Problem first thinking

How to unlock innovation, productivity gain and social value

A few weeks ago, I acquired a desk (an antique Victorian bureau) that was free because it was damaged: it needed a new hinge. I spent a bit of time googling hinges and discovering that bespoke ironmongery is expensive. And then I came up with a simple solution: I used half a brass picture hook to recreate the missing bit of the hinge. It worked beautifully.

Yes, I was smug about this for several days: but that’s not the point of the story. This is an example of ‘problem first’ thinking - a way of working that is critical to both fixing our public services and growing the economy. Let me explain why.

When I tell myself I need a hinge, I’m not thinking about the problem - I’ve already chosen a solution. When I recognise that what I actually need is a mechanism to hold two pieces of metal together, allowing one to rotate, I unlock an imaginative solution.

The state operates at a different level, of course. But precisely because of that scale, when the state acts in a genuinely problem-first way, it can wield extraordinary power.

The iconic example is the Covid-19 Vaccines Taskforce. They didn’t specify a solution - they created a market for a solution. By pre-committing to buy successful vaccines at scale, even though they didn’t yet exist, they helped unlock billions in private R&D and manufacturing capacity.

Most of the time, the state does not operate in this way. It does not buy solutions to problems, it buys activities. It buys them in largely rigid ways, to replicate activities that have happened before. When it wants to do something new, it brings together a team of people to decide what new activities need to happen and then it buys those, with new but still rigid contracts, rules, job descriptions and structures.

All that structure makes change slow, innovation difficult, and productivity gains nearly impossible.

Problem-first commissioning

The good news is that more and more parts of the state are thinking differently.

Social Outcomes Contracts pay providers when they achieve goals for the people they’re working with: like avoiding reoffending, returning to work, or staying healthy. One of the best examples is in Kirklees, where three local organisations are working together to support adults at risk of homelessness. These contracts can be supported by a Social Impact Bond, a financial mechanism that unlocks cash to cover the upfront costs for providers.

Advanced Market Commitments are part of a suite of tools used to procure technology that doesn’t exist yet, or doesn’t exist at the scale or price that’s needed. A great example is the partnership surrounding the National Security Strategic Investment Fund (NSSIF). This starts with a set of ‘Problem Books’ (like this one for cyber security). NSSIF invests in and supports researchers and companies with the potential to solve those problems: getting them to market quicker.

I hear more and more public sector leaders thinking about outcomes contracts, problem books and advanced market commitments: so I have hope.

These processes are different at the technical level, but they unlock similar things:

• Collaboration. Centering the problem acts as a kind of gravitational pull for people with different skills to contribute. That could be at the scale of a multi-disciplinary ‘team around the child’ for social services or a multi-million-pound collaboration between university departments and scale-up businesses. While most research and most public services operate in silos, problem-first structures tend to break those silos.
  
  

• Agility over time. When you buy activities, you freeze the delivery model in the year the contract is written. When you buy a problem, you allow delivery to shift as new knowledge emerges. If a provider invents a faster, cheaper or better way of achieving the outcome, they can use it immediately.
  
  

• Agility between places. In the social space, activities often fail when transplanted from one location to another. When the state mandates a fixed activity model, local teams must twist their environment to fit the process. When the state defines the problem, local teams can tailor the process to fit their environment. That is how you get scalable solutions that adapt rather than decay when moved to new places.

There’s also a wider economic reason for problem-first thinking: it creates the conditions for innovation that spills over into growth.

Many new technologies never make it out of the lab because the commercial risk is too high. A start-up can’t spend three years developing an unproven technology if it has no idea whether anyone will buy it. But when the state says, “If you can solve this problem, we will pay,” that uncertainty collapses. Researchers can secure grants. Companies can raise investment. Engineers can leave their steady jobs to build something new. Charities can design new delivery models.

Problem-first commissioning doesn’t just produce a one-off solution. It can pull entire supply chains into existence, for UK and global markets.

Why it’s hard

Of course: that doesn’t mean it isn’t hard to upend how the state spends its money. Here’s a non-exhaustive list of some of the main barriers.

