A toothpick is a very small thing on which to build an economic policy, and yet for a period in Nigeria toothpicks became part of a much larger argument about industrial development. They appeared on a list of products restricted from accessing official foreign exchange. The intuition was easy to understand. Nigeria has wood, toothpicks are not particularly complicated, so why should scarce foreign exchange be spent importing something that could surely be manufactured at home?

It sounds almost unanswerable until you ask a different question, why, in a country with abundant labour and raw materials, was it cheaper or easier to import the toothpick in the first place? Suddenly the toothpick becomes much less interesting than everything around it, electricity, logistics, finance, machinery, management capability, regulation, scale, supplier networks and the predictability of the wider business environment. That small example captures something I have spent much of my career thinking about.

The Architect Approach

Economic policymaking has an extraordinary tendency to begin with the thing we would like to see rather than the system that would have to produce it. We hear manufacturing is the way an economy grows, so we build an industrial park. Technology companies are booming everywhere, so we build a technology hub. We want jobs, so we announce a jobs programme. Domestic production is required, so we restrict imports. We want farmers to use more fertiliser, so we subsidise fertiliser. And when the desired activity does not become commercially viable, the intervention itself often acquires a life of its own.

Subsidies continue year after year. Protected industries remain protected because removing the protection has become politically harder than introducing it ever was. The infant industry reaches middle age without ever being weaned. A programme intended to create independence quietly becomes the thing on which the industry depends. We build infrastructure that looks impressive from the road but has too little economic activity around it to sustain itself. We measure the number of people trained, the amount of finance disbursed, the hectares developed or the jobs promised, and mistake the successful delivery of an intervention for the creation of a functioning economic system. These are not simply failures of implementation. Often they are failures of diagnosis. We reached for the tool before understanding the garden.

Policymaking and Gardening

Meanwhile, some of the most powerful development stories begin almost the other way around. In Yaba, Lagos, a technology cluster emerged because entrepreneurs, technologists, investors, universities, advisers and institutions were already beginning to find one another. Government did not need to invent the cluster. At an important moment its contribution was remarkably modest, remove a charge that was making it uneconomic to lay fibre optic cable into the area, clear the obstruction, then allow entrepreneurs and investors to do what they were already trying to do.

In Sialkot, Pakistan, thousands of exporting small and medium sized businesses did not wait for the state to provide every piece of infrastructure they needed. They pooled resources and helped create an international airport that connected their businesses more effectively to global markets. These stories fascinate me because nobody commanded the final outcome. The energy came from inside the ecosystem. Public institutions mattered, infrastructure mattered, rules mattered, but firms and people had their own reasons to make the next investment. One useful response made another more likely. This is the phenomenon I have become increasingly interested in throughout my career.

Momentum is an elusive life force of prosperity. It is extraordinarily difficult to manufacture directly, yet once a system begins generating it for itself the consequences can be profound. A successful company creates skilled workers. Those workers move between firms or start companies of their own. Suppliers see enough demand to specialise. Investors gain confidence because they can see businesses succeeding. Training providers respond because employers will pay for skills. Customers become more confident because quality improves. Institutions learn because people actually use them. One turn creates some of the conditions for the next. That is the flywheel.

Can You Keep the Momentum?

But finding momentum is only half the task. Flywheels do not emerge in a vacuum. Nature requires healthy and reasonably balanced ecosystems for positive cycles of growth to emerge, and so do economies. A plant can possess enormous potential, but if the soil is depleted, water cannot reach it, disease overwhelms it or the surrounding ecosystem cannot support it, that potential will not become growth. A business cannot compound indefinitely if power is unreliable, skills are scarce, contracts cannot be trusted, payments cannot move, infrastructure fails, finance does not fit the business model or the rules change unpredictably. This is why Flywheel Economics has two inseparable ambitions, improve the health of the economic ecosystem, and hunt for the points of momentum within it.

The healthier the ecosystem becomes, the more likely productive flywheels are to emerge naturally, and as those flywheels strengthen, they in turn deepen the ecosystem around them. Skills attract firms, firms create demand for suppliers, suppliers develop capability, successful firms attract capital, capital supports expansion, workers see viable careers, institutions become stronger through use, and each cycle makes another possible.

The ecosystem is the context, the flywheel is the productive dynamic we are searching for within it. That distinction has become central to how I think about economic policy. Government should not imagine itself as the engineer standing above the economy designing every moving part. Nor should it imagine that its only role is to stand aside. I increasingly see the policymaker as a steward of a living system, or in the metaphor that runs throughout the Toolkit, a thoughtful gardener.

A good gardener is not passive. Gardeners intervene constantly. They improve the soil, remove weeds, protect young growth, put up fences, introduce new plants, prune, irrigate and sometimes radically reshape part of the garden. But they begin by understanding the place. What grows here already or what is struggling or what is crowding something else out? What does this soil support? Where does the light fall? What would be foolish to plant here, however attractive it looked in somebody else’s garden? The gardener does not confuse intervention with growth. The plant grows. The gardener creates conditions in which growth becomes possible.

