Car-dependency is a bad deal

Is it a good idea to have a transport system oriented primarily around the car? Cars are useful for a lot of things, but is it a good idea for most people to use them for most trips?

This is a practical question rather than a philosophical one. Over the last 70 years, different countries have taken very different paths. Some countries, particularly the US but also New Zealand to a lesser degree, have invested heavily to make driving the go-to transport option. Others, such as Germany and the Netherlands, have taken a different course, with a greater focus on public transport, walking, and cycling.

In a 2015 CityLab article, Eric Jaffe tallies up the results for the US and Germany. Here’s the scorecard:

To summarise, Americans have a transport system that:

  • Requires them to spend more on transport, because they own more cars, travel longer distances, and hardly use zero-cost transport modes like walking and cycling
  • Requires their government to spend more to build roads and provide public transport subsidies to compensate for all the underpriced roads
  • Kills all types of road users at much higher rates
  • Contributes to a serious obesity epidemic, which in turn jacks up healthcare costs
  • Adds greenhouse gases to the atmosphere at a much faster rate.

Basically, a car-dependent transport system is a bad deal on pretty much every level. Germans do plenty of driving, but cars are part of a much more balanced transport system. As a result, their advantages are maximised and their disadvantages mitigated.

As Jaffe observes, differences in transport choices can be observed at pretty much every spatial scale in the two countries. Although Germany has its share of suburbs, people living in them have a much greater range of choices and much less need to hop in the car to do everything.

The data come from a recent comparison of German and U.S. planning approaches led by transport scholar Ralph Buehler of Virginia Tech. Drilling down to the city level, Buehler and collaborators find more of the same driving trends in an analysis of two large metros from each country: Washington, D.C., and Stuttgart.

Both areas have similar economies, labor markets, core populations (roughly 600,000 people), regional planning organizations that outline local transport policies. Yet Stuttgart comes off as less car-reliant than D.C. on all sorts of measures…

What’s especially notable here is that driving behavior in the remote periphery of Stuttgart is about the same as it is in the suburbs of D.C. To wit: the two most car-dependent suburbs of Stuttgart (Nürtingen and Geislingen) have shares of all trips by car roughly equivalent to the two least car-dependent suburbs of D.C. (Arlington and Alexandria): roughly 70 to 75 percent in each place. Meanwhile, walking and cycling account for 6 percent of trips in most D.C. suburbs, while in Stuttgart’s most car-oriented areas these modes still account for more than a fifth of all travel.

So the suburbs of D.C. are basically as car-oriented as the cow pastures of Stuttgart. The map below lays it out pretty clearly:

Shares of trips by car in jurisdictions of D.C. and Stuttgart. (Buehler et al, 2014, International Planning Studies)

But, you ask, isn’t one advantage of American car-dependency that it’s allowed cities to open up vast tracts of land to build new suburban homes? Surely America has benefitted from greater housing affordability as a result?

Yeah, not really. As the Economist‘s House Price Index shows, between 1975 and 2016 real house prices in Germany only rose 9%. In the US, they rose 56%, even accounting for the collapse of a massive house price bubble in the 2000s. So it’s hardly the case that cities need more cars to get affordable housing.

What do you make of the data on transport outcomes in the US and Germany?

The linear city and other science fictions

Last week, urban designer lecturers Dushko Bogunovich and Matthew Bradbury published an article on their vision for transforming Auckland into a “linear city”:

Instead, we suggest a linear, city-region that follows the opportunities and respects the constraints in the landscape. Its central spine would connect many nodes of density, functioning as centres of commerce and production, with high-rise living. There could be 20-odd nodes between Whangarei and Hamilton.

This is what we call the “working city”. In contrast – the “lifestyle city” would be situated on the glorious east coast. We see it as part of the larger “NZ Riviera”, stretching from Whangarei to Whakatane. Here, the world-renowned qualities of Auckland’s superb suburban lifestyle would mature to the level where Auckland would truly become the “world’s lifestyle capital”.

