Nick Kimber highlights the urgent need for better collaboration between Whitehall and local communities
At the beginning of last year, a small team of public sector employees in Sheffield celebrated a fourfold increase in the number of children applying for early years, pre-school literacy support.
What was unusual about their achievement was not so much the exponential increase in take-up as the methods used and the cross-functional and cross-departmental composition of those involved: individuals drawn from the local council (Sheffield’s family hub services), the Mayoral Combined Authority, the Department for Education’s (DfE) Behavioural Insights team, the Ministry of Housing, Communities and Local Government (MHCLG) and the Cabinet Office.
In just 12 weeks the Sheffield team identified the problems, listened carefully both to users and those with existing frontline perspectives, then tested changes on a small scale before further iterating, scaling up or stopping various experiments. The end result has been a series of well-targeted interventions to help children take those vital first steps towards reading and writing. Since then, as part of the Government’s Places for Growth programme, the council, Mayoral Combined Authority and the Sheffield policy campus have doubled down on their collaboration, pioneering new approaches to sharing data that enable all partners to better understand what happens to young people who are not in education, training or employment (NEETs). The approach is putting down deep roots in the region.
These South Yorkshire innovators, it should be stressed, weren’t initially part of a multi-million pound programme with all the bells and whistles that come with that sort of initiative. Some of them were building on a lifetime of expertise working with families, others were walking into a family hub for the first time. Their ways of working involved workshops which allowed them to better define the problem, as well as ‘play, chat and learn’ sessions at schools and nurseries through which they gathered varied perspectives and tested different intervention styles.
The team hasn’t cracked the problem for the whole country – not even for South Yorkshire. But if others like them scattered around the UK could adopt their experimental and innovative approach, chances are that this Government would be well on the way to delivering on one of its national priorities.
Such test and learn methods – as well as co-design and relational ways of working – are already familiar in many areas of local government and the voluntary sector. But with the notable exceptions of the Social Exclusion Unit in the 2000s and the Government Digital Service more recently (at least in respect of its digital content and transactional services), they are not everyday practice in most other public sector organisations and systems. In central government, test and learn is still the preserve of a handful of exceptional teams, pepperpotted across Whitehall and often siloed within the work of larger units.
In my view, the Sheffield example gets to the heart of reform of our public services and provides a template for how central government can deliver rapid change, in partnership with local communities and in ways that make a real and immediate difference for citizens.
The gap between policy and delivery
In October 2024, Chancellor Rachel Reeves announced the establishment of the Test, Learn and Grow (TLG) initiative, partly a programme, partly a coalition for change inspired by the sense that a chronic failing in the UK is the gap between policy and delivery. That was followed by Pat McFadden’s landmark speech in December of that year, which set out this Government’s interest in the ‘how’ as well as the ‘what’, serving as a call to action to public servants to heed Prime Minister Keir Starmer’s oft-stated view that those on the frontline, encountering the same patterns day in and day out as they support citizens, have the best understanding of what works.
The thinking behind Test, Learn and Grow is that by starting small with any intervention and ensuring that you test assumptions and produce real world evidence before investing further time, money and attention in any given course of action, you create more value for citizens more quickly. By involving practitioners alongside policy generalists from the beginning, you achieve better results within policy, regulation and funding. By working alongside citizens and civic society, you increase trust and legitimacy, reducing risk and speeding up delivery.
In this way we can begin to address the scepticism about government services and institutions felt by citizens and communities not just across Britain but across most democracies. The Test, Learn and Grow programme has been shaped by the interactive methods used in design and digital communities, by a long-term trend towards devolution and by the rise and maturing of place-based working as a distinct and valuable practice. It sits alongside and reinforces structural reforms led by MHCLG which seek to stabilise local government and then renew the long-term capabilities of the local state.
Working initially across 10 local authority areas in England, Test, Learn and Grow will help support policy delivery, scaling up prototypes and – crucially – learning about what gets in the way and what the centre can do to reduce barriers. Partners represent a mix of communities: cities, towns, rural and coastal communities. After the initial design phase (including the work above with Sheffield), the programme officially launched in July last year and in June 2026 the Government announced a £10m fund to boost regional test and learn capacity.
