The Radical How

How to run a mission-oriented government like you mean it

15 min readFeb 21, 2024

What would it take to run a government capable of driving progress on big societal missions?

In this post I’ll share some insights from a new report, The Radical How, co-authored by Nesta and Public Digital.

The report uses case studies to describe the ways of working that would be required of an outcome-based or ‘mission-oriented’ government.

I’ll start by situating the report in the wider debate about the ‘how’ of government, before zooming in to its more detailed conclusions.

A new ‘how’ of government

There’s growing interest in the ‘how’ of government, running across party lines and spanning countries from the UK to the US to Australia.

This has been prompted by a range of factors:

The realisation that we’ve forgotten how to build pretty much anything
The way we’re struggling to get traction on today’s biggest challenges, from the collapse of ecosystems to chronic ill-health to regulation in digital markets — and a sense that this is because these challenges are novel/different
A series of technology-related crises — Horizon, Robodebt, the early missteps of Healthcare.gov — that have flared up, against the smouldering backdrop of the polycrisis
The accelerating horrorshow of climate change

This gives rise to the view that we don’t just need more state capacity, but different types of state capacity.

Specifically, we need new ways of governing that are:

Better suited to contemporary challenges — challenges that seem more complex, systemic, and human
Better able to put our most powerful technologies and insights to use

We don’t just need to govern better, we need to govern differently.

Mission-oriented government

In the UK, these debates have found a focal point in the idea of mission-oriented government.

Although it can be a bit shape-shifty, I think this idea of outcome-based or ‘mission-oriented government’ is onto something, and I think it can be defined in a meaningful way.

Without trying to be definitive, here are some behaviours that would characterise a mission-oriented government:

Being long-termist/ambitious about the goals that we are pursuing together, as a society, through the enabling powers of government. And using this ambition intentionally — to breathe life/hope back into politics and to foster innovation.
Acknowledging complexity — admitting that we’ll need to learn along the way, as we did with everyone’s favourite mission, the moon landings. We didn’t know how to get there when we declared the mission, which is a reminder that any mission is partly a mission to learn something.
Focusing on outcomes, not outputs, and being serious about this — using outcomes as an organising principle for the day-to-day business of government.
Being directional, as opposed to agnostic, about the shape of economic growth. Not being shy to say that growth is for something, and should at minimum be consistent with certain outcomes (e.g. deep opportunity, economic security, respecting planetary boundaries).
Being decisively enabling, not top down and not pernickety or bureaucratic. Galvanising actors in the system, whether they are businesses, charities, or individual citizens, but not in a micro-managey way.
Engaging people properly, not through lame/performative consultations. Using deliberative and participatory mechanisms to equip people to participate actively in missions and to pool our collective ingenuity and intelligence

Together, these behaviours would amount to a new way of governing.

Partly in that they imply a different compact between the state, market, and citizen/civil society — one that is more intentional and collaborative.

And partly in that they would make for a different spirit or tenor of government — a different quality of relationship between people and the state. A relationship that feels more empowering and less transactional, and that reminds us that the state is really just a family of institutions that we use to do big, life-improving things that we can only do together.

That’s the intention, anyway. But what would this actually mean in practice?

Practicalities

If you wanted to be an outcome-based or mission-oriented government, how would you need to go about things?

Here are three examples of the practical questions you would need to work through:

1 — What tools would a mission-oriented government use? Maybe, for example, the toolkit of mission-oriented government would include:

Tools that speak to the idea of a state that is enabling, directional, and galvanising — e.g. challenge prizes, social investment/venture-building, investment in civic/digital/data infrastructure, etc.
Tools for meaningful engagement. For example, maybe you’d build new capabilities (skills, techniques, digital tools, institutions) for public deliberation, participatory policy-making, and collective intelligence

2 — What would mission-friendly devolution look like? Maybe, for example, a mission-oriented government would use funding settlements and delegated authorities to help places get behind missions. Maybe localities would become the locus of mission-oriented activity — a way to cut through the silos of Whitehall and bring the work closer to people

3 — How would central government need to function in order to drive missions? And this in turn raises a host of sub-questions:

