Why does it cost so much to build stuff in New Zealand? (and is that OK?)

Big stuff costs a lot, especially complex stuff. But how much should we worry? With more dismaying cost overrun stories, there’s more skepticism, more investigation, and more determination to be better…

Te Waihanga the Infrastructure Commission concluded recently (which may surprise) that New Zealand is on a par with other high-income nations for the cost of building “smaller”, more standardised infrastructure like hospitals, surface rail infrastructure, on-shore wind farms, and electricity generation and transmission. But the Commission found that we are inherently costlier across-the-board when it comes to building large-scale complex things like big motorways through hilly places, and underground rail projects.

There’s a great short podcast and a report [PDF] to dive into this stuff. (There’s also a summary here).

But while NZ isn’t the international laughing-stock of infrastructure development , we should definitely look into how to be better across the board. Because while we’re middle-of-the-pack internationally on mega-projects, that pack is doing pretty badly. A staggering proportion of mega-projects (those big, complex, billion-dollar-budget ones) go wildly over-budget, over-time, with the ultimate product either never delivered at all, or badly under-performing. There’s not even a normal distribution for these kinds of projects’ ultimate costs: it’s “fat-tailed” – so there’s far more likelihood of the extreme cost overruns than you’d expect from a population of things.

There’s a good argument that even though we in NZ haven’t the history of many mega-projects that some bigger countries do, we’re not making well-informed decisions about whether projects are worth doing – because we’re not basing them on good estimates of the costs and benefits.

And as little NZ, we should really worry if lots of even “non mega” projects (especially complex ones like new highways through complex terrain) have “pretty normal” cost overruns. This is all New Zealanders’ money which could go on other important mahi: there’s a lot of kids’ literacy, or cancer treatment, that could be delivered with an “itty bitty 10%” cost overrun on a big project.

So, what to do?

The “How Big Things Get Done” authors explore in depth some ways we can do better estimating for mega-projects – in other words, getting a (more) realistic expectation upfront of how much time and resources something will take to build, in the real world. They talk about using reference class forecasting and cutely named things like “Pixar planning”. This is all useful both for those procuring mega-projects (in NZ, typically the public sector) and those providing them (the engineering and design consultancies, construction firms et al). This conversation about a Wellington Water project’s costs “blowing out” is thoughtful and useful on this too.

But Te Waihanga go further, following the line of thought on government in the public good, focusing on how government can be more sophisticated clients when it comes to procuring infrastructure.

This is really important. Profit-motivated, publicly-listed construction and consulting firms have clear incentives, and it’s typically up to the (government) client to make sure they aren’t taking the p*ss in negotiating the risk and benefit distribution of an uncertain project.

And very quickly, in New Zealand local government especially, we get to “but we’re [too little / turning over too fast / too underresourced] to have that expertise” – a situation exacerbated by the hollowing-out of the public service by private consultancies.

Will the infrastructure reforms (the ones we really hope are coming to scaffold the fast-tracking bill) mean our public sector is more sophisticated as an infrastructure client, or just doing the same kinds of things faster with more PPPs and offshore capital, praying with crossed fingers for better outcomes?

Leave a comment

Your email address will not be published. Required fields are marked *

Sign up for our Newsletter

Unsubscribe any time.