Is Auckland too big?

[This was originally posted in unfinished form this morning. It has since been rewritten.]

Is Auckland too big? Some people are asking that question.

For instance, in a Twitter exchange a while back Radio New Zealand producer Tim Watkin stated:

Others disagreed, observing that we didn’t have any obvious way to make Auckland smaller, meaning that people must have good reasons for wanting to be here:

As it turns out, Auckland (1.6 million people) isn’t the only city where people are debating this question. The San Francisco Bay Area (7.6 million people) has the same challenges, as Kim-Mai Cutler has ably documented. So does Boulder, Colorado (300,000 people). Actual population size is not, it seems correlated with complaints about cities being “too big”.

BOULDER, Colo. — The small city of Boulder, home to the University of Colorado’s flagship campus, has a booming local economy and a pleasantly compact downtown with mountain views. Not surprisingly, a lot of people want to move here.

Something else is also not surprising: Many of the people who already live in Boulder would prefer that the newcomers settle somewhere else.

“The quality of the experience of being in Boulder, part of it has to do with being able to go to this meadow and it isn’t just littered with human beings,” said Steve Pomerance, a former city councilman who moved here from Connecticut in the 1960s.

All of Boulder’s charms are under threat, Mr. Pomerance said as he concluded an hourlong tour. Rush-hour traffic has become horrendous. Quaint, two-story storefronts are being dwarfed by glass and steel. Cars park along the road to the meadow.

These days, you can find a Steve Pomerance in cities across the country — people who moved somewhere before it exploded and now worry that growth is killing the place they love.

Economically, this is looking like an interesting question. Under what conditions can a city be “too big”, and are those conditions likely to hold true in practice?

There are some economic models setting out why and how cities might grow to be larger than their optimal size. (The same models also predict that cities can be smaller than optimal, but I will ignore this case for the moment.) Economist David Albouy and three co-authors investigate the theory of the issue in a November 2016 paper entitled “The optimal distribution of population across cities“.

Albouy et al develop a model of city size that includes two offsetting externalities associated with city size. On the positive side, agglomeration economies, or the economic and social benefits of scale and density. On the negative side, congestion and crowding, which are assumed to increase nonlinearly with city size. Putting it together, they get a picture that looks something like this:

This probably doesn’t make a lot of sense unless you’ve read the paper and sifted through the equations. But it’s pretty simple. If you’re seeking to maximise the net social benefits created by the city, you want it to be size ni (on the X axis). That’s the point at which the marginal costs imposed by the next city resident exceed the marginal benefits that they deliver.

But people will continue to move to the city even after it hits this size, as new residents receive the social average benefit from locating in the city, rather than the marginal benefit. Left uncorrected, the city will grow to a larger size (nm_large on the X axis).

This model shows how cities can grow to be larger than their optimal size, due to congestion costs that increase faster than agglomeration benefits beyond a certain size. However – importantly – it also make a very strong prediction that people will stop wanting to move to cities at a certain point. In other words, this model does not predict that cities will grow without limit: it predicts that they will reach a certain size and then stop growing.

In that sense, this model predicts that city size is analogous to road congestion. Although many roads are above the socially optimal level of congestion, congestion simply doesn’t increase without limit. At some point people decide not to drive any more, as illustrated empirically by Wallis and Lupton:

Within the model, there may be reasons why optimal size differs between cities. For instance, cities may:

  • Have different types of agglomeration economies, resulting in a stronger or weaker case for increased size
  • Have different transport systems leading to different relationships between growth and congestion
  • Be located near more sensitive ecosystems, meaning that growth may be more damaging.

However, on the whole, the predictions made by the model are extremely difficult to reconcile with observed reality, which is that urban growth follows a ‘random walk’ process, with large cities more or less equally likely to grow as small cities. (This is often referred to as Ghibrat’s Law.)

In New Zealand, we can see this at work. Auckland has grown faster than most of the rest of the country over the last century (excluding smaller centres close to Auckland), and it’s expected to continue to do so. Canterbury, which contains one of NZ’s two next biggest cities, is also expected to grow rapidly. If models of optimal city size held true, we’d expect Auckland to be at a disadvantage for further growth:

Statistics NZ’s 2013-2043 population growth projections

As Fujita, Krugman and Venables observe in their great book on new economic geography, the idea that urban growth is more or less random is empirically plausible but tends to upend models like the one I’ve described above. Cities, it seems, are not like roads: They can always fit a few more people in without grinding to a halt. Although it is a common trope, there may be no such thing as a city that is “too big”.

A future for Auckland’s transport network

Lately I’ve been thinking about how to better join the dots between Auckland’s housing challenges and its transport challenges. We’re all familiar with the common stories about Auckland’s problems: Housing is too expensive, pricing young people out of the market and forcing low-income households into crowded or unhealthy accommodation. The transport system isn’t working as well as it could – key roads are congested, public transport is often unreliable due to our mid-century decision to eschew a rapid transit network, and walking and cycling often feels unsafe, again due to policy choices.

But it strikes me that we aren’t yet telling a clear story about how we could solve Auckland’s challenges. This is an attempt to tell some of that story.

It all starts with the street. When Auckland’s suburbs started to get built in the late 1800s, people did a few things to cut costs. One of those was providing long, narrow residential sites without back alleys or many cross streets. This left behind more saleable land while avoiding the need to provide stormwater or sewerage – people simply dug long-drops at the back of their long sites.

The result was a city that has a dearth of streets. The following map compares my neighbourhood in Auckland with my brother’s neighbourhood in Denver, Colorado – his is a bit further from the city centre but otherwise similar. Note the fine mesh of cross-streets and the closely-spaced arterial roads in Denver, and the spidery mesh in Auckland:

When we zoom out the map, the comparison gets even starker. Not only does Auckland lack a Denver-style street grid, it also has a regional transport network full of gaps and pinch-points caused by its position on two harbours.

