Over the last 50 or 60 years, the United States, Australia, New Zealand, and a number of other countries have pursued a “roads first” approach to transport policy. There have been significant public investments in (generally un-tolled) roads, and relatively few investments in competing transport modes.

It’s hard to justify this approach based on preexisting travel patterns. Take Auckland as an example. According to Paul Mees [Transport for Suburbia, p. 21], in 1954 Auckland’s public transport network “accounted for 58 per cent of trips by motorized modes, private transport only 42 per cent. When walking and cycling, which were not surveyed, are taken into account, it is likely that fewer than a third of daily trips were by car.”

However, from this date onward roads – not public transport, and certainly not walking or cycling – have dominated transport spending. Spending on a new system of motorways and arterial roads was considerably higher than spending on other modes that carried more journeys. In other words, public spending to enable car travel did not respond to existing demand – it was intended to shape future demand. (And in doing so, change the shape of the city.)

Another potential justification for disproportionate spending on roads is that it’s just what people wanted. Cars were invented and then cheaply mass produced, people wanted to use them to travel everywhere, so transport agencies had to build more roads.

There is some truth to this. Cars are very convenient for many journeys. But it can also be convenient and cost-effective not to own a car. PT tends to be cheaper than driving to places where you have to pay for parking, and cycling is often quicker than driving, and more enjoyable if there are enough safe bike lanes.

But this argument also ignores other policy factors that shape transport demands. In particular, planning regulations enacted and progressively tightened throughout the 20th century have tended to:

  • Make it more difficult for people to live near where they work, by zoning different areas exclusively for different uses
  • Discourage people from living at medium or high density by limiting building heights and setting minimum lot sizes for dwellings
  • Subsidise the ownership and use of cars by requiring all new buildings to have a significant amount of on-site parking.

But what, then, should we make of Houston, which lacks a zoning code but nonetheless has ended up with lots of driving, low public transport ridership, and a low-density urban footprint? Is Houston evidence that in the absence of planning regulations that distort people’s location choices, people will choose to live at a distance and drive to get around?

Source: NASA Earth Observatory
Houston from space. This image is 100km across. Source: NASA Earth Observatory

In a word: no.

It turns out that Houston is not actually as unregulated as people make it out to be. While the city lacks a comprehensive zoning code that rigorously separates different uses, several other planning regulations (and similar measures) have distorted its urban form and transport choices. A 2005 article by law professor Michael Lewyn identifies four important ways in which planning has influenced transport outcomes in Houston:

  1. Houston enforces a byzantine and quite restrictive set of minimum parking requirements (MPRs). As I discussed last year, these include a parking requirement for bars that defies all concepts of prudent regulation. These requirements make parking cheap, and walking to the shops hard.
  2. While Houston doesn’t formally limit building height, it does establish a minimum lot size of 5000 sq ft (or around 460m2) throughout most of the city. This discourages PT, walking and cycling by increasing the distance between dwellings and discouraging space-efficient typologies like terraced houses and small apartment buildings.
  3. Houston requires streets to be wide, blocks to be long, and buildings to be set back a considerable distance from arterial roads. All of these policies make it dangerous and unpleasant to walk there.
  4. Lastly, new developments in Houston make extensive use of private covenants that restrict uses and building designs. These agreements often simulate zoning, with the result that Houston has similar levels of racial, income, and housing segregation to (zoned) Dallas. Houston has chosen to imbue private covenants with the force of public authority – the city will pay to enforce them even if the people subject to the covenant would rather not.

As a result of these policies, Houstonians cannot make free choices about where to live, where to work, and how to get around. Their decisions are strongly influenced by a suite of planning regulations that, as in many other cities, conspires against density and against non-car travel. Houston’s heavy use of the car is not a natural outcome, but one that has been engineered by policy.

Seen from this perspective, “roads first” transport policies seem less like an exercise in meeting demands, and more of a component of a large social engineering programme.