What management practices align with mission-oriented working? Could missions help to re-energise demoralised public servants, and allow for more management via trust and judgement, relying less on audits and targets?
What is the right architecture for missions? Would you want some big themes that act as umbrellas for a series of sub-missions? And/or would you want to name 15–20 exemplars/‘outcomes of special interest’ that are especially important? And what is your typology for the objectives that would sit under missions? After all, not all problems are equal, so you’ll want a mixed economy — mapping different types of objective to different tools and delivery models
What skills or capabilities do you need to be a mission-oriented government? What disciplines are we light on in the centre of government? What specific institutions are we missing? (e.g. are we missing centres of expertise for those tools listed above?)
What operating models and ways of working would you need to embrace if you want to cuts through the complexity of missions, delivering progress at a pace people notice? How would mission-oriented teams need to be set up; what rituals and processes would be used to run the work; what cultures and mentalities would be needed?

That’s not intended to be a comprehensive list. I’m just trying to show that this is quite a generative conversation. You can see how it starts to populate a fairly broad and arguably quite coherent agenda.

Anyway, as practical as these questions might be, this is all still very high-level.

So let’s jump into the last question on the list as a bit of a worked example.

What operating models and ways of working would a government need to adopt in order to drive decisive progress on missions? Which, as luck would have it is the topic of today’s new report, The Radical How.

From ‘projects and programmes’ to ‘services and missions’

Today’s new report, The Radical How, from Nesta and Public Digital, has a lot to say about the way a mission-oriented government would need to function. It bears most directly on that question of operating models and ways of working, but it also has implications for delivery architecture and capabilities.

The report is packed with insights and is admirably concise, so I recommend reading it in full. The headline, though, runs something like this:

In order to drive progress on big societal missions, central government would need to be reoriented away from functional delivery silos, and output-based projects and programmes, towards outcome-based services and missions.

The report breaks this down into three shifts that would bring this reorientation about:

Organise work around multidisciplinary teams
Embrace incremental, feedback-driven iteration
Focus more on outcomes

The combined intent of these shifts is to make the outcome-based, cross discipline team the unit of delivery, in a system designed to run responsive services and drive progress on missions.

Multidisciplinary teams

To do good work in service of an outcome or mission, you need a range of relevant disciplines working together in unison — from policy to law, design, user testing, and delivery.

You can’t pass the work back and forth between functional silos; that’s too slow and transactional when a problem is complex and its different strands are intricately intertwined.

To drive progress on a mission like home decarbonisation, for example, disparate functions need to work as a team, so that they can reconcile tensions and learn together quickly.

On the UK’s goal to decarbonise home heating, for example, we need to close a huge gap, from currently helping around 35,000 households to upgrade to a heat pump each year, to helping at least 600,000.

To do that, we need to make progress on:

Financial incentives/subsidies (e.g. the design/delivery of the boiler upgrade scheme)
Planning permission — making approvals fast, easy, or just unnecessary
Installation capacity — training up thousands of engineers
Public awareness towards, and understanding of, heat pumps
Price signals (e.g. the ratio of electricity to gas prices)
Technology innovations (e.g. are heat pumps quiet, attractive)
Service innovations (e.g. is installation easy)

Work like this is more than complicated; it’s complex, in that we don’t know what will unlock progress on each of those fronts — and we also don’t know how each blocker/bottleneck interacts with each other.

This means we have to learn as we go (more on iteration in a minute). And this means we have to work together, quickly testing hypotheses and identifying how we can cut through blockers and interdependencies.

That’s why it’s vital to reconcile tensions between different aspects of the problem hour by hour, not in the days or weeks it can take for a Whitehall write-round or to escalate to a Programme Board/Cabinet Committee for a decision.

Tensions and disagreements must be reconciled in the team. To do otherwise is like trying to fly a plane and having to radio ground control for permission each time you need to move the aileron.

Feedback-driven iteration

To make decisive progress on a big societal challenge like home decarbonisation, or preventative healthcare, or equal opportunity for children, you also have to work in quick cycles of iteration.

This is a change from the traditional linear approach to Whitehall programme governance, in which risks are managed in large part by planning for them in advance, or at most by running a time-limited pilot.

As we see in the case studies in The Radical How, the downside of linear planning is that you make the biggest decisions when you know the least about the problem. When you’re working on a complex problem that’s especially problematic. To say that this can be solved with better planning is like saying we should stare harder at the tea leaves.

When applied to policy, linear planning also tends to delay the point at which you try things out.

