides a direct measure of the additional years of cultural activity enabled by the CFF intervention. Table C.11 provides an example based on the example inputs.

Table C.11 Example cultural viability summary showing the difference between CFF intervention and without.

Cultural service	Years viable (no intervention)	Years viable (with intervention)	P10 Years Viable	P90 Years Viable	P10 Years gained	P90 Years Gained	Years gained	P10 Lifetime multiplier	P90 Lifetime multiplier	Lifetime multiplier
Authenticity services	59	91	108	75	49	16	32	1.59	1.10	1.34
Communal services	50	82	97	67	47	17	32	1.83	1.26	1.55
Inspirational and creative services	44	76	89	62	45	18	32	2.07	1.44	1.77
Identity services	40	72	85	59	45	19	32	2.30	1.59	1.95
Knowledge (educational) services	27	64	75	53	48	26	37	3.57	2.52	3.05
Health services	36	69	82	57	46	21	33	2.56	1.78	2.16
Environmental services	0	47	56	38	56	38	47	N/A	N/A	N/A
Capital services	32	67	78	55	46	23	35	2.89	2.04	2.48

By linking asset condition to the continuity of cultural services, the model moves beyond traditional asset-condition analysis and instead focuses on the cultural outcomes that the infrastructure enables. This approach aligns with the Culture and Heritage Capital framework (8.2) where the value of cultural assets is understood through the services and benefits that flow from them.

Outputs

The above heritage science modelling approach will produce two key metrics which can be used to within an evaluation of CFF:

• Number of years of life gained for each asset/ stock.
• Number of years of extension for services produced by the asset/ stock.
• Lifetime multiplier for cultural and heritage services produced by the asset/stock.

Limitations

The model is limited primarily as thresholds for cultural viability will, in many cases, depend on professional judgement. Different practitioners may reasonably take different views about the point at which an asset ceases to support a particular cultural service. This is not a flaw unique to the model; rather, it reflects the reality that cultural significance and cultural functionality are often multi-dimensional and context-specific. The model should therefore be understood as a structured framework for consistent professional judgement, not as a claim to fully objective measurement.

Furthermore, assets within cultural venues are often interdependent. For example, while an intervention may target a specific element such as the roof, the wider building structure may remain degraded and continue to affect the delivery of cultural services. Conversely, repairs to one asset may slow the rate of deterioration in related components.

The model allows for multiple assets to be considered within a single framework. However, further work could explore these interdependencies more explicitly through a systems-based approach.

Annex D: Starter Indicator table

At the request of DCMS, Ipsos have produced a set of metrics based on the theory of change that can support with internal monitoring and analysis. It should be noted that it is not yet possible to populate the full set of metrics as this relies on monitoring returns, which at the point of writing have not occurred. Accordingly, the table should be viewed as an indicative ‘starter set’ of input, activity, and output indicators and is intended for DCMS to further develop, refine, and expand over time.

Level	Metric	Definition	Primary data source
Input	Total grant awarded (£)	ACE grant value per project as per grant agreement	ACE award/grant agreement
Input	Strand (1/2) and project type	Grant strand and high-level works type (fabric/MEP/access/decarb/equipment)	ACE award letter; Full application
Input	Match funding secured (£ and % of total)	Confirmed matched funding at award and subsequent confirmations	Grant agreement (award); Quarterly monitoring (in‑year)
Input	Planned start date (construction)	Planned mobilisation/start on site date	Full application; Monitoring
Input	Planned practical completion (PC) date	Planned completion/PC	Full application; Monitoring
Input	Applicant delivery team time/costs (internal)	Internal client PM time/costs attributable to project	Primary survey (if agreed)
Activity	Works categories delivered	List of work packages delivered (fabric; MEP/HVAC/electrical; access; decarb; technical/production equipment)	Quarterly monitoring
Activity	Procurement route and contract award	Procurement pathway (e.g., open/framework) and award date	Quarterly monitoring
Activity	Contract variations/change orders (# and £)	Number and value of approved variations/change orders	Quarterly monitoring
Activity	Spend to date vs budget (%)	Cumulative eligible spend as % of approved budget	Quarterly monitoring
Activity	Closure/decant periods (dates; partial/full)	Recorded partial/full closures and decant dates during works	Quarterly monitoring
Output	Critical maintenance items addressed (#/£)	Count and/or value of backlog items resolved by the project	Baseline condition survey + End‑of‑project report
Output	Space reopened (m²)	Previously unusable floor area brought back into use (m²)	Quarterly/Annual monitoring; End-of-project
Output	Seats returned to service (#)	Number of seats made usable again	Annual monitoring; End-of-project
Output	Accessibility features installed (list / YN)	New/updated features (e.g., Changing Places, lifts, step‑free routes, wayfinding)	Annual monitoring; End-of-project
Output	Energy/fabric upgrades completed (type; Y/N)	Completion of specified decarbonisation/energy efficiency measures	Annual monitoring; End-of-project
Output	Maintenance plan implemented/updated (Y/N; date)	Whether a PPM/asset management plan is in place and current	End-of-project report
Output	Asset register updated (Y/N; date)	Confirmation that asset registers/O&M manuals updated at handover	End-of-project report

