Child Poverty Practice Accelerator Fund evaluation: final report

5: A Systems Perspective on Sustainability

This section examines what the CPAF evaluation reveals about the conditions required for local approaches to reducing child poverty to be sustained over time. It addresses the evaluation’s research question on sustainability by applying a systems perspective to evidence gathered across CPAF Round 1 and Round 2 projects.

Much policy and evaluation practice is grounded in a linear, cause-and-effect model, in which interventions are assumed to produce outcomes through relatively direct and predictable pathways. While this approach can be useful for assessing delivery and short-term impact, it is less effective in explaining why improvements are sustained in some contexts but not in others. Evidence from CPAF suggests that local action on child poverty operates within complex social systems shaped by relationships, organisational cultures and pressures, professional judgement, data infrastructure, trust, and lived experience. Change within these systems is rarely linear. Instead, progress is influenced by feedback processes that can either reinforce improvement over time or lead to gradual erosion once attention, funding or continuity is lost. This section explores these themes and what they can tell us about the future of CPAF and related projects. It builds upon systems change research conducted already by Scottish Government, available here.

Key Points

How are CPAF projects contributing to system change within local services?

CPAF projects have contributed to system change primarily by strengthening key capacities within local systems, rather than through isolated interventions. Where projects reinforced joined-up coordination, continuity of support, ongoing dialogue and shared intelligence, changes were more likely to embed into routine practice and persist beyond the life of individual projects.

How and to what extent are public and partner organisations working in partnership across CPAF?

Partnership working across CPAF has been most effective where it was supported by clear coordination roles, shared purpose and regular cross-agency routines. In these contexts, collaboration moved beyond information-sharing towards joint problem-solving and collective ownership of outcomes. Where such structures were absent or fragile, partnership working was more dependent on individual relationships and therefore less stable.

How do grant recipients and local partners view working in partnership, and what are the successes, challenges and unintended consequences?

Grant recipients and partners consistently valued partnership working where it improved relevance, reduced duplication and enabled more coordinated support for families. Challenges arose where partnership relied on short-term funding, informal arrangements or unclear data-sharing rules, which increased workload and uncertainty. In some cases, well-intentioned local workarounds on data systems and insights create additional complexity, representing unintended consequences of operating within fragmented systems.

How and to what extent are values, cultures and behaviours changing among those who design and deliver services?

CPAF has supported shifts towards more person-centred, collaborative and learning-oriented ways of working, particularly where co-production and dialogue were embedded as ongoing practices rather than one-off activities. These cultural and behavioural changes were most evident where staff saw that learning was acted upon and sustained. Where dialogue or coordination was time-limited, changes in values and behaviours were harder to maintain.

How effective has collaboration been between and across CPAF projects, including across rounds?

Collaboration across CPAF projects has been strongest where learning, tools or roles were carried forward across time, organisations or rounds of funding. There was ripe opportunity for this in cases where the same local authority received both Round 1 and Round 2 funding for different projects. Where system capacities had been built up prior to CPAF, projects were able to progress more quickly and achieve greater coherence. However, limited mechanisms for formal cross-project learning reduced the extent to which learning could be consistently shared or scaled.

Overall implications for sustainability

Taken together, the evidence shows that sustainable impact depends on whether CPAF projects strengthen the underlying system capacities that support partnership, learning and continuity. Where these capacities reinforced one another, progress was more resilient. Where one or more were weak or disrupted, improvements were more likely to stall and could erode over time.

Representing a complex systems perspective

Taking a complex systems perspective represents a shift in how sustainability is understood. Rather than focusing only on whether individual projects or interventions “worked”, it considers how change emerges over time from the interaction of people, organisations, policies, resources and lived experience. From this perspective, progress is shaped not just by what is delivered, but by how different parts of the system respond to one another, reinforce one another, or work at cross-purposes.

This approach is particularly relevant to CPAF. Projects funded through CPAF do not operate in isolation. They sit within a wider child poverty action capacity that is itself embedded within broader systems of social welfare, public services, labour markets and economic policy. Decisions made elsewhere in this wider system — for example around funding cycles, data governance, workforce capacity or policy priorities — can have indirect but significant effects on what is possible locally. A systems perspective helps make these interactions visible, and helps explain why similar approaches can lead to different outcomes in different contexts.

The systems models used in this evaluation (contained in full in Annex C) emerged directly from the fieldwork, rather than from pre-defined frameworks or theoretical constructs. Through analysis of interviews, observations and project documentation across CPAF Rounds 1 and 2, recurring patterns were identified across very different local contexts. These included common pressures, enablers and points of fragility, as well as consistent accounts from stakeholders of how different elements of their work influenced one another and interacted with the wider system. Methodologically, this involved familiarising ourselves with the diverse data available, pulling put key themes and observations, engaging in group sense-making discussions to draw out recurring dynamics of the systems key informants and family participants observed, and comparing different projects and cases to test assumptions further.

