This pattern is enabled by the handover problem described in Infrastructure Without Culture: infrastructure transferred without the governance culture needed to sustain it embeds the invisibility problem from the first day of operation. This pattern names the structural mechanism that makes that failure so persistent and so hard to interrupt.

Maintenance is systematically underfunded not because its importance is unrecognised, but because its success is invisible and its failure spectacular — and budgets, careers, and political attention all follow visibility.

A canal that does not silt up generates no event. An embankment that holds through a difficult monsoon produces no ceremony. The maintenance workers who prevented a failure last season have no moment at which their success can be photographed or credited. Maintenance done well is indistinguishable, from the outside, from nothing happening.

New construction is the opposite. It has a beginning and an end. It produces something that was not there before. It can be inaugurated, reported, and closed out. This is not incidental to how construction is funded — it is the mechanism by which funding is justified. Engineers who want to do good work pursue new projects because that is where resources flow. Institutions that want to demonstrate impact report on what they have built, not on what they have kept working.

The failure, when it comes, is highly visible. A breached embankment, a silted intake, a pump station that can no longer function — these generate exactly the political urgency that maintenance never could. And that urgency typically produces a response in the form of new construction. The cycle restarts, with the underlying incentive structure unchanged.

This pattern has a documented lifespan. Water management committees established under externally funded projects in Bangladesh functioned while budgets were available and external attention was present. Once the formal project ended, most had disintegrated — not because the infrastructure was unimportant, but because the conditions that made maintenance legible had been removed along with the funding. The self-serving assumption that users would take the initiative to maintain their own situation, recorded faithfully in the evaluation documents, substituted a hope for a governance arrangement.

The field notes put it more plainly: pay the doctor while you are healthy, not when you are sick. The paradox of the incentives of budget availability. Everyone understands the logic. The system continues anyway, because understanding a structural problem and having the institutional means to correct it are not the same thing.

What breaks the cycle, where it is broken, is accountability that is local and continuous — governance arrangements in which the people who depend on the infrastructure are also responsible for its condition, and in which deferred maintenance becomes visible to the community before it becomes visible as a disaster.

Make the condition of existing infrastructure visible on the same institutional terms as new construction — reported, assigned, and credited — so that maintenance competes for attention before failure makes the competition irrelevant.

Where maintenance failures accumulate across generations of projects without changing practice, the mechanism is Governance Cannot Reflect on Itself. Where the inevitable failure generates political pressure for new construction rather than reformed maintenance, the dynamic is Relief Crowds Out Prevention. See also Infrastructure Without Culture