This is exacerbated by the fact that we have recently built out most of the space in most designated motorway corridors. Once the Waterview Connection opens, the motorway network will be largely complete and will probably never be significantly expanded again, at least within the city. Contemplate that, for a moment.

In short, we are a growing city that lacks street space and has extremely constrained ability to add more transport corridors virtually anywhere in the city.

This brings me on to the second part of the story: Cars. Cars are wonderful things. They are the best way to get to the West Coast beaches, and the second-best way to get to urban beaches, after cycling. If it weren’t for home delivery, they would be the only way to buy a refrigerator or a tonne of compost for the garden. But we’re not going to be able to fit an ever-growing amount of them on Auckland’s roads at peak times. We don’t have the space for it.

In saying this, I’m not arguing that we should necessarily fear congestion. Auckland’s existing performance isn’t terrible: the aggregate cost of congestion is right about what you’d expect based on data from large Australian cities, and average commute times are reasonable. But the constrained nature of our street grid and regional motorway network leads me to think that it will tend to increase more rapidly as the city grows. Consequently, we will need to do something differently.

[A brief digression: We will face this problem regardless of where new residents end up living. Banning growth in your neighbourhood and insisting that all newcomers move to Drury will not solve the problem: many of those people will simply hop on the road to commute to jobs in the city or in the growing Auckland airport business park. Similarly, banning growth on the fringes won’t fix the problem either: many newcomers will still need to drive to get to jobs spread around the city.]

This leads directly to the third part of the story: What can we do instead, if the current approach won’t keep working?

Basically, there seem to be three things we can do.

One: We can implement congestion pricing – or, as Jarrett Walker calls it, a decongestion charge – to take the edge off peak-period delays on busy corridors. I’ve discussed this extensively in the past so won’t rehash this discussion here. One point that many people raise, though, is that congestion pricing should be paired with a strong focus on improving alternatives to driving, to allow people to avoid the charge.

Two: We need to improve Auckland’s regional rapid transit network to ensure that it is possible to travel longer distances within Auckland both quickly and reliably. Setting aside congestion pricing for a moment, rapid transit is the only way that we can reliably achieve this. If you want to travel 20 kilometres and get to work on time most days, you’re better off being in a train or a busway service than a car.

Rapid transit improvements are likely to be especially important for making greenfield growth work well. People who will soon be living in Dairy Flat, Whenuapai, and Drury can benefit from the option to access fast and reliable transport options.

However, good rapid transit isn’t simply a matter of building a busway out to the wops. Service integration is also essential. What that means is that buses or trains need to connect with each other at key points, offering easy and reliable transfers between services and access to a wider range of destinations. Interchanges like Otahuhu and Panmure are important, but the city centre is even more important, as it will always be the place where most of the lines converge.

In other words, if we want to make rapid transit work well for greenfields, we also need to sort out what’s happening to buses and trains downtown and in the inner urban areas.

Three: We need to improve Auckland’s urban cycleway network to give people new options for short- to medium-distance trips within the existing urban area. Cycling has a lot of unrealised potential in Auckland (and most New Zealand cities): At peak times on congested roads, a bicycle can get you to your destination faster than a car, and technological improvements (ebikes!) are flattening out the hills as we speak.

Getting more people cycling for everyday transport would go a long way to sorting out the transport challenges associated with new housing development in a city with a fragmented street grid. Every person who rides to the shops or to work is one who isn’t competing for road space and parking space. We will value those people more in the future.

A key barrier to cycling in Auckland is the perception that it is not safe. This doesn’t necessarily dissuade the mid-30s bloke in lycra, but it will keep many schoolkids, middle-aged women, and a whole bunch of other people off their bikes. We can fix this – and get people from ages 8 to 80 cycling – by designing streets better and providing safe cycling infrastructure where it’s most needed.

To summarise: Auckland’s built itself into a bit of a hole, and in order to meet the needs of a growing city, it will have to do things differently. That means congestion pricing (to make the road network work better), a really good regional rapid transit network (to ensure fast and reliable journeys throughout the urban area), and a safe, joined-up network of urban cycleways (to give people more options for shorter trips). This shouldn’t be seen as an alternative that we could pursue once we’re done building motorways: it is now the most realistic way forward for the city.

What do you think Auckland should do in order to address its growth challenges?

Population Growth in 2016

Last week Statistics NZ released their latest detailed population estimates for the country and they show that Auckland is continuing to grow at a rapid pace. Population increased by an estimated 44,400 people, an annual increase of 2.8% making it the fastest growing region in NZ and of course Auckland already started with the largest base. That growth has pushed population in the region above 1.6 million people having only passed 1.5 million about three years ago and at current rates, would see Auckland hit 2 million people in less than a decade. Auckland accounted for 46% of all the growth in NZ.


In 1996 Auckland’s population was 1.1 million so in 20 years has grown by the size of the entire Wellington region after taking 51% of NZs total growth. Auckland is now home to 34.4% of New Zealand’s population.

The graph below shows the change in Auckland’s population over the last 20 years and as you can see, as a total the growth this year is the highest ever seen although as a percentage increase, it was topped by 2002 and 2003. At 2.8%, growth is running ahead of even the high growth predictions released by Stats NZ in early 2015 which if it continues will have serious implications for the timing of many projects.


An even more staggering increase can be seen in the Auckland City Centre – as defined as in the map below. This is slightly different from the traditional definition as it includes the area units of Newton (south of Karangahape Rd) and Grafton West (east of Symonds St) and is what was used by the Ministry of Transport when they were monitoring targets for the City Rail Link.