The patterns we identified were synthesised into Causal Loop Diagrams (CLDs) as a way of representing the system dynamics described in the evidence. CLDs are visual tools that show how different factors influence one another over time. They focus on relationships rather than processes, and on patterns rather than linear sequences of steps. Arrows indicate influence between factors, and simple symbols show whether a change in one tends to increase or decrease another. When influences link together in a circle, they form a feedback loop. Some feedback loops reinforce change over time, while others constrain or counteract it.

Importantly, the CLDs are not intended as technical or predictive models. They do not claim precision or completeness. Instead, they provide a structured way of bringing together what project stakeholders consistently described about how systems behave in practice. Readers do not need to interpret the diagrams themselves to follow the argument in this section. The CLDs that underpin the analysis are provided in Annex C for those who wish to engage with the system dynamics in more detail. In the main body of the report, references to the diagrams are used only to support and illustrate the narrative.

Viewed through this systems lens, the CPAF principles of innovation, partnership and collaboration, evidence-based learning, and person-centred practice can be understood not simply as features of individual projects, but as qualities that must be supported and sustained by the wider system. The evaluation shows that these principles are most likely to endure where they are embedded into the culture, structures and routines of local systems, and most likely to dissipate where they remain dependent on time-limited projects or individual roles.

Rather than being attributes of individual projects, CPAF principles are realised through system capacities — enduring features of local systems that enable those principles to be enacted consistently, even as projects, staff or contexts change. Section 4 shows that impacts, and potential for further impact, have been strongest where systems change has begun to embed. The four capacities set out below describe how CPAF principles were translated into practice across different local settings, and what this reveals about the conditions required for sustainability.

Capacity 1: Joined-up leadership and coordination

Please note: the corresponding CLD and analysis section in Annex C provides further information on this capacity.

This capacity describes the ability of local systems to organise and coordinate action across services, organisations and sectors around shared goals for reducing child poverty. It reflects CPAF principles of partnership, collaboration and learning, and is most visible where there are clear coordination roles, regular cross-agency routines, and shared ownership of priorities and outcomes. As set out in Section 3, this capacity strongly shaped how CPAF projects were implemented in practice.

From a systems perspective, coordination is not a one-off activity but an ongoing process that strengthens or weakens over time. Shared purpose supports collaboration; collaboration enables joint problem-solving and learning; and visible progress reinforces commitment to working together. Where this reinforcing dynamic was in place, coordination became easier and more resilient. Where it was absent or disrupted, coordination was fragile and required repeated effort to re-establish.

Evidence from CPAF shows that this capacity was weakest where coordination depended primarily on informal relationships, short-term funding or individual champions without organisational backing. Staff turnover, unclear remits and competing priorities could quickly interrupt shared understanding and slow progress. By contrast, sustainability was strongest where coordination was treated as a core function of the system rather than an additional or temporary feature of a project.

Projects that built on existing partnerships and coordination structures therefore had a greater chance of success within the funding period. Where partnerships were still forming at the start of a project, a significant proportion of time was often absorbed by establishing ways of working, reducing the scope for delivery and learning within a fixed timescale.

Aberdeen City Council’s Round 2 project to co-design employability services for young parents illustrates this capacity in practice. A dedicated coordination role supported alignment across employability, financial inclusion and participation services, while the organisation facilitating the co-design was a pre-existing trusted partner with a clear understanding of local priorities. A change in the coordination role partway through the project created some disruption, but the breadth of partner involvement meant that knowledge and ownership of the work were sufficiently embedded to allow the project to continue.

Capacity 2: Ongoing dialogue with families and frontline practice

Please note: the corresponding CLD and analysis section in Annex C provides further information on this capacity.

This capacity refers to the mechanisms through which lived experience and frontline insight inform service design, delivery and improvement on an ongoing basis. It reflects CPAF principles of partnership, evidence-based learning and person-centred practice, and was frequently a deliberate innovation within CPAF projects. It includes the meaningful use of lived-experience panels, peer research, and structured feedback from practitioners.

From a systems perspective, dialogue performs a regulatory function. It helps ensure that services remain relevant, ethical and responsive as needs, contexts and constraints change. CPAF evidence shows that dialogue was most effective where it was embedded as a new way of working, with explicit recognition that families, practitioners and strategic partners each bring valuable and complementary forms of knowledge.

Where dialogue was treated as continuous and consequential — with clear routes for learning to inform decisions and changes in practice — systems began to adapt. Stakeholders reported increased confidence that their input was valued and acted upon, and services became better aligned with people’s reported needs. Where dialogue was limited to one-off consultations or information-gathering exercises, its influence was weaker, and confidence among participants was more easily eroded.

