The Canal Zone – Transforming the Precinct between Green Square and Mascot

There is unquestionably a continuing housing shortage in Sydney, leading to some of the most expensive real estate in Australia and the world.

It is a law of modern cities that no urban water feature should stay underdeveloped. The Alexandria Canal was built around 1900, widening Shea’s Creek, a tributary of the Cook’s River, to serve industrial activities like tanneries that lined its shore. Proposals have been floated to transform it to into a Little Venice, but these were spiked for environmental reasons, disturbing the canal would let loose toxic sediments which would contaminate other waterways. So instead it was left fallow, to wallow in its chemical filth. While that may (or may not) have been a wise decision under one set of economic cost calculations, considering the current costs of remediation, the environmental costs of doing nothing and the value of the redevelopment, conditions have changed and the decision should be revisited.

Alexandria Canal, NSW looking north. Source wikipedia.
Alexandria Canal, NSW looking north. Source wikipedia.

The parallel Airport Rail Line opened a century later in 2000, just prior to the Sydney Olympics. It was financed under a poorly structured public-private partnership which initially levied exorbitant station access charges, that the NSW government eventually bought down for the non-airport stations. Nevertheless it has significant excess capacity which could be taken advantage of.  If we assume it eventually becomes a normal line, we can think about how it might be used to promote the kind of growth that is beneficial to the public.

Walking Route from Green Square to Mascot. Approximate half way point in orange oval. Alexandria Creek to the southwest. Map by Google.
Walking Route from Green Square to Mascot. Approximate half way point in orange oval. Alexandria Creek to the southwest. Map by Google.


I walked the corridor this weekend. This Airport Rail Line cries out in pain for an additional stop halfway between Mascot and Green Square. This lands just south of Huntley Street and Bourke Road, at the northern edge of the Alexandria Canal, home today to the redeveloped Mill at Bourke Road.

The distance between the rail stations at Green Square and Mascot is 2.6 km, 32 minutes walking, 3 minutes by train. Typical station spacing this close to the CBD is much shorter. Redfern to MacDonaldtown, e.g. is only 1.7 km. Given 800m is a useful threshold for walking distance to rail stations, 1.6 km (1 mile) is a natural spacing. 1.3km might be a bit close between stations, but it is still longer than the CBD stations (Town Hall to Central is 1.1 km). I don’t know the cost of designing and constructing a new station on this existing line, it is undoubtedly more expensive than it should be, but there is experience with infill, and it is less expensive than a new line on a per passenger basis. I would think the real estate development could cover it.

The historic Mill plus a train station  would make a great community centre for a new precinct, perhaps the Mill District, or the Canal Zone, which would feature more intense residential, office, and commercial development  complementing, and eventually replacing, single story warehouses, auto dealerships, big box retail stores, and light industrial between Mascot and Green Square, lining the area from the canal to O’Riordan Street.

Development in Mascot. Photo by author.
Development in Mascot. Photo by author.

The region is focusing on new Metro Lines while forgetting opportunities that lie immediately at hand, incremental investments in the Trains network which likely reveal benefits well in excess of costs. Mascot and Green Square will soon be built out, and new land will need to be engaged.


I have now been in Australia a year, it’s time for the breakdown: at what does Australia beat the US, at what does it need improvement. Topics listed alphabetially


