Research and analysis

Review of local civil society and community data: summary report

Published 11 July 2025

DCMS commissioned Ipsos UK, the University of the West of Scotland and the Centre for Regional Economic and Social Research (CRESR) to conduct a review of civil society and community datasets, with a focus on assessing the quality and usability of geographical data. The aim of the review was to support DCMS in enhancing its approach to local data. 

This report presents insights from 17 interviews with data and sector experts and a review of 45 civil society and community data sources.

Key findings

Data landscape

Overall, stakeholders described the current civil society and community data landscape in the UK as fragmented and patchy, with data available from numerous sources that are not well joined up. Civil society datasets often include precise geographic information for organisations (e.g., postcode), but there are challenges in relation to coverage (e.g., non-charitable organisational forms and smaller organisations missing from administrative datasets). Community datasets, mainly large-scale and long-running social surveys, are often captured at higher geographic units like Government Office Region or Local Authority District. 

Challenges and limitations

The data review and interviews with stakeholders highlighted challenges related to civil society and community data across three broad themes:

1. The accessibility of data: the research highlighted a number of barriers to access for key data sources, including unpublished data held by government departments or data sources held behind paywalls.

2. The usability of data: a lack of standardisation and the complexity of many datasets meant there were often technical barriers to using data more widely, compounded by skills gaps in the sector and in government. 

3. The completeness of data: there was significant variation in terms of the timeliness and geographical granularity of reviewed datasets. There were also a number of key data gaps identified including a lack of data on organisations’ areas of operation (and impact of the “headquarters effect”) and more limited data on rural areas. 

Looking ahead

Across the sector, the interviews found that efforts are taking place to improve civil society and community data use, including making improvements to geographical data. This included:

• Resources and platforms that showcase data in formats that are more easily usable for non-experts through dashboards and data visualisation.

• The use of Artificial intelligence (AI) to help data owners to improve efficiencies in data extraction and aggregation.

• Publication of large-scale social surveys at lower geographic units.

• Collaboration between sector organisations to improve data quality and coverage and support greater standardisation in certain metrics.

Throughout interviews, participants raised several suggestions for DCMS to help address challenges related to civil society and community data, including: 

1. Convening and stewardship with the sector: there is a role for DCMS to convene key stakeholders across the sector to improve data collection and analysis, as had been done previously through DCMS-led advisory groups. This could include co-designing a set of agreed protocols which could act as standards for use across data collection.

2. Encouraging good practice across government: DCMS could take a lead role in encouraging greater data sharing from government departments in relation to civil society and community data. 

3. Investing in data infrastructure: there is a role for DCMS to invest in data infrastructure, AI, and data dashboards to make civil society and community data more accessible and usable.

1. Introduction

1.1 Background

Through its Missions and Plan for Change, the government has committed to ensuring “every nation and region realises its full potential”^{[footnote 1]}. This includes how the government works with the voluntary, community and social enterprise sector (VCSE), with the Civil Society Covenant aiming to develop “a new model of civil society […] with collaboration that joins up local services and supports people in all parts of the country”^{[footnote 2]}. The focus on place requires a renewed emphasis on quality data at local levels, and builds on existing work across government to improve the geographical granularity of government datasets and official statistics. 

To support these efforts within DCMS, the department commissioned Ipsos UK, the University of the West of Scotland and the Centre for Regional Economic and Social Research (CRESR) to conduct a review of civil society and community data. This builds on a number of wider explorations of VCSE sector data including the Law Family Commission’s review of social sector data and a feasibility study for creating a Civil Society Satellite Account in the UK, the VCSE Barometer, and the Improving Access to and Use of Organisation-Level Data on the Third Sector and Civil Society project ^{[footnote 3][footnote 4][footnote 5][footnote 6]}. What distinguishes this review from other similar work is the focus on identifying and analysing the degree to which civil society and community data can be used to study local conditions.

The Civil Society and Youth Directorate (CSY) within DCMS wanted to understand how to enhance their approach to local data to identify where local needs are greatest, better target place-based programmes and more accurately assess geographical variation and change over time. The key objectives were to:

• Review the quality and relevance of existing local data on civil society and community needs.

• Generate recommendations around the direction of travel for future data improvement activity.

For this project, DCMS defined local data as being any dataset that provides statistics for England or the UK at sub-regional levels (below ITL1 ^{[footnote 7]}, including local authorities, wards, MSOA, LSOA, and Output Areas). The data in scope should cover at least one of two main areas:

1. Civil society sector activity: This includes general demographic information about the Voluntary, Community and Social Enterprise (VCSE) sector, such as size, composition, and activity. It also covers specific policy areas like youth, philanthropy, and social and community infrastructure.