The results are not necessarily stellar. While the city is known for low house prices, Todd Litman points out that Houston is relatively unaffordable for its residents, compared with other large US cities, once transport expenditures are factored in:

Litman US city housing and transport costs chart

Furthermore, Houston’s commuters experience more hours of delay in traffic than most other US cities. New York, on the other hand, looks pretty good. Although it is large and congested, many commuters choose to opt out and take the subway instead. In Houston, they lack that choice:

Litman US delay hours per commuter chart

What do you think would happen if we tried to facilitate choice instead?

Share this


  1. Indeed. All transport investment is made on the ‘decide and provide’ basis, as it should, this is the power of it! It forms our world, shapes our lives. The big danger comes when those responsible for this shaping pretend, especially to themselves, that they’re not doing it.

    Like now.

  2. Thankyou Peter, great post. The whole roads only/suburban development investment programme from the 1950s on has been about private developer and infrastructure company profit, ideology and yes…social engineering. Very interesting also to note the planning controls in Houston that encourage (arm-wrestle) development to go out, and not up.

    In New Zealand, we have Christchurch shaping up (out?) to be our very own Houston. Very mitigating circumstances that go down into the fundamentals of the land itself that Christchurch is built on. Nonetheless, it might now be time for the planners and architecture community of Christchurch to have some fresh discussions with a city that is at long last regaining its mojo.

    1. Maybe -but Christchurch does not have the huge rings and radials of motorways that Houston has -or even somewhere like Raleigh (a comparable mid-sized city to Christchurch in the US). So Christchurch still has a choice on what sort of city it will become.

      Although Christchurch has more potential for out development in the 10 to 25 km radius compared to Auckland or Wellington -it is not as sprawl favourable environment as Houston. To the east -development is stopped by the sea. To the south by the Port Hills, To the north by marshy -liquefaction prone land and the Waimak river and to the west by the airport and the no development noise zone.

      The most favourable conditions for sprawl are to the southwest -the Rolleston-Lincoln-Prebbleton arc in the Selwyn District and to the Rangiora-Pegasus -Kaiapoi arc on the other side of the Waimak in the Waimakariri district. Both districts have some of the fastest growth in the country.

      If Christchurch City Council allowed growth on its urban fringe -roughly 10 to 15 km then these further developments would not be happening as fast. Further if CCC periphery developments were part of a fast transit orientated system (so if NZTA came to the party) then it would massively decrease transport and congestion costs.

      1. Agree Tuktuk that Christchurch is getting its mojo back and it is time we had a fresh discussion of where we are headed as a city : )

  3. Another aspect of the more roads agenda in the US and NZ is the influence of Federal and central government. Freeways (as opposed to older US user pays PT -such as Streetcar suburbs) have been a national project in the US since the 1950s for defence reasons initially.

    In NZ roads funded by central government are sold to the public for their ‘national significance’. Central government if it does economic analysis for new transport infrastructure initiatives uses techniques such as -cost benefit analysis -which does not take into account long term costs/benefits from the way that infrastructure changes future residential and commercial development. For the most part agencies like the NZTA have been ignorant (willfully?) of how transport infrastructure shapes urban environments.

    In simple terms -motorways disperse development – they complement ‘out’ development -this has some advantages as it allows greater competition wrt land-owners -thus urban land prices can be lower and more stable. This was important a hundred years ago when cities were over-crowded, disease ridden and expensive.

    Well designed public transport because of its higher carrying capacity corridors concentrate development -they complement ‘up’ development -this has advantages for inner city revivals, which can be important due to cities becoming so big that low density peripheral growth is not feasible. Also due to growth in inner city service jobs and to the demands of time poor but income rich double income households and probably other factors too.

    As Peter rightly points out the public decisions on what transport infrastructure is provided and what type of urban regulations to enforce has a large influence on the built environment. It determines the sort of lives we live. Collectively it is probably the easiest thing we could change to make the biggest improvement in our collective quality of life.

