Density is destiny baby! Transport infrastructure and local government costs

A new report is out which has a number of “WELL D’UH” findings, and a few “Huh, interesting” findings – but mostly is a relief because it signals us finally starting to look at the real cost implications of our dominant urban form. Here’s some transport highlights…

Here’s the paper’s how, why and wherefore:

 Sense Partners was commissioned by Greater Wellington Te Pane Matua Taiao to build a proof-of-concept model of the costs of infrastructure and urban form across the region.

This report sets out the proof-of-concept model and the preliminary findings. Even at this stage, our results yield valuable insights into how urban planning decisions can impact the infrastructure costs local government faces.

We focus on the financial costs faced by local government, both territorial authoritiesand the regional councils. Our model does include the cost of maintaining the statehighway network across the region, but we present these results separately.

We examine costs across Wellington City, the Hutt Valley, Porirua, and Kāpiti Coast. These are highly urbanised areas that, from an economic lens, function as a single urban area.

It’s worth a read by itself.

It’s also extraordinary that in 2024 we’re just starting to see authorities asking “Hmmm, is there maybe something in the thin smearing of urban area that dominates our region, and these terrifying infrastructure costs we’re facing?”

So, a long-overdue read! We recommend you get in there yourself, noting the big caveat that it’s a proof-of-concept, in many cases using the most readily available datasets which are pretty low-resolution, or partial-picture ones (such as journey to work / journey to school as proxies for movements).

There’s also some stuff that’s probably obvious if you are a spatial nerd or a policy wonk but isn’t clear to us randoms, so if you can help shed light please do in the comments!

Here’s some stuff that stood out for us, on transport. (The three-waters info is even simpler and more compelling in its density benefits.).

Density and proximity lower costs, as does being in the right place

One for the “well D’UH” category, it’s nonetheless great to see it laid out in an unarguable form that these fundamental, universal laws of urban physics also apply to us in the Wellington region. Bookmark this one for your next interaction with a greenfields developer or pro-greenfields councillor!

Our analysis indicates that density lowers the per-dwelling costs. Economies of density mean the higher cost of larger roads and pipes in dense areas is amply offsetby the number of dwellings that can be serviced

“Per-dwelling” matters because local government ultimately pays for infrastructure by dividing the overall cost by units like dwellings. (And yes this is how it’s done regardless of the entity collecting the money, or the entity running the infrastructure). Lower per-dwelling / per-resident / per-physical-business costs means people pay less for the same quality of service. Hooray!

Crucially, even with the more sophisticated and multi-use road corridors (with mass transit, wide and universally accessible footpaths and crossings, bike traffic-lights, good greening, good lighting etc etc), it’s still much cheaper! See “amply offset” – because decent high urban density means you’re serving and taxing lots and lots of households and businesses.

They dig into this further in section 2.3:

Our results indicate density is more cost efficient

They also confirm that costs are sensitive to local context. There are three main determinants of cost in our results.

The first is density. There are economies of density in infrastructure. This means an increase in the number of houses in anarea can be efficiently served by a much smaller increase in infrastructure. This results in lower per-dwelling costs. Even when we account for the higher per unit [per-kilometre or per-dwelling] cost of the infrastructure required, the cost per dwelling still falls. In our model, higher density means higher capacity roads are needed (and more public transport services), at higher per-km cost. Despite this, costs per dwelling are still lower.

The second is proximity. Being close to the central city, or major employment centres, lowers the distance people travel, which lowers the cost to local government. It also tends to shift people to alternative forms of transport, like walking or public transport, which our model indicates are cheaper. High density areas tend to be those in close proximity to the central city, meaning the lower costs faced by local government here are a product of both of these factors.


The third factor that lowers cost is to“be on the way”. [footnote refers here]. This refers to an area that liesbetween a main destination and the origin of many trips. This leads to many trips passingthrough the area because it is on the way. And as a result, those trips end up sharing some ofthe costof the ‘on the way’ area. While the roads would need to be bigger (and more costly) toaccommodate this, economies of scale mean the per trip cost is lowered.Some low-density areas, like Hataitai South (1,334 dwellings per km2), also achieve costs notmuch higher than central city areas (local road costs of $76 per dwelling versus $97 inCourtenay with 2,700 dwellings per km2). The reason for this is Hataitai South lies in betweenthe eastern suburbs and the central city. This means that the roads connecting Hataitai South also connect the eastern suburbs, and thus the cost is shared with those suburbs.