2. Community need and resilience: This includes headline economic and social indicators (e.g., deprivation and wellbeing) and information focused on community connections, such as volunteering, community participation, loneliness, belonging, place attachment, and community cohesion.

1.2 How to read this report

This report provides a summary of insights from a data review of existing civil society and community sources, a series of 17 depth interviews and two workshops with the DCMS CSY team to inform our design and reflect on the key findings. The research took place between December 2024 and April 2025, meaning it reflects the availability and quality of data sources at this time.

We have included charts and tables throughout the report to illustrate findings from the data review. These are based solely on analysis of the 45 data sources reviewed. 

Interviews took place with experts from the VCSE sector, academics and policy experts working on local civil society and community data. Please note, we did not interview frontline organisations or the wider public as part of the external consultation. This means the insights do not reflect their perspectives including on the feasibility of different activities. 

Throughout the report, the people that took part in the interviews are referred to as stakeholders or participants. We have included verbatim comments, but these have not been attributed to individuals to maintain the anonymity of the stakeholders who took part. We have also removed commentary on specific datasets to maintain anonymity. 

1.3 Glossary

Acronym	Meaning
AI	Artificial Intelligence
CIC	Community Interest Companies
CIO	Charitable Incorporated Organisation
ITL	International Territorial Levels
LAD	Local Authority District
LSOA	Lower layer Super Output Area
MSOA	Middle layer Super Output Area
OA	Output Area
VCSE	Voluntary, Community and Social Enterprise
Ward	An administrative division of a city or borough that typically elects and is represented by a councillor or councillors.

2. The current civil society and community data landscape

2.1 Data use across government

Interview participants were optimistic that there was interest from policymakers in harnessing datasets, particularly those on smaller organisations within the civil society sector. This is within the context of the Government’s increasing focus on using more local data to inform decision-making. However, there was a perception that the government’s current use of civil society and community data is limited. They emphasised how engagement with the wider data landscape was not perceived to be particularly strategic or proactive, and there was an opportunity for greater join-up between different departments.

Participants also mentioned they felt government favoured ‘big data’ from large quantitative datasets over qualitative information. Despite this, there was a feeling that statistics are not always used effectively or designed to capture the most useful domains.

“I wouldn’t say that they always use stats to best effect, but more importantly, I think sometimes they look for a stat or look for that type of data when there really isn’t any need for it, and sometimes use the lack of that data as an excuse for not taking action.” 

Stakeholders suggested that data is not always understood well within government due to skills gaps. This can result in stakeholders spending time explaining different data sources to policymakers or sometimes resulting in a negative impact on policy. 

“We will often be asked to help support government departments in their thinking in terms of rollouts of programmes and areas to potentially focus on, and that’s because I think they’re one step removed from the data.”

That said, stakeholders generally felt that it was important that government continued to engage with the sector around data, including consulting non-government experts (see section 4.2).

2.2 Findings from the data review

The data review evaluated 45 sources of data identified as in-scope. This included:

• 23 community data sources 

• 22 civil society data sources 

These data sources were associated with 41 datasets in total: 19 relating to the Community domain, and 22 to Civil Society. Some data sources did not yield any datasets (e.g., research projects yet to produce them), and some produced more than one dataset (e.g., UKCAT Charity Classification project). Table 3.1 below provides a high-level summary of how the themes and datasets map to the Community and Civil Society domains.

Table 1 . The number of civil society and community datasets, by key VCSE themes ^{[footnote 8]}

	VCSE funding	Infrastructure	Youth	Community connections	Community capacity building
Civil Society datasets	20	16	1	1	1
Community datasets	0	1	8	4	11

Note: Datasets can map to more than one theme, hence why row totals sum to more than the total number of datasets in each domain. Three datasets did not map to any of these themes.

Few sources or datasets brought together metrics related to both civil society and community themes, highlighting how domains tend to be handled separately. The exceptions to this are some of the large-scale social surveys such as Understanding Society or the Longitudinal Study of Young People in England which contain a range of measures relating to community characteristics and volunteering behaviour, as well as the Community Needs Index and Thriving Places Index which combine aggregate information on civil society organisations with community characteristics relating to social cohesion.

Interview participants described the current civil society and community data landscape in the UK as fragmented and patchy, with data available from numerous sources that are not well joined up. Although there are rich data sources available, the scattered nature of the data presents a challenge for attempts to build a full picture of the sector and community needs. 