  • Banking – AUS … There are fewer banks in Australia, but they work better in a number of ways. Electronic payments are standard and quick. Debit cards are far more standard than credit cards … and there is less credit card % rebate gaming here. What’s a cheque? There is still enough upset here about bank behaviour that a Royal Commission will investigate. Superannuation (the fancy name for retirement plans) have issues, but I don’t think it’s worse than US pension funding problems.
  • Broadband – USA  … in Minneapolis, eventually we had 2 competing Broadband providers providing 40MBPS for $40/month, in AUS at home, I might as well be on a dialup 300 Baud modem. Netflix can’t even stream consistently at 5:00 am.
  • Bureaucracy (Tax) – AUS … Filing taxes is much easier in Australia. In fact, you don’t have to, since they already took your money, you only file if you want some of it back.
  • Bureaucracy (Motor Vehicles) – AUS … The time to get a driver’s license in Australia was minimal and the experience was excellent. The staff had uniforms, just like airlines.
  • Citizenship – USA  … I was born there, I am a citizen dammit, it is my birthright, who the hell cares about where your grandparents were born. Who can possibly actually know, as opposed to repeat stories we were told. Yet Australia is in knots over it’s ridiculous Parliamentary citizenship crisis, perhaps the world’s stupidest scandal. Given how the court has ruled, North Korea could disband Australia’s Parliament simply by giving citizenship to all current members.
  • Democracy – AUS  … despite the Citizenship crisis, democracy functions better in AUS. Parliament works, no government shutdowns (lack of supply), voting is mandatory, convenient in general (Saturday), and no efforts are made to suppress voting.
  • Food (Groceries) – AUS … There are some things you can’t get in AUS (essentially bagels), and the sushi is highly geared toward salmon, but the quality of the Turkish bread and the baguettes make up for it. The food system is less industrialized than the US, so the quality tends to be a bit higher, but certainly not universally.
  • Food (Restaurants) – USA … The food is price competitive with the US (especially considering no tipping in restaurants and the sales tax is embedded in the price). However the good middle tier (America’s sit-down chain restaurants) seems to be missing. Food delivery is much more common here, though our experience with it is mixed at best: it’s delivered, but is it still food by that point?
  • Government Transparency – USA … The agencies in Australia are trying to be Open and Transparent, but paranoia is a big destroya. In NSW they are afraid (or prohibited) from releasing even basic information like traffic counts on monopoly toll roads, and the “Business Case” is confidential in cabinet. This should all be public.
  • Health Insurance – AUS … Though I pay for private here, since I don’t qualify for Medicare (the national health insurance scheme), it is tax deducted, not outrageously priced, and trivial to get reimbursed for expenses through an app. The Doctors are less well trained on average, until recently many only had Bachelor Degrees.
  • Housing – USA … House prices are absurd in Australia, even accounting for the superior weather and higher demand. Houses here don’t have insulation, have poor AC, and leaky windows (and roofs). They are building lots of housing in Sydney, so the trend may abate some.
  • Measurement – AUS … The metric system remains superior to the imperial system of measures for everything, except arguably temperature.
  • Newspapers – USA … These are dying here in Australia as well as in the US, though cities here still have two competing newspapers, in addition to the national papers. The depth of reporting has been generally gutted even more in AUS than the United States.
  • Post Office and Last Mile Delivery – USA … While we get stuff from AusPost, it’s just not as good as the USPS/UPS/FedEx combo in the US, and they make us go to the post office to collect things that could have easily been delivered.
  • Radio – USA … NPR’s Morning Edition beats the ABC’s Radio National Breakfast. Still, it is charming that they have national traffic reports so I can find out about traffic smashes in Adelaide back in Sydney.
  • Retail and Shopping – USA …  Amazon is just not a real thing here. The shopping centres are nicer in some ways, with better food choices, but the retail selection is a bit smaller. The supermarkets are smaller as well. However the grocery delivery options seem greater.
  • Spectator Sports – USA  … Australia has more sports leagues per capita than anywhere in the world, but how many different football codes are we supposed to care about. NFL (American Football) while unnecessarily violent and slow, is still more interesting to watch due to the forward pass than AFL or Rugby League or Rugby Union. Baseball still beats the unfathomable cricket.
  • Television News – AUS … The Australian morning breakfast shows on commercial TV are even more of a giant commercial than the US shows. However the ABC is better than the US networks, including PBS, for news.
  • Transport (Highways – Intercity) – USA … Australia doesn’t have a complete intercity freeway system, it’s still working on it, the US finished the Interstate System essentially in 1982.
  • Transport (Public Transit) – AUS … The buses and trains work much better in Sydney than most all US cities, despite the complaining and despite the many, many imperfections. There is also a better regional train service here than most of the US. Not that it’s good by European standards or anything, but it runs.
  • Transport (Walking) – AUS … It’s not great here, but it is more walkable. The noise level of Sydney is surprisingly high, I think due to the effectively unregulated motorcycles.
  • Weather – AUS … Almost every day of the year, on a day-to-day comparison with Minnesota, I would prefer to be in Australia.


So by my count: AUS 11, USA 11. This is an incomplete list and imperfect weighting, so subject to change.

There are some things I won’t comment about publicly at this time due to conflict of interest, like public schools, universities, and the visa system.

