Stakeholder interviews.

RO, SM and LP conducted interviews. Interviews were conducted online through Zoom [35] or Microsoft Teams [36]. If appropriate, interviews could have involved two stakeholders from the same local authority.

Interviews explored the context in which FSM auto-enrolment was implemented, alongside the factors that influenced implementation, implementation processes and barriers and facilitators. Interviews also focused on data protection, governance, workforce requirements and partnership with schools. Local authority topics varied slightly, based on the stage of auto-enrolment implementation (see S1 Appendix). Local authority and national stakeholder interview topics were broadly similar. However, the former focused primarily on local implementation, processes and barriers, while the latter included topics such as local vs national implementation and policy recommendations (see S1 and S2 Appendix). Demographic characteristics were also shared by participants during interviews to provide a contextual understanding of the stakeholders implementing auto-enrolment. In interviews with two participants, only the demographics of the primary participant are reported to maintain participant anonymity.

Document collection.

PG conducted the documentary analysis. Documentary analysis was used to explore factors that influence implementation in local authority areas [38,39]. Documents were gathered with permission from local authority stakeholders to support qualitative data analysis (including, but not exclusively those areas that took part in the interviews). For those that did not participate in interviews, information regarding the documentary analysis was shared in webinars and workshops, where local authority stakeholders were invited to share any relevant documents with the study team. Document sharing by interview participants was not compulsory. The type of documents included (but were not limited to) progress updates shared through emails (e.g., from directors and implementing stakeholders), workshop notes (e.g., from data support workshops), meeting minutes and other documents (e.g., data protection impact assessment [DPIA], briefs or business cases). Documents could be either printed or electronic and were collated over the 18-month study period (1st July 2023–31st December 2024). All documents were stored in a secured drive.

Interview data analysis.

Interview data were collected by members of the research team, then sent for transcription by a third party. Once returned, members of the research team proofread the transcripts to remove any identifiable information and to ensure that the transcripts provided an accurate account of the interview. Three transcripts were double coded by members of the research team, who then discussed and resolved discrepancies. Following this, a codebook based on framework constructs was developed, which was used for all future analyses.

Each transcript was coded deductively according to the CICI framework by one member of the research team. Where data did not fit within the framework, they were coded inductively and, if relevant, data were coded into multiple constructs. All data were coded using NVivo 14 software [37]. Following data coding, a framework matrix was developed using CICI constructs and sub-constructs, which was used to explore the data and develop themes. Findings are reported according to the overarching framework constructs and themes within these.

Documentary analysis.

An initial review was undertaken to understand the narrative of FSM auto-enrolment implementation, based exclusively on the documents gathered. Following this, documents were categorised based on the relevance of the document and the information provided. Data analysis was based on Braun and Clarke’s model of thematic analysis [41]. After a detailed review of the documents, deductive thematic analysis was undertaken according to the CICI framework, by one member of the research team. During coding, document quotes were logged chronologically to help observe changes over time. Each document was reviewed by researchers, with quotes, themes and interpretations jointly refined. A framework matrix was then developed, which categorised key findings into constructs and sub-constructs, based on the CICI framework.

Setting: Instability, population and objectives.

All participants reported varying levels of deprivation and poverty within their area. In some, this was relatively new following a period of economic instability. In others, poverty was a longstanding problem which was exacerbated by economic instability, which highlighted the need for local authority intervention. In response, local authorities investigated various opportunities, with FSM auto-enrolment emerging as a potential solution to support residents and increase food security.

“Poverty is something that’s always been around in this country and absolutely within LA19, but… it’s just been brought into a sharper focus in the last year… it’s been hitting a lot more [people].” (LA19)

“With [the] cost of living crisis and people… struggling, there’s a lot of food insecurity. So… it’s about identifying those families who are… [entitled] for free school meals who need that support. And a lot of them do.” (LA12)

Alongside the need to improve food security, one participant reported an increase in families requiring financial aid and highlighted their role in preventing residents from getting into debt. FSM auto-enrolment was perceived to increase financial security, as families would no longer be responsible for providing meals.

“So, people that have never… had to think too hard about choices in terms of budgeting, or cutting back on certain things, or going without, are now having to make those choices… with regard to… mitigating the impact and trying to not get into too much debt...” (LA19)

With circumstances changing due to economic instability, several participants reported an increase in the number of children entitled for FSM, ultimately motivating local authorities to consider auto-enrolment as an intervention.

