Research and analysis

Barriers to data sharing between DLUHC and local councils

Published 14 August 2023

Applies to England

Foreword

The Department for Levelling Up, Housing and Communities (DLUHC), through the Better Outcomes through Linked Data (BOLD) programme, commissioned research exploring local council perspectives in relation to data sharing and linking.

DLUHC has initiated several projects that aim to enrich understanding and shape policies for vulnerable groups by linking data from local councils with data held by central government. While some local councils have already participated in such projects, DLUHC acknowledges that other local councils may face barriers preventing them from signing up to agreements to share personally identifiable (PI) data required to enable linking and have lower awareness of the benefits of data linking.

Therefore, DLUHC commissioned Softwire Ltd to conduct research to identify and understand these barriers, make recommendations on how to address them and facilitate greater participation in future data sharing and linking initiatives. The research also set out to identify what outputs or products resulting from data linking would be considered most beneficial at the local level, to increase the impact of future projects. Finally, DLUHC wanted local council feedback on preferences for central government engagement on data sharing initiatives related to people experiencing homelessness and rough sleeping. As part of this research, DLUHC staff experiences of running data sharing projects were considered for additional context.

The research used a multi-stage qualitative approach through which Softwire Ltd conducted desk research and interviews with key DLUHC stakeholders and with employees in a variety of job roles from 11 local councils across England. Softwire Ltd also ran workshops where participants (from both DLUHC and other government departments) shared their ideas and potential solutions.

The key learnings and recommendations in this report will help DLUHC and local councils to tackle barriers to data sharing practices and facilitate better policy and service provision. DLUHC remains committed to exploring solutions to enable better data usage and communications across central and local government. We are committed to continuing to develop our evidence base on the causes of and solutions to homelessness and rough sleeping, working with other government departments, local councils and expert advisers and charities across the sector to facilitate change.

Stephen Aldridge
Chief Economist & Director for Analysis and Data
Department for Levelling Up, Housing and Communities

Acknowledgements

We would like to thank the many people who generously gave up their time to take part in this research.

We would like to extend our thanks and gratitude to the 11 local councils and the third sector organisations who participated in our research, whose contributions are paramount to the ongoing development of policy and services beneficial to all involved.

This report was produced by Softwire Ltd in collaboration with internal stakeholders at DLUHC, with particular thanks to:

• Ellay Pryce, Delivery Manager
• Afreen Saulat, Lead User Researcher
• Kat Ashton, Lead Service Designer
• Lorraine Gillespie, Delivery support
• Callum Bates, Design support

1. Executive summary

1.1.1. This report presents research commissioned by DLUHC and carried out by Softwire Ltd, to understand the barriers that local councils face which prevent them from signing up to agreements to share personally identifiable (PI) data, necessary for facilitating DLUHC data linkage projects.

1.1.2. The research used a multi-stage qualitative approach through which Softwire Ltd conducted desk research, interviews and workshops with key DLUHC stakeholders and employees across a variety of job families from 11 local councils.

1.1.3. The insights gained from this research will help DLUHC and local councils tackle barriers to data sharing practices and facilitate better policy and service provision in the future.

1.2. Key findings

1.2.1. Barriers to data sharing

1.2.1.1. Softwire Ltd identified 7 barriers to data sharing:

1.2.1.2. Legislation and legal concerns: More guidance, support and collaboration were requested from DLUHC around Data Sharing Agreements (DSAs). Fear of making the wrong decision as data controllers, and varying levels of risk appetite and experience present a significant barrier to some councils in signing DSAs.

1.2.1.3. Resource concerns: Data sharing initiatives were perceived as more burdensome by smaller local councils who have less funding and resource to engage with new data requests from DLUHC.

1.2.1.4. Logistical concerns: Logistical challenges related to collecting, cleaning, and inputting data into multiple systems, lack of IT system integration, compounded by a lack of resources and funding, can hinder local councils’ willingness to participate.

1.2.1.5. Privacy concerns and a lack of clarity on data usage: Lack of clarity and understanding around how personal information will be used presents a barrier to data sharing. Local councils also raised concerns about the wider impacts of participation, such as whether funding outcomes could be impacted or whether data could be used in relation to border enforcement and benefit claims.

1.2.1.6. Lack of clarity on the benefits of national data linking: Local councils expressed uncertainty about the benefits of central government-led, national level data linking versus local-led initiatives. There was also a general lack of understanding the benefit of sharing PI data, as the potential tangible outputs and outcomes are not made clear.

1.2.1.7. Relationship to central government: Trust and confidence are essential factors in the willingness to share data, and smaller councils outside of London often do not have close pre-existing relationships with central government, leading to lower levels of engagement.

1.2.1.8. Relevance to individual council employees’ roles: Individuals working in roles involving data protection were more likely to see the potential benefits of data sharing, while those more removed from strategic decisions were more hesitant and concerned about the logistical impact.

1.2.2. Relevant case studies

1.2.2.1. Through conversations with local councils, Softwire Ltd identified 3 case studies which demonstrated the power of data sharing and highlighted a number of critical success factors, including:

• clarity on objectives and expected benefits of project
• addressing the challenges around obtaining end-user consent for data sharing
• effective collaboration between council and national government bodies

1.2.3. User needs

1.2.3.1. The following user groups were identified as key to successful data sharing initiatives:

• Housing Officer / Housing Manager, who require clarity on how PI data is used and why it is needed, and simple ways of providing data
• Data Protection Officer / Data Protection Manager, who require clear DSAs and assurance that data sharing adheres to GDPR (General Data Protection Regulation) principles
• Senior Legal Services Lead, who require clarity on legislation
• Head of Data / Programme Lead, who need to understand the benefits of data sharing initiatives

1.2.3.2. Softwire Ltd compiled an extensive list of local council user needs, which could help DLUHC to ensure tangible, relevant outputs are delivered through future projects.

1.3. Key recommendations

1.3.1. Drawing on these findings, Softwire Ltd proposes 6 recommendations for DLUHC to consider when seeking local council participation in future data sharing initiatives.

1.3.2. Define responsibility around DSAs (and broader legalities surrounding the initiative): DLUHC should take more responsibility for determining DSAs or assist local councils in doing so.

1.3.3. Communicate initiatives at the right level: DLUHC should work with key players in local councils and relevant forums to ensure that informed decisions are made at the appropriate level.