You plan a policy in theory — in Ministerial briefings, or spreadsheets, or via written consultations, or parliamentary debate — and then you roll it out, and often even legislate, before you’ve tried the policy on for size.

The report visualises the worst version of this kind of waterfall planning:

The alternative is iterative working, which looks more like this:

As the report notes, this isn’t about running more pilots, since pilots still imply a linear way of working; you start with a phase for learning, and then you make a decision/‘roll the policy out’.

Iterative working is different. It means “conducting multiple small-scale experiments at the boundaries of policy and delivery — and doing this permanently.” The goal is to keep learning and adapting.

In the words of the report, “responsiveness is an embedded attribute, not a phase on a timeline.“

Outcomes

Along with cross-discipline working and iteration, it’s important to build teams around outcomes, not around outputs or functions.

The outcome is the north star towards which the team’s disciplines align, and it’s the goal towards which they’re iterating.

Building a team around an outcome is easy to say and hard to do.

It’s common for organisations to take an existing team, and say they’re now working to an outcome, when often in reality the team is still a functional silo, and/or is still working to a pre-existing set of outputs or projects.

This means the team still has lots of hard dependencies on other teams — other functional silos — that mean they can’t make progress on the outcome without seeking permission, which slows work down and inhibits shared learning.

The team also still feels loyal to its inherited portfolio of projects/outputs, as opposed to doing what’s needed to chase down the outcome.

A true outcome-based team needs to work in line with those other two principles — containing the disciplines it needs to make progress, and being able to iterate. It all works together as a system, or a practice of production.

Finally, you want your outcome to be enduring, not ephemeral like a project. As the report puts it, the goal ultimately is to build:

…a permanent team that constantly improves its service in small increments, and constantly increases the scale of the service, as confidence in its approach evolves and grows.

Beyond this, what you’re aiming for is a whole series of these enduring, outcome-based teams. Each containing the disciplines it needs to be meaningfully self-sufficient, so that each team can experiment and iterate continuously towards the outcome.

But this team of teams can’t be rushed; to get there you first need that small initial team to help you develop a deeper understanding of the sub-drivers/sub-outcomes that make-up the problem.

There’s a whole other post to be written about how these teams — sometimes called squads or spheres — work alongside each other, as part of an organisational ecosystem, for example by using certain processes/rituals to keep their work pacey and open.

This gets into the question of the overall architecture for missions and the question of organisational culture, which is a question for another day.

There’s also a post to be written on the challenge of setting outcomes for missions. How do you set a goal that is specific enough to bite and that is also sufficiently stretching but not too stretching. The answer depends on the work you’re doing.

(In the context of moonshot innovation, for example, at Nesta we talk about goals that are ‘almost impossible’. This is helpful when you’re working on an entrenched problem, since it snaps you out of old tramlines/forces oblique lines of attack on the problem.)

The future is here, unevenly

Are these ways of working radical? That’s an interesting question.

In one sense there’s nothing new here. These practices are by now familiar in contemporary organisations; the idea of running a 6 month evaluation, or deciding on a new policy without continual iteration, would sound to many organisations like relics of an earlier age.

In another sense, though, the model I’ve described cuts across the grain of how things are down in government in a radical way.

But that’s not the same as saying that no-one is applying these approaches.

On the contrary, the most interesting sections of The Radical How are case studies of teams in the public sector in the UK who are already applying the principles of cross-discipline teams, iteration, or outcome-based working.

As the saying goes, the future is already here, it’s just unevenly distributed.

For more about the case studies, read the report. But what’s surprising it that we’re not talking about niche examples, or the kind of digital teams that we might expect to be at the cutting-edge.

The first example is the Universal Credit rollout, which in 2013 was the flagship domestic policy of the UK’s Coalition government and was goign terribly — according to the Institute for Government, “heading for nowhere but the rocks”. By 2016, the programme was three years in, had spent £425 million, had gone through five Senior Responsible Owners, and hadn’t delivered a working service to a single claimant.

This turned around when a new team was set up to do the work differently. The team was small, starting with just 15 people; was separate from the old team, based in a different building; was re-anchored to an intent/outcome, not a rigid ‘plan’; was made up of multiple, co-located disciplines; and worked iteratively, starting by testing the new system with just 100 people.

Not only did the rollout steer clear of the rocks — just about — but the new system proved impressively resilient when, during the Covid pandemic, there was a sudden influx of millions of new claimants.