Moving the capital of New South Wales to the west

The capital of New South Wales is currently in Sydney, eastern Sydney, historic Sydney, tourist Sydney, or to speak the language the planners understand, the Harbour City. Parliament meets in a gorgeous building adjacent to the Domain, a large urban park. Government offices are scattered throughout the city and the metro area.

New South Wales Parliament Building
New South Wales Parliament Building

Policy in Sydney has recently engaged around the idea of a 30-minute city, the idea that people can get where they need to go on a daily basis (work, shop, school) in 30 minutes or less by walking, biking, or public transport. (Or that 70% of the people do so, depending on which definition.) This can be achieved through a combination of transport and land use strategies. On the transport side is the question of how fast and how direct the transport network is. On the land use side is the question of where desired activities are located relative to each other. The government of New South Wales is promoting the development of jobs in Western Sydney (and housing in Eastern Sydney) to reduce commuting times and encourage the 30-minute city. This is a noble goal, and the market may move in that direction.

At one extreme we can imagine a completely functionally separated city, where all the homes are on one side of town, and all the jobs are on the other side of town. If the sides are more than 30 minutes apart, there is little that can be done to achieve the goal, though perhaps the connection between the two parts can be made faster or more direct. But since transport networks act to spread out cities physically, it might only induce more suburban development. This functionally separated city is equivalent to the classic monocentric city, with a single dominant downtown surrounded by residential suburbs.

At the other extreme we can imagine a completely functionally integrated city, probably relatively dispersed, where jobs and housing are completely integrated, so there are as many jobs in any suburb as there are workers. There is no guarantee that a worker will be able to find a job next door (or choose it), but the likelihood of finding a job nearby is higher than in the monocentric city

If everything else were equal, from a transport perspective, we would probably prefer an integrated city, as this would place the least strain on the transport network. Moving towards jobs/housing balance is a long held goal, if only weakly operationalized.

But all else is not equal. Employers have an affinity for each other. All the big banks want to be near each other, as do other big companies in various sectors. As does the government. This is what economists call economies of agglomeration.

The government is not just an employer, it is also a major player in real estate markets. It can catalyze development of western Sydney, its Aerotropolis/Parkland City, as it is called in the 2056 Three Cities plans, by moving itself there first.


Cities change with the pre-dominant transport technology. When the capital was established in Sydney in 1788, the dominant technology was animal and human powered, with wind and sails moving ships. Since then, much has changed, and the center of population has migrated inland.

The shape and form of the pedestrian city differs from the rail (trams and trains) city, and  differs from the automobile city. Retrofitting trams into the pedestrian city, and especially automobiles into the pedestrian and rail cities broke much earlier urban functionality, while creating new problems, new opportunities, and new designs. Technology played and plays out differently on greenfields, which could be designed to serve a new transport paradigm.

As we approach the transition from the traditional automobile to the autonomous electric and shared vehicle, with all of the ancillary changes, the opportunity for a new city of the future emerges. This technology will invade existing places, which will need to adapt, and new places which can more fully adopt the new technology. But we also need to keep an eye out for the next transition, whatever that may be (flying cars?), so that what we build now is not soon obsolete.

Transport is not the only shaper of cities, other technologies are also critical, from piped water and sewer, electricity, telephony, elevators, and air conditioning historically, to wireless high-speed internet most obviously today, and robotics coming up shortly.

The new capital will need to orient itself around these new technologies, as well as new extensions of well known technologies, like trains and Metros and light rails and bicycles and pedestrians. This is a huge opportunity, and while I won’t suggest a specific design, I will say it should be forward looking as well as reflective of the changes that have come before. Canberra was an opportunity, but by spreading itself out so much, it foreclose the possibility to effectively use slower modes.

If Daniel Burnham were designing the new capital for Sydney, it might look like this.
If Daniel Burnham were designing the new capital for Sydney, it might look like this.


A government campus for key departmental headquarters and Parliament at the end of the Mall, a now traditional design for capitals, with the vast majority of government offices scattered throughout the rest of New South Wales, could spark development. Access to the new airport and rail lines will provide connectivity to the rest of the state.

Ancillary businesses, not just those serving lunch to government workers, but those dealing with government on a daily basis, will migrate to deal with their public sector clients and customers. There are many sites on the axis between Parramatta and the Blue Mountains that could serve this purpose.

Sydney’s soon-to-be-abandoned historic Parliament House can have a variety of uses, from appropriately sized conventions to space for a museum. Other government offices in Sydney can be sold off, retrofitted for urban housing, or replaced as warranted. The Sydney CBD is thriving, and will continue to without a few thousand additional government workers. But that could be all the difference in success for a new city for Western Sydney.