1. Campaign for the Arts. (2024, December 17). English councils have more than halved culture spending since 2010, new CFTA analysis finds. ↩

2. Sigal, S., & Rowson, J. (2024). Institutional transformation: Growth and change in post-pandemic theatre. The Royal Central School of Speech & Drama. ↩

3. This evaluation plan was scoped out and formalised prior to the announcement of additional funding rounds. The addition of further funding increases the scope of implementing quasi-experimental designs (e.g. DiDs) by increasing the sample size of both the treatment and control group – and thus the statistical power. Further scoping work would be required to fully understand how the additional funding affects the recommendations within this report and whether the first funding Round is appropriate for analysis given institutional change. ↩

4. Note: This has changed for future rounds. ↩

5. Please note that in May 2026, Arts Council England published its new Strategic Framework for 2026, which replaces its previous strategy Let’s Create. This alters some of the marking criteria and we expect future rounds of the Creative Foundations Fund to align with this strategic framework and any future ACE strategy. This should be taken in mind by the evaluator when designing their methodology. ↩

6. DCMS (2026), DCMS Culture Priority places ↩

7. Local level impacts should be considered in reference to the list of DCMS Culture Priority places. ↩

8. Arts Council England - Our NPOs and the Annual Data Survey ↩

9. GOV.UK - Registered Charities in England And Wales - Download Page ↩

10. Office for National Statistics - Business Structure Data Base - UK ↩

11. Office for National Statistics - Annual Gas and Electricity Consumption at the meter level - Great Britain ↩

12. Office for National Statistics - Previous Releases - Business Insights and Impact on the UK Economy Statiistical Bulletins ↩

13. TRG Arts - Purple Seven ↩

14. Whilst ACE made recommendations as to whether applicants should progress to the full application stage, they were obliged to consider full application submitted by organisations who were not recommended to apply. ↩

15. This is largely driven by high variability within outcome metrics, small volumes of funded organisations, and only two time points. ↩

16. HM Treasury - HM Treasury (2020) “The Magenta Book: Central Government Guidance on Evaluation”. ↩

17. We note that detailed technical documentation (for example, costed condition surveys, RIBA Stage 3 design packs and lifecycle/PPM schedules) is typically available for larger projects. Many smaller awards will not hold all of these materials. Consequently, the HS model will be developed on the subset of projects with sufficient documentation. This will limit the HS sample size and means findings should be interpreted as case‑study estimates rather than programme‑wide effects. To mitigate this, it is recommended to: (a) use a tiered modelling approach (see below) to maximise coverage; (b) purposively sample to ensure representation across strands, work types and geographies; (c) use conservative standard lifetimes where appropriate; and (d) triangulate HS results with theory‑based evidence and any available administrative data. ↩

18. Grau-Bové, J., Andrews, M. (2025) General collections demography model with multiple risks. Humanit Soc Sci Commun 12, 889. ↩

19. Sagger, H and Bezzano, M (2023) “Culture and Heritage Capital: An Economic Approach”, Department for Digital, Culture, Media and Sport (DCMS). ↩