Auckland CBD Area Units

As you can see below, the population of the city centre substantially increased to just shy of 47,000, up from just 5,000 20 years ago, a staggering 924% increase. The increase of 5,770 accounts for a full 13% of Auckland’s growth and 6% of all the growth in NZ over the last year. And of course, more people in the city centre has continued to help in making the place more vibrant and liveable. It also is another reason why it’s important agencies like Auckland Transport need to drastically change how they think about transport in the city. Moving to make it easier for pedestrians (and bikes) will be a key factor in improving the lives of the people living in the area, not to mention the other 100k+ that spend much of their time there every day.


Even if you used the more traditional definition of the CBD – the three northernmost areas in the map above, the population is sitting just under 40,000, up 5,000 on last year. As a comparison, the 2012 City Centre Masterplan suggested the CBD would reach not reach a population of 45,000 till 2032.

Looking at population at a slightly higher altitude we can see the changes by local board. The biggest increase comes from Waitemata driven by the growth mentioned above. Following that there has been strong growth in a number of local board areas. Many are areas where there is also a lot of greenfield growth has been occurring, such as Upper Harbour so in that regard one that also sticks as having a strong level of growth is Orakei.


Given the recent elections, the population figures also got me thinking about the levels of political representation we have and whether the ward boundaries need to be changed before the next election.

Auckland has 20 Councillors spread across 13 wards. In 2010 when the city was amalgamated there was a population of 1.44 million meaning for each Councillor there are an average of 72,000 residents. With the population now up to 1.61 million people we’re approaching one Councillor for every 81,000 people, more than twice the number of people per member of parliament. Furthermore, with so much growth in some areas, it’s leading to even more disproportionate representation and this is especially the case in the Waitemata and Gulf area which includes the CBD area.

This graph showing the population per Councillor can look a little busy but does show clearly how much things have changed in Watemata. The area is now home to around 112,000 people and they’re all covered by just one Councillor, Mike Lee. I don’t think that’s fair to either the residents in the Waitemata area or to Mike Lee



The same issue, although to a lesser extent, exists at the local board level with some local boards having vastly different levels of representation. Each local board has differing numbers of members on them. Again, Waitemata is a clear mover although it is still behind the Howick Local Board and Henderson-Massey Local Board on a per population basis. which has the highest population compared. Howick has just over 16,000 people per local board member while Papakura is the lowest of the urban area Local Boards with almost half that at just 8,800 people per member.

One of the challenges is that the amalgamation legislation prevents the number of Councillors being increased like can happen in the rest of NZ. It also requires that Council Wards align with the local board areas so any change isn’t straightforward.

What do you think should happen, change the ward and local board boundaries, allow for extra Councillors and local board members to be elected, a combination of both or leave things unchanged?

Auckland’s urban growth: a historical perspective

The other week, BNZ economist Tony Alexander made an interesting point about Auckland’s current urban growth (via

The plan set out in Auckland’s proposed Unitary Plan to build 422,000 houses will require a rate of building over 25 years that it previously took 161 years to achieve in the City of Sails, BNZ chief economist Tony Alexander says.

In his Weekly Overview report Alexander notes the Unitary Plan’s aim for 422,000 extra houses to be built over the next 25 years requires an annual construction rate 2.5 times higher than the average achieved over the past 25 years. Canterbury achieved 1.5 times post-earthquake.

“Look at this another way. As at the 2013 census there were 509,000 dwellings in Auckland. In 2001 there were about 420,000. The plan is another 422,000 in the next 25 years, in other words doing in 25 years what it took 161 years to do following the designation of Auckland as the country’s capital by the first Governor William Hobson in 1840,” Alexander says.

Now, it’s worth putting some caveats on this. The Unitary Plan allows around 422,000 new ‘commercially feasible’ homes to be built in Auckland – but it doesn’t guarantee that they will be built. Other factors, such as the availability of construction labour or funding for infrastructure, could still pose road-blocks. On the other hand, Auckland’s population might not grow fast enough to need another 422,000 homes, eg if the New Zealand economy declines for a prolonged period.

But Alexander’s comments nonetheless put matters in perspective. Auckland is currently experiencing historic levels of population growth. The city has never been faced with the task of accommodating as many people, in as short a time-frame, as it must at the moment.

[As an aside, I’ve covered the reasons why this is happening in a previous post.To summarise:

  • The majority of recent and future population growth – about 60-65% – is from natural increase, i.e. people being born in Auckland
  • Net migration, principally from overseas, accounts for the rest. Net migration is very volatile – high inflows in one year can be balanced out by outflows in the next.

For the record, commenters complaining about population growth or the origin of migrants without engaging with these facts or providing references for their own views will be deleted.]

Migration Herald

How unprecedented is Auckland’s current and projected population growth, anyway?

Historical data on urban populations provides a window into this question. Here’s one of the key graphs from a 2013 paper by Arthur Grimes and Nicholas Tarrant, showing the growth of New Zealand’s largest five cities in 1926.

Grimes and Tarrant urban population growth

At the start of the 20th century, Auckland wasn’t that dissimilar to Christchurch, Wellington, and Dunedin – a bit bigger, but not by much. But over the course of the century, it diverged, and, what’s more, the rate of divergence seems to have increased. In 1926, Auckland was only 60% larger than Wellington, the second-largest urban area. In 2006, it was 235% larger:

Auckland to Wellington population 1026-2006

It’s common to measure urban population growth in percentage terms. That’s not a bad measure, in a lot of ways, as it captures the degree to which cities of different sizes are experiencing relative changes. Adding 500 people to a town of 10,000 may feel as transformative as adding 50,000 to a city of one million.

But when you’ve actually got to build stuff to meet demand – homes, roads, pipes, busways, etc – the raw numbers matter a lot. When Auckland grows by 1%, it has a much greater impact on the national construction task than 1% growth in Timaru.

