Co-production of Government Social Research: lessons learned from a Welsh Government and Disability Rights Taskforce prototype - Executive summary

In September 2023, the Chairs of the Disability Rights Taskforce in Wales mutually agreed to collaborate with the Equality, Race and Disability Evidence Units. Their aim was to co-produce Government Social Research (GSR). 

The aim of the prototype was to trial co-production as a way of conducting social research alongside representatives of disabled people using the principles of the social model of disability. 

Many members of the co-production team had lived experience of being disabled by barriers in society, which included some of the Welsh Government officials. 

The prototype aimed to test the extent to which co-production can be achieved within a GSR setting and when considering the GSR code. GSR is the professional membership organisation for social research in government. The GSR code is an addendum to the Civil Service Code and includes the core values of integrity, honesty, objectivity and impartiality. This sets out a number of standards that must be met for research to be considered GSR. 

The ambition was to enable decision makers to develop better informed policies and drive better outcomes for people as part of contributing to a more equal Wales under the Well-being for Future Generations (Wales) Act 2015. The prototype was designed to provide learning for the GSR community and policy officials to put ‘involvement’ into practice as one of the act’s sustainable development principles. The prototype would also enable delivery of the mission of the Welsh Government’s Equality, Race and Disability Evidence Units “to improve the availability, quality, granularity and accessibility of evidence about individuals with protected and associated characteristics to fully understand the level and types of inequalities in Wales.”

Following the co-production prototype, the aim of the lessons learned research project was to understand to what extent it is possible for government social researchers to co-produce social research. Additionally, the research aimed to consider any lessons on applying the social model of disability in GSR. Welsh Government social researchers were responsible for delivering the lessons learned research from the prototype. 

The social model of disability focuses on societal barriers rather than impairments, advocating for the same rights for disabled people as their peers. The social model shifts responsibility to society, calling for the removal of barriers (structural, cultural, and discriminatory) which hold back disabled people's participation.

The objectives of the lessons learned research project were:

• to draw conclusions about what level of involvement is feasible at each stage of a GSR project
• to understand the barriers and enablers to co-producing research in GSR
• to understand the added value of carrying out research collaboratively with people with lived experience and expertise of being disabled by barriers in society and incorporating the social model of disability 

In total, 16 semi-structured interviews were undertaken with 5 Disability Rights Taskforce Chairs and 11 Welsh Government officials (which included social researchers and policy officials) between September and November 2025. 

An independent researcher from Co-Production Lab Wales was commissioned to collect views from 4 Taskforce Chairs and undertake thematic analysis of the interviews. The independent researcher then shared the themes identified with the Welsh Government social researchers leading on the lessons learned project. 

Two Welsh Government social researchers led the data collection and thematic analysis from 11 Welsh Government officials and one Taskforce Chair.

A senior research officer who had previously been a member of the co-production prototype led a desk-based review of project documents. 

A literature review was undertaken by Welsh Government’s Internal Research Programme of evidence relating to co-production and disability research. 

An independent peer reviewer was commissioned to sense check research tools before fieldwork started, to facilitate a thematic analysis reflection workshop and to review the draft report. The independent peer reviewer provided advice and support to the project team, particularly around any methodological or ethical issues.

The co-production prototype had 3 phases of delivery, agreed with the Chief Social Research Officer for Welsh Government.

• Phase 1 (to co-design a vision for the research, including establishing the topics for the research and ways of working together). 
• Phase 2 (to co-produce evidence which included co-design and delivery of a research workplan, co-writing of research reports and to jointly brief ministers and disseminate the research findings).
• Phase 3 (co-production lessons learned research led by Welsh Government social researchers to learn from and evaluate the co-production approach and consider additional ways to capture lived experience and expertise and share that learning from the prototype).

The co-production team worked together to co-design and agree the overall prototype vision. This defined co-production as “working in equal partnership, learning how to produce research together, involving Disability Rights Taskforce Chairs without over-burdening them, and publishing reports of a good standard as co-produced and co-authored research”. 

The vision set out the co-production team’s intention to adopt ways of working together that valued diversity, were inclusive, transparent and allowed the team to be creative and learn from each other.

During the scoping stages of the co-production prototype, the team agreed to co-produce 2 research projects based on Disability Rights Taskforce recommendations. These recommendations had previously been identified and proposed to the Welsh Government following the inequalities experienced by disabled people in society during the coronavirus (COVID-19) pandemic. The first research topic explored the potential for establishing a National Centre for Independent Living in Wales. The second research topic explored ways to improve the Disability Confident Employer Scheme in Wales.

The learning from this work can be used for future co-production projects, and also the involvement of people with lived experience in GSR projects. Much of the learning is also applicable to co-producing in different contexts.