In 1908, Australia, then with a population of 4.1 million, decided to relocate to Canberra. Today (2018) New South Wales has a population 7.8 million. As Australia has proven, the political capital need not be the largest city.  In the US, most state capitals are not the largest city: St. Paul not Minneapolis, Sacramento not Los Angeles, Albany not New York, Harrisburg not Philadelphia, Springfield not Chicago, Annapolis not Baltimore, and so on to name but a few.


Albany, New York, another planned state capital district
Albany, New York, another planned state capital district. Source: Flickr

It is time to plan and create a new government precinct, out west, to help spark the development the government seeks. It will bring the government to the people, de-center the government from its locational bubble, and juvenate new places with new ideas.


Sydney University should get a Sydney Metro station

The Sydney Morning Herald reports on the latest update of the planning for the Sydney Metro West line, which will go from the Sydney CBD to Westmead, with stops in the Bays District and Parramatta and other locations (such as Pyrmont, Kings Bay, North Burwood or Five Dock, and Camellia or Rydalmere). None of those locations will be the University of Sydney main campus, despite the fact the University recently announced a new campus at Westmead, on the Metro line, which could eventually match the significance and size of the main campus in Camperdown and Darlington.

The University of Sydney is a large and growing campus, the largest in Australia by some counts, serving about 50,000 students and expected to expand by half again as many over the next 20 years. As a point of comparison, the Parramatta CBD, Sydney’s second CBD, has only 47,000 jobs, and so perhaps fewer daily commuters than the university campus.

Construction is already in process to serve up to 10,000 more students at the University. That expansion will reduce parking, and make it more difficult to drive to campus. Getting those students, and the staff who serve them, into and out of campus safely and efficiently is critical, and will get increasingly challenging as the transport capacity serving campus remains essentially fixed.

The University of Sydney is served directly by numerous buses, and indirectly by several train stations. The most notable of these is Redfern Station, about 8 minutes walk south of the Darlington Campus, but a nearly 25 minute walk from the north end of campus. Central Station is 22 minutes from the north end of Campus and 19 minutes from the Engineering Precinct. Macdonaldtown is 19 minutes from the Engineering Precinct and 22 minutes to the north end of campus. These are not inconsiderable access costs, experienced each way each day by train users, in a region aiming for a 30-minute city.

Sydney Metro West map (2018-03)
Sydney Metro West map (2018-03)

A new heavy rail line going west, but somehow missing the University of Sydney, the biggest market between the CBD and Parramatta, which is both growing on the main campus and due to interactions with the upcoming western campus at Westmead, is an opportunity that cannot be recovered from. Metro construction is about a once per century investment, and getting it right is essential, failure is irredeemable.

The argument against is that the next target station, the Bays District is farther north is true, so a line that picked up both the University and the Bays District would be circuitous, and thus slower for everyone on-board. There is a classic tradeoff between station number and ridership, lowered access times for higher in-vehicle times. So perhaps the same line shouldn’t pick up both stations, at least not in sequence. The Sydney West line could instead serve Central and the stations west of the Bays District, and Pyrmont and the Bays District dealt with separately.

One alternative would be to run a separate line to the Bays District, Pyrmont, Barangaroo, and downtown, and on to Zetland. While the Bays district (White Bay Power Station) has development potential, it remains speculative. Google for instance has passed on the site for its Australian headquarters. This map shows an earlier version of this alternative that should still be on the table.


Planned route of the 2008 West Metro, which may be indicative of the future Sydney Metro West. Click to enlarge. (Source: Railway Gazette“//””) Via:
Planned route of the 2008 West Metro. (Source: Railway Gazette) Via:

From the transit agency perspective, the Bays District (just east of Rozelle on the second map) offers value capture possibilities. No doubt under the right negotiated framework that is true, and it would not just be a giveaway to crony developers. However, I suspect the University of Sydney is also an entity that would contribute towards the construction of a station serving its main campus and which could provide a high speed service between the main campus and its new campus in Westmead.

The good news is such a line is unlikely to be locked in before an election intervenes, so there is at least an opportunity to revise it over the next few years before serious money is committed and tunneling begins. Comments are open until May 18.