So with that in mind, here’s a graph showing the last 115 years of urban population growth in Auckland, plus the projections for the next 30 years or so. I’ve pieced it together from a couple of sources – Grimes and Tarrant’s numbers, including their unreported population figures for 1901, Stats NZ’s population estimates for 2006-2015, and Stats NZ’s population projections for 2013-2043. (These figures don’t line up perfectly, but they’re still broadly comparable.)

Here’s what we’re looking at. Between 1996 and 2015, Auckland added over 20,000 new residents a year – the fastest-ever increase. The only period that comes close is the late 1960s/early 1970s, and that turned around pretty quickly.

The range on future population projections is wide. At the low end, Auckland would “only” add around half a million people over the next three decades. At the high end, it would add another million people. However, the mid-point of the range, which is reasonably reliable over multi-decade periods, would see the growth trends of the last 20 years continue.

Auckland historical and projected population growth

Now, this could potentially be a very good thing. The last 20 years of urban development has been, on the whole, positive for the city. It’s injected new life back into the city centre, created demand for a wider range of housing and transport choices, and given Auckland – for possibly the first time ever – great dining options. None of that would have been possible in the absence of population growth.

But growth also creates challenges. The city’s post-war growth model – low-density zoning and car-centric development on the fringe – no longer works. While there are still greenfield sites that can be subdivided, they’re located in less desirable places – Dairy Flat and Drury aren’t comparable to Browns Bay or St Heliers. And fringe locations built around the car are becoming increasingly expensive to connect in to the rest of the city, due to the need to expand road capacity across pinch-points. All the cheap motorways have already been built:

Cost per added lane-km chart

In other words, the city will have to change to accommodate growth. That means changing the way we regulate and build the housing market – the Unitary Plan has made a useful contribution. It also means rethinking the way that we invest in and manage our transport system. If the cost to develop transport corridors is going up, then we need to make sure that we get the best use out of them.

This means looking more closely at how congestion pricing can balance demands between time periods. It means choosing to dedicate more corridor space to high-capacity modes like rapid transit, and, in reasonably dense areas, cycling.

Packed Train

In addition to better enabling Auckland’s future urban development, there are a range of other benefits to this approach. People who have better options to get out of the car tend to be more physically active and healthy. Town centres where more people walk, ride, or bus up are often more vibrant and economically successful than those where people must navigate a maze of carparks. And, in generally, more connected places with better street life tend to be more enjoyable for everyone.

Here’s hoping the next phase of Auckland’s growth takes us in that direction.

Devonport population growth and stagnation, 1891-2013

Following on from a previous post, this is a quick review of population growth in Devonport over the last 125 years. These figures are for the former Devonport Borough, which was created in the 19th century and persisted until 1989 when it was merged into North Shore City. The borough only really covered the southern half of the Devonport peninsula – the northern half, including Seacliffe, Hauraki and Bayswater, was part of the Takapuna Borough.

The map below shows the Devonport Borough as it looked in 1899:



And here’s how the population has changed (or not) since 1891:

Devonport's Population

I’m surprised that Devonport’s population has been completely flat for the last 70 years. The population on census night 1945, of 11,662, was still 318 people higher than the population on census night 2013, of 11,346.

Zoning controls, have certainly played their part in limiting the number of people who can live in a very desirable coastal area. There are probably a few more houses in Devonport today than there were 70 years ago, but any growth in household numbers has been cancelled out by there being fewer people per household.

Unfortunately, I don’t have long-term data for the northern half of the peninsula. It was part of the Takapuna Borough, and as the name suggests that included a number of other suburbs as well. Since 1986, though, the northern half’s population has risen from 9,251 to 11,862.

Given that the population has (at best) increased modestly in the last 30 years, Devonport is quite lucky to have such a high quality ferry service today. Tourists and other Aucklanders visiting Devonport help to support this service, and of course they also contribute to traffic on Lake Rd, especially on weekends.

A bit of intensification around the ferry terminal and the wider peninsula would support further transport upgrades, such as more frequent ferry sailings and upgrades (or widening) to Lake Rd. Under the Proposed Unitary Plan, though, the opportunity to add more homes and people close to the ferry looks very limited.

Again, the northern half of the Devonport peninsula does have some growth on the way. As per the RCG Development Tracker, Ryman are building a retirement village and Ngati Whatua O Orakei will be redeveloping ex-Navy land for apartments. There’s also a proposal to build apartments at the Bayswater Marina, which seems like a perfect location – right next to the (less frequent) Bayswater ferry, coastal amenity, and few immediate neighbours meaning very little downside for existing residents.

Central Auckland population growth, 1891-2013

I’ve been investigating how different parts of New Zealand have grown in the last 125 years, inspired by a 2013 Motu study. As part of that, I noticed that there’s actually pretty good data for ‘central’ Auckland, i.e. the isthmus (other parts of Auckland are a bit trickier, but hopefully I can work through that). The 2010 amalgamation of all the old Auckland councils still seems quite recent, but those councils had actually only been around since 1989.

Before that, we had a system of boroughs and counties which had survived quite intact since the 19th century. And it turns out that the Eden County, and the various boroughs within it, was almost identical to the 1989-2010 Auckland City boundaries, what Stats New Zealand now call the “Central Auckland Zone” of the Auckland Urban Area. The only difference is that it excluded Otahuhu.

If that all sounds like too much of a mouthful, here’s a map of Eden County – I’ve just used the figures for that, but adding in Otahuhu.



And here’s what the figures look like:

Central Auckland's Population v2

By 1891, there were already more than 50,000 people living in central Auckland, which grew pretty steadily until 1971. The population hardly budged from 1971 to 1991, actually dropping slightly at one point, but since then it’s been on a major growth path, faster than at any time previously. My guess is that we’ll keep up this growth for a long time to come.

