The quiet case for boring infrastructure

Infrastructure gets attention twice: when it is built, and when it fails. The decades in between are where the value actually accrues, and they are almost impossible to write about.

The failure mode is not dramatic. It is a slow substitution of capital spending for operating spending, made one budget cycle at a time by people who are each behaving reasonably within their own horizon. That produces three compounding effects:

• Deferral looks free in the period it is decided, because the asset does not visibly change.
• The cost lands outside the deciding horizon, usually on a successor.
• Interdependence turns a local deferral into a systemic one, which is the part almost nobody models.

Deferral is rational for any single decision-maker and irrational in aggregate. The asset outlives the people who chose not to maintain it, which is the whole problem in one sentence.

That figure is less damning than it looks. Deferring once is normal. Deferring structurally, so that the deferral becomes the plan, is what produces the eleven-fold cost multiple at the far end. See our note on decision memos for why the horizon problem is a structural one rather than a failure of judgement.

Someone always does. The question is only whether it is the operator now or the public later, and it is almost never asked in the room where the deferral is decided.

Horizon	Who decides	Who pays	Typical gap
Annual budget	Operations	Nobody yet	0 years
Rate case	Regulator	Ratepayer	3 to 5 years
Asset life	Nobody	Successor	20 to 40 years

The asset outlives the people who chose not to maintain it.

The three-row table above is the entire argument. Everything else is elaboration, including the part where someone points out that deferral is a polite word for a decision nobody wrote down.

The interesting cases are the ones where an operator does the right thing and is punished for it, because the cost lands inside the reporting period and the benefit does not.