Setting and purpose

On 19 May 2026, a roundtable was held as part of the Centre for the Edge initiative, at the offices of Demos in Whitehall. The Centre for the Edge is a collaboration between Kinship Works and the Joseph Rowntree Foundation. This was the third roundtable in the series; the first two explored the risk that the evidence standards used in government narrow our field of view, and the institutional monoculture that has taken hold in the public sector.

This third discussion explored the differences in worldview and mentality between organisations at the centre of government and those innovating around the edges. What specific differences do we see, as reflected in the language people use? And what can be done to stop these differences from being a barrier to institutional renewal?

The roundtable brought together practitioners, policymakers, senior public sector leaders, academics, and innovators from across central and local government and civil society. Two provocations were shared in advance:

• The centre is from Mars, the edges are from Venus, by James Plunkett, Kinship Works

• Dancing in the messy middle: institutional immunity and the challenges of bridging, by Sophia Parker, JRF

The discussion was held under Chatham House rules; in the write-up, comments are unattributed, with the exception of the opening remarks.

Summary of discussion

Sophia Parker (JRF) chaired the discussion and opened with a short overview of the Centre for the Edge and an introduction to the theme of worldviews. Frictions between the centre and the edges often seem to go beyond one-off misunderstandings, or frustrations about bureaucratic processes, and instead reflect deeper and more systemic divergences in worldview. These differences are then reflected through, and complicated by, imbalances of power between the centre and the edges.

James Plunkett (Kinship Works) opened the conversation by sharing headline findings from a discourse analysis carried out by Kinship Works. The analysis revealed in new detail the different vocabularies that are used between organisations at the centre of public life and those innovating around the edges. Two corpuses of documents were assembled - one from organisations in and around the centre (HM Treasury, the National Audit Office, government departments) and another from organisations innovating at the edges. The vocabulary in each was then compared, exploring what differences in language might reveal about differences in mentality.

The starkest divide was between a logic of production at the centre - deliver, implement, roll out, milestones, fidelity, pilots, RCTs - and a logic of community and ecosystem at the edges; a language of movements, field-building, practice, and relationships. These seem to be not just different vocabularies but incommensurable ways of conceptualising the work being done, and the systems within which the work takes place.

The word "accountability" appears at near-parity in both corpuses, but the surrounding language reveals two different conceptions of what accountability is. At the centre, we see words like audit, inspection, compliance, control, and fraud. At the edges, we see power-sharing, deliberation, legitimacy, and democracy.

There is also an imbalance of recognition and awareness. The labour of translation between centre and edge seems to be unevenly distributed: the edges write extensively about how to engage the centre, while the centre barely registers the edges.

The emotional register of the two corpuses also differs profoundly: at the centre, there is a vocabulary of anxiety, error, failure, and risk, while at the edges people talk about hope and inspiration. The centre tends to focus on how something can be done, while the edges emphasise why.

A handful of words appear at roughly equal frequency in both corpuses - transformation, innovation, partnership, governance, wellbeing - though these sometimes carry different meanings. Perhaps these shared terms offer a patch of common ground for productive conversations.

The analysis led to three categories of recommendations. First, creating more space for the centre and edges to work together on areas of divergence - accountability, procurement, risk, translation - rather than each ‘side’ leaving the other to bear the weight. Second, developing shared frameworks on issues where there are different conceptions; methods for scaling, for example, or the question of what constitutes good evidence. Third, making the public sector itself more pluralistic and tolerant of difference - use more mixed-discipline teams, more permissive procurement methods, and creating more space for unconventional institutional forms to survive and thrive.

In the main discussion, ten themes emerged.

1. Centre and edge as relative positions, not fixed places

A recurring observation was that "centre" and "edge" are relative positions, not fixed identities. As one participant put it: "sometimes we're the edge, sometimes we're the centre, depending on who you're asking." A large local authority might feel ‘edgy; in the context of its wider city-region, but seem thoroughly bureaucratic to local residents. A strategic authority that experiences itself as innovative might look like central government to a small community organisation.

One participant offered the image of an onion, with successive layers of "centre" and "edge" nested inside one another. This matters because it speaks to the different roles organisations play in the system - for example, a Strategic Authority can win the trust of central government because it 'looks like a mini government', but really it is thinking quite differently and sees its role as stewarding a local ecosystem of innovators.

This suggests that the gap in worldviews is not a divide between two camps but is reproduced at every scale of the system, and that the same individual can find themselves on different sides, depending on context. It also means that institutions sitting 'in between' the centre and edges - combined authorities, strategic authorities, multi-academy trusts, large local authorities with experimental cultures - can play distinctive translation roles, because they are required, by their position, to look both ways.