The 5 areas covered by the learning are: 

• the co-production process and application
• power and relationships
• co-production and involvement through GSR research stages
• the social model of disability and accessibility
• individual and institutional learning

The literature shows there is no one definition of co-production and the evidence captured through the prototype showed that individuals have a different interpretation of what co-production means. Co-production methodologies can take different forms and vary in how they are applied and the levels of participation. In this prototype, individuals had different previous experiences of co-production. Developing a joint definition and vision to provide a common understanding of what co-production means was critical for this prototype so everyone had a shared understanding of how individuals involved could work together in a co-productive way.

The findings showed doing a co-production project can be challenging, and co-production team members need a core set of skills and experience to do it. Experience in deliberation, facilitation techniques and stakeholder engagement were seen as enablers. Equipping everyone to work in a co-productive way through training, shadowing, coaching and learning from best practice could help to build relationships and aid a common understanding about how to deliver a co-production project in the future. 

For this prototype, a project remuneration package was co-designed and agreed by the co-production team, providing learning opportunities and research experiences. This was in place of financial payment due to the potential for a loss of welfare entitlement and income due to UK tax and welfare benefit rules. Although the remuneration package was reported to result in a higher time commitment overall, it was well received and provided opportunities for the team to apply existing or develop new skills and knowledge. However, the prototype found that not offering financial payment could limit engagement for some in future co-production projects. Practitioners should also consider factoring in adequate time to agree remuneration arrangements within project timescales. 

For this prototype, co-production as an approach for delivering social research was resource intensive and took more time than a standard GSR project. In this prototype, it was challenging to accurately anticipate the time and resources that were needed to deliver, especially given that co-production was an iterative way of working. The prototype required a great deal of goodwill, commitment and patience to sustain participation. It is proposed this is considered by practitioners looking to co-produce research in the future.

There was an increase in research scope during the prototype where primary qualitative research was carried out instead of initially planned evidence reviews. This led to more demands on people over a longer period than anticipated and meant not all intended outputs were delivered. Future projects should consider spending more time at the early co-design stage to solidify the research scope and agree this in any terms of reference. 

This prototype was felt to have an impact on individual wellbeing at times, given the volume of work and the relationships involved. The co-production team had built in wellbeing check in points at the end of each project meeting as a result. It is proposed that future projects need to actively consider the wellbeing of individual co-production team members throughout project delivery and this needs to be a shared responsibility amongst group members. 

There was a need to clarify how decisions would be made to ensure transparency and maintain trust. Estimating timescales were particularly challenging given the goal was for everyone to mutually agree any decisions taken. For future co-production projects, it is proposed that decision-making protocols should be made clear in any terms of reference.

The prototype needed to be iterative and responsive to decisions that were taken along the way. Future projects should consider a flexible project management approach. Seeking an independent project manager to lead the project management could potentially help reduce burden, particularly for co-production members with project management responsibilities.

The literature shows that relationships are core to success in co-production and evidence showed that more time was needed to establish mutual trust and relationships at the start of the prototype. Asset mapping was used as a tool to explore the reciprocal benefits of co-production. Individuals could express what they each wanted to contribute and what they wanted to take away from the prototype. The prototype found that carrying out asset mapping was useful in understanding individual people’s motivations and developing reciprocal relationships. It is proposed that time is spent establishing relationships and asset mapping at the start of any future co-production projects. 

Overall, an equal power balance was not fully achieved for this prototype and the power balance varied through the different project stages. Given that the responsibility for project management and resources lay with Welsh Government social researchers, this had an impact on the power balance which sometimes led to tensions within the co-production team. It is proposed that any future co-production projects should recognise some of the potential power imbalances in a co-production project from the start and throughout with everyone being open about the reasons for them. 

Co-production was experienced unevenly across participants and across research stages, with some feeling empowered and some excluded. This was the case for Taskforce Chairs, Welsh Government researchers and others involved in the prototype. Some saw progress in power sharing, others felt government structures fundamentally limited partnership. As with this co-production prototype, it is suggested practitioners seek to understand these divergences as it can provide crucial evidence about the conditions under which co-production can succeed or fail. 

The prototype aimed to co-produce across all research stages, however the level of co-production varied across the different stages of the GSR research cycle, particularly as timescales extended. Open and honest communication about time and resources were important to understand the reasons for adapting the collaborative working style across different stages. During the analysis and reporting stages of the prototype the co-production team agreed to adopt a more consultative approach to reduce some of the workload for the group, with Welsh Government social researchers taking more of the lead. Co-production projects should consider what is or is not possible to co-produce and to manage expectations within the group from the outset. 

