City engineers have turned our downtowns into places that are easy to get to but not worth arriving at. – Jeff Speck, Walkable City

A few weeks back I prepared a cross section of the cbd that revealed an interesting land value gap that seemed to correspond to the disurban environment caused by the mini-motorways of Hobson and Nelson Street. In an upcoming post I will look more closely at these streets, in particular their potential for redevelopment, but for now I wanted to compile some background material on the history of one-ways.

Land value gap. Auckland CBD.

Understanding one-ways first starts with the design paradigm that required them. One of the few studies on the subject, “Downtown Streets: Are We Strangling Ourselves On One-way Networks?”, by Walker, Kulash and MacHugh, sets the stage:

For many years, traffic engineers were mandated to “move as much traffic as possible, quickly as possible,” often resulting in degradation of movement for other modes of travel. The unequivocal movement of the motor vehicle through a downtown network was of paramount concern; all other modes of travel took a back seat. Effectiveness of the network was measured by the amount of delay a motorist would encounter on a given street segment or intersection during either the morning or afternoon peak hours.

The authors document the typical transition of American downtowns beginning with the ‘pre-Freeway era’, where streets served a multitude of users and modes and most of the city’s cultural, social and civic activities took place in the historic centre. During this time most of the workers lived within a short distance to work as well (think Auckland streetcar suburbs).

Looking south along the west side of Hobson Street from Fanshawe Street. ( ‘Sir George Grey Special Collections, Auckland Libraries, 4-2052’)

The ‘Freeway Proliferation Era’ encouraged city workers to move farther from their work.  “As downtown workers began to seek out less expensive housing in the suburbs, the mode balance on downtown roadways that had existed for years began to move toward facilitating the speedy entrance and exodus of commuters. Downtown streets began to be converted to one-way travel to facilitate this expedient movement into the city in the morning and out in the afternoon.” Below is an example of the technique used to accommodate traffic into and sometimes through a cbd/downtown.

Typical solution: one-ways. Charleston, South Carolina. (source: Meagan Baco)

The Post-Freeway Era reached its peak in the 1980s, when corporate offices chose to locate in suburban edge settings. In addition to the one-ways, increasing traffic also gobbled up surface and ground floor real estate in the form of parking which contributed to “blighted, empty streets and boarded-up storefronts, devoid of life after 6 pm.”

Multiple-lane, high speed, one-ways. Land use responds with the highest value: parking.

Finally, the Reemerging Era is where we are today with an increase interest in both residential and business activities locating in traditional city centres. This urban renaissance has lead to a proliferation of one-way re-conversions beginning in earnest in 1990 according to student researcher Meagan Baco.

As urban observer Alan Ehrenhalt notes, in the Return of the Two way Street, cities across America are coming to the conclusion that one-ways should go as documented in this story from Vancouver, Washington:

In the midst of a severe recession, Main Street in Vancouver seemed to come back to life almost overnight. Within a few weeks, the entire business community was celebrating. “We have twice as many people going by as they did before,” one of the employees at an antique store told a local reporter. The chairman of the Vancouver Downtown Association, Lee Coulthard, sounded more excited than almost anyone else. “It’s like, wow,” he exclaimed, “why did it take us so long to figure this out?” A year later, the success of the project is even more apparent. Twice as many cars drive down Main Street every day, without traffic jams or serious congestion. The merchants are still happy. “One-way streets should not be allowed in prime downtown retail areas,” says Rebecca Ocken, executive director of Vancouver’s Downtown Association. “We’ve proven that.”

In addition to Vancouver, Washington, ‘hundreds’ of North American cities have already made one-way conversions, including: Baton Rouge, Berkeley, Dallas, Green Bay, Portland, Sacremento, San Francisco, San Jose and Miami. The success of these efforts are increasingly being documented, like this story from Canada:

“It was somewhat controversial at first, but I would say now that, without exaggeration, people are 90% in favour,” said Brian McMullan, the city’s ebullient young mayor. A prominent local businessman came up to me the other day and said, ‘I didn’t support it from the start, but this is the best thing you’ve ever done.’