Embedding dialogue with families and frontline workers within governance structures was particularly important for sustainability. Doing so reduced reliance on individual projects to surface insight, and increased the likelihood that learning would continue beyond the funding period.

City of Edinburgh Council’s Round 2 project provides a clear example. The project embedded both lived-experience and staff panels into the development of anti-stigma videos, ensuring that content was shaped by a wide range of perspectives and refined over time. While the staff panel was newly established, the lived-experience panel built on an existing, trusted partnership with The Poverty Alliance. Although the films were still in development at the time of writing, there was strong shared confidence that the process itself was influencing how partners worked together.

Capacity 3: Trusted, continuous support in communities

Please note: the corresponding CLD and analysis section in Annex C provides further information on this capacity.

This capacity concerns the ability of services to maintain a consistent, trusted presence with families over time, and to respond to needs in a holistic way. While this often relates to geographic communities, it can also apply to communities defined by shared characteristics or experiences, such as families where a household member is disabled. This capacity reflects CPAF principles of partnership, collaboration and person-centred practice.

Systems analysis of CPAF projects shows that trust, local knowledge and informal networks build cumulatively. Consistent presence — for example, a welfare rights adviser attending the same community venue at the same time each week, or an income maximisation worker embedded within a specific service — increases familiarity and accessibility. As relationships develop, positive experiences are shared informally, strengthening referral pathways and reducing barriers to engagement.

Continuity also supports professional confidence. Where clinicians, teachers or third-sector staff could rely on known referral routes and trusted contacts, referrals became more routine and timely. Over time, this reduced duplication and improved coordination across services.

Short-term funding, staff turnover and gaps between projects disrupted this capacity in some CPAF areas, leading to loss of trust and local intelligence. Re-establishing presence required significant additional time and effort. The evidence suggests that continuity is therefore a precondition for prevention and efficiency: protecting frontline roles and relationships over time supports earlier intervention and reduces longer-term costs.

South Lanarkshire Council’s Round 1 project with Home Start illustrates this capacity well. An income maximisation worker was embedded within Home-Start, providing parents with a named and trusted point of contact. Support was offered in familiar settings and through flexible modes of contact. This approach increased parents’ confidence in seeking advice and enabled Home Start staff to make referrals knowing that families would receive timely, collaborative support.

Capacity 4: Shared intelligence and rules that enable action

Please note: the corresponding CLD and analysis section in Annex C provides further information on this capacity.

This capacity relates to how information about families, services and need is created, shared and used across the system. It includes data systems, consent processes and policy frameworks that determine what information is visible, who can access it, what can be shared lawfully, and how confidently services can act together. It reflects CPAF principles of innovation, collaboration and evidence-based learning, and when functioning well supports more person-centred practice.

At a practical level, this capacity is shaped by a set of everyday tools and processes that have a significant influence on what is possible in practice. These include:

• Customer Relationship Management systems, which are used by councils and partner organisations to record contact with families, track referrals and monitor actions taken. They could be specific to one council service, operate across several or all with a customer-facing element of their work, or even be shared systems with third sector organisations as well;
• Dashboards or trackers, which are developed to bring together data from different sources to provide an overview of need, service reach or outcomes. For example, they may include council-held data on rent arrears or council tax reductions, alongside council-level data from national sources such as benefit claims from the Department for Work and Pensions (DWP). These are often represented in the form of a map with layers of information that can be seen individually or together and help with decision-making. They also often enable the download of specific datasets to suits different needs and purposes; and
• Data-sharing and consent arrangements, which govern how information can move between services and be used to coordinate support. These arrangements influence what might be on a CRM system or dashboard or tracker, as well as who can see it and for what purpose use of specific data is lawful. It may be that use of data for one purpose is lawful, while for another it is not.

When these elements are aligned and trusted, they enable services to work from a shared understanding of need, coordinate responses more effectively, and follow through on support in a coherent way. For families, this can reduce repeated explanations and improve timeliness of support. For practitioners and partners, it supports confident decision-making, reduces duplication and strengthens learning over time.

The CPAF evaluation found that this capacity has significant potential but is often fragile. CRM systems are frequently fragmented across departments or organisations, use different data standards, or cannot communicate with one another. Dashboards and trackers can be costly to maintain, dependent on third-party suppliers, or require substantial manual input. Uncertainty about lawful data use and inconsistent interpretations of data protection legislation can make staff cautious, limiting data sharing even where it would be appropriate.

Local teams often develop workarounds to address these challenges, such as parallel datasets or bespoke reporting processes. While helpful in the short term, these approaches add complexity and are difficult to sustain. The evaluation suggests that many of these barriers are structural and cannot be fully resolved through local action alone.