“It’s patchy, it’s atomised, it’s held in multiple different places.”

“Collectively, the information is probably there, but you have to go digging for it and there’s no complete picture.”

2.3 Civil Society data

The civil society organisation datasets vary considerably in their coverage of organisation types (e.g., social enterprises, charities, sports clubs). Stakeholders described how part of the challenge stems from the lack of a clear, consistent definition of what constitutes ‘civil society’. Perspectives vary on which types of organisations and activities should be included within this term. This is compounded by the informal nature of much of the sector, such that many smaller civil society organisations do not appear in administrative or regulatory registers. However, there are several datasets that attempt to provide as broad a coverage in this regard as possible, in particular, Find that Charity, National Lottery Grants data, and 360Giving.

Civil society datasets often include precise geographic information for organisations (e.g., postcode) which can be linked to a postcode lookup file and subsequently to many of the UK geographies of substantive interest (e.g., OA, LSOA, MSOA, LAD, WARD).

2.4 Community data

The community datasets, mainly large-scale and long-running social surveys like the UK Household Longitudinal Study, are important sources of information on volunteering, informal help/care, general wellbeing, and trust/satisfaction with local area or neighbours. The unit of analysis in most of these datasets is adults or households, though a small number capture information on young people (typically younger than sixteen years old). For example, the UK Household Longitudinal Study captures information on both adults and young people.

In general, the community data is often captured at higher geographic units like Government Office Region, Country, or Local Authority District – see Figure 1 for a breakdown of geographic coverage by data domain. Community datasets have geographic variables present more often than civil society datasets. However, the latter are more likely to have postcode information, which can then be linked, through appropriate lookup files, to a wide range of geographic units. There is little-to-no potential to link community data to other geographies due to a lack of suitable linkage variables. This is understandable given the need of many large-scale social surveys to preserve the anonymity of individual respondents.

Figure 1. Distribution of geographic units by data domain (from smallest to largest geographical unit)

Geographical unit	VCSE	Community
Postcode	13	0
Output Area	0	2
LSOA	0	4
Ward	1	1
Local Authority District	1	4
Region	1	5
Country	0	3
N/A	6	1

Number of datasets = 41

There are clear differences in the availability of datasets by domain: Figure 2 demonstrates that 15 of the 22 civil society datasets are available as open data, compared to 3 of 19 community datasets. The latter usually require registration with an appropriate data controller (e.g., UK Data Service, Office for National Statistics Secure Research Service). In many cases, the only post-registration task is to complete a project application. Access to the data for non-commercial purposes is provided in a reasonably short time frame. 

Figure 2. Accessibility of datasets across civil society and community data

Access status	VCSE	Community
Closed: By Request	1	0
Closed: Inaccessible	3	0
Closed: Paywall	2	1
Closed: Safeguarded	1	5
Closed: Secure Setting	0	7
Closed: Special Licence	0	3
Open	15	3
N/A	0	1

Number of datasets = 41

2.5 VCSE policy themes

The number of variables/fields of interest to VCSE policy themes areas in the civil society datasets is lower than in the community datasets. This reflects how civil society datasets are often derived from administrative data containing limited numbers of variables/fields, while the community data are often social surveys with a large battery of questions for each topic (e.g., on volunteering). 

Some themes are covered more extensively by the reviewed datasets than others (see figure 3 below). Data relating to ‘VCSE funding’ (e.g., information on grants, local government funding) and ‘Community connections’ (e.g., volunteering, wellbeing) are covered by over one third of the datasets, with ‘Community capacity building’ (e.g., social connections and trust, satisfaction with local area) the least covered.

Figure 3. Coverage of VCSE policy themes across civil society and community data

Policy theme	Number of reviewed datasets
VCSE funding	20
Infrastructure	17
Community capacity building	12
Community connections	9
Youth	5

Number of datasets = 41

The smallest geographic unit in the datasets varies considerably by theme. As shown in figure 4 below, VCSE funding and Infrastructure are the themes where the datasets have either the smallest unit (postcode) or no geographic information at all (N/A). On the other hand, the Youth and Community capacity building themes tend to have information on government office region or local authority district. It is a more mixed picture for the Community Connections theme, with some datasets covering small geographies (output area) and others relating only to country level.

Figure 4. Geographic units in datasets by VCSE theme (% of datasets)

Policy theme	Postcode	Output Area	LSOA	Ward	Local Authority District	Region or country	N/A	Total
VCSE funding	63	0	0	5	0	5	26	99
Infrastructure	65	0	6	6	0	6	18	101
Community connections	0	22	22	0	0	44	11	99
Youth	0	20	0	20	20	40	0	100
Community capacity building	0	8	25	8	25	33	0	99

Number of datasets = 41. Rounding means not all categories sum to 100%.