1.3.4. Articulate benefits of data sharing and linking and alleviate concerns over data use: Local councils need to understand the benefit of data sharing to them and to the end user and be reassured that concerns over the future use of data are taken into account.

1.3.5. Reciprocal data sharing and improved communication around outcomes: DLUHC should share outcomes of initiatives with local councils, so they can see the tangible benefits of data sharing.

1.3.6. Raise logistics, systems and data challenges into programme level initiatives: DLUHC should consider programme-level initiatives to improve system interoperability for data sharing, to simplify processes and reduce administrative burden.

1.3.7. Conduct further research for additional insights: DLUHC should consider conducting additional research, including with more local councils, to better understand the problem space and the extent to which each of the findings is applicable.

1.4. Report structure

1.4.1. The following sections are structured as follows:

• Chapter 2 presents the policy context for the research and the research objectives
• Chapter 3 sets out the methodology along with the strengths and any limitations of the approach or caveats that should be considered when using the findings from this research
• Chapter 4 presents findings on the barriers to data sharing experienced by local councils
• Chapter 5 presents case studies that include high level detail of data sharing projects currently live at local councils
• Chapter 6 presents user needs that emerged from pain points identified through research
• Chapter 7 concludes this research report exploring local councils’ views around data sharing and linking and highlights recommendations for future consideration.

2. Introduction

2.1.1. This chapter summarises the background to the study and the objectives of the research. It also gives an overview of the remainder of the report.

2.2. Policy context

2.2.1. DLUHC commissioned this research as part of its programme to better serve vulnerable populations. Through the BOLD programme, DLUHC aims to demonstrate how adults with multiple disadvantage can be better supported by safely and securely linking and improving the data held on them by government. BOLD’s initial focus is on reducing homelessness, supporting victims of crime, reducing substance misuse, and reducing reoffending, with DLUHC leading on the pilot for reducing homelessness.

2.2.2. Groups most likely to experience homelessness include people with mental health problems, substance misuse issues, and those who have experienced domestic violence (see Statutory homelessness in England: July to September 2022). Linking data across organisations involved in providing support in these sectors can lead to improved services and better prevention strategies.

2.2.3. Local councils and government agencies play a critical role in addressing homelessness, with policies and services aimed at reducing the number of people experiencing homelessness and improving their access to shelter, healthcare, and support services (see Homeless Link’s “What causes homelessness?” web page. To inform policy and service improvements, there is a growing need for better data sharing and collaboration between local councils and government agencies, as well as with other relevant organisations (as explained by the Institute for Government in their summary document on Data sharing between national, devolved and local government).

2.3. Research objectives

2.3.1. The aim of this research was to provide recommendations to increase local council participation in data sharing and linking initiatives by:

• identifying barriers preventing local councils from sharing data
• understanding what motivates local councils to share data
• understanding how data shared between local councils and DLUHC can be better facilitated
• identifying the outputs from data sharing which would be useful to local councils
• understanding local councils’ preferences for central government engagement on data initiatives related to homelessness and rough sleeping

2.3.2. The research project involved speaking with internal stakeholders at DLUHC to understand the context for current and previous data sharing and linking projects, as well as interviewing 11 local council representatives in various job roles across England. The roles included Legal Services, Head of Data and Strategic Data Leads, Data Protection Officers and Managers and Housing Officers or Managers, some of whom were involved in agreeing DSAs with DLUHC in the past and others who had some experience around data sharing and linking more broadly. The full research methodology is provided in Chapter 3.

2.3.3. Research questions

2.3.3.1. In addition to the main objectives, the research aimed to shed light on the questions outlined at a high level below. A full list of research questions can be found at the end of this report.

• What factors affect attitudes towards data sharing and data linking? These could be related to resource, organisational structure, data linking experience or other characteristics of the local area.
• What factors act as barriers to data linking? These could be the type of data, the reasons for data linking, privacy concerns and logistical issues.
• What experience do local councils have of data linking?
• What preferences do local councils have for communications and support over data sharing initiatives?

3. Methodology and research approach

3.1. Phase 1: Scoping and central government interviews

3.1.1. Desk-based research

Softwire Ltd initially took steps to contextualise the research through desk research. This involved:

• reviewing engagement logs between DLUHC representatives and local councils to demonstrate difficulty in efficient communication
• reviewing existing projects to gain understanding of what data linking involves and how it may be used in the future
• looking at existing secondary research to understand people who are homeless and sleeping rough and for whom these data linking initiatives are aiming to improve outcomes

3.1.2. Central government stakeholder interviews and workshops

Softwire Ltd held online interviews and workshops with stakeholders across DLUHC and some other central government departments as suggested by the DLUHC project team to increase researchers’ understanding and inform interviews with local councils. The interviews covered the following themes:

• perspectives on the BOLD programme thus far and learnings
• what may be preventing local councils from sharing PI data with DLUHC
• why PI data is important and the vision for future data sharing initiatives
• any recommendations on how to facilitate working with local councils
• the research approach to take to avoid duplicating existing work with local councils

3.1.3. Due to project timelines and challenges around re-engaging local councils to further ideate around solutions, further collaborative exploration with local councils is recommended.

3.2. Phase 2: Local council interviews

3.2.1. Recruitment

Through conversations with central government stakeholders, Softwire Ltd identified the roles within local councils most likely to provide valuable insights based on their involvement in the decision-making aspects of data sharing. These included:

• Data Protection Officers / Managers
• Data Governance Officers / Legal Leads
• Housing Officers / Managers
• Homelessness and Rough Sleeping Leads

3.2.2. An item was included in a regular DLUHC newsletter to local councils inviting them to participate in an online survey regarding barriers to data sharing, which also included the option to sign up to participate in further research interviews. To maximise uptake, any local council employees involved in data sharing were invited to participate even if their role was not included in the above list. This yielded 15 responses but not all of those were willing to participate in interviews. After leveraging separate introductions made by DLUHC staff (which resulted in a higher likelihood of participation due to pre-existing relationships), five 60-minute interviews were secured and completed by the first week of March 2023.

3.2.3. To secure more participants, DLUHC assisted Softwire Ltd in outreach by placing further recruitment surveys into newsletters, making more direct introductions to councils and their partners, and mentioning the research project during a monthly call with local council colleagues (followed up by further communication with participants). These efforts increased the number of local authorities interviewed to 11.