There were positive examples of lived experience of being disabled by barriers in society informing the main stages of the GSR research cycle. These examples ranged from lived experience adding value to research design, ethics and scoping to undertaking fieldwork, data analysis and reporting, which were deemed by the group to have improved the overall quality of the research. However, the extent to which some members of the co-production team were able to get involved was also partly determined by research skills and expertise. The prototype set out to draw conclusions about what level of involvement is feasible at each stage of a GSR project, and the evidence demonstrated that this can be incorporated effectively at every stage. 

The prototype showed that it is not possible to achieve joint decision making in quality assurance of GSR research reports, as sign-off has to be the responsibility of GSR researchers. Some co-production team members reflected having a GSR quality assurance process worked well to provide a useful framework and format for publication. As sign-off lay with GSR researchers, a theme around ‘who holds the report?’ emerged. This theme questioned how far final outputs can really be jointly owned and to what extent these sign-off arrangements shaped levels of power and influence.

The GSR code values of objectivity and impartiality were found to be helpful for this prototype. The values set a clear expectation for neutrality when designing and delivering activities across research stages and provided a useful benchmark for reflection and validation. The principle of impartiality was sometimes challenging to reconcile with the ambitions of the co-production team to ensure the research was as impactful as possible. It is proposed that conversations about reflexivity and positionality are included in future co-production projects to recognise individual lived experiences that could impact on the co-production team’s ability to align to these values.

Co-producing with people who have lived experience of being disabled by barriers in society was deemed to have improved the overall credibility and impact of the research. There was evidence of added value of lived experience informing the research focus in terms of the issue being contextually relevant and the identification of topics that required more evidence. It can provide real benefits in identifying legitimate issues that require research to fill any evidence gaps. It is proposed that involving a wide range of people with lived experience can be beneficial and provide valuable validation in evidence planning.

Questions were raised about whether the prototype went far enough to embed diverse lived experiences and consider intersectionality. In the future, it is suggested that where necessary and appropriate, the co-production team could be designed with the representation of the relevant intersections of society in mind. If this is not practical, then varied and intersectional lived experience could be considered in the research methods of a project or within other forms of collaboration.

Co-production team members were positive about the application of the social model of disability within the prototype and that it had gone further than other projects they had previously been involved in. Steps were taken to actively explore and remove individual barriers as far as the project would allow. This involved regularly asking co-production team members to inform the team of any accessibility requirements. 

However, the co-production team needed to work through barriers to apply the social model of disability when interacting with processes beyond the team's control. Working in an accessible way that adhered to the social model of disability could sometimes take more time. However, it was a crucial element of the prototype, and an ongoing and active practice following good ethical principles to ensure participation.

Accessibility was deemed to have supported new knowledge and insight to enter the process. It shaped the level of authority and influence individuals could have in the social research process. It is proposed that practitioners consider how to better apply social model principles to social research and that it is actively encouraged in future projects. 

Some co-producers reflected on their individual learning as a result of the prototype. Evidence from the research demonstrated positive examples of people in different roles learning from one another and overcoming any challenges as they arose. Some acknowledged they had gained a deeper understanding of disability rights, particularly in the current policy context of issues relating to disability in Wales. Others reflected they have gained a practical insight into research skills, GSR processes and the social model of disability. 

The prototype increased confidence and resilience for some members within the co-production team, particularly when working on complex research projects that required members to consider diverse perspectives and find collective solutions. 

One of the aims of this prototype was to reflect on current infrastructure for co-production and the application of the social model of disability in the Welsh GSR context. The findings from this prototype have highlighted opportunities for institutional learning and change. This is already and will continue to be considered under the Well-being of Future Generations Act where it clearly states that ‘involvement’ is one of the crucial ways of working to achieve the act’s well-being goals. It will also be considered in relation to the Equality, Race and Disability Evidence Units strategy to mainstream equality, inclusion and involvement in Welsh GSR and other analytical professions.

Co-producing social research was found to be a more ethical way of working, reflecting the ‘nothing about us, without us’ principle of the disability rights movement. This was described as a response to historical injustices experienced by disabled people in research. Practitioners should consider the right of people who have lived experience of protected characteristics to be involved in studies that are about them. It is proposed this should be considered in research scoping and that involvement can take many forms, from steering group membership, to peer review or full co-production. 

This prototype attempted to apply co-production principles as far as possible through each research stage. The opportunities for equal power dynamics and joint decision making varied throughout the research stages and for different activities and tasks. Looking back, labelling the prototype as ‘co-production’ sometimes caused unrealistic expectations for the co-production team. It should always be considered if and how to ‘co-produce’ or ‘co-design’ and when it is more appropriate to ‘consult’ or to ‘inform'. One of the main learnings for practitioners to consider from this prototype is to really understand the level of decision-making they are permitted to share at each stage of the process and be clear about that from the outset.