Growing the conditions for change
The big danger of Test, Learn and Grow is that pilots stay in the realm of the exceptional, never quite moving to lasting, system-wide improvements. So how we ‘grow’ the work is essential. We are thinking about this in three ways.
Rapid diffusion
How do you share the learning about the Sheffield team’s achievements in a way which means organisational practices proven to work can be adapted and adopted in, say, Blackpool or Brent?
One problem is that while places which invest in evaluation and socialising learning are, in effect, creating a public good, they are generally not incentivised to do this. They have to do so on top of the hard work of delivering core services. That’s why it rarely happens. National government could help provide these system-wide incentives, look for new ways to experiment and be an active participant in this learning.
We need to give this role in learning a name, create reasons to be involved and build a strong platform to support it. Nick Pearce and Gavin Kelly have written about the Beveridge Report and explain how it is often misunderstood as the work of a single heroic reformer in Whitehall. On the contrary, they point out, Beveridge was built on the foundations of decades of experimentation – often at the local level – and of seeing in devolution (both to the nations and to regional and local government) an opportunity to drive change and a more pluralist approach to policy development.
The new £10m Test, Learn and Grow Capability Fund will provide up to 20 locations with £500,000 each over two years to build their own public sector innovation capacity. Rather than treating these places as mere testing grounds for Whitehall, the fund empowers an initial cohort to act as regional convenors. Their resources will grow local skills, run new test and learn projects and diffuse insights across wider regions and sub-regions.
Historically, most pilots designed and funded at the centre are top-down and one-way – something decided on in Whitehall and tested in local areas before being rolled out nationally. We need to turn this logic inside out, with central government rolling up its sleeves and working alongside localities and communities, learning together and applying the lessons in partnership.
Formal rules and frameworks
While they may be useful and necessary for big top-down public sector projects, formal rules and frameworks do not necessarily encourage a Test, Learn and Grow approach.
Take the public sector business case process, intended to help secure value for money and drive better decision-making. The current guidance (the Green Book) and the way it is interpreted works in the appraisal of a certain category of project and programme. Specifically, it is suited to relatively predictable and controllable environments, where both the inputs and intended outcomes are largely fixed over time and there is a clear moment of ‘completion’. This could be a physical infrastructure project like a bypass. In such cases, spending significant time upfront to develop a business case based on comprehensive assumptions about requirements makes sense – investing this energy upfront mitigates financial and delivery risk later on.
Where the current Green Book norms (what they say but also how they are interpreted and applied by practitioners) work far less well is for complex and unpredictable system-level change, where the inputs and outcomes are moving targets, where there are high levels of uncertainty and no end point when the programme is truly complete. Many, if not most, initiatives to radically improve policy, services and outcomes for citizens fall into this category. All test and learn efforts do. Crucially, these challenges are often more about the way the Green Book is used rather than the guidance itself. The Green Book is not just about calculating a straightforward ratio of value of benefits to value of costs. It is about carefully defining the outcomes that we want to achieve and identifying the best projects or policies to achieve those outcomes. It involves developing a theory of change, assessing deliverability and risk and ensuring ‘strategic fit’ with other government initiatives.
Rather than focusing upfront efforts on predicting all inputs, costs and benefits over the phases of any system-level programme, we would be better to assess whether these programmes are clear on outcomes, hypotheses, the nature of the opportunity, alignment of stakeholders and the availability of the right types of resources. These principles which are already central to good Green Book practice and the five case model but not widely practised. Reform is in train, with significant recent changes to simplify the Green Book, to ensure it takes into account place-based priorities and a major programme to simplify controls through Project Reset. Moreover, the Government has just released an update to the Magenta Book by introducing a dedicated new annex on test and learn. This is a massive shift that supports the collection of robust yet proportionate evidence at every iterative stage of delivery. Embedding these steps is key, as is going further so that non-digital policy and delivery is supported by frameworks that acknowledge agility and iteration.