3.2.4. While areas were self-selecting, the participant group included councils from 7 of England’s 9 regions, with a mix of urban and rural areas, covering areas including villages, small and medium towns, and cities within and without London. Seven participants were from local councils which had previously chosen not to sign DSAs relating to homelessness data, while 2 were from those who had previously agreed to data shares.

3.2.5. Approach

Softwire Ltd conducted a total of 11 semi-structured in-depth interviews of 60 minutes each with those local council employees during the last week of February and the first 2 weeks of March 2023. All interviews took place in English and always had at least 2 people present from Softwire Ltd, with one person asking questions and the other taking notes.

3.2.6. All the interviews were conducted remotely, through DLUHC’s Teams platform and interview notes (captured on an online whiteboarding tool) were uploaded to DLUHC’s SharePoint for later access. Due to technical issues, it was not possible to record interviews.

3.2.7. Remote interviewing allowed for access to a wider range of local councils, with interviews at times that were convenient for participants. All councils interviewed already had access to equipment for communicating via Teams.

3.2.8. Consent

Prior to conducting interviews, Softwire Ltd obtained participants’ informed consent and explained the intended use of their data. Participants were made aware that they had the right to stop the interview or decline to answer any questions they were not comfortable with.

3.2.9. In addition, Softwire Ltd sought permission from participants to use direct quotes in this report, making it clear that any findings would be shared anonymously, respecting the privacy of those involved.

3.2.10. Discussion guide

Interviews followed a discussion guide Softwire Ltd developed based on the research brief and informed by interviews with central government stakeholders. This was intended to ease the participant into thinking about the topics outlined below:

• their job role within the wider context of the council
• the current state of homelessness and rough sleeping in the council area and how this has impacted the council’s response
• current data collections around homelessness and the methods employed to bring together these data
• their thoughts on sharing PI data relating to people experiencing homelessness and rough sleeping with central government departments such as DLUHC
• how they envision data sharing practices evolving in the future
• recommendations for how central government departments and local government can facilitate data sharing effectively
• understanding the opportunities and barriers that local councils see with respect to data sharing and linking
• improving DLUHC processes and communications off the back of that
• keeping in mind the impact of initiatives on the end user

3.2.11. Softwire Ltd chose this approach to allow for flexibility in the responses of participants who, due to coming from different local councils and job roles, were likely to have different contexts to draw upon, and because a semi-structured approach has the added benefit of allowing for comparison at the analysis stage.

3.3. Phase 3: Analysis

3.3.1. To ensure participants were content with the use of their quotes within the context of the report, Softwire Ltd reached out to all Phase 1 and Phase 2 participants to offer them the opportunity to review their content before publication, as previously they had only been asked for consent to share any learnings internally. It was made clear that any findings would be shared anonymously, respecting the privacy of those involved.

3.3.2. The notes from all Phase 2 interviews used a template designed around the topic guide to enable thematic analysis, with notes tagged based on the research questions that were outlined. These tags were used to mark out specific themes that were seen arising over time. Notes were then placed into thematic clusters, to provide supporting evidence for these resulting themes. Interview notes were tagged with the name of local councils, to show how many times a theme came up and what were perhaps more peripheral learnings. All the interviews were completed before Softwire Ltd finalised the thematic analysis.

3.4. Limitations of the research

3.4.1. There are several limitations to consider when interpreting the findings of this research.

3.4.2. Limitations of desk research

In this area of research, there was limited information about the concerns that local councils have with sharing PI data on people experiencing homelessness in their areas. Whilst some deductions could be made through analysing existing information and through secondary research done by partner organisations such as Shelter UK, there was no single source of truth that was aimed at uncovering the concerns that local council colleagues may have had.

3.4.3. Sample size

The research was conducted with eleven local councils, which may not be representative of all local councils across different regions. Only 3% of all local councils in England were interviewed. Despite the small sample size, common themes were raised across the group, indicating that these could be relevant to a broad audience. However, where a single local council raised an issue, it was less clear whether this could apply more broadly, and to what type of council.

3.4.4. The limited number of local councils involved in interviews also affected the ability to answer certain specific research questions. For example, the research cannot be used to identify patterns in attitudes towards data sharing based on local council characteristics, such as size or geographic location. This could affect the generalisability of the findings to a broader range of local councils.

3.4.5. Self-selected sample

As the local councils that participated in the research were self-selecting, it suggests they were more willing or motivated to share their thoughts and experiences about data sharing. This may mean that the sample is over-representative of those with stronger views (of either predilection) or those local councils with fewer resource constraints. To mitigate for this, interview times were kept flexible and a variety of recruitment methods were used. However, this potential source of bias should be kept in mind when interpreting the results.

3.4.6. Self-reported data

Often only 1 or 2 people from a single local council were interviewed. As such, it is important to flag that the qualitative data collected in the research is based on self-reported responses from a select few at local councils, which may not always accurately reflect the council’s actual practices or experiences. This potential source of bias should be considered when interpreting the results.

3.4.7. Timeframe

The research was conducted up to a certain point in time and may not reflect the current situation or changes in the circumstances of the local councils involved.

3.5. Steps taken to ensure reliability

3.5.1. To ensure the reliability of our study, several measures were taken while conducting remote interviews with local council colleagues.

3.5.2. Conducting remote interviews

This allowed for a wider pool of participants to be recruited from different locations, reducing the potential for regional bias. It was ensured that participants were given a range of times for their interview slots, so that they could choose a time when they felt most comfortable and were in a space to provide honest feedback.

3.5.3. Stating impartiality

Softwire Ltd introduced themselves as a third party conducting research on behalf of DLUHC, to put participants at ease and to reassure them of the interviewer’s impartiality. This helped build rapport with participants from local councils and allowed for more honest feedback.

3.5.4. Ensuring anonymity

Participants were informed that findings would be collated and thematically analysed and in cases where direct quotes were used, that these would be anonymised. Again, this helped reassure participants and allowed them to share their honest thoughts and feedback.

3.6. Project management

3.6.1. Softwire Ltd presented its initial, high-level findings to the original group of stakeholders on a fortnightly basis, ensuring alignment with the findings and answering any questions.

4. Findings: Barriers to data sharing

4.1.1. This section presents the findings on attitudes towards data sharing expressed by local council employees and barriers identified through local council interviews.