3. Challenges and limitations

The data review and interviews with stakeholders highlighted several key challenges related to civil society and community data. While the granularity of geographic data was important to participants, they explained how this was part of a wider set of challenges that limit the usability and value of existing sources. This covered challenges across a data use journey including: 

1. The accessibility of the data
2. The usability of the data
3. The completeness of the data 

In addition to these limitations, interviewees referred to a wider lack of data skills and expertise across the VCSE sector especially in smaller organisations, reflecting similar perceptions of skills gaps in government as described above. This was seen to be exacerbated by capacity constraints that impact their ability to collect, share and analyse data. With limited staff and budgets, many struggle to deliver their core services, and are unable to dedicate time and resources to data management and interpretation. As such, where advanced analysis is taking place within the sector, it tends to be concentrated in larger organisations. Stakeholders emphasised how this places limitations on the development of data sources within the sector, with a reluctance to put additional burden on small organisations. 

  “I’m very aware that we can’t ask for the voluntary and community sector to give up too much of their time.”

“I think there’s a big question for me which is about the kind of ethics of care to the sector. Just because they’re community driven, you can’t expect them to do everything.”

Interviewees emphasised the importance of recognising the potential burden of additional requests on small organisations, instead developing a reciprocal approach to data collection and use. They described how approaches to data collection and insight generation should be mutually beneficial for participating organisations. This may be in the form of benchmarking data, individualised reports, or other decision-making support resources. 

“It’s often top-down rather than bottom-up, and this is hugely tricky. So, this is not an easy thing at all, to collect data. But if we started by thinking about, ‘what is it that an average community voluntary and community sector organisation needs in its data?’, ‘how do we provide them with that kind of information?’ rather than thinking about, at the top-down level, ‘what do we need to know about those organisations?”

Participants also perceived a general under-appreciation of the value of data in the sector. Datasets are often seen as a ‘nice to have’ rather than a strategic resource to generate insights and drive decision-making. Thus, there were calls for sector bodies and funders to play a greater leadership role in promoting the importance of data and evidence.

3.1 How easily can data be accessed? 

To make use of civil society and community data, researchers and policymakers need to be able to easily access relevant sources, while maintaining data security and safeguarding processes. The research highlighted how there are often barriers to access for key sources. This includes:

• Several data sources held by Government departments which are not made public. This includes HMRC’s collection of Gift Aid information, although summary statistics are published. ^{[footnote 9]} Stakeholders found this frustrating as access to these sources would help build a more comprehensive picture of civil society. 

• Data sources held behind paywalls. While participants understood the need to charge for access to some data, they felt this created a barrier for the sector. For example, the data review highlighted how the Thriving Places Index, NCVO Almanac Accounts Data, and the Community Needs Index all charge an access fee. 

• Large-scale social surveys that are available for use by researchers through special licences. The application forms for access can be intimidating to a stakeholder unfamiliar with the use of these datasets, especially those only available through the ONS’ Secure Research Service / Integrated Data Service. This can dissuade consideration of them for use in civil society research.

• Data sources that require users to meet exclusive criteria or to adhere to permissions in how they store and use the data, which users often need to be highly resourced for.

“There’s potentially a lot of data that’s sort of hidden or underutilised … So I think data about charities and other civil society organisations that would exist in various administrative data sets.”

“There’s some data that’s off limits, which can be quite frustrating.”

3.2 How easy is the data to use? 

Once a dataset has been accessed, there can be challenges for collating, managing, and analysing the information available. This includes technical barriers resulting from how data is stored or shared, and a lack of standardisation across datasets which limits the ability to link or compare data. 

3.2.1 Technical barriers

Participants consistently reported that financial data is stored in individual PDFs creating barriers to collating and using the data. They felt that aggregating this information would be helpful to provide a more accessible and comparable format. However, there was acknowledgment of the challenges of rectifying this, due to a lack of capacity and resource.

“Financial accounts are the number one area or dataset where you can anticipate, understand, and assess risk and yet we can’t actually get to it.”

Stakeholders also felt that datasets are often complex and hard to use. They recalled having to download data in numerous spreadsheets or in block, which they felt were impenetrable unless they held the skills of a data expert. This was also highlighted in the data review. For example, the Free Accounts Data Product by Companies House, which contains the financial returns of CICs and other incorporated forms of civil society, presents a technical challenge in downloading large files (often 2GB or larger). It also requires the user to apply programmatic approaches to processing the data (e.g., a bespoke Python package for converting .ixbrl format to .csv format). These technical barriers meant stakeholders felt data was not always being used effectively, as those without the skills, training or knowledge are often unable to work with or interpret data. 