“We’ve seen a huge rise in FSM from about 17% in 2017… [to] 28.5% at the moment… That’s a really big rise in a short period of time and a lot of children.” (LA07)

While economic instability impacted households throughout England, it was participant knowledge of their population that furthered interest in FSM auto-enrolment. Several participants reported higher than average levels of deprivation, unemployment and FSM eligibility. In one area, FSM eligibility was thought to be considerably higher than the national average. In another, a participant reported the highest level of FSM eligibility on record, having observed a marked increase over recent years due to the cost of living crisis.

“We are a deprived area with a lot of unemployment, one of the most deprived areas in the country.” (LA05)

“LA02 has some of the highest… means tested free school meal eligibility in London.” (LA02)

“We’re at the highest level of FSM we’ve ever been… without a doubt there is [sic] pressures from cost of living, etc.” (LA07)

Several participants highlighted issues related to inequalities during interviews, describing the characteristics of pupils registered through FSM auto-enrolment, which were verified through email correspondence. Further exploration of FSM auto-enrolment in other areas revealed additional inequality-related barriers, including cultural and language barriers which affected the current parent/carer application-only registration method. Several participants also reported a diverse range of languages within their area, with a notable percentage of families who had not self-registered as speaking English as an additional language. In the context of FSM registrations through auto-enrolment, several participants highlighted a greater proportion of registrations from families speaking English as an additional language, compared to the proportion in the self-registration group.

“Out of that group [auto-enrolled pupils], 79% of them came from Black, Asian or multiple ethnicity backgrounds... 50%... didn’t speak English as a first language… 89% [were]... from lone parent families.” (LA14)

“67… [pupils that] came through were… unsurprisingly from more disadvantaged cohorts, higher than our general population level. So… there’s clearly something there about culture, language or something we’re not getting.” (LA07)

“What really stands out for me is how this… clearly demonstrates that the application-only approach disproportionately disadvantages our Black, Asian or Multi Ethnic population [and] how our ability to use this data can correct this inequity.” (LA14 [email])

Overall, high rates of deprivation, unemployment and FSM eligibility, in the context of economic instability, emphasised the need for local authorities to support disadvantaged residents.

Driven by the need to improve the financial and food security of residents, participants continued to learn more about FSM auto-enrolment through presentations and FixOurFood webinars. Ultimately, this led local authorities to implement auto-enrolment as an intervention that sat within local programmes that could reduce inequalities on a local level. In some areas, FSM auto-enrolment was aligned within current local authority strategy priority areas.

“It [FSM auto-enrolment] came to our attention through… FixOurFood, the head of public health was really keen to push this forward under… the [local authority programme].” (LA03)

Often, the decision to implement FSM auto-enrolment came from senior leaders, who recognised the potential benefits of the programme, and how this could meet the needs of their population.

Implementation context: Interest and perceptions.

The majority of participants emphasised the need for continued internal leadership support, as this permitted implementing stakeholders to acquire the necessary resources and support from other teams to take the programme further.

“Definitely the senior members of the CMT [Corporate Management Team] were crucial to this [FSM auto-enrolment] happening and being successful, definitely, because as soon as we hit a hurdle… they would be... advocating for this to go through.” (LA03)

The degree to which senior leaders were involved varied, though generally, leadership support was either a facilitator, which allowed implementation to progress efficiently, or was an essential component, without which implementation could not have occurred. Both interviews and documentary analysis highlighted the importance of leadership support.

“Having a lead member and director of children’s services who is encouraging staff to move forward with auto-enrolment implementation has allowed the process to move forward more quickly.” (LA07 [workshop notes])

“Me and [name 1] weren’t of a responsible level that we could make the decisions.” (LA11)

“The fact that the process requires many teams and a council to work together, often for the first time, dealing with complex issues, where there is clear governance and strong sponsorship from senior leaders in the council, we have found that any barriers and blockers are able to be resolved.” (NS01)

Leadership involvement also helped distinguish FSM auto-enrolment as a priority, which helped mitigate team capacity concerns while giving staff the permission to bring teams together and progress implementation.

