The art of the possible in public infrastructure
Public infrastructure is the most permanent thing a society builds. It deserves our most imaginative thinking, and our most disciplined.
Governments come and go. Strategies are rewritten every few years. But a water treatment works, a rail line, a port, a power station or a fibre backbone will outlast almost every decision made about it. Public infrastructure is the closest thing a society has to a promise made in concrete and steel to people not yet born.
That permanence is why it deserves imagination. And it is why imagination, here, must be disciplined.
The old phrase, reread
Politics has long been called the art of the possible: the craft of getting done what can be done, given the constraints. In infrastructure the phrase is usually heard as a limit. Be realistic. Lower your sights. Build what the budget allows.
We prefer to hear it the other way round. The art of the possible is the art of expanding what can be done by thinking more clearly about the constraints. A constraint understood precisely is often a constraint that can be redesigned. Most of the hard limits on public infrastructure are not physical. They are institutional: how projects are conceived, prepared, governed, paid for and maintained.
From asset to service
Infrastructure is usually discussed as assets: kilometres, megawatts, megalitres, units. Citizens experience none of these. They experience a service. Does water arrive, and is it safe? Does the train come, and when? Does the clinic's power stay on during the afternoon?
Designing from the service backwards changes the questions. It moves attention from the opening ceremony to the fortieth year. It asks who will operate the asset, how performance will be measured, who will be told when it falls short, and what happens next. It treats maintenance not as an afterthought but as the main event, because over the life of an asset, it is.
Citizens never experience an asset. They experience a service, every day, for decades.
The invisible layer
Every physical network now has a digital twin of sorts, whether or not anyone designed it: the data about its condition, its use, its failures and its repairs. In many places that layer is fragmented across contractors, departments and spreadsheets, and nobody owns it as a whole.
Imagine treating that information layer as public infrastructure in its own right. Designed at the start of a project rather than bolted on at the end. Owned by the public authority rather than lost when a contract ends. Open, where openness is safe, so that citizens, researchers and oversight bodies can see how the service is performing.
This is also where the new tools become genuinely useful. AI that predicts a failure before it becomes an outage, that prioritises maintenance by consequence rather than by complaint volume, that reads thousands of inspection reports no team has time to read. None of it works without a well-governed information layer underneath. And wherever that layer touches personal information about citizens, POPIA applies, including section 71 on decisions made solely by automated means. A public system that decides about people must still be able to explain itself to them.
Preparation is where the future is decided
The most consequential decisions about a piece of infrastructure are made before anything is built: what problem it solves, which option is chosen, how risk is shared, how it will be governed. Preparation is unglamorous and chronically under-resourced. It is also where imagination has the most leverage, because changing a design on paper costs a fraction of changing it in the ground.
A society that wants better infrastructure should invest disproportionately in the thinking that comes first: clear problem statements, honest option analysis, realistic operating plans and governance designed before the first contract is signed.
Design with the people who will depend on it
The people who will rely on a piece of infrastructure for decades are rarely in the room when it is conceived. Consultation, where it happens, often arrives late, when the main choices are made and the only question left is where the fence goes. Imagine inverting that: involving users while the problem is still being defined, when their knowledge of how the service actually fails can still change the design. A commuter knows where the connection breaks. A clinic manager knows which hour of the day the power matters most. That knowledge costs almost nothing to gather, and it is usually ignored.
Institutions are infrastructure too
The capacity to conceive, prepare, procure, govern and maintain is itself infrastructure. It is invisible, it decays without investment, and when it fails, everything built on it fails too. Strengthening a public authority's operating model, its decision rights, its data and its skills, may do more for future infrastructure than any single project.
This is where our essay on the institution as an operating system meets the physical world. The bridge is only as good as the institution that maintains it.
What we imagine
Imagine public infrastructure planned around the service a citizen will feel in the fortieth year, not the ribbon cut in the first. Imagine an information layer owned by the public and designed on day one. Imagine authorities able to show, at any moment, how their networks are performing and what they are doing about it. Imagine preparation treated as the most important phase, and funded accordingly.
None of this requires a miracle. It requires imagination applied to the dull parts, and the discipline to see it through. That combination, applied patiently, is what the art of the possible looks like when it is built.