Incidentally, Auckland, Wellington, Christchurch and Dunedin all had quite similar populations at the end of the 19th century. It was only in the 20th century that Auckland became so much larger than any of our other cities.

So why did central Auckland stop growing in the 70s and 80s? New Zealand as a whole had lower growth rates in this period – lower migration, perhaps. Auckland had lower growth as well, and most of the growth that did happen was out to the north, west, or (especially) the south. From some other Stats data (and note that Otahuhu has shifted back to south Auckland for this one – population around 11,000)

Total Auckland Population v2

Another interesting point here is that central Auckland dominated population growth in the region until well into the 1950s. It was after 1956, or maybe even 1961, when the southern, western and northern areas started to take off.

The “consumer city” and urban growth in Christchurch

This article was originally posted on Making Christchurch, a group blog set up by Barnaby Bennett in the wake of the 2011 Canterbury Earthquake, at the invitation of Transportblog commenter Brendon Harre.

Why do cities grow and change? And how can cities harness those dynamics?

Last month, I took a look at agglomeration economies, which describe the productivity and innovation gains arising from urban scale and density. The advantages that cities offer for production have underpinned urban success throughout history.

Economic productivity is important. To paraphrase Paul Krugman, in the long run, productivity growth underpins our ability to consume more of everything from electronics to healthcare, and to have more of the non-economic things that make life enjoyable. All else being equal, people tend to move towards more productive places in search of higher living standards. But economic productivity isn’t the only thing that matters for wellbeing — or for growth and change in cities.

Urban economics tackles urban amenities

For a long time, people assumed that cities offer advantages for production but disadvantages for consumption. This assumption, which shaped a lot of economic analysis and policymaking, was understandable. After all, modern cities first arose as manufacturing centres at a time when manufacturing was a dirty business. People could get jobs in the city’s “dark Satanic Mills”, but they had to suffer bad air, choleric water, and high crime rates to do so.

But things appear to have changed over the last half century, at least in developed countries. The bad aspects of cities, such as crime and pollution, have improved, and the good parts have also gotten better. Cities have become attractive for consumers as well as producers.

A pioneering 2000 paper by Ed Glaeser, Jed Kolko and Albert Saiz explored these dynamics. They argued that the availability of “four critical urban amenities” would shape future urban growth:

  • The availability of a rich variety of consumer goods and services —  which, in the era of and the iTunes store, means “non-tradables” like restaurants, live bands, bars, and dating opportunities
  • Aesthetics and natural settings — in other words, the quality of the city’s architecture, public parks, natural environment, and climate
  • Good public services such as schools and low crime rates
  • The quality and speed of transport systems — cities that make more destinations accessible are more likely to be attractive to residents.

In their view, the rise of the “consumer city” opens up other pathways for urban growth. If cities want to attract new residents and businesses, they don’t have to focus only on providing “producer amenities” like convention centres. Supplying great “consumer amenities” can also foster ongoing vibrancy and growth.

Here, I want to look at the prospects for New Zealand cities, and Christchurch in particular, to become successful “consumer cities”. I’m going to focus on the first two dimensions — variety in goods and services and aesthetics and natural settings — and leave a discussion of transport for a future post. (Public services are a bit less relevant to urban growth in New Zealand, as education and law enforcement are run by central government.)

Goods and services

New Zealand cities are coming around to the importance of consumption. It hasn’t always been thus. In the middle of the 20th century, when my parents were growing up in Auckland, the country was firmly in the grips of what historian James Belich called the “tight society”:

homogenous, conformist, masculist, egalitarian and monocultural, subject to heavy formal and informal regulation. There were no licensed restaurants, little weekend shopping, one supermarket (opened in Auckland in 1958) and a very limited range of goods and foods to buy in the shops and unlicensed restaurants that did exist… School milk was free, but you had to drink it.

A lot has changed since then, economically and demographically. While the wholesale deregulation of the 1980s was not an unmixed blessing, it certainly expanded the consumption choices available to New Zealanders. A more liberal migration policy brought new migrants with, thank goodness, new cuisines. And since the advent of mass-market international air travel, Kiwis returning from OEs have come back with new ideas for things to do in cities — from rock bands to restaurants to cycle lanes.

The result is a favourable climate for the adoption, invention, and proliferation of a variety of goods and services in cities — especially when it comes to bars, restaurants, and entertainment.

Christchurch has been instrumental in shaping a key part of the hospitality market: beer. When I started to be able to afford to drink nice beer in bars, the best thing on tap was often from craft breweries in Christchurch like Harrington’s and Three Boys. Their success has fostered competition: craft brewing has since taken off in Wellington and, more recently, Auckland.

In short, New Zealand cities have potential, but they may have to do a few things differently in order to fully realise it.

The first is simple: get some of the barriers out of the way. For example, minimum parking requirements can be a major impediment to opening new restaurants and bars, or converting old warehouse space to retail. They often require restaurants to devote more space to parking than to dining areas, which can be the kiss of death for hole-in-the-wall eateries.

The second is to understand — and take advantage of — positive feedback loops between population density and consumer amenities. Neighbourhoods with more people support a greater variety of consumption choices. While density isn’t for everyone, cities need some medium-to-high density, mixed use neighbourhoods to supply a rich variety of urban goods and services.

In medium-sized cities, city centres have traditionally filled that role. As the following population density maps (darker blue = higher density) show, that’s an area where Christchurch lagged behind Auckland and Wellington even before the earthquakes. The destruction caused by the 2011 Canterbury Earthquakes has created an opportunity for revitalisation on different lines — but government bungles seem to have delayed the process. It’s important that it get back on track.