2. The trouble with bridges; who builds them, and where to?

The metaphor of a ‘bridge’ from the centre to the edges came up repeatedly, and was also challenged. When we talk about building bridges, we tend to imply we are building a bridge ‘to the centre’; the centre holds the dominant worldview, and the bridging is towards this view. The edges are asked to explain themselves, to translate their work, to make themselves legible to a worldview that is not their own. This is exhausting, asymmetric work, and something is lost in the translation. The radical underpinnings of edge thinking often do not survive the journey, leaving the centre with what one participant called "the fruits and flowers of innovation, but not the soil in which the innovation was grown."

It was suggested that the metaphor of dancing might serve better than bridging - a relational, two-way exchange in which both parties move. But even that may be too optimistic. For some edge organisations doing genuinely paradigm-shifting work, the most useful posture may be not to dance at all, but to shield and protect their work from the centre, until it is less fragile. The challenge for the centre is to recognise that even an offer of partnership can be assimilative, and that allowing edge work to remain unbridged may sometimes be the most useful thing it can do.

A helpful perspective came from a participant with a background in negotiation and peacebuilding in post-conflict zones. There, the lesson was that bridges must be built with precision, to the right people, or the right parts of the system. The "centre" in any negotiation will be made up of camps: those who are quietly pro-change and who have internal legitimacy, those who are sympathetic but are afraid to ‘come out’, those who are ideologically committed hardliners, and those who are hardline only for tactical reasons. Building a bridge to the wrong person achieves nothing; building one to the right person can unlock everything. So maybe the key is to find the ‘edgy’ people in the centre, and to work with trusted connectors - like peace-building’s mediators - to hold space in between, lowering the transaction costs and the risk of engagement. The example of Audrey Tang's "horizontal ministry" work in Taiwan was offered: it created a third space that made it safe for those in the centre of government to engage with ideas from the edges, in a way that gave both sides confidence.

3. What it takes to work in the messy middle

A lot can be learned from work that has happened in the messy middle - the spaces where big institutions and edge innovators are already trying to work together. One participant shared insights from their work in this space.

First, collaborations between the centre and edges have to focus on 'doing real work'. You cannot just have workshops to improve relationships. You have to bring the centre and edges together around tangible work - something everyone wants to see happen, and name the barriers together. Greater Manchester was offered as an example of leaders across the system working together to deliver concrete work, and in the process naming and overcoming the barriers that were stopping them. One important lesson: often it turned out the barriers had been self-imposed - people were following rules that turned out to just be habits or conventions. This realisation unlocked people’s agency, even without changing rules.

Second, it is helpful to frame the work as an experiment, and not as 'reform'. The language of experimentation takes the pressure off, and gives participants a sense that they are finding a way forward together.

Third, you have to ground the work in shared purpose, and this starts not with technical or institutional questions but with the motivations that brought people to public service in the first place. This is where you find common ground, even between people with divergent worldviews.

One further observation: this work is very emotionally demanding. The people leading change - particularly those in the centre making themselves vulnerable to new ways of working - often need personal support. One-to-one coaching is a small but significant ingredient in the most successful messy-middle programmes.

4. Theoretical underpinnings and tensions with relational work

Some contributors noted that the centre's dominant worldview is not just a reflection of habit or culture; it has a basis in theory. New Public Management, and behind it public choice theory, gave the centre a coherent operating logic for thirty years; the language of audit, inspection, performance, compliance, and delivery flows from this. It can be helpful to make this explicit, so that we can ask: where are these theoretical underpinnings still serving us well, and where are they not?

In this context, the notion of "delivery" came under scrutiny. At the centre, public service feels like ‘delivery’ and the system is there to 'implement policy'. But for people in front-line public service, their work does not feel like delivery at all; it feels like building relationships with people. Seen this way, policy can be enabling and it can be constraining, but either way it is secondary to the relational core of the work. As long as "delivery" remains the master verb of the centre, no amount of relational language at the surface will close the gap. As one participant put it, "you cannot deliver a relationship".

5. The importance of daily practice to worldviews

Another participant argued that worldviews are not stable abstractions. They are aggregations that arise from daily practice, so they depend on who is in a room on a given day, and they are shaped by the way people, and especially leaders, behave. Daily practice produces, and changes, worldviews: this means practical questions like the cadence of meetings, the order in which people speak in plenaries, hiring decisions, feedback mechanisms, the transparency of decisions, all matter a lot. If you change these things, worldviews can start to shift in response. This means we can shift worldviews not (only) through argument but also through practice. Worldviews accrete, slowly, from the way people work together.