“Some of the staff… were more, ‘Oh, this can’t be done. We haven’t got that capacity’.” (LA03)

“It’s because I’ve got the seniority in the organisation. When you just say we’re doing something people were just like, ‘Yeah, alright’... Whereas I think in [other local authorities] they had a data team that were a bit more junior and… nobody quite had permission to just say, ‘Yeah, let’s go and do it’.” (LA14)

Initial perceptions toward FSM auto-enrolment varied. In some areas, there was scepticism as to whether auto-enrolment was necessary, which may have been related to the general risk tolerance of participants or the local authority overall. However, both interviews and analysis of documents revealed that, as participants began to observe the success of the programme (both within their area and others), they felt the fear of risk subside.

“People talked about those councils that are risk averse, actually, just having access to this information of what we’ve been doing and the success that it’s brought will start to remove some of that fear.” (LA07 [workshop notes])

“There was quite a lot of… initial scepticism… It didn’t seem to a lot of departments that it was something that needed to be fixed. And also a lot of schools I think also felt that they had pretty good systems in place… but obviously for a lot of schools it has proven to bring in a lot of funding.” (LA02)

“It didn’t feel achievable and then felt very achievable.” (LA14)

Other local authorities were optimistic from the beginning. Several participants reported changing attitudes throughout implementation, particularly once specific barriers were overcome.

“At the time members were really keen, my managers were really keen, and it [FSM auto-enrolment] was just seen as a bit of a win-win really.” (LA22)

Full implementation.

Participants in areas that had implemented FSM auto-enrolment reported that implementation was intensive and contained several stages detailed in the following sections. Fig 2 below provides an overview of the FSM auto-enrolment process.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Free school meal auto-enrolment process flow.

Implementation timescales are not representative of all local authorities and vary considerably based on barriers and facilitators. *The October census (also referred to as the Autumn census) is a round of statutory data collection from schools that includes data such as free school meal eligibility, absences and special educational needs.

https://doi.org/10.1371/journal.pone.0339477.g002

Depending on the type of data being used, the order in which initial data matching was undertaken varied, relating to existing consent to use different data sources. If participants used council tax or housing benefit data, data matching was used in the beginning to identify and remove those already receiving FSM. This was then followed by writing to parents/carers to give them the opportunity to opt out. In instances where local authorities had deemed consent from parents, participants were able to finalise the list of pupils not receiving FSM more quickly and issue opt-out letters.

“Because we had deemed consent from parents filling in application forms… we could… extract those people already receiving free school meals [and]… finalise our dataset.” (LA02)

“Revenues and Benefits sent the data across… people that were claiming council tax and housing benefit. And we cross-matched that data with the existing free school meals claimants so we could eliminate those… Then we sent out those [opt-out] letters.” (LA03)

If universal credit data was used (i.e., data identifying families in receipt of means-tested, universal welfare support), several participants reported sending opt out letters in the beginning.

“The difference in the universal credit data… [is that] we started with the opt out process… because we didn’t have… deemed consent [from parents] to process or to check the eligibility or register the child, so [parents]… needed their… opt out opportunity.” (LA02).

Irrespective of the order in which datasets were consolidated, all participants offered families an opportunity to opt out. Despite numerous families being offered this opportunity, the numbers opting out were either very low or non-existent.

“There was one opt-out, they only said they don’t want to be enrolled in free school meals when they replied.” (LA17)

“And massive, massive support from parents for this programme, which probably reflects the fact that very few parents have also opted out.” (NS05)

Of those who opted out, various reasons were given, but mostly related to a lack of understanding of FSM eligibility or the auto-enrolment process. One participant reported four ‘opt outs’ in their area due to home-schooling, and another participant cited a food allergy as the reason one family opted out. One father misunderstood the opt-out letter, not realising that eligibility depends on the parent the child lives with..

“[The father] thought that the letter sent to him was in error… the parents were divorced and… the father didn’t qualify for benefits, and therefore he thought that his child did not qualify for free school meals.” (LA13)

Once families that wished to opt out were removed, several participants reported running the remaining potentially entitled families through the national government DfE eligibility checking service [42], before sending the list of entitled pupils to the team within the local authority responsible for FSM registrations.

“We processed this data through the DfE’s eligibility checker to get the final set [of entitled pupils].” (LA17)

Alongside identifying additional entitled families, participants reported that the eligibility checking process had the unintended consequence of identifying families that were almost entitled and thus may still require support given the current entitlement threshold. As such, one participant reported using this additional information to develop wider local welfare/support policies.