Funding and accountability
Another challenge lies in the UK’s funding and accountability frameworks. Currently, these are overwhelmingly constructed along departmental lines on narrow parts of the citizen’s service journey. Too often the broader outcome is missed, users find services disjointed and we intervene later than we should to help people with complex needs. For the same reasons, those on the frontline have only limited flexibility on how to do things differently and are constrained in how they respond to what service users need. We must place more emphasis on improving awareness of these cultural barriers, as well as seeking mechanisms to tackle them.
With this in mind, TLG is collaborating with MHCLG to work with localities and explore how we can pool budgets to give more flexibility to the frontline, making it easier for local places to join up and apply a test and learn approach. This intervention is one of a portfolio of initiatives aiming at integrating and joining up services at the local level, alongside a new outcome framework for local government, Integrated Settlements (a simplification of funding) and programmes like Changing Futures focused on supporting citizens experiencing multiple disadvantage. These examples show how the existing framework is evolving rather than setting up new systems from scratch. Integrated Settlements, incidentally, will enable Strategic Authorities to operate with increased local flexibility and more streamlined accountability. Furthermore, new measures in the English Devolution and Community Empowerment Bill aim to enhance local accountability and scrutiny alongside increased mayoral powers.
A coalition for change
If test and learn is to take root, the centre should reflect especially hard on how regulation can promote learning while at the same time providing the necessary assurance. One step might be for regulators to build a picture and tell a story of what is working well and not so well in a locality, and why.
Existing practices like the public sector business case, the funding framework and regulatory inspection have a rationale. They exist, like GK Chesterton’s fence, for a reason: erected with good intentions and put in place in response to legitimate concerns (the safeguarding of vulnerable citizens, maintaining spending within budgets and ensuring accountability for public expenditure). But there’s a growing case that some of these practices now obstruct rather than support public goals. In complex domains – most of what we are dealing with when we’re thinking about public services – risk is better managed by iterative planning based on real world information.
I have spent 15 months as a civil servant – having spent 15 years in local government – and am beginning to gain a basic fluency in my ‘second language’. There is much that is different between life in local government and the wider public sector, but what is clear is that everywhere in Whitehall you will find good, caring civil servants motivated by public service. The problem is that the change they want to see is still largely dependent on individuals taking action themselves, within a system that is often risk averse. Change will not come through a chain of command or traditional levers, certainly not at the pace it is needed. It will require a movement, not just of the 500,000 people employed by the UK Civil Service, but of the 5 million public servants across the country and the millions more working across civic society.
The good news is that there is already a plethora of networks – from the Better Way Network, to One Team Gov and New Local’s community power network – sharing remarkably similar values and principles. Test, Learn and Grow recently launched its own complement to this (further information at the end), convened by a series of high-profile public servants from central and local government whose involvement makes a powerful point about the two-way bridge between Whitehall and communities and the need for collaboration. They include Beth Russell, Second Permanent Secretary at the Treasury, Catherine Howe, Chief Executive of Dorset Council, and Emma de Closset, Chief Executive of UK Community Foundations. Janet Hughes, former Director General of Public Sector Reform at the Cabinet Office, has also been one of our convenors and is intending to rejoin after her sabbatical in the autumn.
My Home Office colleague Paul Morrison (also one of our convenors and formerly Chief Executive of the Planning Inspectorate) recently set up a network of more than 100 Whitehall directors, all interested in new ways of working and collaboration. Through it, he’s championing a mindset he calls “Coming Home”, a bottom-up approach that aims to connect all members of the senior civil service to a place to which they have an emotional connection, using their experience and networks to add value to a community they love. The Pride in Place programme provides a similar opportunity for the whole of central government, not just those working directly on a ‘programme’, to connect to the 200 or so neighbourhoods who will be involved, so that as a civil service we think ‘neighbourhood and community’ first.
Change is an emotional project, not just a technical one. You can’t enact change from behind a desk and you can’t solve problems through deft drafting without experience of, and a feeling for, the matter at hand. Remember the words of the woman who gave life to arguably the most positive representation of a public servant this century (and a local government officer no less): Amy Poehler. No doubt with Leslie Knope at her shoulder, she wrote that “The doing is the thing. The talking and worrying and thinking is not the thing.”
Nick Kimber is Director of Public Service Reform at the Cabinet Office, seconded from local government. You can join the TLG community on LinkedIn.