6. The central issue of accountability

The discourse analysis had flagged accountability as one of the sharpest sites of divergence. Participants agreed that different interpretations of accountability were a key source of friction. Several contributors noted that today’s dominant model of accountability - heavily reliant on compliance and targets - has proven its limits, especially in complex domains where we see widespread gaming of targets and "accountability theatre". Some organisations in the public sector are now exploring a greater focus on learning; asking people "what have you learned today?", "how did you learn it?", and "how have you adapted your practice as a result?" The challenge, however, is that today's compliance-oriented approach to accountability runs down from Parliament through ministers. These issues therefore touch the heart of the way accountability plays out in a parliamentary democracy. This does not mean we can't make progress, but it does mean we need to take this issue seriously. To make progress will require shared contributions from people at the centre and the edges, rather than just leaving accountability as a challenge for people in the centre to solve.

7. Outmoded tooling for information and intelligence

Many of the tools that are used to manage the relationship between the centre and the edges (procurement processes, contracts, accountability mechanisms, reporting templates) are still designed for a paper-based world. There is therefore a significant engineering dimension to this challenge, and this is more tractable than the relational and political aspects of the work. We could get a long way by redesigning these processes, information flows, and forms of shared intelligence for a digital world.

8. The question of power

Several participants said we need to talk more directly about the underlying distribution of power. One contributor noted that, when the going gets tough, the centre can walk away from a partnership and survive, while many community organisations depend on the centre for their existence. If the language of collaboration, dancing, and bridging is going to mean anything, the edges will need independent sources of power - financial resilience (e.g. endowments), organisational scale, and their own sources of political legitimacy. The edges need enough power to be a genuine partner, and not just a local delivery arm of the state.

Another contributor pointed out that different parts of the system hold different sources of legitimate authority. Central government holds democratic legitimacy from national elections, and local government from local elections. Community groups have authority through their proximity to the people they serve. Each of these sources of authority are legitimate, and each is partial. We should acknowledge these as parallel legitimacies that need to be reconciled - sometimes through devolution, sometimes through deliberation, sometimes through co-production.

A further perspective was that power often hinges on the definition of value. In commissioning, evaluation, and reporting, the centre defines what counts as success, and this definition then cascades down, becoming progressively thinner and more impoverished as it passes through layers of the system. Community organisations end up telling a thin "output" story to satisfy the centre's metrics, while a richer story of what actually happened - relationships being built, places being transformed, lives being changed - is lost. In these ways, power imbalances flow through daily procedures, and this has a profound impact on what it feels like to work at the edges.

9. The radical vision of the edges and the risk of co-option

On the question of power, it was noted that many edge actors do not just see their work as "innovative" or "relational" but as explicitly anti-capitalist. This is the political frame within which their work is grounded, and it envisages a fundamentally different type of world. The corollary is that the centre's worldview is seen to be liberal, or neoliberal. Naming these frames matters not because it’s helpful to have abstract arguments, but because if we’re not honest about these differences we will misdiagnose the divergences we’re seeing. We will also risk the centre adopting edge language while failing to embody edge practice, and this will be experienced by some at the edges as a form of "epistemic violence". So we cannot gloss over the fact that many on the edges are working along these more transformative, regenerative, redistributive, justice-driven lines, while the centre is more focused on amelioration and augmentation.

10. Loss, and the politics that lie ahead

Some participants worried that the conversation was too modest for the moment we are in. One contributor with long experience in central government argued that the worldview at the centre is not just out of date but is collapsing. There is, they argued, no longer a coherent shared worldview in the centre - no stable set of national goals, no agreed account of what the state is for, no theory of state capacity to replace the one we have lost. This has shown up repeatedly in major policy and fiscal moments - the only sense of strategic direction had emerged by accident from a thousand small decisions, and not from any deliberate vision. Another participant challenged this view, arguing that there is a coherent worldview at the centre, but it just goes unspoken; it serves as "the water we swim in".

Either way, the implication is that the work of articulating what the centre is for, now, at a national level, remains to be done. The upside is that this leaves space for a new vision of the centre’s role; the downside is the risk that this space could be filled with something more authoritarian. The risks and opportunities are civilizational; this is not just a question of delivery, or bureaucratic effectiveness. The energy in discussions like this one should reflect the scale and urgency of this challenge.

On these closing themes of power and crisis, a practical suggestion was that we should take seriously the emotional weight of this work, and the sense of fear and loss it carries. For people at the edges, the worry is that their work will either lose the battle, or else will be diluted and co-opted as it becomes mainstream. For people in the centre, meanwhile, there is a different and equally fraught sense of loss as the system breaks down - a fear of losing professional identity, and losing an operating logic that has framed whole careers. Both losses are real and should be acknowledged.

These themes, related to power and politics in the context of institutional renewal, will be the focus of the next outputs and roundtable in the Centre for the Edge series. Beyond this, the initiative is open to partners, and to seeding practical experiments. As Sophia Parker noted in closing, the discussion had come full circle to where previous roundtables had landed: the real work happens not in abstract conversations, but in experiments, on real problems, between people in the centre and those at the edges.