“The process in itself is worth going through to find those families who are struggling but not… [entitled] as well, because we will build policies around that.” (LA07)

In areas with close geographical boundaries, several participants underscored the challenges with identifying pupils who live in one area but go to school in another. In these instances, identifying eligible pupils depended entirely on sharing data across boundaries, which revealed inefficiencies in local implementation.

“The big limitation is doing it on a local authority by local authority basis, that assumes that children are going to school in the borough in which they live, and therefore you need to start sharing data across borough boundaries.” (NS05)

“Because in [region] we have a lot of cross-border movement… [and we] only hold the data of children in our schools. We don’t hold data of children who… are out of borough and we have an awful lot of movement.” (LA07)

As FSM auto-enrolment implementation progressed, those leading the implementation were required to communicate with several teams and organisations. Internally, several participants highlighted the need to regularly update stakeholders on progress. Several participants also emphasised the importance of frequent communication with schools, which allowed local authorities to identify pupils that had recently moved into the area.

“Quite a cross council kind of push. And through that period just going back to brief directors on where we had got to, brief members on where we had got to. And so, yeah, there was a bit of comms involved in just keeping everyone on side.” (LA07)

“We can only automate those that we know about. So schools have very much become the partner in this… and we have developed good relationships.” (LA22)

In addition, the majority of participants shared their knowledge and experiences with other local authorities, occasionally through FixOurFood webinars, which were attended by local authority representatives at various stages of implementation. During these workshops, stakeholders discussed barriers and solutions with their peers and members of the FixOurFood research team. Several participants also emphasised the importance of receiving support from those further along in the process.

“We’ve found through the FixOurFood meetings that although our neighbouring boroughs are at the start of their process, we’re really happy to work with them, we’d like to form some kind of data sharing cluster with them, which could then be spread across London.” (LA13)

“The fact that they [other local authority] are so much further down the line to be perfectly honest, and [name] is so willing to help. The [FixOurFood] toolkit and obviously the co-production and collaboration with other local authorities has definitely helped.” (LA01)

By maintaining internal and external communication channels, those implementing FSM auto-enrolment could keep internal stakeholders engaged, learn from experienced local authorities and identify additional pupils throughout the year. Similarly, by maintaining relationships with schools, local authority stakeholders could receive feedback specific to the regional context in which local authorities were implementing auto-enrolment. For example, one participant highlighted the need to translate communications to various languages to meet the needs of their population. Another participant reported that schools within their area liaised with parents on their behalf, as schools would likely have greater knowledge of the family situation. This relationship with schools led to a greater number of letters reaching the intended recipient.

“We were speaking to schools throughout autumn, we got some feedback from them about how to implement, including things around getting letters translated into different languages that are spoken across the borough.” (LA13)

Barriers – capacity, governance and data processing.

Several participants reported that they were initially concerned they lacked the resources to implement the programme, particularly regarding staff workload, with one participant reporting the impact of recent financial difficulties and how this hindered their ability to bring in additional support to increase capacity.

“Obviously their [stakeholder] nervousness around capacity and the volumes of the work… ‘It won’t work and we won’t have time, we won’t have the capacity’.” (LA03)

“The main issue has been the resourcing of the project… A few years ago, there might have been money to be thrown at it in terms of project teams, etc, we’re not in that place anymore, we’re in a very different financial position.” (LA01)

Whether staff capacity was a barrier depended on current workload, or if a specific role existed to lead implementation. If a role did not exist, implementation relied on the support of additional teams or external consultancy services, rather than creating a new role solely for auto-enrolment.

“We are… an additional resource to try these things [additional projects]… that’s the nature of our roles.” (LA14)

Several participants reported variation in the ability of local authorities to undertake FSM auto-enrolment from a capacity perspective, both now and on a recurring basis as workloads (and thus stakeholder capacity) vary over time.

“I think some local authorities still don’t necessarily have the capacity for this kind of work, and every local authority is quite different and has different funding and capacity available to them.” (NS02)

Several participants highlighted that FSM auto-enrolment was, and could be, impacted by the financial context in which the programme was implemented. For instance, one participant expressed concerns around implementing the programme in areas of high deprivation, where they could identify a large number of children. In these areas, schools could face financial difficulties, as the funding for newly identified pupils may not arrive until the following year.

