Last week three industry bodies, Civil Contractors New Zealand, Infrastructure New Zealand and Water New Zealand, released a report called The Cost of Stopping, written by economist Shamubeel Eaqub. It put a number on something most people in the sector already feel in their bones: New Zealand starts, pauses and cancels infrastructure projects with a regularity that wastes money. The figure was $11.8 billion over 25 years, and it came with an interactive calculator at costofstopping.nz that lets you estimate the cost of stopping any given project.
The basic claim is true. Stop-start construction genuinely is expensive. When a project pauses, crews disband, equipment demobilises, and prices climb in the gap before work resumes. Eaqub is one of the country’s clearer economic communicators and he is right that the assumption “stopping saves money” deserves to be challenged. None of that is in dispute here.
The problem is the calculator, and it is the kind of problem that an economist of Eaqub’s calibre would normally be the first to spot. The tool forgets the first thing an economics course teaches you. Opportunity cost.
The money does not disappear
When a road is paused, the money allocated to it does not vanish into the harbour. It services debt, or it funds a different project, or it stays with taxpayers who do something else with it. Every one of those uses has its own return. So the real cost of stopping a project is not the value of the project. It is the value of the project minus the value of the next best thing the money does instead.
The calculator books only the first half of that equation. It totals up sunk cost, resumption cost, escalation and what it calls deferred public benefit, and it treats the entire benefit of the stopped project as a loss. The other side of the ledger, the benefit of whatever the money funds in the meantime, never appears. That is not a small omission. It is the whole of the discipline that separates an economic estimate from an advocacy number.
Deferred benefit is a dial, not a measurement
The deferred public benefit line is the one that should give any reader pause. Once you decide that “we didn’t get the benefit of this project” counts as a cost, you can run the argument on anything. Every road never built, every rail line only ever sketched, every harbour crossing that lived and died in a feasibility study has a foregone benefit attached to it. And because that benefit never has to survive contact with reality, you can set it as high as you like. There is no budget constraint on imaginary projects. A cost category with no upper limit and no counterfactual is not a measurement. It is a rhetorical dial.
This is also why the tool only ever runs in one direction. Nobody has built a “cost of building the wrong thing” calculator, even though New Zealand has plenty of those too. The asymmetry is the tell. Auckland Light Rail, which the surrounding coverage offers up as a cautionary tale, was paused precisely because its business case had collapsed and its cost kept climbing. Stopping it was the disciplined call, not the failure. Yet feed it into a tool that can only output a cost of stopping, and it still produces a number that frames the decision as waste. The worse the project, the more fictional its deferred benefit, and the calculator cannot tell the difference.
What the honest version looks like
There is a real cost of stopping, and it is worth measuring properly. The key distinction the tool blurs is delay versus cancellation. For a genuinely good project that is merely delayed, there is a legitimate and calculable cost. The stream of net benefits gets pushed back in time, and because a dollar of benefit in 2035 is worth less than one in 2030, that delay has a real present-day price. That is sound economics and Eaqub knows it cold.
But two conditions have to hold. The project must have cleared the bar in the first place, meaning its net benefit actually beat the alternative uses of the money. And the cost is a timing effect, the value of bringing benefits forward, not the gross benefit number the calculator implies. Strip the tool back to projects that were worth doing and were only delayed, count the time value rather than the full benefit, and net off what the money did in the meantime, and you would get a far smaller and far more defensible figure than $11.8 billion. For context, that headline works out at under $500 million a year against national infrastructure investment now running well above $20 billion a year. Annualised, it is closer to a rounding error than a crisis.
The good report underneath
It is worth saying that the recommendations buried beneath the calculator are sensible. Ring-fencing maintenance budgets so basic upkeep is not the first thing cut, committing to funded multi-year pipelines so firms can plan, and requiring a formal cost assessment before a project is paused are all reasonable asks. That last one is quietly revealing, because it concedes the thing the calculator denies: that stopping is sometimes the right answer, and the way to find out is to actually do the assessment rather than assume the cost.
That is the frustrating part. There was a good, modest report here about the genuine waste of policy whiplash, and it has been packaged inside a one-way tool and a headline number engineered to sound alarming. The framing has the fingerprints of the people who commissioned it, three bodies whose members are paid to build. It does not have the fingerprints of the economist whose name is on the cover, which is exactly why it reads as a commission rather than his analysis. Shamubeel is better than this calculator, and on a quieter day he would tell you so himself.
This is a Newswire analysis of The Cost of Stopping, a report by economist Shamubeel Eaqub commissioned by Civil Contractors New Zealand, Infrastructure New Zealand and Water New Zealand. We covered the report’s headline findings in our earlier story, and the interactive calculator is at costofstopping.nz.
Do you think the cost of stopping is overstated, or is policy whiplash the real waste? Have your say in the comments below.
This article was written by AI, briefed to report the facts, hopefully without some of the bias people bring to the job 🙂
Leave a Reply