New infrastructure technologies, such as localised sewerage and water systems, super-efficient solar panels, internet and electric cars, mean that any new urban settlement is not necessarily reliant on expensive centralised infrastructure systems. We no longer have to get our power from the South Island or by burning fossil fuel, and we don’t have to drive two hours to work.

If this sounds a bit like science fiction, that’s because it literally is science fiction. The ur-form for the linear city – and its most complete expression – is a 1975 utopian science fiction novel by Ernest Callenbach: Ecotopia.

Ecotopia imagined an environmental utopia in a future US West Coast that had seceded from the rest of the country. Urban space and economic life have been upended: the new nation has pursued radical decentralisation and sustainable living.

Bogunovich and Bradbury don’t go as far as Callenbach in calling for an end to investment capital, radical downsizing of central government, and a ban on all cars, but they do harp on many of the same themes when it comes to transport and urban form. In the book, for example, San Francisco has been downsized to a mere village, its population spread out into “minicities” on rapid transit lines:

the great concentrations of people in San Francisco, Oakland, Portland, Seattle, and even the smaller metropolitan areas began to disperse somewhat. New minicities sprung up in favorable locations, with their own linkage necklaces of transit lines: Napa, on its winding, Seine-like river, at last pollution-free; Carquinez-Martinez, stretching out along rolling hills dropping down to the Strait; and others throughout the country.

Bogunovich and Bradbury echo Callenbach’s language when they speak of a central transit spine connecting “many [small] nodes of density”. It’s a seductive idea. But, as public transport guru Jarrett Walker pointed out in his review of Ecotopia, it’s an intrinsically unworkable one from a transport perspective:

A gleaming high speed rail system delivers his hero through a transbay tube to an intimate, shrinking village called San Francisco. But real transbay tunnels and high speed rail require major cities to create the demand around their stations. Those cities need the big infrastructure of power and water and transit. That infrastructure may sometimes require cutting down some trees, accepting the impacts of a dam, building densely where somebody already lives, or creating space for efficient movement on a street that could otherwise have been a park, a creek, a kiosk, a gathering place.

The contradictory, fantastical nature of Bogunovich and Bradbury’s vision becomes even more apparent when we consider the real-world examples they cite for Auckland to follow. These places, they argue, combine a low-density linear form with highly efficient rapid transit and natural amenities:

Frankfurt is a famous example of a super-efficient city that consists of more than 70 local authorities. It prides itself on its inclusion of agriculture into the metropolitan fabric, its first class, evenly distributed, recreational green open spaces, and international airport amidst a forest, which serves three major cities.

Other famous models of successful, decentralised and polycentric development are metropolitan Munich and the urban region of the Ruhr. Both cover large areas, include plentiful open spaces, and have managed to contain urban sprawl in the form of a coherent polycentric pattern.

Let’s take a look at these places. Here’s a map of the Ruhr region. According to Wikipedia, the region is home to 8.5 million people – over five times as many as in Auckland. From end to end, the main urban corridor – from Duisberg to Dortmund – is around 80 kilometres long. That’s about the same as the distance from Pukekohe to Silverdale.

So if we wanted Auckland to be more like the Ruhr, we would have to increase the population of urban Auckland fivefold. That’s a level of intensification far, far beyond anything contemplated in the Unitary Plan.

Source: Wikipedia

Source: Wikipedia

We run into similar problems with Frankfurt and Munich, which are roughly comparable in population to Auckland but considerably denser. Charting Transport has helpfully published comparative data on population-weighted densities in Australian and European cities. (Population-weighted density is the most accurate measure of density – it measures the density of the neighbourhood the average resident lives in.) According to that data, Frankfurt is twice as dense as Melbourne, and Munich is almost three times as dense. (Auckland and Melbourne have pretty similar densities.)