4.1.2. The research identified 7 separate types of barriers to data sharing:

• legislation and legal concerns
• resource concerns
• logistical concerns
• privacy concerns and a lack of clarity on data usage
• lack of clarity on the benefits of national data linking
• relationship to central government
• relevance to individual council employees’ roles

4.1.3. These will be discussed in detail below. Core issues include a lack of clarity on how data will be used, privacy concerns, and perceived benefits to and administrative burden on local councils.

4.1.4. It is important to note that attitudes towards data sharing are complex and not binary, as evidenced by the conflict between the preference of some local councils to be involved from the outset, and the resource concerns raised by other local councils. Differences in attitudes are influenced by factors including the data maturity of the organisation (as measured by the Government Maturity Model) and the perceived benefits of data sharing schemes.

4.2. Legislation and legal concerns

4.2.1. A significant barrier to data sharing is identifying appropriate legal gateways and legislation to underpin DSAs.

4.2.2. Where legal representatives at local councils were consulted, the majority requested that DLUHC take a more leading role in determining the legislation for the DSA as they were reluctant to take on this responsibility alone, particularly due to resource pressures. Others noted that there is no one-size-fits-all approach to legal determination, as each local council is responsible for what they agree to do with data, and there are varying risk appetites across councils.

4.2.3. Concerns over the legal bases for data sharing included a variety of views over whether explicit end user consent was required for data sharing. Some local councils highlighted the difficulties in obtaining end user consent while others questioned if it was necessary for the nature of the data collection and intended purposes.

4.2.4. As data controllers, local councils are responsible for reputational risk and the potential of fines from the Information Commissioner’s Office. While some councils felt unable to find a suitable legal gateway for sharing PI data, fear of getting things wrong and risk aversion deterred others from participating in data shares.

4.2.5. “You tell us which legislation they can have the data under and write it into a robust data sharing agreement for [Data Protection Officers] to review.” (Senior Housing Officer)

4.2.6. In both cases, councils felt that more support from DLUHC to help navigate these legislation challenges would be welcome.

4.2.7. DLUHC has advised that as data controller, local councils are responsible for this determination. However, it may be possible for DLUHC to ensure that more detailed DSA templates are provided to local councils.

4.3. Resource concerns

4.3.1. Staff in operational roles at local councils expressed concern about the impact of data sharing on their workload. The extra effort required to clean PI data before agreeing to sharing it is significant, as is the time required for reading, understanding and obtaining sign-off for DSAs, a process affected by inconsistency in management due to staff turnover.

4.3.2. Local councils have expressed concerns about the workload and GDPR compliant systems required to collect and clean data for DLUHC. They require more clarity on the logistics of the process and the time required to collect the data. Local councils feel that the burden of data collection is falling on them, and they require support from DLUHC to streamline the process.

4.3.3. “Even if we agree to [share more data] and we understand why they need it in a specific format, we can’t get it into the format due to effort and resource.” (Data Protection Manager)

4.3.4. Local councils also reported that while improvements had been made recently, outputs from data collections were not always returned to them a timely manner, thus reducing their usefulness and leading to questions on the necessity of each data initiative, given the time and energy that local councils invest in each collection.

4.4. Logistical concerns

4.4.1. Lack of consistency and interoperability between different local council data collection and storage systems create a logistical barrier to data sharing as significant effort and resources are often needed to reconcile data across systems.

4.4.2. The case management systems used by local councils are often different from those used by their supporting organisations such as charity delivery partners. It was reported that the systems are not integrated well and “don’t talk to each other.” This mismatch necessitates duplicate data entry and data cleaning, which is time consuming and frustrating. It is understood that software systems in use include Jigsaw, Civicas and Locata, but there may be dozens of systems in use by local councils and their partners for managing their homelessness and rough sleeping data.

4.4.3. As data is collected by different staff members or outreach teams for different purposes, the format required is not always consistent, which can lead to upload errors. Data errors relating to system uploads require manual intervention which is time consuming for overstretched local council employees. The same systems and data challenges emerged in interviews irrespective of the data collection that was being referring to. Participants also mentioned that data is often held outside dedicated IT systems, such as in spreadsheets.

4.4.4. To address this, there were some suggestions around using existing systems such as DLUHC’s online data collection system to submit data and aggregate it, which could then be sent back to local councils. Alignment across IT systems would alleviate some of these issues. Furthermore, the lack of IT systems integration makes updating case data, as well as sharing data with supporting organisations and with central government challenging.

4.5. Privacy concerns and a lack of clarity on data usage

4.5.1. Local councils reported that unless there is clear and specific information on how personal data will be used and stored, which other government departments’ data it will be linked to, and what the intended outputs and outcomes are, it is hard to make decisions and commit to sharing data.

4.5.2. Participants raised end user privacy concerns as a significant factor in making these decisions, with hesitancy to share personal information due to apprehension about who would have access to it in the future and how they could use the data. Worry about how unanticipated data use could impact vulnerable people, and the subsequent reputational damage, presented a barrier to local councils participating in data sharing initiatives.

4.5.3. Given that some data collected by DLUHC is used when assessing funding proposals, local councils also mentioned that information on whether and how any findings from analysis of linked data would inform future funding would be welcome.

4.5.4. To address the above concerns, councils expressed preferences for clear protocols and boundaries for data sharing to ensure there is transparency in how data is collected, stored, used and disposed of, following the Caldicott Principles. While much of this would be covered in DSAs, it is important to avoid ambiguous and broad language around the intended use of data as this can lead to uncertainty, particularly for those in Data Protection Officer and Legal Services roles.

4.5.5. “[The] DSA is key - maybe local councils needed more information, practical full paragraphs around exactly what it all means - to inform policy, to inform strategy, possible individual outcomes - proper understanding is needed and clarity in this.” (Legal Lead)

4.6. Lack of clarity on the benefits of national data linking

4.6.1. Some local councils were aware of the benefits of data sharing and linking at a local level but want a deeper understanding of the additional potential benefits and tangible value from national data linking, to justify sharing PI data.

4.6.2. They also felt that addressing the needs of people experiencing homelessness should primarily be the responsibility of local support teams, leading to questions about why DLUHC needed access to PI data, as the role of PI in data linking had not been made clear.

4.6.3. If local councils were to better understand the types of outputs possible from the DSA with DLUHC they may be more motivated to participate. Ideally DLUHC would work in collaboration with local councils to determine exactly what they believe to be a useful output.

4.6.4. There was also a lack of understanding of the benefit of gathering personal data more generally. i.e., how more context about a person could enhance understanding of individual cases (compared to only looking at aggregate data) and how it could ultimately result in service improvement and policy change.