Disclosure: The author works for the University of Sydney and could benefit if such a station were opened. However given the timeframes, and where the author actually works on campus, such a benefit is small, and probably less than the amount of time spent researching and writing this blog post. This is not necessarily the view of the University of Sydney, Transport for NSW, or anyone else.

Sydney’s New Metro: A bonanza for Sydney residential and commercial property buyers or not?

I was interviewed by Clarissa Sebag-Montefiore for an article on Curtis Associates website: The Sydney Metro. A bonanza for Sydney residential and commercial property buyers or not? My quotes below:

Professor Levinson adds: “One reason, I am told, that the government chose the Metro over trains for the new lines is that because it is a different technology, it will be easier to manage separately from Sydney trains (and will be privately operated under contract). It was a technology choice to achieve a policy aim of breaking the existing bureaucracy and labour unions.”

“To help ensure the political separation, the tunnels for the Metro line are just a bit too short, and the tracks too steep, for double-decker trains to use. This could have been remedied at little or no cost, providing future technological flexibility, but the government want to reduce the flexibility of future governments,” he continues.

“Flexibility would have been a potential benefit from providing compatibility. On the other hand, separation has some value from a reliability perspective, problems on the train lines should not cause problems on the Metro (except for crowding where people have a choice between the two), and vice versa.”

As Professor Levinson, professor of transport at the University of Sydney points out: the “game is the same. Just more territory is brought into the game.”

“Since the lines and stations for the Metro North West and City/South-West are already set, the landowners have already realised their price appreciation,” he says. “There is still a small fortune to be made on the Metro West line, since where the stations land is still not set.”


Full Interview below:

What are the problems with the current [Public Transport] system?

First, the rail system in Sydney, though like all systems engendering complaints, is actually really good, reflecting on the genius of Bradfield’s original rail plan. The evidence for this is the high public transport mode share in Sydney compared to other Australian (and similarly sized US) cities. That said, it is far from perfect. The signal system on the rail lines should be modernized to increase safety and throughput. The bus network needs to be completely rethought and streamlined, so the routes don’t go hither and yon. More on-street right-of-way should be designated exclusively for buses so they move faster and are less likely to be stuck in traffic. Tap-on, tap-off should be off the bus (at the bus stop) so that the buses can board and alight more quickly. The system as a whole needs more capacity in places. The rail stations should be modernized with more exits so the access time to and from stations is shorter, and so there are lifts for each platform.

Why has it taken this long to find a solution?

There have been fits and starts on expanding public transport in Sydney for decades (nearing a century on Bradfield’s plan, which is still not built out). The best theory I heard is the city exhausted itself with all the over-building for the 2000 Olympics, and then couldn’t get anything going until recently. These lines are on the maps and have been for decades, so it’s a question of money and willpower. The recent government has been far more keep to use the private sector for financing, and using asset recycling than previous governments. The advantage to private financing and control is that the infrastructure is sort of “off-the-books”, so since it is privately funded, users pay directly, as opposed to being intermediated through the political layer. This makes it easier to charge users more. This is especially the case for toll road construction as opposed to untolled motorways, but could be applied to public transport as well.

What has been confirmed / what are we still waiting to find out?

Sydney Metro Northwest is well under construction.

Sydney Metro City and Southwest, replacing the T3 Bankstown Line, is in engineering and early construction. What happens to service at the end of the existing T3 line (beyond Bankstown, e.g. Yagoona and Birrong) that is not served by Metro is unclear.

Sydney Metro West is in early planning stages, and the line has not been set, but some of the stations are locked in.

What are the challenges facing the new metro? Critics have pointed out that it is flawed – why?

The issue of Metros vs. Double-Decker Trains is a technology choice which has some practical tradeoffs. Metros have fewer seats (1 deck) and more doors, but a greater capacity, because they can run more frequently (because they can board and alight much faster, and because they are automated with modern signal controls). So for a short trip, fewer seats might be fine, since you save time. For a long trip, not being able to sit in rush hour, but standing for say 40 minutes, might be a bit tiresome for some people (noting that this is common in transit systems around the world). So one of the questions is whether this (Metro) technology is right for this (very long) corridor. I suspect there will be grumbling from people who want to sit and can’t.

One reason, I am told, that the government chose Metro over Trains for the new lines is that because it is a different technology, it will be easier to manage separately from Sydney trains (and will be privately operated under contract). It was a technology choice to achieve a policy aim of breaking the existing bureaucracy and labour unions.