“The financial strain of paying for those… extra school meals every day between October and April next year wholly sits with the school and their current budget.” (LA20)

To alleviate the potential financial pressure on schools, this local authority provided interim financial support. This participant also emphasised the importance of using the October census date (See Fig 1) as the deadline to ensure the correct funding allocation and timely receipt of funding, eliminating the need for local authority intervention.

“The cut-off date was the 5th [October] this year. If we had registered people on the 6th then the school is legally obliged to provide that free school meal, but… funding would not be in the budget until April.” (LA20)

The financial position of the local authority influenced several aspects of implementation: (a) whether the local authority could support schools as an interim measure, (b) whether the local authority could commission external services to manage the project and (c) whether stakeholders would be willing to undertake new work, in the context of recently increased workloads caused by staff redundancies.

“Another economic factor at local council level is the ability of the council to be able to provide resource to project manage the opt out auto award process.” (NS01)

“The local authority is bankrupt, you know, we’re going through multiple teams across the council, having to go through budget cuts, staff redundancies, all those sort of things which presents additional pressures to the local authority, which directly and indirectly impacts on whether people want to take on new work.” (LA09)

Although FSM auto-enrolment can increase future funding for schools via pupil premium, one participant highlighted that auto-enrolment does not bring in any income directly to local authorities (where the majority of implementation tasks lie). This view was shared by another participant, who suggested that ‘resourcing costs’ could impact how local authority stakeholders engaged with the process. In one area, as school meal budgets were devolved to schools, schools procured their meals from different sources, and so the cost of the meals varied. Consequently, government funding may not cover the full cost of the meal (currently £2.61 for most of England), as one participant suggested, and schools would be required to make up the difference, which could be particularly problematic in heavily deprived areas where many pupils could be identified. Further, delayed payment could impact some schools more than others.

“This process [FSM auto-enrolment] does not drive income for the council at all. It may drive income for schools through increased free school meal registration and… [pupil] premium funding down the line, but the expenditure of the process is not aligned to the income in an immediate cost benefit.” (NS01)

“Resourcing costs… whether… there’s been an investment of budget in managing that programme, how many staff hours it might take… I think there’s been some reticence to engage with it [FSM auto-enrolment].” (NS05)

“LA20… devolved school meal budgets to the schools themselves… So, every individual school… they commission and procedure their own school food… [and] school meals range from £2.30… to £3.25 per meal… So schools were still subsidising all the free school meals for the children who were already registered.” (LA20)

Several participants found sourcing welfare data to be challenging, and experienced that data controllers were wary of sharing data because of confusion relating to the legalities of data use.

“There were some sort of internal barriers around hesitancy in terms of how we use up people’s data. Is it a change of use? Is it, are we complying with GDPR [General Data Protection Regulation]?” (LA13)

One participant revealed considerable difficulties engaging the welfare stakeholders involved in data matching, due to the perception that all entitled children were already registered. This participant expressed frustration that this assumption persisted despite data to the contrary, which led to substantial delays.

“The main barrier we had was literally engaging with our revenue and benefits team, and that was predominantly driven by, like I said, there was a bit of arrogance around thinking that there [wasn’t] a problem…” (LA09)

Several participants also experienced barriers when accessing universal credit data from the national government Department for Work and Pensions (DWP). However, in one instance, the local authority had an information sharing agreement with the data controller (DWP), which gave stakeholders the confidence to use the data without seeking explicit permission.

“I know that some local authorities have raised a bit of an issue about using data that’s come direct from DWP… because we haven’t directly collected it… [but the information sharing agreement] basically says that we will use this information for welfare purposes, and as far as we’re concerned at LA11, free school meals is a welfare purpose.” (LA11)

On several occasions, database access depended on the type of local authority, which, for those undergoing a restructuring process, changed the permissions for data access. For other, non-unitary councils, this meant obtaining the information from the data controller (the DWP or other councils).

“[The] main barrier so far is being a non-unitary Council so we don’t hold the datasets required and would need to gain the information from DWP or local borough and district councils.” (LA08)

“Formally we were two tier and so… [we need to be] quite sure of the potential consequence of DWP removing the council’s licence to access Searchlight to administer our localised benefits system.” (LA20)

Once stakeholders had sourced the data, there were further barriers relating to data consolidation and processing, which subsequently caused delays and increased costs, as stakeholders commissioned additional support to extract the data from different sources.