• Price. The challenge isn’t that outcomes-based approaches cost more, it’s that we don’t know what the price should be. Set it too low and providers collapse; too high and it looks wasteful. Of course, this missing the point: when the state buys an outcome, it is also buying the trial-and-error that produces cheaper solutions in the end. Nevertheless, outcomes commissioning requires different kinds of price and contract mechanis, - you can read about ‘relational contracting’ here.
  
  

• Timing. Activity-based contracts pay as you go; outcome-based contracts pay at the end. That creates serious cashflow strain, especially for charities and small providers who can’t absorb months of risk. Social Impact Bonds were supposed to be a magic bullet for this cashflow problem but they’re challenging to deliver not least because of…
  
  

• Non-realisable savings. Many social outcomes don’t create cashable savings. If someone doesn’t reoffend, society wins, but the prison budget doesn’t shift because the fixed costs remain. Much of the true value sits outside departmental spreadsheets. Outcomes can be socially vital yet fiscally invisible, at least in the short term. As Demos and others have argued, we probably need to find a new mechanism for ring-fencing prevention spending.
  .

• Culture. Moving from activity-led management to outcome-led management is not a tweak but a reversal of how systems think. There was plenty wrong with the 2011 Health and Social Care Act, but one of the core ideas Andrew Lansley had was to shift away from process targets to a focus on outcomes: asking the NHS to improve 5-year cancer survival rates not manage how long people had to wait at A&E, fro example. Performance collapsed*. The whole operating culture was built around processes. Culture change of that scale takes years, and public systems are wired for fast accountability.
  
  *this was of course partly because everyone was distracted with a reorganisation

Political self-sabotage
The biggest problem of all, though, is politics. Politics keeps boxing us into activity-based targets and input commitments, even though those very targets limit creativity, constrain growth and hold down productivity.

I used to teach a course for prospective parliamentary candidates on how to talk about policy. Policy people love talking about the ‘what’: I will create a pupil premium. Normal people only connect with specifics if you give them a ‘why’: Every child deserves a great start in life.

The problem is you can’t just stop at that abstract goal - you sound vague or implausible. You need a what to go with the why.

But you can’t say, “I will take an iterative approach with an open commissioning model to develop effective solutions to this problem.” You would be bundled off the stage.

So politicians make specific activity promises. And worse than that, because they’re pushing ideas that need to work on the side of a bus, or a leaflet, they tend to choose the activity with the strongest heuristic link to the outcome they want: what sounds like it might solve the problem, rather than what actually solves the problem.

We end up with commitments to have more nurses, not healthier people. More prisons, not safer streets. More deregulation initiatives, not more productive firms.

I don’t have a neat solution to this political physics problem. The incentives are deep, the campaign cycle is short, and our information environment rewards the simplest possible story. But if we want a state capable of problem-first thinking, we will eventually need a politics capable of sustaining it.

Getting started

Nevertheless, even if we can’t see the whole solution, we can still get started.

I am a big fan of the government’s Test, Learn and Grow programme, which is doing precisely that. TLG is working in local areas across England to work with councils, frontline workers and residents. Instead of prescribing a service model from Whitehall, they start with a real problem - like low uptake of family-hub services or pressures in homelessness support - and rapidly prototype different approaches. Promising ideas are iterated and scaled; weak ones are dropped.

This matters because it creates space for safe experimentation, rather than locking services into a delivery model that may already be outdated. It builds solutions around local context instead of forcing a national template to fit everywhere. And it proves that problem-first thinking can happen within the state, not just by commissioning someone outside the system.

If you’re outside the TLG scheme, you too can start thinking in this way!

A simple first step is just to map and broadcast the problems you actually want solved. Problems aren’t just problems: they’re demand signals to innovators, but only if we talk about them. This is hard in the private sector where the last thing most companies want is for their competitors to know their struggles. But if you work for the state, or for a charity - there is no reason to be open and honest about what you need.

I don’t mean a formal and regulated Market Engagement Process through your procurement team. I mean publishing a structured, clear Problem Book that explains

• The problems you face

• The value associated with those problems

• What you wish you could buy if only it existed

And if your strategy department doesn’t already know what your biggest problems are, is it any wonder you’re not solving them?

Ideally you’d be able to make advanced commitments to buy solutions - to derisk investment. But even the problems are useful, as a signal and the start of a conversation.

If we don’t start telling innovators what problems matter most to us, they’ll just build another app for faster food delivery. And none of us want that.