3.2.2 Lack of standardisation

The lack of standardisation in the definition of key terms, consistent question wording, geographical boundaries, and publication approaches makes it difficult to compare and link different data sources. For example, stakeholders noted that there are several regulators who collect relevant data on civil society and community needs, but that they collect it differently. 

“I’d definitely say that fragmentation and the fact that it’s quite difficult to get the different data sources to talk to each other is a hindrance.”

Participants observed a lack of consistency in geographic coverage within datasets, for example varying between local authority, regional or ward level. This makes it difficult to compare different parts of England using data across sources or make comparisons over time if geographical boundaries change. Participants felt that geographic data needs to be as granular as possible to gain a comprehensive understanding of a place.

“I think there’s a challenge around what level the data is available at. There are some things that are only available at local authority, others only available at regional levels. And when we’re trying to understand the specifics of civil society at place level, I think that can be a challenge if the data isn’t available at that level specifically.”

Variations in accounting conventions and how information is recorded in financial documents also reduce the comparability of data as it cannot be guaranteed that information is always being reported consistently across different organisations (or within the same organisation over time). This includes differences in how organisations account for income and expenditure. There is also often a lack of distinction in how grants and contracts are recorded.

3.3. How complete is the data? 

Once a dataset has been accessed and captured in a useable format for analysis, limitations in the data may remain as a result of the chosen methodology, frequency of data collection and resulting gaps in the data available. 

3.3.1 Methodological limitations

Participants noted concern about the methodologies some data sources used, in particular those that rely on self-reporting. They described how this can reduce the accuracy and reliability of a source due to: 

• Challenges with recall.

• Inappropriate respondents (e.g., those without full knowledge of an organisation).

• Voluntary nature of participation in data collection efforts, which can lead to incomplete datasets or potential biases.

“I think there is something about how useful is the data? So it’s that the categories might be useful, but if everyone’s ticking every box, then it then no longer becomes as useful.”

Although there was a recognition of the time required for qualitative research, participants suggested greater use of qualitative techniques could increase the nuances and explanatory power of existing data sources. For example, they felt it could help give a greater understanding of how services are delivered locally, especially where quantitative data showed gaps in services.

Other potential methodological limitations included: 

• Small sample sizes. Results may be less credible due to a larger margin of error, and sub-group analysis becoming meaningless with limited sample sizes. 

• Questions regarding panel quality. For example, are the right people being sent social surveys to complete.

• The use of estimates and proxies throughout data sources such as indices.

Stakeholders described how they could find ways to overcome or recognise these methodological limitations. For example, drawing on multiple sources to triangulate and link data, conducting additional data collection to fill gaps where resources allowed, or recognising the limitations of the data in any research outputs. Interviewees questioned whether those without knowledge of these challenges would factor this into their analysis or reporting, potentially resulting in misleading insights. 

3.3.2 Frequency of data collection

Participants described how key sources can be ‘outdated’ by the time they are published, sometimes reflecting the nature of how annual returns are submitted, and social surveys are processed and released by data controllers. Other sources also take considerable time to be published due to the scale of data collection. Stakeholders who have worked in the sector for a long time suggested this has become more of a challenge recently. They described increased demand for ‘live’ data on different regions and communities, during events like the Covid-19 pandemic. 

The lack of timeliness is confirmed using analysis for the data review: all except one of the community datasets had a most recent collection date of 2023 or earlier. The situation is better for civil society datasets, where 12 of the 21 datasets reviewed had been collected in 2025. Although some records (e.g., National Lottery grants, charity registrations) are more up to date than others, there is often a significant gap between when a dataset is made available and the time period the data refers to. 

Figure 5. Timeliness of published data across civil society and community data

Year	VCSE datasets	Community datasets
2015	1	2
2019	0	2
2020	0	1
2021	1	1
2022	2	3
2023	1	9
2024	4	1
2025	12	0
N/A	1	1

Number of datasets = 39

3.3.3 Data gaps

Stakeholders highlighted several gaps in current sources that limit the usability of data, often focused on the organisational definitions used in civil society sources. This included gaps around: 

• Community Interest Companies (CICs). CICs are not required to register with the Charity Commission so are not found within this data. This has consequences for the comparability of social enterprises and charities, and much less data are collected on the former by Companies House than the latter by the Charity Commission. 

