Back in July former World Bank urban planner Alain Bertaud and his wife Marie-Agnes, a fellow professional in the field, came down to New Zealand at the invitation of the NZ Initiative and the Minister of Finance’s office to deliver a series of talks on urban economics. He had a number of thought-provoking things to say to urbanists of all stripes – a message that was very much in line with Transportblog’s core principles and big ideas.

He looks much happier in person

While Bertaud is sometimes cited as a proponent of low-density urban sprawl and motorway development, his arguments about urban development were nuanced and thoughtful. The Bertauds are, after all, urbanists themselves. They have chosen to live in vibrant, dense, and diverse cities – Paris, Washington, D.C., and lately New York. (That’s a revealed preference if I’ve ever seen one – they certainly don’t live in Houston!)

In addition to seeing Bertaud’s talk , which is available online here (pdf), I was lucky enough to sit in on a smaller discussion section with other professionals in the field. I took three key messages away from the talk and the conversation.

First, cities are labour markets. We often forget this fact, even though it’s the reason we have cities at all. Cities are the physical expression of agglomeration economies, or the productivity advantages of locating near other people and businesses. In Bertaud’s view, ensuring the efficiency of urban labour markets means ensuring that people can access a large number of jobs from their homes.

As a result, he argued that urban and transport planning should aim to keep down commute times. He recommended looking at two key measures – first, the number of jobs available within a 30 minute drive, and second, the number of jobs available within a 45 to 60 minute public transport journey. Here, for example, is his analysis of commute times in Singapore and the US.

Bertaud travel time graph

Bertaud didn’t recommend any specific policies to reduce travel times, although he spoke positively about Singapore’s use of demand-responsive road pricing and development of an expansive metro network to reduce average travel times. As a transport economist there are a couple of key observations I’d make on the topic:

  1. Building more roads is not a good way to reduce travel times. Induced demand – people driving more or moving further out of town in response to new road capacity – usually eats up the forecast travel time savings. In short, people travel more but they don’t travel any faster. If you want to actually reduce driving times, the only way to do it is to introduce road pricing.
  2. Cities with reasonable densities and an underdeveloped public transport network – like Auckland! – are in a good position to improve employment accessibility through investments in rapid transit networks and better bus networks. Fortunately, Auckland’s pursuing this approach.
  3. In light of induced demand, the best way to improve the accessibility of jobs may be to simply make things closer together. The efficiency of dense urban environments is often underrated. For example, although the roads in downtown Manhattan are far more congested than Houston’s, Manhattan’s effective labour market is much bigger simply because everything is so close.

Second, we must plan for the cities that actually exist, not the cities that we wish could exist. Bertaud presented an excellent graphic to illustrate this point. It showed four kinds of cities – three that exist and one that does not (and can not):

Bertaud urban structure graph

Most cities that exist today are what Bertaud calls “composite cities”, meaning that a significant share of jobs are located in the CBD or in major centres, while other jobs are scattered around in industrial parks, neighbourhood shops, etc. In this city, people have a range of different travel needs. Many people need to get to large-scale, high-density employment hubs, which are efficiently served by rail lines and busways, while others are better off driving to more dispersed employment locations.

In short, real-world cities require a range of transport solutions, and they will not function well if one mode is unreasonably neglected. We don’t have to go far for an example of the perils of mode bias: the remarkable renaissance of the Auckland city centre, and its increasing contribution to New Zealand’s economy, would not have been possible without reinvestment in the rail system and the development of Britomart.

City centre screenline survey results show public transport accounts for all growth in inbound trips over the last two decades

However, Bertaud criticised what he described as the “urban village” model, which hypothesises that if employment is dispersed evenly throughout neighbourhood centres then people will travel only to the nearest centre. This is a seductive idea – it promises to reduce travel distances by distributing employment. Unfortunately, it doesn’t exist in the real world because people have complex travel needs. Even if one member of a family chooses to live near where they work, their partner will often have to commute to a job further away.

We see this flawed idea pop up from time to time in New Zealand from advocates for suburbanisation. For example, people sometimes argue that we could reduce congestion by stopping growth in the city centre and relocating it to Manukau central instead. Aside from the fact that we tried this before and it failed, decentralising employment would only increase congestion from all the cross-town trips and reduce the efficiency of Auckland’s labour market.

We don’t have to go far for an example of the failures of the urban village model. Christchurch lost its city centre in the 2011 Canterbury Earthquake, and employment dispersed throughout the city. Although the jobs have decentralised, the city now suffers from higher congestion as it can’t run efficient bus services or provide enough road capacity.