“We don’t even have a database which would connect some of these key data sources together… there is always waiting time and there is a charge associated with it for developer time to write the script that extracts the data.” (LA17)

“The requirement to match together different types of databases is flawed because there will always be an element of fuzzy matching or inaccurate matching of data.” (NS01)

As participants faced various data-related barriers, this led some to consider whether they could re-use data for the purpose of FSM auto-enrolment. However, stakeholders felt that they had received inconsistent responses from the national government regarding permissions to use specific datasets. Subsequently, these inconsistent responses caused further stakeholder apprehension and delays. In one instance, a participant contacted their local Members of Parliament for clarification, but did not receive a clear response.

“I know… [name] wrote to the DWP and applied for a reuse of data and then they were rejected…” (LA19)

“There is no consistent application of legal basis… and that has caused significant variation in a council’s accessing the benefits data source in order to be able to identify entitled families.” (NS01)

“I drafted those written questions recently and suggested them to those MPs [members of parliament] and naturally was eagerly awaiting the answers in the hope that we would see in very clear, stark, explicit language that local authorities were free to use data… but there was still that caveat around… data being used within the boundaries of existing legislation.” (NS03)

In some instances, stakeholders would not begin exploring FSM auto-enrolment, due to the fear of pushback from the national government, as highlighted by one participant: “[Local authorities] are not even prepared to begin looking because of fears of reprisal from the national body” (NS01). Of all the barriers to implementation, this participant suggested that data concerns were the barrier that stopped intent, and thus could not be overcome by sufficient leadership support. Another participant had a similar view, suggesting that implementation ultimately relies on access to the data.

At the time of this research, national government has provided clarity regarding the legality of FSM auto-enrolment; however, during the early stages of implementation, external organisations began to raise legal concerns in some areas, including the lawfulness of FSM auto-enrolment implementation. This led to local authority stakeholder apprehension, as many were implementing the programme for the first time and therefore did not fully understand the legal implications.

“I think information governance, it’s a scary thing for people and I think it is always a barrier because we are all rightly very worried about doing the wrong thing.” (LA14)

“We’d started hearing from [name] in [organisation] that the DfE had asked them to stop publicising it [FSM auto-enrolment].” (LA11)

Interviews and documentary analysis showed that stakeholders supported one another throughout implementation. Those that were contacted by government organisations shared their concerns with others, which increased apprehension in other areas that had not been contacted by government organisations.

“We’ve already had issues with the IG team and the legal team anyway and they were nervous about this project going forward, getting this [feedback from other local authorities] has just added to that nervousness, so I’m unsure [whether this will impact future implementation].” (LA03)

“The push back and inconsistency at the national level makes us feel like we are doing something wrong or illegal when in fact all we are trying to do is support children to receive a fundamental necessity of food to which they are perfectly entitled to.” (LA20 [email])

One participant suggested that legal confusion may have arisen as DWP guidance could be open to interpretation. Proceeding with implementation then, could be influenced by risk tolerance, where some would seek clarification from the DfE and DWP, and others would be more confident in their legal position.

“The difficulty is that DWP are not precise enough in their wording… [it] is not specific enough in the legislation or the guidance that we are issued with as a local authority.” (LA20)

Several participants reported that adversity to risk generally stemmed from information governance teams and related to the opt out and eligibility checking process. Additionally, information governance stakeholders raised concerns relating to the legal basis for data processing and General Data Protection Regulation [GDPR].

“Obtaining details of benefits a family receives without their knowledge (because they haven’t made a request) may breach other rights… [and] having an opt out option does not sufficiently ensure that individuals are given the opportunity to withhold consent to such information about them being shared.” (LA06 [email]).

Several participants highlighted that the decision to proceed with implementation, despite resistance from information governance stakeholders, often relied on senior leadership support.

“We had a lot of sway from above… and I think if we hadn’t have had that support… when we did hit the barriers of the information governance and the legal information, I think that we probably would have faltered at that point.” (LA03)

“Sometimes you feel like there’s a bit of reluctance… the legal stuff and our data governance team, there’s been a few concerns coming from them as well… My manager [name]... thinks it’s a great idea [FSM auto-enrolment]... So, it’s good to have that support.” (LA12)

As information governance was a considerable barrier to implementation, one stakeholder shared this with their senior leadership, who consequently initiated discussions with national government representatives, aiming to clarify the government’s legal position.

