What next for 3 waters?🍋
Improving diagnosis and policy for NZ's drinking, waste and stormwaters
NZ’s outgoing Labour government legislated major “3 waters” reforms, to transfer drinking, waste and stormwater assets and responsibilities from local councils to newly created Water Services Entities (WSEs).
While the final shape of the government to emerge from the 2023 NZ election remains unclear, it is clear that a majority of voters cast their votes for parties committed to ditching those reforms.1
Major reforms creating long-lived organisations to address long-run problems should ideally be developed with wide consultation, careful scrutiny, and receive bipartisan support. The 3 waters reforms fell well short of that ideal. In particular, pushing the relevant legislation through Parliament under Urgency was incompatible with those criteria.2
Is there a 3 waters problem?
It’s easy to mount an argument that local government ownership and responsibility for water infrastructure is broken, by pointing to specific examples, such as:
In 2016, contamination of drinking water supplied by two bores near Havelock North resulted in an estimated 5,500 of the town’s 14,000 residents becoming ill with campylobacteriosis, 45 hospitalisations, and possibly contributed to 3 deaths.
As at 2020, 20-30% of Wellington’s 2500km (or so) of water and wastewater pipes were already passed their use-by date. They are cracking, bursting and breaking.
Auckland’s 2023 intense rainfall was the largest rain event since 1869. The sheer amount was just too much for the city’s stormwater pipe network to cope with.
Starting from such examples, only a short jump in logic is necessary to conclude that “we need more regulation, as council self regulation has failed” and “we need to spend more, but councils don’t, won’t or can’t spend enough”. It’s a far bigger jump from there to “we need to take responsibility away from councils”, and “4 (or 10) new entities featuring complex financial engineering and multi-level co-governance is the best or only solution”.
While I am convinced there are problems with water services in NZ, I am not convinced by these logic jumps. I don’t think the 3 waters reforms were based on a clear problem definition, let alone that they represented the country’s best or only solution.
A new government means a new chance to revisit these issues and, hopefully, do a better job at devising solutions worthy of public and political consensus. In that spirit, I’d like to offer my two cents’ worth on some of the issues that drove the design of 3 waters.
Everyone faces weak incentives for long-term investment
Water infrastructure is in a poor state because councils (its owners) haven’t been spending enough money to maintain and renew it. That’s pretty uncontroversial at the NZ-wide level. (As an aside, some councils have done a much better job than others. It would be interesting to understand why.)
But why is it that councils don’t spend enough?
A political economy explanation is that there are low political returns to spending on long-lived out-of-sight infrastructure. Libraries, parks, heritage buildings, roads, rubbish collection, and just about anything else performs better in the votes-for-dollars-spent calculus. Consider, for example, a mayor facing re-election. They are almost invariably better off spending the next million dollars on something visible (and better suited to a ribbon-cutting ceremony) than on an aging pipe that might — and only might — burst before election day.
A more standard explanation is that local councils are reluctant to spend more because current ratepayers don’t want to pay higher rates. While that’s likely to be true, it’s also the case that ratepayers’ interests are reasonably aligned with those of councillors. Ratepayers can use a library or park tomorrow, while they might reasonably assume they won’t be around to receive any benefit from a pipe that lasts 75 rather than 50 years.
In principle, the cost of long-lived infrastructure is best shared between current and future users. Debt is an ideal mechanism for charging future users for future benefits. So why doesn’t this problem get solved with council borrowing? It seems there is a problem with debt — both ratepayers and councillors like borrowing, but would rather spent it on things with nearer-term benefits. Accordingly, councils tend to get in hock up to their debt limits, while long-lived out-of-sight infrastructure remains underfunded.
If we can’t trust councils to look after the interests of future generations, then who should we trust?
Central government? They face very similar incentives to local government. Politicians’ time horizons are firmly focused on the next election. Arguably, these incentives are blunted somewhat if they are members of longer-lived political parties.