Third, it’s important to ask whether planning regulations are restricting development in areas that are accessible to employment and amenities. Economists in New Zealand have spent a lot of time talking about Auckland’s metropolitan urban limits while paying little attention to regulations in the rest of the city. Bertaud argues that limits on density, such as building height limits or minimum lot sizes, can price out the poor from accessible areas. Incidentally, this may be happening in Auckland – my research found that poorer people tend to live further from employment hubs and commute longer distances as a result.

Bertaud said that when he was advising developing-world cities on planning policies, he’d often start by creating a map of the minimum lot size required by existing rules and estimating what share of the city’s population could afford that amount of land. Here, for example, is his map of floor-to-area ratios in Mumbai, India, which shows that in most parts of the city people are required to buy 1 square metre of land for every 1 square metre of dwelling they want to build. As land is quite expensive in Mumbai, this is basically a policy that requires the poor to get out or build illegally:

Bertaud Mumbai FAR map

It would be fascinating to see a similar map for Auckland if anyone wants to have a go…

If you went to Bertaud’s talk, what did you think of his ideas?

Share this


  1. Planning employment & accommodation as a module is not new and the idea falls apart when the employment segment vanishes. It is quite common overseas and here Tokoroa, Kawerau, Otahuhu etc

  2. It’s strange how all the sprawl advocates jump on Bertaud as agreeing with their hypothesis of urban expansion and mass road building when often he’s saying the opposite.

    Auckland’s geography requires a dense city. It’s just not feasible to afford the transport infrastructure for hypersprawl with so many geographic constraints.

  3. According to the first graph Auckland not far off from NY in population density?!? I assume it’s true but I wonder how the data was sliced. It doesn’t seem to match up with following slides very well.

    Interesting how south Auckland is as densely populated as the central suburbs. But that’s probably because of Polynesian families with 10 people in one house.

    I follow lots of The NZI reports and find their some of their ideas radical for NZ and yet seem to make sense. They are very pro market and not interfering people in how they use their property.

    1. The density measure Bertaud uses is simple average density. I discussed how this measure can be misleading in a recent working paper ( and advocated an alternative measure that more accurately measures the density experienced by a city’s average resident.

      While comparable figures on population-weighted density across different countries are hard to come by, the available data indicates that NYC is in fact several times more dense than Auckland.

      1. Also it is always important to be clear what is meant by ‘New York’ for example. Clearly Manhattan is extremely dense, but large swathes of Long Island and other parts that are included in the administrative legal entity called New York are empty. It is important to know exactly how much of a city is being measured when comparing densities.

  4. That’s why our local council should reduce bureaucracy if we are to rebuild the single story house to a multi-story apartment. At the moment the consent cost is expensive and the process is complicated and the result is uncertain.

  5. Peter I’m interested with Bertaud also mentioned the other reason that road building doesn’t ultimately make for efficient agglomeration effects along with induced demand mentioned in 1. above: The fact that driving dominant pattern is always dispersive; so extending a city to fit ever more people and employment on a driving dominant model will extend the physical distances much more than expansion through intensification and therefore perforce increase inefficiencies?

    The push for all this new housing way out at Warkworth, Pukekohe, and Helensville, is not efficient whatever movement technology is used.

  6. With those spatial patterns I imagine city A would have expensive individual infrastructure projects (e.g. underground systems) but on an overall basis it would be much more cost effective than B or C which needs high levels of provision everywhere

    1. Most cities are now firmly C. Even those that are old enough to have once been only A, like Paris or London, are now surrounded by heaps of auto-age dispersed habitation. So, both those cities with their extensive underground and overground rail systems still run vast quantities of buses in the outer suburbs. Of course both are also now making huge investments in new rail projects to connect the outer bits of the composite more efficiently.

      After all Crossrail in London, like the CRL in Auckland, is a tunnel project in the centre, but really it’s about connecting the outer reaches of the rail system to and through that centre and with each other. Also just like the CRL.

  7. Dharavi, the most disgusting and filthy hell hole on the planet. Amazing that human beings can lower themselves to living in such squaller.

    1. Dharavi is an amazing wealth generator actually. People from extremely poor rural India shift there as a stepping stone to a way more prosperous urban life.