For the visual learners, here’s a randomly selected neighbourhood several kilometres from the Frankfurt city centre. Observe how this kind of medium density would be totally illegal under existing Auckland planning rules:

Frankfurt neighbourhood

So rather than making the case for a sprawled “linear city”, Frankfurt, Munich, and the Ruhr illustrate Jarrett Walker’s point that population density is necessary to obtain efficiencies in infrastructure provision, including well-utilised rapid transit. Those cities have developed intensively where there is demand to do so, especially in inner-city suburbs. As data on infrastructure costs for low- and high-density developments in Auckland shows, this can save money:

CIE and Arup Auckland infrastructure costs by density

Bogunovich and Bradbury’s problems in distinguishing between science fiction and reality get worse when they start discussing Auckland’s existing urban form and infrastructure. They argue that:

Being located on a land-bridge, Auckland has mainly grown in the northern and southern directions. After 100 years of growth and amalgamation, it has grown into a linear conurbation some 70km long. By 2040 it could be 150km long. This is not bad news; linear cities are famously efficient.

Are they really? As Bogunovich and Bradbury concede, Auckland already has a relatively linear urban form. If this does indeed improve efficiency, shouldn’t we already be reaping the benefits in terms of lower house prices and more efficient transport outcomes?

Or, to put it another way, isn’t continuing to do the same thing and hoping for a different outcome the very definition of insanity?

In response to this concern, Bogunovich and Bradbury say that they want to continue doing the same thing – urban expansion into nodes up and down State Highway 1 – but differently in an unspecified way:

Growth is already happening along this corridor anyway – witness the boom in Te Rapa, Pokeno, Silverdale and Warkworth. However, this development is haphazard, exacerbating traditional urban sprawl and commuting distances. It also relies too much on expensive and vulnerable infrastructure.

This is also very problematic: they don’t provide any specific explanation of how their linear city would differ from the one that actually exists. This has serious cost implications. As Auckland Council found when devising a “Future Urban Land Supply Strategy” last year, urban expansion is expensive. They are expecting network infrastructure costs to rise to $100,000-$200,000 per dwelling for greenfield development.

In Bogunovich and Bradbury’s vision, “distributed, small scale, clean, green and smart infrastructure” would bring down these costs. This, again, echoes the science fiction world of Ecotopia. But without details – or better yet, costed and implementable plans – “technology will transform the way we live!” is an empty slogan. It means nothing.

Their discussion of the transport and labour market implications of a 150-kilometre long linear city “that extends at least from Wellsford and Helensville to Pokeno and Orere Point” is equally unsatisfying. They state that considerable horizontal expansion will lead to lower, not higher, transport costs: “we don’t have to drive two hours to work”.

For this to work, it would require people in the outer nodes to work locally, rather than commuting to other areas of the city. That would represent a significant change from the way that Auckland (and every other large city) works. At present, people who live further out commute longer distances, on average:

Trip Length residential 2013

Previous attempts to decentralise the city have not changed this pattern, because it is intrinsic to the way that urban labour markets work. As former World Bank urban researcher Alain Bertaud observes, normal cities involve people commuting between a lot of different points, which enables the agglomeration economies that make cities work. An “urban village” model, in which everybody commutes short distances to the nearest “node”, occurs in planners’ dreams but never in real life:

Bertaud urban structure graph

This isn’t to say that Bogunovich and Bradbury’s ideas are all bad. Given Auckland’s geographical constraints, there is a good case to build a better rapid transit network focused on key corridors with high demand. That’s exactly what Transportblog has proposed in its Congestion Free Network, and it’s what Auckland Transport is planning to build:

AT Rapid Transit Network 2015-2045

Enabling more housing in areas that have good transport accessibility is also a good idea. In fact, that is exactly what the Unitary Plan’s Regional Policy Statement says should happen:

2. Enable higher residential densities and the efficient use of land in neighbourhoods:

a. within and around centres and within moderate walking distances from the city, metropolitan, town and local centres

b. in areas close to the frequent public transport routes and facilities

c. in close proximity to existing or proposed large open spaces, community facilities, education and healthcare facilities

d. adequately serviced by existing physical infrastructure or where infrastructure can be efficiently upgraded.

But, as I’ve explained above, the vision of Auckland as an exclusively “linear city” simply isn’t grounded in reality. It may be fine as science fiction, but it would fail in practice.