  4. So Auckland’s transport and housing problems are largely the result of policy decisions that have been made over the past 50-60 years and despite the problems these policy decisions have caused many are still proposing that we can solve the problem by continuing with the same policies. The definition of insanity comes to mind…

  5. Essentially; that was then, this is now. Roads first is over. It now has to be Transit and Active next and predominantly. Sixty years of roads only has led to the imbalanced city we now have. The great news is that this additional not substitutive, in other words we will still have all the roads and driving we have now but be able to grow the next phase of the city on the back of the currently underdeveloped complementary Transit and Active modes.

    This is the zeitgeist. We need our institutions to catch up with the times, to become conscious of their path dependencies and propensity to group think.

    1. The great news is that this additional not substitutive…

      Unfortunately we live in Auckland, where an attempt is being made to prove otherwise. At great cost. And hence we are currently falling behind all comparable cities in apartment builds.

      Auckland city allows only limited outward expansion and expects these enforced constrictions to cause substitutive growth in transit rich areas. This is impossible.

      We have moved from an orthodoxy of roads first to an orthodoxy of transit first, without ever pausing to consider the middle ground.

      1. So many things wrong with your claims…

        First off, Auckland is currently experiencing MASSIVE outward growth. You don’t seem to have been in Kumeu, Silverdale or Papakura much recently. What is PLANNED for these areas in the Unitary Plan to add to housing there will eclipse even that sprawl, so your first statement is demonstrably wrong already.

        Then the claim that transit is now the orthodoxy – well, it may be promoted by a lot more people than it used to, but the reality (barring the occasional win like CRL*) is that the SPENDING – again, facts, not claims – still massively prioritises roads and motorways. Again, your statement is demonstrably false, as this blog, and govts own budgeting and project announcements regularly show.

        *And even the CRL investment by govt was back-ended with a roads announcement (East West Link) that was MORE expensive than the rail.

        1. I’ve looked at those areas, but unless I look at them in comparison to other cities it is just admiring the view. We are expanding slower than almost every comparable city in Australia or Canada – places that have been subject to the same investment and house price rises as ourselves. Auckland adds standalone houses at less than 1/2 the typical rate of contemporary cities and apartments at about 1/4 the rate.

          So “MASSIVE” compared to where?

          Yes, large areas are “PLANNED” for the future, but not now. Auckland has a housing deficit that is expected to grow and grow until it reaches a maximum in 2018. Then the deficit is planned to continue to be a deficit until about 2027. This is an unacceptably slow expansion, because rent rises until 2027 are unacceptable.

          Australia has a housing deficit that is going to transit into a surplus in 2018, simply because they are building outwards twice as fast and upwards four times as fast.

      2. “We have moved from an orthodoxy of roads first to an orthodoxy of transit first…”

        With due respect to your other points, this just isn’t correct. If you look at transport policy in NZ or Auckland, it’s impossible to find evidence of a “transit first” orthodoxy – most of the money is being spent on roads.

        Similarly, if you look at planning policy in detail, it’s a mish-mash. Some parts of planning policies in large NZ cities attempt to enable or encourage transit-oriented development – e.g. town centres and limited areas of medium-high density residential zones. But they are counterbalanced by other rules that discourage and prevent transit-oriented development – MPRs, density controls, etc.

        As for the MUL, I’d observe that it’s there in large part to manage the costs associated with extending *road* infrastructure to suburban areas and widening roads in built-up areas to cope with the additional traffic. It undoubtedly has unintended negative consequences. But I would argue that it is best seen not as a means to corral people into TODs, but to manage around the intrinsic unaffordability of roads-only urban transport policies. If Auckland had been built around rapid transit rather than motorways, outward growth would be cheaper.

        1. I won’t dispute you on transport spending, rather offer evidence based on land policy. But even there I must admit my claim of transit first* orthodoxy is likely unjustifiable hyperbole.