To help ensure the political separation, the tunnels for the Metro line are just a bit too short, and the tracks too steep, for Double-Decker trains to use. This could have been remedied at little or no cost, providing future technological flexibility, but the government want to reduce the flexibility of future governments. See: link

Flexibility would have been a potential benefit from providing compatibility. On the other hand, separation has some value from a reliability perspective, problems on the train lines should not cause problems on the Metro (except for crowding where people have a choice between the two), and vice versa.

Sydney is the first Australian city to build a metro system. What does this mean for Sydney as a city? And for its inhabitants?

Not much. Note: Melbourne is also building a Metro. (Metro Tunnel). The distinctions between trains and Metro will not be terribly significant for most  users, they will just see it as new and old trains.  Now in the corridor that gets new service (especially the Northwest) this is new service (replacing buses rather than trains), so should increase people’s willingness to take transit for certain city-oriented trips. It will also encourage development in the corridor around stations.

What does this mean for property owners / property developers? Which suburbs will suffer and which will benefit? And how is current uncertainty over the lines affecting property / developers / homeowners? How does “value capture” come into play?

Value capture is way to help finance transport infrastructure by using property value appreciation to offset construction costs. This is not done systematically in Sydney, but should be. There are various techniques. (See my value capture study from about 9 years ago, still true)

Is the metro a game changer? i.e will it give massive buying opportunities for previously low density suburbs? Will only the long term gamblers benefit?

The game is the same. Just more territory is brought into the game. Since the lines and stations for the Metro NW and City/SW are already set, the land owners have already realised their price appreciation. There is still a small fortune to be made on the Metro West line, since where the stations land  is still not set.

One cabinet paper stated the Metro West project would trigger a high-rise boom, from the sale of development rights around a dozen new underground stations – what does that mean for standard of living?

It means people who want to live in new high-rises near Metros (and the evidence is there are many such people) will have more opportunities and pay lower prices, and people who don’t will be largely indifferent. The neighbours of those new high rises will suffer more traffic, but have better restaurants and shopping.

Will Metro West deliver on the promise of housing, jobs, and business opportunities?

Jobs will come from construction, but Sydney is pretty close to full employment now, so if that remains so, it will attract workers for these jobs who otherwise would be working on something else, driving up the costs of construction and increasing inflation.

Sydney will not become significantly larger due to MetroWest (i.e. it won’t cause more babies to be born or increase national immigration, it might keep a few people in Sydney who otherwise would have gone to Perth or Hobart or wherever). However development with Sydney will likely concentrate around new stations to take advantage of the convenient accessibility the system provides, so station areas will attract housing and jobs that otherwise would have been more dispersed.

How does the new metro reflect on the building of new roads, such as WestConnex, largely funded by the tax payer? Why wasn’t the Metro West rail project wasn’t considered as an alternative to WestConnex.

I don’t know. The answer to these questions is usually “Follow the money.”

Up or Out: Travel Demand and Thirty Minute Cities

Adapted from Levinson, D. and Krizek, K. (2017) The End of Traffic and the Future of Access. Network Design Lab. Cross-posted on the ITLS Thinking Outside the Box  blog.

Each technological advance in mobility over the past 200 years increased the size of metropolitan areas. The ability to go faster, either owing to new technologies or more completely deployed and deeply connected networks, allowed people to reach more things in less time. The Underground drove the expansion of London, streetcars did the same for many American cities,[a] while trams and trains made Sydney, Melbourne, and Brisbane among others, and highways have exploded the size of cities everywhere. Historically, the time saved from mobility gains was reflected mostly in additional distance between home and the workplace, maintaining a stable commuting (home to work) time.

Will autonomous vehicles follow the path well worn by earlier technologies?

Fast, driverless cars that allow their passenger to do things other than steer and brake and find parking impose fewer requirements on the traveler than actively driving the same distance. Decreases in the cost of traveling (i.e., the availability of  safe in-vehicle multitasking) makes travel easier. Faster roads arise because of capacity gains from vehicle automation (due both to closer following distances and narrower lanes, even more practical with narrower vehicles fit to serve the single passenger they usually carry). Easier travel means increases in accessibility and subsequently increases in the spread of development and a greater separation between home and work, (pejoratively, `sprawl’), just as commuter trains today enable exurban living or living in a different city.[b] Autonomous automobility reinforces the disconnected, dendritic suburban street grid and makes transit service that much more difficult (as if low density suburbs weren’t hard enough). People will live farther ‘Out.’