In fact, the examples chosen by Bogunovich and Bradbury make that very clear. Auckland’s low-density, linear urban form has led to our current housing affordability and transport problems. The German cities, which are much more densely populated, have been more successful in avoiding those problems. Emulating them would mean allowing more mid-rise housing to be constructed near the centre, not less!

Postcard from Freiburg

After Paris we spent a day and two nights in Freiburg before heading on to Switzerland to meet Stu Donovan and go skiing. By the time you read this, we will be on the slopes! Here are a few thoughts and impressions of the place.

Freiburg is a small city of about 220,000 people near the Black Forest in southwestern Germany. It’s gained an international reputation for being one of the most sustainable cities in Europe. What this means, in practice, is that it seems to combine the best of both big cities – on-street vibrancy, especially around the historic downtown – and small towns – quiet, safe neighbourhoods that are good for young families.

According to our hosts, the city’s cycling mode share – for all trips, not just commuting – is roughly equal to its car mode share. About 30% of trips are on bicycle, and 30% are in cars. The remainder is split between the city’s light rail system and walking.

But it hasn’t always been thus. Freiburg was headed down the road to motordom in the 1950s and 60s. But, like many Dutch cities, it took a different direction in the 1970s, expanding its tram system rather than pulling it out, and starting to build cycleways. They’ve got 420 kilometres of bike lanes, including separated cycling highways, 36 kilometres of tramways, and expansions in the works.

Freiburg’s public transport system has roughly the same level of patronage as Auckland – almost 80 million trips a year. (And, apparently, a farebox recovery ratio of over 80%.)

So far, so good. But Freiburg is growing quite rapidly – a consequence of its attractiveness as a place to live, I suspect – which is putting pressure on housing. Rents are apparently rising, and the city council (which is led by the local Green Party) has to plan for this growth. Freiburg is already a midrise city – three to five storeys in most places – so it is debating its next steps.

In line with past practice, this means a bit of up and a bit of out. New apartments are being planned for some parts of the city, and the city’s negotiating with farmers to buy some more land at the edge. Hearing about the options, and the constraints, I was struck by how familiar it all would seem in Auckland. The farmers were holding out for higher prices for their fringe land. Heritage preservation policies in the old parts of the city make it difficult to build up there. It’s not possible to develop in biodiversity areas or in the Black Forest. Minimum parking requirements add significant cost to new housing.

Wait, minimum parking requirements? In the sustainable city of Freiburg?

Apparently, yes. MPRs in Germany were imposed by a federal law in 1939 – perhaps the earliest example on the books, which would make them literally a Nazi invention. They are still generally in force today, although there’s been a trend to devolve control over minimums to local governments. In Freiburg, the minimums have recently been reduced from one parking space per dwelling to 0.6 spaces per dwelling. (In Berlin, by contrast, they have been removed entirely.)

However, there have also been some innovative ways to get around MPRs. For example, the suburb of Vauban was developed as an energy-efficient, car-free neighbourhood. Requiring one carpark per home would have undermined the concept pretty badly, so parking was instead provided in three garages at the edge of the development.

However, one of those garages was not actually constructed. Instead, everybody in Vauban who didn’t want a car bought into an association that had the option to develop the underlying land for carparks. In effect, rather than buying a carpark, they bought insurance against the possibility that they would need one. Apparently this saved car-free households around 18,000 euros. And the third parking garage still hasn’t been built – the land is still being used as a playground.

The word “association” seems to be an important one when it comes to housing development in Germany. Germany law and tax policies make it easy and advantageous for people to club together to develop multi-family dwellings. A lot of the apartment buildings and terraced houses in Vauban were built this way. Basically, get together with some friends, buy a bit of land, and hire an architect to design an apartment building for you. You’ve got to comply with some high-level rules about bulk and form, but other than that the design is up to the occupants.

I would be interested in learning more about this model. It could be really compatible with New Zealanders’ aspirations for housing – we have a preference for designing (and even building) our own houses, but Auckland’s high land prices make it costly to build standalone dwellings. Midrise apartment buildings could fill the affordability gap, but the current developer-led model doesn’t suit everyone. So perhaps look to Germany for an alternative model?