          1. Geoff, sometimes your ignorance just blows me away. “Amazing that human beings can lower themselves to living in such squaller” Did you ever stop to think why these people live in the slums – it’s not like they wouldn’t all rather be living in a villa in Ponsonnby if they had the chance! There are probably some highly intelligent and talented people living in those slums. Far from “lowering themselves” some are in fact striving to increase opportunities for their own advancement and the advancement of their families. How about showing some compassion rather that chastising people for something that you clearly do not understand.

  8. What cities are most effective at keeping their average commute times down, but without ruining the city? Phil Hayward and his fellow bunch of sprawl lovers will point to Indianapolis, Cleveland or various other semi-abandoned rustbelt cities as having nice and short commutes but they hardly seem good models for Auckland.

    1. By the looks of things, Singapore’s been doing pretty well with a combination of demand-responsive road pricing and rapid development of a subway system. They only opened their first subway line in 1987, and now they get 2.8 million riders a _day_. Amazing how much change can happen in a generation!

    2. Just on physical geography alone the landlocked cities of the North American flat and featureless plains are totally irrelevant to Auckland. But add the fact that his only metric for access; driver travel time, works best in places hollowed out of economic activity and people and it’s easy to see just how daft their thinking is.

  9. Most cities may work like pattern ‘C’ for employee hunting, but is it possible that pattern ‘D’ is a real place for school aged children, pattern ‘B’ is where family lives & pattern ‘A’ is how tourist see a city.

    1. I think you’ve hit the nail on the head. Cities are full of people with a range of different travel needs and preferences. That’s why they work like model “C” in practice. Different people need to get to different places at different times and for different reasons. As a result we recommend a “horses for courses” approach that offers a variety of modes for different needs, e.g.:
      * Good walking and cycling facilities to enable people to make local trips (e.g. retail and school trips) efficiently and safely
      * Rapid transit infrastructure to efficiently serve major employment nodes and bridge constraints in the network (as the Northern Busway has done for the Harbour Bridge)
      * Frequent bus networks to serve a range of trip types
      * Well-maintained roads to serve the many different trips that are hard to do efficiently on other modes (e.g. commutes from West Auckland to East Tamaki).

      1. I attended Bertuad’s talk in Christchurch and one important aspect of his talk was he could quantify the importance of transport. Stating “In Korean cities, a 10% increase in the number of jobs accessible per worker corresponds to a 2.4% increase in workers’ productivity. Additionally, for 25 French cities, a 10% increase in average commuting speed, all other things remaining constant, increases the size of the labor market by 15 to 18%. In the US, Melo et al. show that the productivity effect of accessibility, measured by an increase in wages, is correlated to the number of jobs per worker accessible within a 60 minute commuting range. The maximum impact on wages is obtained when the number of jobs accessible within 20 minutes increases; within this travel time, a doubling in the number of jobs results in an increase in real wages of 6.5%. Beyond 20 minutes of travel time, worker productivity still increases, but its rate decays and practically disappears beyond 60 minutes.

        These papers demonstrate that workers’ mobility –their ability to reach a large number of potential jobs in as short a travel time as possible, is a key factor in increasing the productivity of large cities and the welfare of their workers. Large agglomerations of workers do not insure a high productivity in the absence of worker mobility. The time spent commuting should, therefore, be a key indicator in assessing the way large cities are managed.”

  10. Alain Bertaud also saw housing affordability as a type of mobility. That workers should be able to afford housing close to where they work or need to be. He wanted quantifiable targets such as mean house price to household incomes measures. He spoke about elasticity of land supply and that it was not the duty of planners (or nimbies) to dictate density. He was agnostic over which was better ‘up or out’. He also seemed agnostic regarding transport mode.

    It is on the housing affordability front that I see New Zealand and Auckland having the most challenges. When a teacher can’t afford to house their family in a city you have to wonder about the future of that city ?

  11. I don’t understand the argument saying that induced demand is bad. If the mean commute time of a road is 40 minutes and you double the width of the road and induced demand means that commute time remains 40 minutes then it sounds like you have wasted your money. But in fact there are now twice as many people doing that commute so induced demand is GOOD.

    The argument saying that it is bad is like saying “we don’t like to build more roads because the people use them.”

    1. The problem is that roads are justified on the basis that they will lower congestion not on the basis that they will move more people at a similar level of congestion.