          Drury, Runciman, Paerata, Pukekohe & Swanson are available for transit centric expansion yet are prevented from doing so. Runciman and Swanson are planned to be outside the MUL indefinitely; the others are to be developed after our current investment boom has finished. Our land policy is to expand slowly into distant car-centric exurbs in Silverdale, Hobsonville and Pokeno – hardly transit friendly. And the MUL’s “unintended negative” consequence is a land cost rise across the entire city orders of magnitude more per dwelling than the road cost it alleviates. (I reckon for the MUL to be used to facilitate city densification, we need to mitigate the cost of the MUL. Currently I estimate the required subsidy to be $30,000 to $50,000 per apartment.) The MUL is, like the “MPRs, density controls, etc.”, an unmitigated impediment to new development. So to adjust my hyperbole – we have moved to a nimbyism first orthodoxy.

          * In my defence certain lobby groups that associate strongly with the merits of transit orientated growth are either ambivalent to or wholly supportive of the MUL.

          1. huh, Swanson is currently having hundreds of sections created right next to the station as part of the Perihana development. Houses should be starting to go up soon. From memory further development of the area to the south of the station and even some of those already being created are affected by rules surrounding the protection of the Waitakere Ranges and foothills. Some of that is likely nimbyism but not all of it. Further given the state of the likes of Christian Rd, they would need millions spent on them to enable any kind of serious development, they barely pass muster as rural roads. On the other side of the tracks most of the greenfield land is a golf course

          2. Matt L,

            Yep, the Perihana development is a great example of what could be achieved – density housing built next to a railway station. Unfortunately the nimbys have prevented anything being built on that gently sloping land across the street from Perihana. The MUL is fixed at the very station edge.

            The golf course was until the recent 2015 Unitary Plan modifications afforded the possibility of becoming housing, but that was changed. And all the large sections inside the MUL were changed in late 2015 from single dwelling (which allows the possibility of infill and densification) to large lot which blocks everything.

            The nimbys have won, the local councillor is Penny Hulse.

            An enlarged car park has been added to spare all the local nimbys, ultra low density car centric lifestyle blockers, the indignity of parking the 4x4s on the street.

  6. I think given the geography of our big cities and the demand for more inner city growth (0-10km radius) i.e Auckland Isthmus and Wellington City District (not Hutt valley or Kapiti coast) then we should not only be looking at the impediments to going up -setback rules, minimum parking requirements and so on. But also look at changes we could make that would facilitate/encourage going up. More infrastructure for Transit and Active modes of transport would help.

    I think we also need to look at the contiguity problem for going up. Alan Evans writes several chapter about this in his book Economics, Real Estate & the Supply of Land. This is the problem that our urban land has been traditionally been broken up into 1/4 acre and smaller parcels of land. Intensification of those size parcels gives poor outcomes -ugly infill housing. Acquiring larger blocks is very time consuming and expensive for developers for better housing options -apartments -row houses etc. Compulsory acquisition -something the Productivity Commission has suggested has perhaps as many fish-hooks as the problems it is trying to solve.

    I think there is a possibility for contiguous neighbourhoods to voluntarily alter their planning rules to give themselves better development opportunities than infill housing without needing to be expensively bought out by developers or bulldozed by a heavy handed ‘compulsory acquiring’ government. I think given the opportunity by planners and councils some neighbourhoods would voluntarily convert from traditional stand alone housing to Euro-blocs and maybe other high density forms as dictated by local opportunities taking into account, geography, economics, local preference, transport infrastructure etc http://greaterakl.wpengine.com/2015/09/09/guest-post-1-billion-fletchercrown-housing-development%E2%80%8A-christchurch-cbd-part-5/ and as discussed in bottom-up development here http://greaterakl.wpengine.com/2015/10/09/where-does-top-down-meet-bottom-up/

  7. I have changed my mind about the futility of cycleways like the Upper Harbour Drive idiocy that caused a traffic jam for the sake of a few cyclists. Now I understand the point of them. We close off a lane and put in a cyclelane to create a long queue of cars pumping out exhaust. Then we can use those delayed cars as justification for road widening with a huge benefit cost ratio. As the title says Roads First is a mistake. We need cyclelanes to create congestion then build the roads second. It is the start of a whole new era of road building!

Leave a Reply