• Unregistered charities and Charitable Incorporated Organisations (CIOs). Unregistered charities represent a set of organisations that are charitable in law but do not have to register with the Charity Commission if their annual gross income is less than £5,000. These are a subset of “under the radar” CSOs active across the UK on which only sporadic, ad-hoc data are collected on e.g., through receipt of grants from National Lottery or charitable foundations.

• Smaller, unregistered community groups and voluntary organisations that may not meet criteria or thresholds for inclusion or do not want to be discoverable.

• Other types of organisations that deliver civil society functions but may not meet specific definitional criteria (e.g., churches or libraries).

“It’s almost as if there’s a big missing chunk of the sector that’s actually really difficult to get at.”  

One mitigation suggested was to require all organisations to register with the Charity Commission regardless of size. This would result in a more nuanced mapping of the sector. However, stakeholders expressed concerns about asking for more information from small organisations with limited resource, and felt this would require a significant legal and administrative overhaul to reach organisations who currently are unregistered.

“I think exempting small organisations from submitting annual reports isn’t right. I understand why because I mean it’s deregulatory but actually, I don’t think it’s helping to defend that part of the sector and it’s, and it’s not helping to understand small charities.”

Stakeholders also recognised geographical limitations of current data sources including: 

• A lack of rural data, with gaps limiting understanding of the dynamics of rural communities, compared to urban environments. 

• The impact of the “headquarters effect”, where organisations such as charities are only required to register their main office location which may not reflect where its work takes place. 

The “headquarters effect” is a longstanding issue in geographic analysis of civil society organisations which has a number of implications: 

• The location may refer to the address of an individual and not the organisation itself.

• The address does not refer to an operating location but rather an office / postal location.

• The reporting of a single address likely undercounts the number of locations where an organisation operates.

• General difficulties with the accuracy of this information over time (i.e., it is often only reported once per year). 

“The area of operation data is a bit off. it’s not that it’s not easy to work through, it’s not intellectually difficult. It’s just that, I mean you can’t take this stuff at face value. The geographical data is sort of not very good.”

Stakeholders said they were taking efforts to rectify this. For example, the Charity Commission for England and Wales requires charities to report one or more UK and/or overseas locations where they operate. However, this in part relies on organisations accurately sharing their areas of operation. There are wider initiatives that have sought to address this issue: 

• The “Improving Access to and Use of Organisation-Level Data on the Third Sector and Civil Society” research project has created a linkage between a comprehensive list of more than 300,000 UK civil society organisations and the Business Structure Database (BSD).^{[footnote 10]} The BSD is a snapshot of a subset of data from the Inter-Departmental Business Register (IDBR) containing information on the operating location of businesses in the UK.^{[footnote 11]} This includes incorporated civil society organisations and therefore offers the potential to analyse where these entities operate (e.g., location of charity shops) and not just where they are registered.

• 360Giving, which provides access to grants made to organisations and individuals by UK grant making bodies (including Government departments), recently amended its collection of geographic data by adopting a “Best Available Data” approach. This ensures that grants are assigned to the location where the funded work will be conducted, and if not then to the location of the beneficiaries of the funded work.

The “headquarters effect” is a known issue but it is important not to overstate it. For example, the majority of registered charities report operating within a single local authority and therefore analysing data at this geographic scale can often be valuable for understanding patterns in the registration and dissolution of these organisations.^{[footnote 12]} In addition the nature of charitable / social purpose activities has changed in recent years, with a greater variety and number of services provided online. As such, the analytical aim will determine whether the “headquarters effect” is a minor or significant limitation of understanding where civil society organisations operate.

4. Looking ahead

4.1 Emerging trends and innovations

Across the sector, our interviews found that efforts are taking place to improve civil society and community data use, with a focus on increasing the usability by filling gaps within current datasets. This suggests there may be an opportunity for efforts to improve the accessibility of existing sources. 

4.1.1 Improvements to the usability of datasets

Artificial intelligence (AI) is helping data owners to improve efficiencies in data extraction and aggregation. Stakeholders spoke of using AI to extract granular financial information from sources like the Charity Commission, where they previously would have extracted data manually. This can increase the speed of updating records, and leaves resource for other improvements to data, such as sample sizes.

Efforts are ongoing to showcase data in formats that are more easily usable for non-experts through dashboards and data visualisation. For example, 360 Giving’s GrantNav and GrantVis platforms allow users to search by location to see how grants are distributed. This includes being able to see what funders operate in the area and details on the types of organisations receiving funding, including their income levels and the age of the organisation.