“I have taken this up with our Director of Transformation who is trying to unblock this nationally. National government are looking at areas where national processes are creating barriers to local government work and he is keen to get this resolved.” (LA12)

Facilitators – internal and external support and collaborations.

Participants recounted three overarching facilitators that helped them overcome barriers: implementation support, external collaborations and document sharing.

Several participants emphasised the role of bringing relevant stakeholders together early in the process, which was often achieved by gaining support from senior leaders with the authority to mobilise teams. Guidance on which people needed to be involved came from other local authorities.

“Make sure that you get a really good coverage of people at the right level from the right services… [somebody from] a legal perspective… information governance… And then clearly you need… teams that are actually doing the job, so your revenues and benefits team, your customer services team… “ (LA19)

“In terms of the whole approach, it was very good that we had support from the top... I think that helped a lot in terms of mobilising people and steering this from the very beginning.” (LA17)

While senior leaders were involved in bringing teams together, others were critical in overcoming barriers, with one participant highlighting the importance of councillor engagement to influence teams.

“This particular elected member, as well as working with us, is also highly respected as an elected member within the rest of the council and across all other elected members, so that also helps to have a person that is influential politically.” (LA09)

Two participants found that having a small project team helped, as this meant that they did not have to rely on other people/teams.

“We were able to get things done quite quickly. There [weren’t] too many people that we were relying on.” (LA13)

“We’re a small team. So we can kind of have those [FSM auto-enrolment-related] discussions quite quickly.” (LA01)

Communication and knowledge sharing with other experienced local authorities also facilitated implementation, particularly when there were legal challenges, as this provided initial guidance followed by reassurance that stakeholders were implementing the programme correctly.

While support from internal colleagues facilitated implementation, local authorities were limited by their FSM auto-enrolment knowledge. As such, the majority of participants sought process guidance from experienced local authorities, many of which led to the development of new external collaborations and sharing of relevant documents.

“We did have a conversation with [local authority] and we had a conversation with other people about the processes they went through and the challenges that exist.” (LA09)

“Collaboration with other local authorities has definitely helped… we are kind of all learning from each other and more than happy to share best practice or… lessons learned… that has been key to it.” (LA01)

Often, implementation support came from FixOurFood webinars, which gave those implementing auto-enrolment the opportunity to receive tailored feedback from peers at a similar stage of implementation.

“Through early last year it was about… attending webinars, getting that information, trying to have a better understanding of what the issues are, what we might need to think about... tapping into the regional support that’s there.” (LA12)

As local authorities connected through FixOurFood webinars, some learned that neighbouring authorities were implementing FSM auto-enrolment, and used the opportunity to establish relationships with those who understand local challenges. For example, local authorities in one region used this as an opportunity to share information to better address local challenges, which ultimately led one local authority to formalising this arrangement through a data sharing agreement. By sharing data, this allowed local authorities to identify and check eligibility for children whose address had changed and check whether they were entitled for FSM.

“The nature of London is that boroughs are quite small, people don’t live within their borough boundaries, people cross over and some don’t actually realise that they’re doing it, and so being able to share information across two boroughs was helpful.” (LA13)

In situations where local authorities needed guidance, they made use of an available FixOurFood FSM auto-enrolment toolkit, which included templates and example documentation from experienced local authorities.

“Our process was very much influenced by the toolkit that’s in place as well as the [local authority] and [local authority] process.” (LA02)

Further, if and how stakeholders would use the FixOurFood toolkit depended on their stage of implementation, with some having implemented FSM auto-enrolment before the toolkit was developed. In these situations, stakeholders would use the toolkit retrospectively to check that they had included all the essential steps. Similarly, those earlier in the process found that the toolkit provided an overarching view and critical steps to implementation and in some instances, answers to specific questions.

“The toolkit does reflect how we’ve done it, and certainly what our intended path is going forward, it lines up very well with the toolkit.” (LA20)

“I think what the toolkit does is it shows you no matter where you are in the council… there are some parts that you can’t do it without.” (LA17)

Several participants found that example documents provided by experienced local authorities within the toolkit helped facilitate implementation, as this gave local authorities a starting document which they could adapt to their local context. The documents provided were considered essential to implementation, and included example legal documents, DPIAs and privacy notices.

“We sort of followed [local authority’s] process. We got the information from them. We got the letters from them. We tweaked the letters to suit us.” (LA03)