Introducing the Flywheel Economics Toolkit

The launch of the Flywheel Economics Toolkit therefore felt deeply personal to me. It is the culmination of much of my career, years spent working on economic reform, studying development, observing programmes succeed and fail, talking to businesses, investors, policymakers and practitioners, and repeatedly returning to one deceptively simple question, how do we create the conditions in which prosperity begins to reinforce itself? We teach economists to think instinctively about aggregate demand and supply, the Phillips curve, comparative advantage, market failure, incentives and externalities. These ideas become part of their mental furniture.

My ambition is that Flywheel Economics becomes part of that furniture too. I would love a generation of policymakers to think instinctively in terms of ecosystems, flywheels, binding frictions, capabilities, incentives and self sustaining momentum. I would love this way of thinking to sit at the heart of future industrial policies and sector master plans. Before building the industrial park, understand the industry. If you must subsidise the input, understand the constraint. Before protecting the firm, understand what would eventually make it competitive without protection. In creating a national training programme, understand why employers are not already paying for the skills they say they need. Before announcing the jobs, understand the commercial relationships capable of sustaining them. Before copying somebody else’s technology hub, automotive strategy or special economic zone, understand what actually made the original ecosystem work.

Policymakers should ask who is already investing their own money, time and reputation, what relationships are beginning to reinforce one another, what is stopping the next turn, which functions genuinely require public action and which should remain with firms, workers, investors and communities. They should ask these questions not because we happened to get lucky and appoint an unusually insightful minister, permanent secretary or adviser, but because this is how they were trained to make economic policy.

The ultimate test should not be whether a programme survives. It should be whether the ecosystem is stronger because the programme existed. Are firms reinvesting? Are workers acquiring valuable skills because employers will pay for them? Is there evidence of suppliers emerging and are customers returning? Will investors commit their own capital? Are institutions becoming more capable? Is the system beginning to solve problems that previously required outside intervention? Is one productive response making the next more likely? Eventually, the wheel should turn less because government, a donor or an adviser is pushing it, and more because the system has acquired momentum of its own.

None of this means government disappears. Courts do not need exit strategies. Neither do basic regulation, public safety, macroeconomic stability, core infrastructure, competition policy or many forms of social protection. A healthy ecosystem can depend on strong permanent public institutions. The distinction is between those enduring public responsibilities and the exceptional support introduced to change an economic system. A subsidy may be justified, but what capability is it buying and when should the industry require less of it? Protection may be justified, but what learning, productivity or investment is taking place behind it and what would allow the protection eventually to fall? Public finance may be needed, but is it crowding in independent investment or replacing it indefinitely? A training intervention may be valuable, but are employers eventually willing to hire, co finance and shape the training themselves? This is where the three year question in the Toolkit comes from.

Imagine the special programme ends, the donor leaves, the minister changes, the project office closes or the exceptional subsidy is withdrawn. What keeps happening? Who finances the next cycle, maintains the capability, coordinates the firms or trains the next worker? Who has a commercial, institutional or social incentive to continue? If the answer is nobody, perhaps we created an intervention rather than a flywheel. If the answer is that firms, workers, investors, suppliers and institutions now have stronger reasons and greater capability to carry the activity forward, something more important may have happened. Prosperity has begun to develop roots.

What the Flywheel Economics Toolkit Is and Isn’t

I also want to be humble about what we have launched. The Toolkit is a first iteration, not a finished doctrine, and I hope it never becomes one. If Flywheel Economics begins from the premise that economies are adaptive systems, then the Toolkit itself must be allowed to adapt. It needs to encounter reality. Policymakers need to use it on decisions that matter. Businesses need to tell us where it misunderstands how they behave. Economists should challenge its assumptions. Development practitioners should expose the questions that are missing.

We will discover places where the tools are too complicated, concepts that need sharpening, distinctions that matter more than we realised, and assumptions that reality simply refuses to honour. Over the coming years I want the Toolkit to be tested, criticised, refined and improved across countries, sectors and institutions. If a future edition looks exactly like this one, I suspect we will have failed one of our own principles.

I am enormously grateful to Favour Ogbadu, who has played such an important role in developing and challenging this work. I am also grateful to the other members of the Flywheel Economics team, Yasmeen Aliyu and Elizabeth Nwachi; our associates, including Daniel Simpkins, Daniel Ticehurst, Deep Biswas, Motselisi Matsela and Veronica Masubo; and the wider Flywheel Economics community. Their experiences, case studies, questions and disagreements have helped shape and strengthen the work.

It may have started as my vision, but it has taken on much broader ownership, and that is exactly how it should be. My ambition now is simple but large, that one day designing an industrial strategy without first understanding the ecosystem and searching for its flywheels will feel as incomplete as doing economics without thinking about incentives, supply and demand. I want a policymaker opening a blank page to begin almost instinctively with three questions, What garden am I standing in? Where is the momentum? What must I do so that this economy can keep creating prosperity long after I am gone? If those questions become ordinary, we will have changed something more important than any individual policy. We will have changed what good economic policymaking means.