This brief post does not address the traffic and street design dynamics of one-ways that make them so dangerous for pedestrians and repulsive to urban life.  Instead, I thought it would be best to document how common the practice of retrofitting has become and where it sits in a historical context. Taking a different look at street design, mobility and accessibility (Link v Place) questions the entire premise of our modern transportation planning (What’s the point). It is increasingly becoming evident that designing our city, streets and neighbourhoods for people (as opposed to cars) provides significant value and resiliency.

From the first report mentioned above  I recreated an interesting diagram that depicted the premise that there is a “livability” dividend that can be realised if we are prepared to sacrifice a few seconds of time from a vehicular trip.  Turning one-ways back to two, like shared spaces, signal priority for pedestrians, and providing for transit and cycling is increasingly being seen as a “no-brainer” for cities intent on remaining relevant in the 21st century.

New paradigm: modified from Walker, Kulash, MacHugh.
Share this


  1. Ironically the one way streets in Auckland probably have the highest density of residents anywhere in the country, yet as you say their public amenity is abysmal. I live off Shortland Street and basically have never ventured further West than Albert Street except in a bus/car–and when going to Ponsonby by bike or foot I go via the Wynyard Quarter. I doubt I am alone in being walking distance from this part of town yet avoid it, and it’s a shame because in effect half the city is a wasteland. In my view the two-laning of these streets and an immediate diet to provide amenity for all users, be them pedestrians or cyclists needs to be a top priority for the council. In some ways I’d actually see this as a higher priority than the upgrade of Quay Street, which although also an unnecessarily barren and windswept area, has been partially saved by its presence next to the Harbour.

  2. Very good. It strikes me that among the cheapest and most effective urban design moves that Christchurch should do is not reinstate its awful one-way system.

    Other really cheap yet effective ways of helping to make that city a desirable and therefore successful place again are; keeping the city streets narrow, enforcing slow speeds, and protecting a full Transit ROW for future development.

    Undoing the blight of Hobson/Nelson is also an obvious fix for Auckland. It would either require a revolution in thinking at NZTA [because of the need to re-jig the motorway ramps], or a top down change in their brief to force them to acknowledge the huge influence their programmes have on place and a requirement to include those outcomes in evaluations.

  3. To play devils advocate, Nelson St in particular drains traffic out of the central motorway junction. If we are to reduce that draining capacity even a small amount, that could increase the frequency with which traffic backs up along the off ramps and into the CMJ itself.

    So how do we sell such a concept to NZTA?

      1. Not good enough for them I expect, even the same number of lanes would have lesser throughput at intersections due to move phasings, and reduced capacity for turning movements.

    1. There is quite a long queuing capacity for Nelson on the flyovers, no? An alternative would be to convert to 2-way at Cook St, or even reduce the number of lanes increasingly towards to the water’s edge. Currently there are 3 right turning lanes onto Sturdee. If the 2-way network was reintroduced, Hobson would be distributing a lot of this local movement.

      1. Yes quite a lot, but how can you justify a reduction to NZTA? I can imagine they’ll say “tailbacks onto the CMJ are already a problem so we cannot possible reduce queuing capacity whatsoever”.

        So what next?

        1. It will need to be modelled, it isn’t absolutely certain that the current set up is the most efficient for traffic flow anyway… will depend on phasing etc….

          But the point of the post is that traffic flow cannot remain the only value in these decisions… lovely flowing traffic and low value city is a shitty outcome, and arguable what we have now in this part of town. It is a privileging of far away places over potentially high value city place.

        2. One implementation techniqiue, as described in the research paper- is called ‘nibbling’. Start with one segment at a time. Prove the world wont end.

    2. I wonder how much extra lane changing happens as a result of the road being so wide. People who pick a lane on the off ramp because it is shortest only to have to cut across lots of traffic on the race down hill? Also on Hobson St I have in the past found that traffic bound for SH16 often gets held up by cars queuing to get on SH1 due to them getting in the wrong lane then holding things up while trying to force their way in to the correct ones.

      I guess the way to sell such a concept is on the safety and economic benefits.

        1. I talked to an NZTA road engineer recently who said he wouldn’t care if a project added 5 minutes to the journey time of a SOV as for him the most important thing was safety. He also added that people shouldn’t be using SOV to get to work anyway due to how inefficient it was with things like road space.