Additional efforts are underway to standardise factors like data collection and impact measurement. Groups have facilitated efforts to agree on shared measures by developing common frameworks and metrics. For example, related to developing a standardised approach to capture demographic information about young people. It has been noted that efforts are still premature, and discussions tend to move slowly. However, these initiatives have the potential to help organisations to standardise their data collection methods and demonstrate impact more uniformly.

4.1.2 Improvements to the comprehensiveness of datasets: capturing data at smaller geographic units and funding flows

A number of the large-scale social surveys have mapped their data to lower geographic units than originally collected. For example, The UK Household Longitudinal Study now links its survey responses - which include information on volunteering, loneliness, youth employment, and neighbourhood cohesion – to census output area (OA), lower super output area (LSOA) and middle super output area (MSOA). The Participation Survey now generates boosted samples every three years so that meaningful analyses of survey responses can be conducted at local authority level (LAD). The Community Life Survey has also been published at the local authority level for the first time.

There are a number of innovative initiatives aiming to capture the funding flows to civil society organisations in more granularity and volume. The UK Grantmaking annual report is a large-scale data collection and analysis collaboration between 360Giving, Pears Foundation, The Association of Charitable Foundations (ACF), The Association of Charitable Organisations (ACO), UK Community Foundations (UKCF) and London Funders.^{[footnote 13]} The report gathers information on grants to UK civil society organisations from a broad array of sources including government, corporate foundations, donor-advised funds, companies, lotteries and charitable foundations.

Capturing funding flows from local and central government has traditionally been difficult to do due to technical and cost issues. A number of projects have sought to address this gap. The “Civil society data partnerships: open data, grants data and the funding base of the third sector 2004-2015” research project used computational methods (e.g., web-scraping and text reconciliation) to gather, clean and link data on local authority funding (grants and contracts) to civil society organisations. The “Improving Access to and Use of Organisation-Level Data on the Third Sector and Civil Society” project has addressed a similar gap by extending the data collection to include all public procurement contracts listed on the UK Government’s Contracts Finder database. The result is a single spreadsheet containing tens of billions of pounds of local and central government contracts awarded to civil society organisations.

Other organisations are developing innovative approaches to data collection and sharing, which aim to be more reciprocal and community driven. For example, work is being conducted to develop mutually beneficial ways of incentivising organisations to share their data and, as a result, build a more complete dataset. This may include giving individualised reports for participating VCSE organisations. 

“We’ve just launched […] individualised reports. So an organisation, by filling it in, they get something back as well, it matches their response to the sector as a whole.”

Stakeholders thought that the next trend in civil society and community data use could involve analysing data to map and predict trends. They suggested that by using data in a more proactive way, policymakers could identify challenges in advance or model different scenarios based on projections of current datasets. Stakeholders have already begun exploring the use of Multilevel Regression and Post-Stratification (MRP) data, predicting figures based on national-level results. There is optimism that, by monitoring the effects of policies and other social indicators, predictive techniques could foresee things like public unrest.

“Instead of just getting data for 2025 we can also say, but this is what the state of the sector will look like in terms of finances, workforce and volunteering in 2026 and 2027.”

4.2 Suggestions for DCMS

Throughout interviews, participants raised several suggestions for DCMS to help address challenges related to civil society and community data. 

4.2.1 Convening and stewardship with the sector

Participants believed that DCMS should take a leading role in convening key stakeholders across the sector to improve data collection and analysis. They cited previous examples of DCMS-led advisory groups that included academics and sector experts. Stakeholders expressed a willingness to take part in an exercise like this again, and stressed that DCMS’ leadership is crucial for driving meaningful change. Participants felt that without the presence of DCMS, efforts to improve data will continue to be fragmented.

“I think DCMS, or central government, undervalues or underappreciates the value of just its convening power […]  if DCMS said we want to do something and we want to bring everyone together, people would jump at the chance.”

“There was the data and evidence working group that existed beforehand. It absolutely made everything that I did so much easier. I had a network group of contacts who we could talk to about research.”

Likewise, participants believed that DCMS should provide stewardship for civil society and community data. They suggested DCMS could partner with stakeholders to create a set of agreed protocols which could act as standards for use across data collection, or a common set of core questions to repeatedly collect data on, to collect a set of indicators which signify the health of the sector. Similarly, they argued DCMS should encourage the collection of granular geographic data to support place-based policymaking, such as through the Community Life survey. 