Private enterprise? Opinions on this tend to vary by political orientation. Those fond of more government often rail about the short-termism of company CEs, matched by those favoring more private enterprise, who see only short-termism of politicians. Suffice to say that a combination — private infrastructure service providers supervised by well-designed & well-run regulatory agencies — can work reasonably well. Examples include electricity and telecoms in NZ, and water supply in France. That said, a switch to the private supply of drinking, waste and/or stormwater might be beyond the pale for a majority of NZ voters.
Special purpose entities? Yes, in theory. Tricky, in practice. The aim is a governance structure in which decision-makers’ incentives are aligned with those of future generations. But you don’t want decision makers who are disconnected from current service users either. Service customers, quite reasonably, expect an prompt response if the water doesn’t flow, the toilet doesn’t flush, or the basement is flooded. And they expect decision makers to face consequences, should they fail to respond adequately.
The 3 waters reforms would have created 10 (originally 4) special purpose entities. I would have expected the governance design to focus on the incentives faced by decision makers, including accountability for poor performance. Instead, it appears to have been driven by the desire for “balance sheet separation”, a form of financial engineering, along with a desire to transfer a degree of decision-making power to Māori groups.3 The upshot was a structure in which decision makers would have been thoroughly shielded from the short-run consequences of failing to deliver for their water-services customers.4
Financial engineering doesn’t change the underlying economics
If councils can’t borrow to fund water infrastructure, then who can? The financial engineers’ answer seemed to be highly-leveraged WSEs with complex and somewhat opaque organisational structures, and atrophied ownership and accountability linkages. Lenders would, apparently, have high trust in the ability of these WSEs to service and repay large levels of debt.
All this ignores the underlying economics. Lenders won’t lend significant amounts of money without security for the loaned capital. Security requires being able to repossess specific assets, take ownership stakes in defaulting borrowers (e.g. as part of liquidation process), extract a slice of future income flows (called factoring in commercial contexts), or call in guarantees from a rich and reliable external party.
The WSE legislation explicitly ruled out asset repossession and private ownership stakes, leaving only factoring and external guarantees. Factoring would have involved creating a right for a lender to levy a charge on the customers of a defaulting WSE. This situation is a pretty rough deal for customers, as they face the costs of paying off the debts of a failed WSE (whose failure they had little control over), at the same time they presumably will be on the hook to recapitalise its replacement.
During 2023, central government realised that it would have to act as guarantors for WSE debts. Financial engineering — despite driving arguments for, and the design of, WSEs — was exposed as essentially useless.
A stark lesson of economics is that if something is costly, then someone, somewhere, sometime has to pay. It’s always best to be explicit.
Understanding economies of scale and density
Another of my frustrations with the 3 waters reforms was a simplistic appeal to economies of scale5 as a rationale. These economies were declared to reduce the underlying costs of infrastructure and service provision for the WSEs. The claimed cost reductions were implausible large, as can be seen in this diagram.6
These claims flew in the face of global evidence. For example, a 2017 World Bank study found that:
On average, the analyzed aggregations [of Water Supply and Sanitation Utilities] have had no effect on cost and various other performance indicators when compared with similar utilities that did not aggregate.7
The proponents of the reforms appear to have confused economies of organisation scale (i.e. larger organizations being more efficient) with economies of production scale (i.e. organizations producing a larger quantity of services are more efficient). They perhaps thought they were getting the latter by pursuing the former.
Moreover, they overlooked likely diseconomies of scale, which, for example, arise from increased coordination and information costs when organisations get large and unwieldy. Looking at the map above, it is easy to envisage such diseconomies in the geographically fragmented Entity C, which stretches from Farewell Spit to East Cape.
Taking a step back, I think there is reasonable case for economies of scale in water collection/treatment/storage systems and sewage treatment plants. But many of these facilities are already shared where it is practical to do so. Lower and Upper Hutt, for example, already share water collection and sewage treatment. These economies of scale are real, but already realised. Gisborne and Motueka will require their own independent systems.
Beyond those facilities, water infrastructure is mostly about distribution networks — the pipes that distribute and collect the various waters. Such pipes, like fibre broadband, are subject to economies of density.8
Economies of density in distribution networks arise from population density. Such efficiencies are likely to have already been realised. They are not under the direct influence of those drawing the boundaries of WSEs.