4.7. Relationship to central government

4.7.1. Local councils reported insufficient engagement and a perceived lack of understanding from central government. Out of the councils engaged in the research, the larger local councils with more resources had forums in place to facilitate discussions around data sharing, while smaller ones lacked the resources to engage in such discussions. It was suggested that for larger councils, especially those based in London, the proximity to central government can strengthen relationships due to more frequent interactions and more available resource. This suggests that smaller local councils, especially those in rural parts of the country, may struggle with their relationship to central government, creating an imbalance in levels of engagement.

4.7.2. “You’ll get more active engagement if you go to the big councils versus the shires.” (Data Protection Manager)

4.7.3. Trust and confidence are essential factors in the willingness to share data, with council staff noting that existing trusting relationships do exist between local and central government, particularly where central government representatives have previous experience of working with local councils. Similarly, the research found that local councils have close relationships with organisations where many staff members have previously worked in local councils and are therefore perceived to have a better understanding of the context of any challenges faced. This was found to be the case with London boroughs and the London Office of Technology and Innovation (LOTI).

4.7.4. “LOTI is co-funded by [local councils] - plus other bodies - it is perceived to be a sector organisation and doing stuff for us versus [central] government doing stuff to us.” (Strategic Intelligence Lead)

4.8. Relevance to individual council employees’ roles

4.8.1. Those who are more involved in data strategy and have a better understanding of data protection are more likely to see the potential benefits of data sharing, while those who are more removed from strategic decisions may be more hesitant and concerned about the specifics and logistics. The power of PI data in informing policy was mentioned by a Head of Data, but smaller local councils may be more concerned about the logistical and resource challenges associated with implementing such an ask.

4.8.2. Housing Officers and Data Protection Officers (DPOs) are often involved in these conversations, but they may not always have the strategic oversight to see the potential benefits of data sharing beyond their day-to-day responsibilities.

5. Findings: Case studies

5.1.1. Through interviews with local councils, Softwire Ltd identified several existing data sharing and linking initiatives. These case studies can be used by DLUHC as examples to highlight the tangible outcomes that data sharing and linking can have as part of communicating with local councils less familiar with this type of activity, including the benefits to end users. The case studies also provide a useful template of tried and tested initiative methods that can be repeated in future.

5.1.2. Due to the limited interview time with representatives (60 minutes), and the number of topics covered in interviews, case studies could not be delved into more deeply for the purpose of this research. The findings below are therefore quite high-level but could be enhanced through further interviews with the councils highlighted below.

5.2. Changing Futures programme: Using personal data sharing to inform service improvements

5.2.1. Objective

To improve the delivery of wound care provision for vulnerable people in Sheffield, reduce A&E attendance, and enhance overall efficiency of the healthcare services.

5.2.2. Approach

Sheffield City Council collaborated with the NHS to share and analyse personal data to understand patterns of A&E attendance by vulnerable people. The data included demographic data and reason for attending A&E for ‘vulnerable people.’ The data sharing was governed by strict data protection protocols to ensure privacy and confidentiality, including ‘active consent.’

5.2.3. Outcome

The data analysis revealed an increase in A&E attendance on Monday afternoons due to wounds becoming infected at the weekend, indicating a potential issue with wound care provision over the weekend. To address this, the council coordinated wound care provision in the city centre over the weekends, resulting in quicker treatment for vulnerable people and a reduction in A&E attendance, as well as more efficient use of healthcare resources.

5.2.4. Key success factors

• effective collaboration between Sheffield City Council and the NHS
• successful sharing and linking of personal data
• identification of the root cause of the issue through data analysis

5.3. Implementing Empowering Communities Integrated Network System (ECINS) for improved data sharing and service provision

5.3.1. Objective

To improve data sharing, with the aim of providing more efficient and effective support and service provision for vulnerable individuals in Redbridge.

5.3.2. Approach

Redbridge Council is aiming to deploy the Empowering Communities Integrated Network System (ECINS), which was instigated by the Mayor’s Office for Policing and Crime (MOPAC) and is already being used by several London local councils, by the end of March 2023. The system will be used to share information between everyone supporting vulnerable individuals in Redbridge, enabling better optimisation of service provision. ECINS is a cloud-based national platform that allows for consistency and scalability in tracking vulnerable individuals’ journeys across the country.

5.3.3. Outcome

It is expected that ECINS will improve support for vulnerable people, particularly those in temporary accommodation and for “no second night out” cases. The use of person profiles, actions, and cases will enable better tracking of vulnerable individuals’ journeys and provide a more joined-up approach to service provision. Additionally, the system will be used for integrated community management, improving safety in the community.

5.3.4. Key success factors

• using an existing, tested system
• clarity on objectives and expected benefits of project upfront

5.4. ‘Routes off the Street’ initiative

5.4.1. Objective

To provide a more integrated approach to substance abuse and mental health service support through the ‘Routes off the Street’ initiative in Camden.

5.4.2. Approach

Camden Council is using a portal to share data between various organisations involved in substance use treatment and mental health support through the ‘Routes off the Streets’ initiative, a more integrated approach to service provision. Data processing and end-user consent to data sharing are key considerations within the initiative. While in the council representative’s view, explicit consent does not need to be obtained since it is processed under public task and statutory duty, this is not a universally agreed upon position, with some staff being more cautious in this area.

5.4.3. Outcome

The use of a portal to share data between organisations will enable better optimisation of service provision.

5.4.4. Key success factors

• successful implementation of the portal for sharing data between organisations involved in service provision
• identifying that explicit end-user consent may not be required when data is processed or shared under the ‘public task’ legal basis

6. Findings: Identification of local council user needs

6.1.1. Services designed around users and their needs are more likely to be used and more likely to help achieve the intended outcome, thus meeting their policy intent. Local council user needs, grouped below by job role, were determined from conversations with eleven local councils and aim to give DLUHC a starting point from which to design new data initiatives. Note that not all roles were reflected in all councils, and that role titles vary across local councils.

6.2. Housing Officer / Housing Manager

6.2.1. Participants in these roles were responsible for engaging with partner organisations such as charities and completing data collection activities. They are close to the end user. People in these roles noted that to make decisions around the best use of their limited time, they needed to understand the benefits of data sharing for the local council and for the end user.