Aesthetics and natural settings

Christchurch, like many other New Zealand cities, has some intrinsic aesthetic advantages as a result of the natural landscape. Here’s the view west across the city:


Photographer: P. Stalder

New Zealand’s environment has always drawn migrants, who often come for the landscape and live in the cities. Take, for example, novelist Eleanor Catton’s description of what drew her family to Christchurch:

I grew up on the South Island of New Zealand, in a city chosen and beloved by my parents for its proximity to the mountains — Christchurch is two hours distant from the worn saddle of Arthur’s Pass, the mountain village that was and is my father’s spiritual touchstone, his chapel and cathedral in the wild. For many years while I was growing up my parents did not own a car. We rode around town on two tandem bicycles and one single (a source of considerable embarrassment to me at the time) and at weekends we would occasionally rent a car in order to drive into the alps, and go hiking.

However, urban aesthetics also matter — even if you go tramping on the weekend, you still spend your weekdays in the city. This is an area where Christchurch has some strengths and some challenges.

From the start, Christchurch had a reputation as a “garden city” as a result of its large public parks and street trees. Although the idea of parks as a city’s “lungs” is less salient today than in Industrial Revolution cities, parks and street trees are still public amenities. They make people better off simply by existing in the vicinity.

The earthquakes seem to have opened up some opportunities to enhance the garden city — particularly along the Avon River, where many houses have been red-zoned. In addition to the central government-promoted Avon River Precinct, some community groups are calling for a forest park out to New Brighton. Others have simply gotten on with creating something new.

The built environment, however, is more problematic. Central government oversaw the demolition of over 1200 buildings in the city centre in the years following the earthquakes, including many of the city’s historic buildings.

Source: Wikipedia

This has been controversial and at times acrimonious. As I am neither an architect nor a Cantabrian, I’m not in a good position to weigh in on the debate. But as an economist from Auckland I’d observe that heritage buildings have a definite public value  — not one that trumps all other costs, but one that should be accounted for in decision-making. At the very least, it would be smart to replace any demolished buildings with more attractive and usable ones.

Prospects for population growth in Christchurch

Thus far, I’ve considered — in a thematic way — some of Christchurch’s challenges and opportunities as a “consumer city”. But what would success look like?

Let’s take a look at a few data points. First, here’s a chart showing Statistics NZ’s latest (2015) regional population projections. The Canterbury region, which includes Christchurch and its satellite towns, is projected to grow faster than all regions other than Auckland over the next three decades.

In other words, Stats NZ expects Christchurch to be relatively successful at attracting and retaining people. But look at the range on their estimates: the city could grow faster than Auckland, or it could hardly grow at all.

Without digging into Stats NZ’s forecasting methodology, it’s difficult to say why they’ve picked such a wide range. But perhaps it reflects uncertainty about the future attractiveness of Christchurch as a consumer city. Wages in Christchurch tend to be lower than in Auckland and Wellington, meaning that urban amenities potentially have a stronger role to play in fostering urban growth.

A second data point. A few months ago, I took a look at the sources of Auckland’s population growth. I found that natural increase accounted for the majority of growth but that net migration — more people arriving than departing — fluctuated wildly from year to year. Here’s what the picture looks like for Canterbury:

Source: Stats NZ Infoshare

Net migration to Canterbury has followed a very similar trend as net migration to Auckland — the peaks and the troughs coincide remarkably well. However, the troughs are just a little bit deeper in Christchurch, as substantial numbers flow out of the region in a bad year. In Auckland, by contrast, net migration seldom turns negative.

Net migration will always be a bit of a rollercoaster in New Zealand — it’s followed a boom-and-bust cycle for a very long time. But it’s possible — with the right combination of a resilient economy and good consumer amenities — to reduce the depths of the troughs and raise the height of the peaks. It might not be an inspiring mission statement for a city, but perhaps it’s the right one for Christchurch.

Auckland Population change heat map

Included in the Housing NZ presentation describing what changes they wanted for zoning in the Unitary Plan that I wrote about the other day was this interesting graphic showing the level of change in the population of Auckland from the 1996 to each of the census’ since.

Change in Population Density 1996-2013

That bright red spot above the city centre is one of the is the result of thousands of people shifting into the area and stats NZ estimate that the growth has been continuing with the population of the city centre now over 40,000, up from just 5,000 in 1996. This growth is one of the key reasons the city centre has become more vibrant in recent years – and will continue to do so.

City Centre Population - 1996-2015 2

The bright red around Albany, Henderson and Botany will represent large swathes of suburbia being being developed on greenfield lands.

Unitary Plan Fantasies


In this post I discuss two related questions that concern common “fantasies” about the Unitary Plan, specifically:

  • Question #1: To what degree has Auckland’s density changed during the last few decades?
  • Question #2: To what degree does the balance of brownfields/greenfields development in the Unitary Plan differ from the past?

We might be able to agree on answers to these two questions. Why? Well, they are positive questions, insofar as they refer to attributes, i.e. density and brownfields/greenfields development, which are able to be subject to empirical measurement and testing.

Ideally people would agree on answers to important positive questions before moving onto normative questions, because the latter are not empirically testable. An example of a normative question would be: “How much weight should we place on the preferences of existing homeowners versus potential homeowners? I hope the difference is obvious; normative questions tend to be gnarlier.

It’s often helpful to separate positive from normative statements. People can often vehemently disagree on the answers to normative questions, while still agreeing on the answers to positive questions. Hence, in this post I will try to provide clear answers to two important positive questions that seem to be frequently misunderstood by those who oppose the Unitary Plan. Rest assured that I hope to tease out some of the important normative questions in more detail in a subsequent post.

Question #1: To what degree has Auckland’s density changed during the last few decades?