However, concomitant with automation is the emergence of the sharing economy, with at least some people transitioning from today, where the typical Australian owns their own car, to mobility-as-a-service (MaaS) — automated taxis. This is more likely in larger, central cities where taxis are common, auto ownership is already difficult, and parking scarce and expensive. In this world, while the total cost of travel drops as vehicle ownership costs disappear, the cost per trip might rise, as the cost of ownership is allocated to each trip. This reduces travel demand.

Driverless cars which can be summoned on-demand allow people to avoid vehicle ownership altogether. This reduces vehicle travel, as people will pay more to rent by the minute than exploit the sunk costs of vehicular ownership. By saving total expenditures on transport, more funds are available to pay for rent in cities, and more trips are by walk, bike, and transit. People who seek the set of urban amenities (entertainment, restaurants, a larger dating pool) will find these amenities increasing in response to the population. The greater value in cities with the new more convenient technology leads to more and taller development. (Hence the use of the word ‘Up.’)

At first blush, ‘Up’ and ‘Out’ appear to be contrasting scenarios; they are not exclusive, however. More people living in the outer suburbs or exurbs does not mean fewer people live in cities, because the overall size increases (with more people overall). Sydney for instance, is expected to grow from just over 5 million to about 8 million people over the next four decades.

Similarly, as the cost of travel decreases, people will be more willing to live in locations far from where they work. At safe speeds of 160 km/h on freeway lanes exclusively dedicated to automated vehicles, the commuting range expands widely. From Sydney in this new world, Newcastle can be reached an hour on road, and Kiama and Katoomba are even nearer.

Sydney planners have recently proposed the benchmark of the “30-minute city“, the idea that most people can find work, school, or daily shopping within 30 minutes of their homes by walk, bike, or transit. The threshold of 30-minutes is roughly equal to today’s one-way commute in Sydney (actually 35 minutes), shorter by car (26 minutes), longer by train (62 minutes) according to BITRE. The long times by train are because trains are designed to serve longer distance trips, and focus on the Central Business District.

The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape. By David M. Levinson and Kevin J. Krizek.
The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape. By David M. Levinson and Kevin J. Krizek.

The 30-minute city can be achieved through a combination of transport and land use strategies. On the transport side is the question of how fast and how direct the transport network is. On the land use side is the question of where desired activities are located relative to each other.

If the 30-minute city is defined for walk, bike, and transit as the relevant modes, with mobility-as-a-service easily available on-demand, the Up Scenario works best, though getting one-way commuting times for train users down from 60 to 30 minutes is a large ask. In contrast, the Out Scenario can continue to enable a 30-minute city for privately owned autonomous vehicles so long as jobs don’t centralize further in downtowns.

The interplay of AVs and road pricing is especially important. While autonomous vehicles may eventually double or quadruple road capacity, total demand will rise as well due to population growth, so long as people continue to work, shop, and play outside-the-home at today’s rates, even more if traditional patterns of induced demand hold.

It is quite possible that sharing remains a niche while most people choose to own their own cars — the ‘Out’ scenario dominates. Thus, exurbanization and AVs better leverage newly available capacity. But, in the absence of pricing, and with cheap energy, there is little to discourage tomorrow’s privately owned AVs from circulating empty on the road network rather than pay for high prices of parking, and thereby slow travel for everyone else. This possible outcome is so obviously bad, it suggests road pricing or similarly effective regulation in some form is likely.


[a] See Levinson, D. (2008). Density and dispersion: the co-development of land use and rail in London. Journal of Economic Geography, 8(1), 55-77 and Xie, F., & Levinson, D. (2009). How streetcars shaped suburbanization: a Granger causality analysis of land use and transit in the Twin Cities. Journal of Economic Geography, lbp031.

[b] For more on this reasoning, see Chapter 11 in Levinson, D. and Krizek, K (2008)  Planning for Place and Plexus: Metropolitan Land Use and Transport. Routledge.

Fixing the Intersection of Broadway and City Road

Satellite Image of Improvement of Broadway - City Road Intersection
Satellite Image of Improvement of Broadway – City Road Intersection

In the second half of 2017, I supervised a first year undergraduate student Tingsen Xian on an independent student project to redesign the intersection of Broadway and City Road in Sydney.