6.2.2. It is important for them to understand the need for PI data and how data will be used – especially that it will be used ethically, and not affect funding - so that they can feel more motivated to partake and more confident in facilitating data sharing conversations internally, including with DPOs.

6.2.3. As people in this role are responsible for other data-related activities such as statutory collections, they need to be confident that their efforts are being rewarded, to incentivise them to partake in additional data initiatives. Multiple sources of data and varying reasons for data requests means that people in these roles also need better ways to capture data and to transfer it across systems in order to avoid double keying and subsequent data errors.

6.3. Data Protection Officer / Data Protection Manager

6.3.1. Participants in these roles were responsible for the legalities around data sharing and linking arrangements. They review DSAs and uphold GDPR principles.

6.3.2. It is necessary for people in these roles to be involved from the outset of any data sharing projects so that perspectives and concerns can be shared early on. Information Governance forums need to be used so that staff in this area do not have to advise in isolation around something as high risk as personal data sharing, and so that they can hear the perspectives of other representatives. In addition to the use of forums, data protection staff in local councils need DLUHC to contribute towards determining the legislation in the DSA, and to consider obtaining third party legal counsel. This would ease the workload of data protection staff and provide some comfort.

6.3.3. It is vital that information on how data will be used (immediately and in the future, and by which government departments) is articulated to staff in data protection roles through more specific Data Protection Impact Assessments (DPIAs) prior to DSAs. This is a legal requirement and is needed in order to ensure that the project will align with GDPR principles. For similar reasons, staff in these roles also need to know whether data subjects have provided consent in line with GDPR principles.

6.3.4. To ensure that data sharing documents are returned in a timely manner, staff in data protection roles need DLUHC to ensure that they are written in plain English, with acronyms kept to a minimum.

6.4. Senior Legal Services Lead

6.4.1. Participants in this role were members of forums with multiple council representatives such as Information Governance Group for London (IGfL), worked with other legal services representatives in determining DSAs, and were involved with initiatives at a more strategic level. This role generally appeared more senior than the DPO role, but this differed across councils.

6.4.2. Like staff in data protection roles, legal services staff need to know about data sharing initiatives early on. As well as for providing strategic support to DPOs, this will allow legal services staff to liaise with other legal representatives to hear multiple perspectives around risk appetite.

6.4.3. Staff in these roles need to understand the specifics of the possible legislation so that they can determine whether GDPR rules are being abided by. Additional information on the benefits of sharing PI data and its intended immediate and future data uses is also needed so that informed consultation can be confidently given to others within legal teams.

6.5. Head of Data / Programme Lead

6.5.1. These roles were less task-oriented and more strategic. Responsibilities vary, but all staff in this area had experience working with data sharing and linking initiatives.

6.5.2. As with other roles noted above, staff in these roles need to be involved early on. This is so that they are fully aware of what is being asked of their teams in relation to logistics. They also may need funding to support these initiatives so that they can assign the relevant resource.

6.5.3. To feel confident in supporting data sharing projects, staff in these roles need to understand why PI data is required and what the benefits will be to their organisations and to the end user.

7. Conclusion and recommendations

7.1.1. This research successfully met its objectives in discovering the current processes around data linking, understanding the barriers for local councils to sharing data, identifying the specific people involved in providing data, and identifying potential useful outputs from data sharing.

7.1.2. The research also uncovered potential beneficiaries of data sharing outside of DLUHC (such as councils themselves, supporting partners such as charities or other central government departments) and explored local councils’ preferences for central government engagement around initiatives designed to support people experiencing homelessness and rough sleeping.

7.2. Summary of findings

7.2.1. As noted previously, 7 distinct barriers to data sharing were identified:

1. Legislation and legal concerns
2. Resource concerns
3. Logistical concerns
4. Privacy concerns and a lack of clarity on data usage
5. Lack of clarity on the benefits of national data linking
6. Relationship to central government
7. Relevance to individual council employees’ roles

7.2.2. Softwire Ltd also identified the key roles that need to be involved in agreeing a data sharing project along with their specific user needs, which if met are likely to increase the likelihood of any initiative being successful:

• Housing Officer / Housing Manager, who require clarity on how PI data is used and why it is needed, and simple ways of providing data
• Data Protection Officer / Data Protection Manager, who require clear DSAs and assurance that data sharing adheres to GDPR principles
• Senior Legal Services Lead, who require clarity on legislation
• Head of Data / Programme Lead, who need to understand the benefits of data sharing initiatives

7.2.3. From understanding the barriers and user needs, Softwire Ltd have developed 6 recommendations for DLUHC which should help better facilitate data shares with local councils, as detailed below.

7.2.4. Further to the specific recommendation, DLUHC needs to be aware that implementing data sharing initiatives can be a time-consuming process, and it may take several years to get these initiatives in place. By recognising the challenges and taking a deliberate, strategic, and collaborative approach to data sharing, central government can maximise the benefits of sharing personal data whilst building trust and mitigating the risks as the initiative unfolds.

7.3. Recommendations

7.3.1. The recommendations below are high-level recommendations, obtained from suggestions made by interviewees. It is Softwire Ltd’s view that further research is required to fully flesh out solutions to the issues raised. DLUHC should contemplate and consider each recommendation area, and then stress-test and prototype solutions in working groups including DLUHC and local councils.

7.3.2. Recommendation area 1: Communicate initiatives at the right level

7.3.2.1. Local councils noted that for data initiatives to be successful, it was crucial that the correct people were involved at the right stage of the project, as notifying people in the most relevant roles would mean project time was being used efficiently, and that decisions were made at the correct level of seniority. In addition to this, allowing for decisions to be made more collaboratively would motivate local councils to take part, and ties in with recommendation areas 3 and 4, which are related to the reciprocal benefits of data sharing initiatives.

7.3.2.2. To enable this, DLUHC should focus on clear communication with local councils. This needs to happen both at a holistic level, by facilitating a collaborative working relationship between DLUHC and local councils, as well as on an individual process level, by ensuring DLUHC provides support to councils if they struggle with specific tasks (e.g., DSAs, as in recommendation area 2).

7.3.2.3. DLUHC should meet relevant council employees at the start of an initiative, to ensure that the right stakeholders and key decision makers are engaged in the process. This will likely require the engagement of Homelessness Leads and Information Governance Leads (such as DPOs / Managers and Legal Leads) as well as Housing Officers / Managers.