West Lothian Council’s Round 2 project demonstrates both the potential and the limits of this capacity. Through close collaboration between analysts, frontline staff and a lived-experience panel, the project developed a dashboard that was repeatedly refined to reflect practice reality and support targeted action. The involvement of The Improvement Service, operating at both local and national levels, was critical in enabling learning to be shared and structural issues to be surfaced. At the same time, wider data fragmentation limited what even a well-designed local solution could achieve.

For this capacity to support sustainability, shared data infrastructure and clear, consistent rules for data sharing are required. National leadership is needed to address structural barriers, provide clarity on lawful data use and support interoperability across systems.

How the capacities interact to support sustainability

The systems analysis shows that no single capacity is sufficient on its own. Sustainability emerges where capacities reinforce one another over time, as illustrated by the feedback dynamics set out in Annex C:

• Joined-up leadership and coordination are strengthened by trusted frontline intelligence;
• Trusted and continuous support is easier to maintain when supported by shared data and organisational routines;
• Ongoing dialogue ensures that strategy and practice remain aligned with lived experience; and
• Shared intelligence systems support confidence, collaboration, integrity and learning across the system.

Where one capacity weakened, the effects were felt across others, helping to explain why progress could stall even if individual projects are well delivered.

Taken together, the evidence in this section shows that CPAF has contributed to meaningful system change within local services, particularly by strengthening partnership working, supporting more person-centred and learning-oriented approaches, and enabling greater coordination across organisations. These contributions are most visible where CPAF projects reinforced existing system capacities or created the conditions for those capacities to develop early and consistently. At the same time, the evaluation highlights clear limits to what CPAF projects can achieve in isolation.

Additionally, collaboration between CPAF projects — across local areas or across rounds — was most effective where people, tools or coordinating organisations connected learning over time. However, the evaluation found no consistent, programme-wide mechanisms for cross-project collaboration. Project leads expressed strong appetite for this, and each round benefits from at least one event to draw people together. However, beyond this learning was shared unevenly and often depended on individual initiative rather than design.

This systems perspective helps explain variation across CPAF projects. Projects operating within stronger pre-existing systems were able to progress more quickly, collaborate more effectively and adapt to challenges. Others used CPAF funding to begin building these capacities but faced constraints arising from funding timescales, workforce pressures and fragmented data infrastructure. Cultural and behavioural change among staff was evident where dialogue and learning were continuous and acted upon, but was less likely to persist where these practices were time-limited.

Overall, the evaluation suggests that CPAF has been effective in demonstrating how local systems can work differently to reduce child poverty. Its longer-term impact depends on whether the capacities it has helped to strengthen are protected, reinforced and connected over time, and on the extent to which national and local policy frameworks support this continuity.

Implications for future CPAF and wider child poverty strategy

Viewed through a systems perspective, the findings from this evaluation suggest that future CPAF investment should prioritise system viability alongside innovation. Sustainable progress in reducing child poverty depends less on the delivery of individual projects, and more on whether the underlying capacities that enable coordination, continuity, dialogue and shared intelligence are intentionally supported and allowed to strengthen over time.

Key implications for the design, funding and evaluation of future CPAF-funded work include:

• Design for continuity, not only pilots. Multi-year support for coordination and frontline roles is critical to sustaining trust, shared learning and effective partnership working. Where continuity is disrupted, systems repeatedly lose capability and require additional time and resource to rebuild. This points to the value of longer-term funding approaches and early consideration of how system improvements will be maintained beyond the life of individual projects.
• Embed co-production as routine practice. Ongoing dialogue with families and frontline practitioners should be resourced and expected as part of business-as-usual decision-making, rather than confined to discrete project activities. This can be supported nationally through funding requirements and guidance, and locally through investment in skills, facilitation capacity and clear routes for learning to influence decisions.
• Invest in shared intelligence as enabling infrastructure. CPAF evidence highlights the need for coordinated national leadership on data-sharing, consent and interoperability, to enable local systems to act confidently on shared insight. Without this alignment, local innovation risks being constrained by fragmentation and reliance on fragile workarounds. Organisations operating across local and national levels, such as The Improvement Service, have an important role in supporting this alignment.
• Commission and evaluate for learning and durability. Evaluation approaches should place greater emphasis on whether systems are becoming more adaptive, coordinated and resilient over time, rather than focusing solely on short-term outputs or attribution. This includes supporting longer-term evaluation horizons and building local capacity to monitor and reflect on system-level change.

Summary

Taken together, these implications point to a shift in emphasis from funding discrete interventions towards supporting the conditions that allow effective practice to endure, adapt and spread. By strengthening the system capacities identified in this section, CPAF and related programmes are better placed to support lasting reductions in child poverty.