The answer to this question is simple: In the last 10-15 years the population density of Auckland has increased. In this working paper, Peter quantifies the density for various New Zealand cities, which are summarised in the following table. We see that Auckland’s population-weighted density (i.e. the density at which the average resident lives) has increased by around one-third (33%) in just over a decade.

As Peter discusses in this post, increased density is consistent with other empirical data. When we look at population growth in Auckland, we find that the population of central areas, especially the city centre, is growing faster than other places in the region. Waitemata (which covers most of what we refer to as the “Isthmus”) stands head and shoulders above the rest in terms of population growth, both in total and relative (%) terms, as shown below.

The increase in density observed in central areas doesn’t seem to be caused by regulations on urban expansion. Instead, Auckland seems to have grown denser primarily because there is increasing demand from people to live and work centrally, i.e. as a result of people’s preferences. Research by Arthur Grimes, for example, has found that Auckland’s central areas have become much more valuable relative to less central areas, as illustrated in the figure below.

This change is significant, and is mirrored in cities elsewhere, such as Amsterdam (NB: Amsterdam has always controlled urban expansion, providing further evidence to suggest that controls on urban expansion are not behind changes in the relative values attached to centrality). Increasing density in Auckland are also consistent with the experience in Sydney and Melbourne, as illustrated in the figure below (NB This figure is taken, incidentally, from the excellent ChartingTransport website). Here we see that density in both Sydney and Melbourne increased by a similar % to that observed in Auckland. 

So from where I’m sitting the answer to the first question is fairly clear: Over the last 10-15 years or so Auckland has become a much denser place, and it’s become denser because more people and firms want to locate in central areas. As far as I know the sky hasn’t fallen on our heads. Indeed, from what I can tell Auckland has been doing relatively well of late. 

In this context, the imposition of regulations preventing intensification would seem to have the following impacts:

  1. Reduced development and higher property prices;
  2. Fewer people and jobs being located in central areas;
  3. Increased urban expansion, with associated infrastructure, congestion, and energy costs; and
  4. Transfer of wealth from those who have less to those who have more (further reading).

The likes of Richard Burton, Dushko Bogunovich, and David Seymour may argue that the costs of regulations preventing intensification are outweighed by the benefits, e.g. maintaining the “character” of inner-city suburbs.

I know of no quantitative evidence to show this is the case. On this basis I think it’ fair to say that their claims are unsubstantiated, at least in quantitative sense. I note that recent changes to the RMA (passed, incidentally, with the support of the ACT Party) places a higher bar on the economic evidence needed to support restrictions on development. In the absence of such evidence, and given the large body of quantitative evidence that demonstrates the costs of regulations that prevent intensification, arguments against intensification would seem to be rather flimsy. I can only hope that the IHP agrees.

Question #2: To what degree does the balance of brownfields/greenfields development in the Unitary Plan differ from the past?

The answer to this question is hinted to in the previous discussion: In the last two decades most development has happened within the existing urban area, i.e. brownfields. More specifically, development has been split 71% and 29% between brownfields/greenfields respectively. Data supporting this analysis is summarised in the table below, which is extracted from the Development Strategy published by the Auckland Council (available here).

The historical percentage of brownfields/greenfields development is similar to that enabled by the Unitary Plan (60-70% and 30-40% for brownfields and greenfields respectively). At this point I think it’s worth highlighting a rather extraordinary exchange from Peter’s recent post on the linear city (source).


To explain:

  • “Brian” asks Duskho Bogunovich (who works for Unitec and has publicly criticized many aspects of the Unitary Plan) what proportion of Auckland’s historical growth has been accommodated within the urban area (“brownfields”) and what proportion has been outside (“greenfields”); and
  • Duskho replies with “I don’t know” but then suggests a ratio of 1 part brownfields to 5-10 parts greenfields. Converting this into percentages would imply that Dushko believes 9-17% of historical development has been brownfields, with the balance in greenfields.

Dushko’s numbers are at odds with the data presented above. Indeed the data flips his percentages around completely. Now, in Dushko’s defense this particular question asked about the last *30* years whereas the data presented above goes back only *20* years. On the other hand I can’t see this ratio changing too dramatically though even if we went back one more decade.

The key takeaway message from this exchange is that 1) Dushko doesn’t know the actual brownfields/greenfields ratio and 2) the data which is available suggests a brownfields/greenfields ratio that is at odds with his intuition. I personally would expect that those who oppose the Unitary Plan, such as Dushko, would spend some time familiarizing themselves with the empirical evidence, especially when such evidence is crucial to the argument they are themselves advancing.

Keep this issue in mind when you consider another one of Dushko’s comments (source):

But forcing massive intensification inside Auckland cannot fix the housing crisis anyway … The city must grow both ways – up and out – to allow the land and housing market work properly. And getting the ‘up/out’ ratio right is crucial … this ratio for Auckland is probably 1:2. That is, 1/3 should be growth by intensification, and 2/3 by growing out (new suburbs; satellite towns; redistribution to the outer region – Waikato and Northland). Sadly, the council, in its ‘compact city’ ideological zeal, managed to get this ratio exactly the opposite – 2:1. The ‘70% fantasy’. This is PAUP’s fatal flaw. That’s why the Plan is a dud. And will never be implementable. Unless we use the North Korean approach.

In Dushko’s world, Council via the Unitary Plan is “forcing massive intensification” that is at odds with the “right ratio” for intensification. Dushko’s sees evidence of “ideological zeal” and “fantasy”, ultimately concluding that the PAUP is “fatally flawed” and a “dud”, which will not be able to be implemented unless we resort to North Korean style policies. Hyperbole much?