At one corner of this intersection is Victoria Park (lower left) and the University of Sydney (just off site), at another is the Broadway Shopping Center. This intersection has a high pedestrian count, high bus count, reasonably high car count, is very wide (befitting the name “Broadway”), and has long delays, especially for pedestrians. The proposed alternative removes the free left turn and porkchop island on the southwest corner,  gives more space to pedestrians, buses (red), and bicyclists (green), and less space to cars, and the signal retiming reduces total person delay by 1.5% (a lot for pedestrians, while increasing it somewhat for car users), and sends the right incentives. The revised layout is shown in the image.

You can download the full report with more graphics, tables, and yes equations here: broadway-city-road.

[Obviously there are simplifying assumptions in any engineering analysis, and limited measurements and time to conduct the study, but I think the results are better than official results which don’t consider pedestrian delay when timing intersections. It suggests professionals should be able to do a lot better than they have done here.]

Australian License Plates

In the era before widespread electronics, the common activity for kids in backseat of a car during a road trip was to collect license plates. At some point I had a sticker book with plates for each US state, and when I saw a new out of state plate, I transferred the sticker to the appropriate page. This sticker book is long gone.

You could learn about the country that way, in a sense there was a spatial distribution following a gravity like process. Maryland came first, followed by nearby states like Virginia and Pennsylvania, and DC. Texas or California would come up occasionally, though not as often as New York. But finding an Hawaii or Alaska plate was practically impossible.

Australia is simpler, with fewer plates, and on one stretch of road in Sydney, I scored a bunch. In short I learned

  • New South Wales is the First State (Sorry Delaware)
  • Western Australia is the Golden State (Sorry California)
  • Queensland is the Sunshine State (Sorry South Dakota)
  • Victoria is the Education State. It used to be the Garden State (Sorry New Jersey)

I did not see Tasmania (Explore the Possibilities), South Australia (The Festival State), Northern Territory (Outback Australia), or ACT (Canberra – The Nation’s Capital) here, but everything you ever wanted to know about Australian Vehicle Registration plates can be found at wikipedia. The designs and slogans change over time. None are as cool as DC’s Taxation without Representation plates.

I will just say, some of the newer US plates are far more attractive.

George Street, Sydney

A section of George Street (Park St. to King St.) (map) has recently been transformed into a transit mall in preparation for Sydney’s new L2 light rail line (City and Southeast) which will open in a few years. The construction in Sydney has been disruptive to traffic, to pedestrians, to retailers, (who will undoubtedly complain so long as there is compensation for complaining), but at least one part of it is mostly done. Not all of it is: the station is missing, the catenary is missing, [update: the power supply is in the ground] the intersection tracks are missing. Well ok, it’s mostly not done, but at least this part is open for the holidays.  The section looks lovely. I took some pictures yesterday.

The Washington Avenue Transit Mall in Minneapolis could have/ should have looked like this, without a nasty fence down the middle and all those signals.

George Street Mall next to QVB.
George Street Mall next to QVB.


All is well and good, but then you see a car, driving on this pedestrian paradise. Wherefore art thou, car? Then I realize, this is not my beautiful transit mall, it is a shared space-ish thing.

Does everyone know that? Is the white line just a temporary accommodation, or long-lasting feature. (In fact the white tape seems to be failing already.

Car vs. Stroller. I did not stage this. Apparently Sergei Eisenstein was here.
Car vs. Stroller. I did not stage this. Apparently Sergei Eisenstein was here.



The car was paused for the unattended stroller. (the parents are just out of shot, and their pre-school child retrieved the stroller shortly after the photo). It turns out, there is one remaining driveway on this block associated with the Hilton Hotel, who for some reason was allowed to keep this exit. There is also a driveway on the next block too for Westfield.

Car exiting Hilton onto 'pedestrianized' George Street.
Car exiting Hilton onto ‘pedestrianized’ George Street.

I am sure the planners did the best they could, but I really hope there is a longer term solution here than allowing cars access to these buildings off of this street.

And clearly everyone recognizes there is a problem, as a worker has been posted at all the cross streets. Still the number of both “accidental” and “intentional” acts of car driver on crowd violence in the past year is enough that this is likely to remain an issue without more concrete solutions.

Traffic cones and high-viz vests will keep us safe.
Traffic cones and high-viz vests will keep us safe.