7.3.2.4. Before initiating data sharing projects, it is recommended that DLUHC gathers information on existing forums, to ensure that conversations are had at the right level. Forums such as the IGfL or the GMPS Group (the Greater Manchester Public Sector Group) provide access to multiple local council representatives and allow for a wide range of views to be heard. This makes better use of existing relationships and enables collaborative deliberation on the scope of the initiative and the details of legislation. Taking a top down (forum utilisation and area legal leads) versus a bottom-up approach (Housing Officer engagement) may result in improved resonance and efficiency around DSA alignment and subsequent data shares.

7.3.2.5. Working with local councils via forums to determine the broader initiative goals could facilitate DLUHC establishing smaller initiative pilots, iterating on them, and using the outputs to support further conversations with other local councils. It will then be easier to embark on further reaching programmes, with the potential for national rollout.

7.3.2.6. DLUHC should also use the early engagement phase to clearly articulate to the local council what their expected responsibilities and likely resource requirements are, to alleviate any concerns in this area.

7.3.2.7. It is recommended that communications with local councils are conducted through email conversations and face-to-face meetings and workshops. An easily accessible communications channel should be set up, likely in the form of a dedicated email inbox where all relevant queries can be directed. A two-way communication channel will also go some way towards addressing the issue of reciprocity as in recommendation area 4.

7.3.3. Recommendation area 2: Define responsibility around DSAs (and broader legalities surrounding the initiative)

7.3.3.1. Local councils felt that closer collaboration with DLUHC in determining legal documentation was needed. As the risk around data sharing sits with local councils as the data controller, DLUHC should support councils by providing more clarity and support in defining the details of DPIAs and DSAs. Local councils struggling for resource would particularly benefit from this. Previously, DLUHC’s approach towards engagement with local council housing officers and DPOs has meant that the responsibility around determining the appropriate legislation was challenging.

7.3.3.2. As close collaboration will require more resource investment from local councils, Softwire Ltd recommends that DLUHC engages with them as early as possible, to help with planning. Additionally, councils can be incentivised to invest in the relationship if it feels reciprocal. I.e., has clear benefit to them and to the end user (as in recommendation areas 3 and 4).

7.3.3.3. The approach of collaboratively devising the details of DSAs has proven successful in existing relationships between local councils and government departments such as the Ministry of Justice (MoJ), or other public bodies such as the Metropolitan Police. Data sharing projects which instead took an approach of arranging a call and sending over legal documents were not sufficient for many local councils to be able to make informed and confident decisions around the suitability of DSAs shared with them by DLUHC.

7.3.3.4. A more collaborative, equal approach to determining the scope of the project which also allows for conversations around the specifics of both the DPIA and DSA legislation, what the data will / will not be used for (now and in future) and exactly who will have access to the shared data will result in more clarity and comfort for local councils. Using collective intelligence and gathering diverse perspectives from multiple representatives at Information Governance forums (as in recommendation area 1) will ensure that responsibility for DSA alignment and risk is not falling on one data protection representative and gives DLUHC an opportunity to address a variety of risk appetites collectively, instead of less efficient council by council conversations.

7.3.3.5. Providing third party legal counsel around the legal gateway for data sharing and data linking could provide additional comfort for local councils who as data controllers are carrying the burden of important decision making, especially decisions as important as personal data sharing. For example, there was mention of ICO (Information Commissioner’s Office) fines as an outcome for making the wrong decision. A nationwide approach to clarify the position around legal gateways would therefore be beneficial and provide consistency across councils.

7.3.3.6. The differing views from legal professionals at different local councils around end user consent and whether this is necessary also needs further investigation. Again, this is something that could be raised with an Information Governance forum to source diverse perspectives and to determine the steps required to address concerns. For example, whether a new privacy notice needs to be served, whether end user consent needs to be gained, or whether the initiative falls under public task and statutory duty.

7.3.4. Recommendation area 3: Articulate benefits of data sharing and linking and alleviate concerns over data use

7.3.4.1. Local councils reported limited resources and other delivery challenges for meeting existing and new data requests and may struggle to justify the work involved if the need for PI data and the benefits to the council or data subjects of proposed initiatives are not clear. As local councils perceive a risk to themselves and their service users due to the legal and privacy concerns discussed in the findings, it is particularly important that benefits are made clear, so councils can make informed decisions on participation.

7.3.4.2. Softwire Ltd recommends that DLUHC ensures that specific benefits of any data sharing and linking initiatives are well articulated and that DLUHC consults local councils to ensure that these are comprehensive. These benefits need to be more specific than ‘for policy and service improvement.’ DLUHC should be transparent about any other consequences that data sharing might have and explain how the data will be used, now and in the future. For example, as some data collected by the department is used when assessing funding proposals, it is important to emphasise where this will not be the case.

7.3.4.3. It is recommended that conversations around this recommendation area should be had through the channels identified in recommendation area 1, especially through the use of a dedicated email inbox, as this will reassure local councils of the subject matter expertise of the people with whom they are communicating as well as keeping lines of communications around data initiatives separate to other communications such as those related to funding bids.

7.3.4.4. Projects which have successfully secured local council participation have clearly communicated the technical background to the steps required as well as the expected outcomes. Demonstrating the processes and outcomes from previous case studies may help to engage local councils beyond the details of the DSA. Especially where staff at local councils are generally more involved in individual cases of service users (for example, those in Housing Officer roles or those who use data for operational purposes), it is important to articulate the practical impact of conducting research using PI. DLUHC should also emphasise if findings are expected to be used for strategic purposes as opposed to operational, to alleviate the concerns around impact on individual service users.

7.3.4.5. It is recommended that DLUHC creates an information pack on the types of service user that may benefit from the research, with flow diagrams to explain the steps in the data sharing project which must be realised before these benefits may become apparent. Encouraging collaborative conversations on this topic between and within local councils is also recommended, as some staff within local councils may already have an understanding of how the use of PI data can be beneficial.

7.3.4.6. “PI [data] is so important - so you know the thing you’ve done has driven the change you’re trying to create… [by seeing] individual stories versus macro numbers.” (Programme/Data Lead)

7.3.4.7. “You need PI to know how to devise and deliver the support in a more responsive way working with nebulous cohorts - [or it’s] hard to define what’s happening and why.” (Programme/Data Lead)

7.3.5. Recommendation area 4: Reciprocal data sharing and improved communication around outcomes

7.3.5.1. DLUHC should collaborate with local councils at the outset of data sharing initiatives to better understand what outputs would be useful to them, as well as to communicate the expected outcomes of any data sharing initiatives. Timelines of expected outputs should be created, as these will allow local councils to make better informed decisions around their participation.