  12. There isn’t a lot new in what he says. It’s mostly just classic planning theory. The point about the geography of poor people, though, requires some comment. First of all, poor people end up living far away from employment centres even in the most unregulated cities. It may be an opposite pattern – living in the central city while employment goes to the suburbs – but it is hardly limited to cities with growth boundaries. In fact, growth boundaries can keep employment areas from spreading even farther. As to affordability, if you wait for the real estate market to accommodate low-income housing, you’ll wait a long time. Almost inevitably, low-cost housing requires some kind of market intervention, and it has for every major city I know of, because it’s an area of market failure. There may be good reasons to not have growth boundaries, but these aren’t two of them.

    (I wish he’d get a picture that doesn’t make him look like some nazi executioner.)

  13. Although car use has a range of negative externalities associated with health, air quality, and global warming (which is economist-speak for “stop driving so much!”), my argument here is _not_ that induced traffic is bad because people are driving more. To quote: “In short, people travel more but they don’t travel any faster.” In other words, building more road capacity may move more traffic, but it will not reduce travel times to work – i.e. it will fail to deliver on the key metric that Bertaud recommends.

  14. The “urban village” model does exist in a sense. I am thinking of regions of Germany, Switzerland, Netherlands etc, where discrete small towns have nucleated themselves in close proximity, and the result is a web of interconnected centres which function as a bigger entity. The key to success is of course, good public transport connectivity between these centres, such that it is not perceived as a penalty to live in one and work in another. And not relying exculsively on congested roads to provide those links.

  15. Dave B what you are talking about is a polycentric model not the urban village model. The urban village model is where one village has nothing to do with the other villages -it is not interconnected. It doesn’t exist in modern times it only existed in medieval times….

  16. But if twice the number can access a given area in the same time -that is an improvement. That is more agglomeration. More interactions. A bigger labour market and so on.

  17. If twice the number of travelers can acess a given area in the same time (through whatever transport mode) that is an improvement. That is more agglomeration. More interactions. A bigger labour market and so on.

    I think Bertaud’s tool of looking at how many places you can access in a given time using different transport modes is really useful. Not to choose a particular mode bias but to better design the system.

    1. Yes that’s right. But the observed phenomenon of induced demand says that with roads it will not remain ‘at the same travel time’ for long, as this causes the new m’way to suffer from congestion which delays movement for all. Also expanding the numbers by the motorway/new suburb model inevitably increase the distances between all locations per person, which must, overtime, increase travel times. Eventually the distant new suburbs just add too many new drivers to the previously sufficient motorway to stress it into failure [same occurs for parking too].

      It is simply not economically nor physically possible to keep widening motorways and providing ever more parking to try to keep driving times from ever distant places efficient. Bertaud does refer to this when he says that keeping people and employment closer is another form of travel time improvement. And because the destinations within a city are always changing that can only occur through a more compact model with spatial efficient movement systems of all modes.

    2. Auckland this decade will be a good case study in this. The billions being spent on the urban motorways will result in this kind of failure by success. Waterview; SH20 and SH16 will so incentivise driving that local roads and parking are all about to come under huge pressure. AT is already spending too much money and we are loosing too much public amenity to road widening all around this mega structure. And the m’way system will become hostage to new forms of snafu generated by the opening of various filters that currently ration sudden additions onto the network. 20, 18, + 16 will funnel waves of traffic far too efficiently into the CMJ and this singularity will not cope.

      Am I right about this likely outcome? Because of the failure, or refusal, to do the sensible thing and build the NW busway at the same time as the widening there, there won’t be good alternatives to driving on this route, the rail system will be at capacity without the CRL, so the only chance that volumes remain moderate is a general economic downturn, and a fall in the dollar, sending fuel prices higher etc. So the failure here is not in m’way building in itself, but in only building m’ways. It’s naive and pigheaded, and a tax on everyone.

  18. Patrick Bertaud said in the talk in Christchurch that the problem with cars is their appetite for real estate. Due to Auckland’s geographic real estate constraints this is a real problem. I can understand the importance for a mixt mode transport system for that reason. I would wish that we had a calm non-partisan, non ideological discussion on how best to achieve quantifiable measures like mobility and affordability.

    The problem I see is NZ is heading towards a one party partisan ideological state so this sort of grown-up discussion is not going to happen. The Wellington fly over is the latest salvo in this process.

    1. That’s exactly the conversation Transportblog is promoting. I’m glad you’re actively participating and I wish everyone else was as well!

  19. Very good blog! Do you have any hints for aspiring writers?
    I’m hoping to start my own blog soon but I’m a little lost on everything.

    Would you suggest starting with a free platform like WordPress or go for a
    paid option? There are so many options out there that
    I’m completely overwhelmed .. Any ideas? Many thanks!

Leave a Reply