20. Bond S, and Worthing, D. (2016) Managing Built Heritage: The Role of Cultural Values and Significance, 2nd ed., John Wiley & Sons. ↩

21. UNESCO (2003) “Convention for the Safeguarding of the Intangible Cultural Heritage”, United Nations Educational, Scientific and Cultural Organization. ↩

22. P. Kaszynska, D. Coyle, E. Dwyer, R. Lawton, P. Riganti, S. Watson, M. Dâmaso and Y. Wang (2021) “Scoping Culture and Heritage Capital Report”, Department for Digital, Culture, Media and Sport (DCMS). ↩

23. Sagger, Hand Bezzano, M (2024) “Embedding a Culture and Heritage Capital Approach”, Department for Culture, Media & Sport. ↩

24. Historic England (2025) “Culture and Heritage Capital Taxonomies”, Historic England. ↩

25. Sagger, Hand Bezzano, M (2024) “Embedding a Culture and Heritage Capital Approach”, Department for Culture, Media & Sport. ↩

26. What constitutes existing evidence and what will be considered further evidence collected within v) will depend on the detailed design of the evaluation. ↩

27. Better Evaluation - Contribution Analysis, updated 2021. ↩

28. Collier D (2011) Understanding Process Tracing. Political Science and Politics 44, No 4 ↩

29. DFID’s Approach to Value for Money (VfM) - Department for International Development. Initially, the framework only included economy, efficiency and effectiveness, with equity considerations added on to the framework by the NAO. ↩

30. GVA is the total value of goods and services less costs of intermediate inputs used to create them (e.g. raw materials, parts, services) ↩

31. DMCS - Applications of non-use values in the context of culture and heritage report - 7. Takeaways for the CHC Programme and future research ↩

32. This build on recent DCMS research which explores how non-use values can be robustly included within a CBA: DCMS - Applications of non-use values in the context of culture and heritage report ↩

33. Formally, the L2 is the ratio of the mean squared error for the synthetic control and the mean squared error for the simple average of control units. ↩

34. Nowogońska, B., & Kurowska, A. (2025). Management of the risk of building degradation. Engineering Failure Analysis,179, Article 109801. ↩

35. Pourebrahimi, M., Seyed Rahman Eghbali, & Ana Pereira Roders. (2020). Identifying building obsolescence: towards increasing buildings’ service life. Structural Survey, 38(5), 635–652. ↩

36. Sagger, H and Bezzano, M (2024) “Embedding a Culture and Heritage Capital Approach”, Department for Culture, Media & Sport. ↩

37. RICS (2022) Planned preventative maintenance. 1st edition. London: Royal Institution of Chartered Surveyors (RICS). Effective from 1 April 2022. ↩

38. RICS (2022) Planned preventative maintenance. 1st edition. London: Royal Institution of Chartered Surveyors (RICS). Effective from 1 April 2022. ↩

39. O’Connor, P. D. T., & Kleyner, A. (2011). Practical Reliability Engineering: Fifth Edition. ↩

40. British Standards Institution (BSI). (2015). BS 7543:2015 Guide to durability of buildings and building elements, products and components. London: BSI RICS (2016). Life cycle costing. 1st ed. London: Royal Institution of Chartered Surveyors. Effective from 1 July 2016. ↩

41. British Standards Institution (BSI). (2008). BS ISO 15686-8:2008 Buildings and constructed assets — Service-life planning — Part 8: Reference service life and service-life estimation. London: BSI. ↩

42. Mills, E.D., 1994. ‘Design and Building’. In: E.D. Mills, Building Maintenance and Preservation: A Guide for Design and Management. 2nd ed. Oxford: Butterworth-Heinemann. ↩

43. British Standards Institution (BSI) (2024). BS EN ISO 41001:2018+A1:2024: Facility management. Management systems. Requirements with guidance for use. London: BSI. ↩

44. British Standards Institution (BSI) (2012). BS 8210:2012 Guide to facilities maintenance management. London: BSI. ↩

45. Pourebrahimi, M., Seyed Rahman Eghbali, & Ana Pereira Roders. (2020). Identifying building obsolescence: towards increasing buildings’ service life. Structural Survey, 38(5), 635–652. ↩