7.3.5.2. In relation to initiatives involving PI data collection, understanding of both the expected outputs and the tangibility of the outcomes that data shares can lead to are limited. Better communication of the outcomes that data sharing and data linking can lead to for the council and for the end user is required. Adding more context around data sharing initiatives would help to reinforce the wider goals of the initiatives and provide evidence that data sharing can lead to positive and tangible change.

7.3.5.3. DLUHC should take a lead in collating case studies of successful data sharing initiatives, as these will aid in allowing local councils to understand the potential benefits. As well as sharing case studies where there have been positive outcomes for service users, it is recommended that DLUHC shares information on case studies where there have been cost savings, as this may resonate with local councils.

7.3.5.4. As local councils reported that outputs from data collections were not always provided in a timely manner, DLUHC should take a more active role in ensuring that data collections across the department result in practical outputs for the local councils which provide them. Creating a culture in which the timely return of outputs is a priority will go some way towards motivating local councils to partake in data initiatives in the future.

7.3.6. Recommendation area 5: Raise logistics, systems and data challenges into programme level initiatives

7.3.6.1. Local councils spoke of logistical and technical challenges including mismatched IT systems, lack of system integration, manual inputting being time consuming, having to clean data and fix errors on upload, and a number of practical complexities, with integration across local council and partner organisations being a common pain point. System and data challenges are not isolated and the challenges that emerged were consistent across local councils regardless of the data collection to which they were referring.

7.3.6.2. To address these challenges, it is suggested that a full systems and data flow audit is considered to better understand the projects in flight which require data provision from local councils to DLUHC. This would have to be part of a wider programme level digital transformation change, that would require a more in-depth discovery, with a stronger technical focus, as noted in recommendation area 6.

7.3.6.3. Discovering the extent of the systems in use by local councils and the challenges experienced around data sharing on a logistical level may enable DLUHC to adjust data collection asks to further support local councils in relation to their resource struggles.

7.3.6.4. Raising the system and data challenges into a broader programme level initiative will also ensure that logistical challenges are not just solved in a siloed way but instead by looking at the challenges faced by local councils, their partner organisations and DLUHC as a whole.

7.3.6.5. “From a big data perspective, some things would need to be satisfied: We need data input consistency (a form maybe), good quality data and a spreadsheet you can put it all in. All the homelessness data would be held in a system dedicated to capturing this information, with a common/standard form [and] all councils would need to use this… Then, press a button to upload it to [the] DLUHC system. People will be capturing all sorts of data, so DLUHC needs to identify the fields [that they need] … how often they need it, and better understand how hard it is for local councils to provide this data.” (Data Lead)

7.3.7. Recommendation area 6: Conduct further research for additional insights

7.3.7.1. While this study provides high-level answers to the research questions, it is recommended that the research is continued with follow-up deep dive interviews as well as with additional local councils to better understand the problem space, the extent to which each of the findings is applicable, and the specific steps required to realise the benefits of the recommendations.

7.3.7.2. Deep dives are recommended into areas where data sharing projects have been initiated in the past. While the case studies in this report highlight that there are many examples where data linking has led to positive outcomes for service users and for local councils, further analysis of what does and does not lead to a successful outcome will provide a better understanding of how to approach data initiatives in the future. Follow-up interviews with the local councils involved in each of these projects are recommended, as more detail on individual studies and more opportunities for shared learning may motivate other local councils to participate in similar initiatives. A focus of the deep dives should also be the financial benefits of each case study, as this could help to motivate staff in local councils to allocate resource for data sharing projects.

7.3.7.3. While the legal issues surrounding data sharing came up in this research, it is clear that this also fed into other barriers. It is recommended that further research includes an analysis of legal bases which councils rely on for data sharing, and what factors affect these. This research did not uncover the extent to which it was the legal issues themselves as opposed to the perception of legal barriers which prevented most local councils from participating in data shares. Follow-up interviews with councils who have previously not participated in data shares with DLUHC will help to address this gap.

7.3.7.4. Recommendation area 1 noted that DLUHC should make use of existing forums to determine initiative goals at the outset of each project. While this research uncovered a few forums, it is recommended that further research is done in-house and externally, to identify forums which represent councils across the country.

7.3.7.5. In addition to this, a detailed technical discovery is recommended, as noted in recommendation area 5. As many councils noted technical barriers to data sharing, interviews on the specifics of what this entails may enable DLUHC to identify quick wins (for example, where many local councils are using a single software provider). A detailed technical discovery could uncover why systems do or do not integrate and whether there is a potential for rationalisation of data collections where differences in requirements are creating additional work for local councils. This research should explore the extent to which local councils are aware of the solutions available to them and better ways in which DLUHC may be able provide solutions for councils who are willing, but may feel unable, to participate in data sharing initiatives.

7.3.7.6. It is likely that research involving more local councils will reveal further areas of consideration, such as around cyber security. The research should assess how common each new concern is across local councils before deciding where further investigation would be beneficial.

8. Full research questions

8.1. Attitudes

Apart from general attitudes towards data linking, are attitudes to data sharing affected by:

• local councils being part of a combined authority
• size of homelessness and rough sleeping teams and other resource-related factors
• role of legal team
• previous experience of linking projects
• size of homeless population

8.2. Barriers

Are barriers to linking data related to:

• the type of data
• the purpose for which the data will be used
• who the data would become available to
• concern over the privacy of individuals
• concern over the immigration status of individuals
• concern over individuals being unable to give consent
• perception of the benefits of data sharing
• simplicity of linking processes by central government

8.3. Experience

• Do local councils already do any linking between their different datasets, or have they done so in the past? If so, what do they use this for?
• Do councils have operational or strategic reasons for data sharing?
• Can local councils give examples of times when a project has been proposed by central government and has been met with a positive response (success stories)? If so, how strong was central government involvement and how much guidance was given?

8.4. Opportunities

• What data would local councils find beneficial to link together?
• How does this differ between homelessness teams and rough sleeping teams?

8.5. Engagement

• What are local councils’ preferred ways of being contacted?
• Are there dedicated staff for dealing with these communications?

8.6. If there are any particular areas of interest, these may be approached iteratively to allow for deeper dives.