2 waters or 3? Where are the economies of scope?
Economies of scope arise when it is cheaper overall for one organisation to undertake multiple activities, than for multiples organisations to do those activities separately. My favourite example is a US manufacturer that makes both snowmobiles and jet skis, keeping its production facilities busy year round. A competitor making only one of those products needs would have an idle production line for half the year (without a super-sized storage shed).
I think there are clear economies of scope for 2 waters, that is, drinking and wastewaters. Individual dwellings mostly need to be connected to both, and drinking water volume is a reasonable proxy for wastewater volume, allowing for a single metering and charging regime.
There are few synergies — and opportunities for economies of scope — between these two and stormwater.
Stormwater infrastructure has some very different characteristics. The primary aim of stormwater systems is to control and prevent flooding. Regional councils currently have responsibility for flood control, so there is an economies of scope argument to group stormwater and flood control. A secondary aim is waterway health. And responsibility for that also sits with regional councils.
There is also an economies of scope that groups stormwater with the other responsibilities of district councils. Stormwater infrastructure is typically built in conjunction with roads, and district councils are responsible for the majority of roads. Wider stormwater management options involve controls on land use, a topic at the heart of district council responsibilities.
I don’t have enough information to resolve the regional vs. district council question, but to my mind both have a better claim to economies of scope for stormwater than do the WSEs.
Where to from here?
The 3 waters reforms are (forgive the pun) dead in the water. But the problems that inspired them remain to be dealt with by the incoming government. I can only hope the replacement policy will be properly grounded in theory & data, be better explained to the public, and attract bipartisan support in Parliament. That might ensure a longer life than that achieved by the 3 waters reforms.
So, let’s kick off the public debate.
By Dave Heatley
The National Party and ACT Party manifestos are specific about revoking the 3 waters legislation. The position of the NZ First party is less clear, though Winston Peters, its leader, has been outspoken in his criticisms of the 3 waters project. He reportedly called it “retarded theft”.
The Water Services Entities Act 2022 was passed under Urgency in November 2022. Supporting legislation, including major changes to the original Act, were passed under Urgency in August 2023, by a government about to face the polls. Many other bills were passed under Urgency the same week, including a major rewrite of the the Resource Management Act. IMHO Urgency is an inappropriate mechanism for passing complex legislation addressing long-run problems. Might I humbly suggest that the next Parliament adopt a new rule: Every Act passed under Urgency is automatically sunsetted after 6 months. That is long enough for replacement legislation to go through the normal Parliamentary processes, including scrutiny by Select Committee.
The 3 waters reforms mechanisms for transferring decision-making power to Māori groups also suffer from poor design.
That said, I don’t think these governance arrangements would have survived more than a few episodes of delivery failure.
Economies of scale occur when it is cheaper for one organisation to produce 2xN units of a good than for two separate organisations to produce N units each. Diseconomies of scale is the opposite.
Department of Internal Affairs (2022). Transforming the system for delivering three waters services. Summary of proposals.
Michael Klein (2017). Statistical Analysis Global Study on the Aggregation of Water Supply and Sanitation Utilities. World Bank, Washington, DC.
Economies of density are cost savings resulting from spatial proximity of suppliers or customers. Typically, higher population densities save on distribution costs, leading to lower unit costs.
Eric Crampton has done a lot of good work on replacement models for 3 waters. Read about it here https://www.newsroom.co.nz/on-water-reform-national-and-act-are-sailing-on-same-course
Kia Ora Dave
Great article!
One little nuance I would add in under the heading “Everyone faces weak incentives for long-term investment”.
There is a class of investors who, if given the opportunity, will self-select into owning regulated utilities. Pension funds and others who welcome the regularity of a regulated return do in fact have the incentive to favour long-term investments. There are looking to invest in long-term assets that match their long-term liabilities.
In New Zealand, we have seen offshore pension funds investing in regulated utilities: First State acquired Vector’s and Maui Development Corporation’s gas assets, QIC and AMP have stakes in PowerCo and Canadian pension funds now own a lot of local cell phone towers.
Ngā mihi.
Peter Wilson,NZIER