“I wish someone would sort of come out with a manifesto that said, this is what civil society data should look like. These are the ten core indicators that we should actually invest in spending money on to make sure that they’re really good quality, they’re robust and reliable. Then these are the next ten that are the nice to haves and we might collect them every other year.”

4.2.2 Encouraging good practice across government

Participants suggested DCMS could take the lead role within government to encourage collaboration and greater data sharing with stakeholders. This includes:

• Influencing the accessibility of data sources that are held by the Government.

• Working with stakeholders to collate a comprehensive list of key data sources and where they should be used to encourage good practice.

This could help to develop a stronger data landscape, which in turn could lead to better informed policymaking. Participants felt that DCMS was best placed to lead this effort within government, due to their responsibilities for civil society and volunteering.

“Their role is setting out, bringing together a group of people to establish a vision and, a use case, a vision for civil sector data and understanding its uses and where it would be most valuable. And then using their position as a central government department to try and influence and shape that, but also supporting researchers to make use of it in a way that is complementary.”

Likewise, DCMS could help with capacity building within VCSE organisations in their data collection efforts. This could include providing support to organisations to reduce the burden of data collection and help improve skills for data collection, so that there is better good practice within the sector.

4.2.3 Investing in data infrastructure

Participants felt that there was a role for DCMS to invest in data infrastructure to make civil society and community data more accessible. This includes making it easier to access datasets without incurring high costs. The data review identified a number of datasets that are only available commercially, although often summary tables for these sources are available.^{[footnote 14]} One option could be to offer a trial / sample of the full datasets so the sector could explore their utility and value for their own projects. This could include co-ordinating efforts to purchase access to the datasets and make them available for use by approved stakeholders, perhaps through a bespoke multi-use licence.

Likewise, there were suggestions on how DCMS could help to increase the usability of key datasets. This includes helping to develop new data dashboards or funding data collection efforts by civil society stakeholders to develop their own. These tools could present key metrics in a more accessible and interactive manner, serving as a way to address skills gaps and encourage greater use across the sector and government. Participants pointed to existing dashboards like 360 Giving, MyCake, and those run by regulators like Sport England, as potential examples to replicate. 

While acknowledging potential concerns regarding new technology, some participants argued that there is a role for DCMS to lead on using AI to help manage civil society and community data more effectively. Stakeholders said they would value support with: 

• Using large language models and AI to process and examine vast amounts of civil society data held within complex datasets.

• Collating information on currently hard to access files e.g. financial PDFs more efficiently.

Likewise, participants suggested there could be cost-effective solutions for DCMS to support data sources held by other organisations to make them more effective. This includes: 

• Looking at ways to improve panel quality.

• Applying boosts to surveys.

• Exploring where additional qualitative data collection could be used to provide local nuance.

• Developing an approach to capture data on organisations like CICs and currently unregistered charities.

1. No. 10, Plan for Change: Kickstarting Economic Growth ↩

2. Department for Culture, Media & Sport, 2024, Civil Society Covenant Framework. ↩

3. The Law Family Commission on Civil Society; Pro Bono Economics (2023, January). Unleashing the power of civil society. The Law Family Commission on Civil Society. ↩

4. Franklin, J.; Jemal, J.; Kenley, A.; Larkham, J.; Martin, J. (2024, November 13). A feasibility study for a Civil Society Satellite Account. Pro Bono Economics & Economic Statistics Centre of Excellence for the Department for Culture, Media, and Sport. ↩

5. Nottingham Trent University, VCSE National Data and Insights Observatory. VCSE Barometer Survey. ↩

6. Mohan, J.; McDonnell, D.; Clifford, D.; Rutherford, A.; Rahal, C. (2022, September – 2025, September). Improving Access to And Use of Organisation-Level Data on the Third Sector and Civil Society. University of Birmingham, Third Sector Research Centre. UK Research and Innovation. ↩

7. International Territorial Levels (ITL) ↩

8. These themes were agreed with DCMS during the project scoping phase. ↩

9. HM Revenue & Customs publish summary UK charity tax relief statistics annually. ↩

10. The University of Birmingham (2024, December 12). ONS SRS Metadata Catalogue, dataset, Civil Society Organisations - UK ↩

11. Integrated Data Service (1997, January 1 – 2022, December 31). Business Structure Database, UK. ↩

12. McDonnell, D., Mohan, J., & Norman, P. (2020). Charity Density and Social Need: A Longitudinal Perspective. Nonprofit and Voluntary Sector Quarterly, 49(5), 1082-1104. ↩

13. UK Grantmaking 2025. ↩

14. For example, NCVO publishes summary tables of their annual Almanac data. See 2024 data. ↩