    3. Maybe I’m being unnecessarily petulant but I think this is a situation where AC and AT should tell NZTA to sod off – it’s our city centre and we’re trying to maximise it’s value to residents and businesses.

      If efforts to reinvigorate the entire western swathe of the city centre incurs a few seconds of delays for vehicles then so be it – that’s likely to be an economic trade-off worth making.

      And from what I understand NZTA only have jurisdiction to the first signalised intersection off the highway, after which the RCA can do as they wish.

      1. Agreed Stu. I think AC/AT need to find a way of quantifying the benefit though to show that it really does outweigh the (relatively minor) delays for traffic.

        1. Four simple steps:
          1. Use a hedonic regression model to estimate the negative impact of traffic on property values (noise ,vibration etc).
          2. Then calculate how a reconfigured street layout would reduce traffic volumes entering the city centre
          3. Then use relationship in step 1 to estimate the property value uplift attributable to the scheme
          4. Then compare the lane value uplift quantified in step 3 to the increased travel-time.

          Bingo, AC/AT can then send a Xmas card to NZTA telling them to sod off.

  4. A quote from above: “A year later, the success of the project is even more apparent. Twice as many cars drive down Main Street every day”.

    Is this a good outcome?

    1. This is a retail consideration. Typically one-ways screen shop visibility in one direction. I’m not sure this matters in Auckland.

  5. A big issue with this will be the Sky City carpark.
    Entrance will have to be totally rebuilt, that may well have to be paid for by the council.
    North of Victoria St should be easiest and pursued promptly. Really need quality pedestrian links from Midtown to the viaduct that avoids the awful Fanshawe St viaduct, and sorting Nelson and Hobson will be a big help.

      1. Interesting to note the poor pedestrian in that pic. What looks like a through footpath seems only to give access to yet more car-parks. Suddenly there is a random gap with no path, hopeless! At least this side is a bus interchange though!

  6. Great post, although I do find the one-way versus two-way debate generally a little … simplistic?

    I remember a presentation from Jan Gehl where he showed an image of one way streets in Auckland and concluded that we had too many. Only problem was that many of the streets highlighted were narrow little lane ways that you could not get two lanes in even if you wanted to, e.g. High Street and Lorne Street.

    In my mind the key thing is the overall street environment. If the mobility dividend (and there is one) associated with one-way streets is reinvested in narrower street widths, shorter cycle times, and longer pedestrian green phases then maybe the street is better off; I’m not sure.

    So my preferred “nibble” would not necessarily be to make both Hobson/Nelson two-way, but by first trying to reduced the width (two lanes max) and overall street environment. Brisbane has a one-way street system downtown and its generally not very nice – but in many places it does result in very quick signal cycles and good pedestrian phases.

    I just fear that we risk throwing the baby out with the bathwater on the one-way issue, when maybe a more incremental change would actually a) cost less and b) get us to a reasonable level of street amenity.

  7. I don’t think you can make this change incrementally. it has to be done all at once in terms of signal phasing. else you will cause plenty of other problems. I don’t think it is a “no brainer”as I am not sure if some commentators understand traffic signal phasing. 2 way = more delay for drivers AND pedestrians. this is a fact of traffic signal phasing that can’t be ignored. you can do all the modelling you want, making all the incorrect assumptions you want but it won’t make 2+2=5. one way is the most efficient IF you are talking about moving vehicles. this is undeniable. it also means lower ped delays in some cases. however, if you can prove it is more efficient by OTHER measures, then you can go down that path. also we can’t discount the safety issues of backing up traffic onto the motorway. NZTA/AT operate their network in collaboration and there can’t be an us and them mentality.

    I would be very wary using land value arguments unless you have a proven model that isn’t based on erroneous assumptions like certain other transport models. such models are highly subjective based on assumptions made.

    I think the CRL would need to be finished(and some more high frequency bus routes) before you can do this in order to give people more realistic options and get people out of their cars. once you do that, then you 2-way and tell people to put up with the congestion or take the train. a good thing is that buses don’t really use these roads so won’t be impacted much.

    I would love to see a altitude graph for that slice in the land value image to see if gradients along there have any relationship with land value.

Leave a Reply