Some “whoa, gosh” things

OMG Porirua city centre. Oh babe. You need an intervention

If you’ve visited Porirua city centre recently, this is arguably in the “well D’UH” category but the degree of it is “Whoa. Gosh”.

Look where Porirua’s city centre is – way up in the extreme corner with Te Horo for crying out loud!

Intuitively, it makes sense given the sheer amount of road – you’d need a lot of people to meaningfully bring down the cost-per-dwelling.

Porirua’s city centre: so. much. road… so few inhabitants

If there was a starker and more prosaic reason to hit both the “too much road” and “too few people” sides of this yukness – ie convert some of the insane amounts of road space in Porirua, and get more people living in the centre pronto – we’d love to hear it.

What trashes our local roads and starts potholes?

Given how much road all councils have, and how few people and valuable things to help pay for looking after it, it’s useful to ask what’s doing the most damage to roads.

Heavy vehicles causing 93% of pavement (road surface) costs, while doing just 7% of the vehicle-kilometres! 93%! WOW.

For those making the case that less general traffic (ie shifting from cars to more fresh-air travel) saves us all road maintenance money … well, seems like that’s a teeny drop in the bucket. However, what about my home-reno or home-build project with a bit of civil works – so means a digger, truck-and-trailer unit and a concrete truck will be traversing my local roads? This is a filthy big negative externality (cost I’m imposing on everyone else)…

An “urgh, we should definitely sort that out” thing

There’s a filthy big skew in the incentive landscape that councils face, and it affects significantly their choices about roads vs development.

It comes from a potent combination of two things. One, which Te Waihanga discusses at length (e.g. here), has been more pithily dubbed the “low rates agenda“. The second, what the GW report’s authors call a “quirk” of our transport funding system: that central government (ie the taxes of the whole nation) pay for building and – crucially – maintaining state highways in anyone’s area, whereas local roads must either be paid for by entirely local taxes (rates) or funded by going cap in hand to central government for a proportion of the costs (usually around 50%).

In other words: where there’s a state highway (or a prospect of one) in your area, that makes councils far more willing to put more homes near it, where there’s typically nothing else except the big road, versus putting them in places where there’s already lots of good things to do.

Combine this with the Wellington region’s very poor record at master-planning and structure-planning greenfield developments so they’re actually halfdecent places to live, and this incentive means one thing: sprawl. Developments that are financial vampires because there’s too little productive landuse to pay for their infrastructure, where you’ll probably die of boredom or loneliness if you can’t drive loads every day (producing heaps of emissions) to do important human stuff.

Some “looking forward to a better version” things

Data. The report’s authors go to some pains to point out they’ve had really limited data on things like transport in particular. The census for example, is a very crude tool for unrstanding people’s movements. It’ll be fantastic to get more insight using finer-grained data – like the Household Travel Survey and research from places like the NZ Centre for Sustainable Cities.

Maps. The cartography is also a bit weird. Perhaps envious of the amazing 3D graphs we rave about, they’ve decided to present the information in 3D but with the much lower-resolution data, there’s whole suburbs obscured.

Unpack? We’d also like to see digging in deeper to what makes for expensive roads. We know that retaining walls are absolutely filthy expensive, even little tiny ones, and they are essential for roads through hilly areas. What’s the level of density which could help a road like Onslow Gorge Road, or many of the roads in Korokoro, or the Porirua hills, to wash its own face? How important are they – does it make sense for the rest of the city to be happy subsidising them?

LOS? We’d also love to learn more about the “reduce the level of service” aspects, discussed briefly in the report but mostly to do with three waters (septic tanks in Paekākāriki, for example, lower the local three-waters costs to council). How much could we gain in roading maintenance by making more roads narrower – more stretches that are “negotiated two-way”, or 30-40kph operating speed, or removing slip-lanes at intersections, for example – and slowing traffic down a bit?

What questions do you have about the report, and what struck you? Tell us in the comments!

Leave a comment

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

Sign up for our Newsletter

Unsubscribe any time.