Especially when one considers the empirical data. Put simply, the Unitary Plan simply is proposing to continue long-established trends in Auckland’s urban development, which have resulted in steadily increasing density with a 70%/30% brownfields/greenfields split.

People like Dushko might argue that we would be better off if changing these trends. I’d disagree but, hey, let’s have that debate. It’s fair game.

What doesn’t seem fair game is for people like Dushko to criticize Council’s Unitary Plan and suggest it represents a “radical” change from the past, when in most respects it’s business-as-usual. Perhaps the only way the Unitary Plan can be described as “radical” is that it provides for only 80,000 new homes to be developed over coming decades, when official population projections suggest we will need approximately 400,000.


I started this post by posing two “positive” questions, to which I have since suggested the following answers:

  • Question #1: To what degree has Auckland’s density changed during the last few decades? Auckland has become 33% denser since 2001. This change appears to be driven more by the growing desire of people and firms to locate centrally, rather than regulatory controls on urban expansion. The increase in density observed in Auckland, and the increasing value placed on central locations, is consistent with trends observed in cities overseas, such as Sydney, Melbourne, and Amsterdam; and
  • Question #2: To what degree does the balance of brownfields/greenfields development in the Unitary Plan differ from the past? The last two decades of Auckland’s developent has seen a 71% to 29% split between brownfields/greenfields development respectively. This data seems to be at odds with the views of many people that oppose the Unitary Plan, who argue that Council is forcing “intensification” and a “compact city” on Aucklanders.

What do you think is fact or fantasy when it comes to the Unitary Plan? And on that note, what is your fantasy for Auckland. In 20 years time would you prefer to be 1) more dense; 2) less dense; or 3) about the same as now? Vote below.

Tauranga: Sunshine, sprawl, and high house prices

Back in 2014 I wrote a short paper exploring how population density had evolved in New Zealand and Australian cities. Among other things, the paper provided a rough estimate of the degree to which various cities were going “up” or “out” – i.e. whether population growth was increasing or decreasing the density of the neighbourhood that the average resident lives in.

Based on the data, you could divide New Zealand cities into a couple of different categories:

  1. Cities that are growing slowly or not at all, e.g. Dunedin, Whangarei, Gisborne
  2. Cities that are growing and becoming increasingly dense, principally Auckland but also Wellington to a slightly lesser extent
  3. Cities that are growing primarily by spreading out, e.g. Hamilton and Tauranga
  4. Christchurch, where normal urban processes were disrupted by the 2011 Canterbury Earthquake and the slow rebuilding effort since then.

There is an interesting comparison to be drawn between Auckland and Tauranga. They are both port cities with stunning natural environments, lots of sunshine, a fondness for urban motorways, and high growth rates. Tauranga is obviously much smaller, with less than 1/10th of Auckland’s population.

But whereas Auckland was the city that went “up” the most, Tauranga went “out” more than any other NZ city this millennium. Between 2001 and 2013:

  • The population of Auckland’s urban area grew by 23%, but its urbanised land area* only expanded by 11%
  • Tauranga’s urbanised population grew by 27%, while its urbanised land area expanded by 25%.

[* Defined as Census meshblocks with more than 3 residents per hectare. This isn’t a perfect measure as it tends to exclude industrial areas.]

In other words, Tauranga’s urban population expanded proportionately to its population, allowing it to remain a low-density suburban city. In order to accomplish this, the city opened up substantial new greenfield areas to the south, west, and east:

Tauranga urban growth 2001-2013

The future seems to herald more of the same. Tauranga-Western Bay of Plenty’s 2013 SmartGrowth Strategy identifies some opportunities for “possible intensification”, but largely commits to outward growth along motorways:

Tauranga SmartGrowth Strategy map

Tauranga’s sprawl serves as a useful “counterfactual” scenario for Auckland – does an abundant supply of greenfield suburbs necessarily result in cheap housing?

Perhaps. But it doesn’t seem to have worked out that way in Tauranga. According to the Demographia housing affordability survey, which compiles a range of useful data but is rather weak on interpretation of that data, median house prices in Tauranga-Western Bay of Plenty are currently 8.1 times higher than median household incomes – an increase from 6.8 the previous year. This is rather high by national and international standards.

Moreover, house prices in Tauranga appear to have followed a broadly similar trend to Auckland, with a run-up in the 2000s, several years of flat or falling prices after the 2008 Global Financial Crisis, and rapid price inflation in the last year or two.

What does this tell us about housing markets? Three things, I think.

The first is that people are willing to pay higher prices to live in cities with desirable amenities like harbours and sunshine. This shouldn’t be a surprise to anybody. We pay more to eat in restaurants that offer better ambiance and tastier food. Why wouldn’t we pay more to live in nice places? And Tauranga, like many other New Zealand cities, is undoubtedly an attractive location:

Source: Wikipedia

Source: Wikipedia

The second is that greenfield land supply is not necessarily a solution for house price inflation. Tauranga is less than one-tenth the size of Auckland and its house prices are already high relative to local incomes. Adding greenfield land supply hasn’t prevented or reversed previous price increases. In larger cities, where fringe locations are much less of a substitute for desirable central locations, it’s likely to be even less effective.

The third is that Tauranga (and many other New Zealand cities) may have to rethink their approach to housing policy.  This is especially true for cities experiencing rapid growth. According to Statistics NZ’s latest (medium) population projections, Tauranga’s population is expected to increase by a further 43% over the next three decades. So the pressure on Tauranga’s housing market is likely to continue unless something changes.

In this context, it’s worth asking a few critical questions:

  1. What is a reasonable expectation for house prices, given geographical constraints, environmental and man-made amenities, and demography and demand?
  2. If greenfield land supply isn’t sufficient to enable growth without accelerating prices, what other policies are needed? For example, how can planning policies facilitate choices of dwellings in various places at various price points?