The second example given is the Future Farming and Countryside Programme, which has faced the daunting challenge of reforming farming subsidies post-Brexit.

True, it’s still called a programme, but this has been a model of contemporary working. The team has embraced a ‘test and learn’ philosophy, adapting both policy and delivery as they go, and working in the open.

Rather than the car crash that we might have anticipated for such a complex initiative, the work has had the almost unprecedented experience of being celebrated for how stakeholders have been consulted; the front page of one industry newspaper declared: “Defra responds to farmers’ feedback”.

The third example is the UK’s Test and Trace service during the Covid pandemic, which is a complicated story with critics as well as supporters. Still, the report shows how the team around Test and Trace made remarkably rapid progress on a breathtakingly complex challenge. It stresses the importance of modern technology infrastructure and open data, which helped disparate actors adapt in a fast-changing environment.

What these case studies show is that these methods might, in a sense, be radical, but they’re not unproven. As the report says, the ‘radical how’ is about promoting:

…methods and processes that have been shown to work, multiple times, at scale. They are the default for many of the world’s most successful companies. However, the occasions where they have been deployed are rare in government. These occasions have come about thanks to exceptional leaders, exceptional circumstances, or both.

The challenge is that, right now, these examples aren’t just rare, they happen in spite of the system, not because of it.

Hence running these teams, or working in them, is often exhausting, leading to burnout. Meanwhile the ways of working associated with crises like Horizon— huge, monolithic projects, run to linear governance with little iteration — remain tragically common.

The report ends by crystallising the lessons with a worked through example. How could we apply the principles of responsive, mission-oriented working to the task of helping people across the UK to decarbonise their home heating systems?

Rather than setting up a programme, with a delivery plan, Committees, and write-rounds, an alternative would be to:

First set an outcome — something of the form ‘X million households will be helped to upgrade to low carbon heating by 2030’
Spin up a small, cross-discipline team, combining policy and delivery, with a remit to work to this outcome
Give the team twice-weekly, even daily, Ministerial contact, and make sure they’re set up to resolve tensions and disagreements within the team, not by having to write round to colleagues across Whitehall
Run a series of bounded experiments to test and learn quickly
Once you better understand what’s needed to achieve the outcome — maybe it’s a slicker process for planning permission; or easier installations; or a simpler subsidy scheme — set up a team of teams to work to these sub-outcomes
Publish real-time data on the headline outcome and the sub-outcomes, and hold yourselves accountable at a monthly, Covid-style press conference (Ok, that last part is my addition)
Establish dedicated structures and processes to spot and resolve dependencies, optimising for the pace at which these are reconciled

Those are the headlines; the report expresses this better and with more of the all-important nuance. This includes rich detail on those case studies and a list of ten more granular principles for doing this work well.

I hope that was helpful, or at least thought-provoking. Thinking back to the overview I started with, my main takeaway is that there’s a lot of important thinking to be done.

At Nesta and BIT we’re mulling how we can contribute to these important conversations about the ‘how’ of government. What’s clear is that we need fresh approaches if we’re going to cut through the knot of complexity, intractability, rolling crises, and fatalism that we seem stuck in.

There are some lessons to share from our own journey at Nesta, after we reoriented our work around missions around three years ago. And there are lessons from over a decade of work by the Behavioural Insights Team to embed empiricism and evidence-based testing into policy-making.

But most imporantly this needs to be a broad conversation, bringing in the many, many people who have been doing this work for years, even decades.

Three questions in particular are on our minds as we think through how we support this conversation, which I’ll write more about soon:

How we can bring more sharpness/definition/sense-making, for example to clarify and codify the tools, techniques, and operating models of a responsive, outcome-based/mission-driven government
How we can help by connecting people who are doing this work, creating peer-support networks to help people learn from, and support, each other, and creating legitimating environments
How we can help to make this work more robust through evaluation, both to guard against ‘mission-washing’ and to hold our collective feet to the fire by asking: are these methods working?

As always, that’s all written in a spirit of thinking in the open and with even more thanks than usual to the many smart people working on these topics, including the folk at Public Digital. I know we’d all be really interested in reactions, builds, and critiques.

If you’re interested in these debates, you can follow me on Medium or Blue Sky or support my writing on Substack. And here are two posts I wrote recently along similar lines: How to govern human and How to solve wicked problems. Oh, and there’s still my book, End State.