Paying for pipes

Dr Eric Crampton
Local Government Business Forum
20 July, 2022

If there were no better way of solving the problem, one could forgive how central government has approached water infrastructure reform.

Some councils have done a very poor job in maintaining their pipes. For a council at its debt limit, making up for deferred maintenance, or laying out new pipes to enable more apartments, townhouses, and subdivisions, can be difficult. If debt cannot exceed 280% of revenues, and council is at its limit, it cannot borrow more than 2.8 times the revenue that the infrastructure supports.

Effectively, infrastructure that will last for decades has to pay itself off in under three years through a combination of developer contributions and new rates or user fees.

It makes no sense.

But it does help to explain the very convoluted governance and ownership arrangements over amalgamated water assets. They seem designed to achieve balance sheet separation, so that the new entities can issue debt that is not tied to the debt capacity of the underlying councils.

A water entity clearly owned by council, or by a set of councils, would have a harder time issuing debt to fund new infrastructure if that debt were tied to council main balance sheets. If bondholders and ratings agencies expect that a council would use its main balance sheet to bail out its water service provider, that provider’s debt will count against council debt limits.

But there seems an obvious and better way of enabling infrastructure renewal and expansion.

According to investment analysts Charles Schwab, only about a third of American investment-grade municipal debt is financed through general obligation bonds of the sort that New Zealand councils use.

The majority, about two-thirds, are revenue bonds.

A revenue bond is not backed by Council’s main balance sheet. Instead, the bond is paid off by revenues from the project that it finances – over an appropriate period. Those who benefit from an infrastructure project are the ones who ultimately cover its cost.

And it is hardly just an American innovation. It is also how the Auckland Harbour Bridge was funded. A revenue bond raised the funds, financed by a toll on bridge users until the bond was paid off.

It sets a far better dynamic. There is more reason for ratepayers to oppose an infrastructure project from which they see no benefit if they bear the cost of it through their rates. If the cost instead is covered by those who benefit from it, there are fewer reasons to object.

Revenue bonds also bring commercial discipline. The bonds are riskier than general obligation bonds because if the project fails, the bondholders wear the cost. If a large bondholder is looking to put up a lot of money for a large infrastructure project, that bondholder will be very diligent in assessing the risks.

Riskier projects will require a higher interest rate. But that is hardly a detriment for the model. The higher interest rate itself tells us something important that we might not otherwise have found out: the project might be too risky to be worth it. Commercial discipline in assessing these risks can be valuable.

So too can the monitoring services that a large bondholder can provide. If the revenue stream from an infrastructure project, over decades, depends on the infrastructure being appropriately maintained, bondholders will demand provisions in the bond ensuring that that maintenance is undertaken. It would be harder for councils to defer maintenance on water pipes, and shorten their effective lifespan, if bondholders are watching.

If infrastructure funding and financing seems to be the real problem that the government’s three waters programme is trying to solve. It isn’t water quality: a new regulator is already set for water quality, and will be monitoring regardless of whether amalgamations go ahead.

It is rather more likely that central government fears being stuck with the cost of bailing out irresponsible councils who prioritised flashy above-ground projects over basic infrastructure maintenance. It is also likely that central government sees the critical importance of water services in enabling the urban growth agenda. It is rather rare that electricity or telecom infrastructure prevents new apartment towers or new subdivisions from being built. Water is the holdup.

But forced amalgamations and messy governance structures are a rather poor solution. Central government could set a simple piece of legislation enabling councils to issue revenue bonds with no recourse to council main balance sheets. The legislation would have to make it credible that there would never be a bailout of a revenue bond from a council main balance sheet. One way of achieving it would be to set hefty penalties on councils that bailed out a revenue bond: what economists sometimes call a boiling-in-oil contract.

And the model would enable a lot more than just water. It could be used to fund other council facilities as well – so long as the infrastructure actually passes a commercial assessment.

If investors believed that a toll on a new Mount Victoria tunnel could pay off the cost of the tunnel over fifty years, Wellington Council could issue a bond to dig the tunnel. And if investors declined to fund the project, because they couldn’t see any way that the tunnel could cover its costs, wouldn’t that information also be valuable? It could help councils avoid sinking money into white elephants.

The New Zealand Initiative will be doing more work on infrastructure funding and financing over the months ahead, looking at the history of revenue bonds in New Zealand and how they could again assist.

They seem an awful lot simpler than what the government is trying to do with the Three Waters programme, more likely to be effective, and far less politically fractious.


To read the Local Government Business Forum newsletter click here.

Stay in the loop: Subscribe to updates