Car-dependent suburbanisation and the Justice Potter Stewart ruling …

“The Los Angeles freeway system has … changed the way people think about the urban landscape and therefore about the metropolitan area. Earlier models of transportation created a multi-centered way of life served by widespread patterns of movement, but development of the freeway system is the critical stage in the formalisation of that process. Greater Los Angeles is not a series of suburbs in close proximity to one another; it is, rather, their integration into a fluid system of exchanges, of which the freeway system is the most important nexus … In an area of sprawling suburbanisation and hundreds of randomly attached communities, the freeway serves to evoke a sense of clarity and sharpness, to delineate and integrate urban space. It has created a new sense of space.” [D. Brodsly, ‘L.A. Freeway’, 1981, p.23]

“If I lose my car, I lose my livelihood – period. All I do is drive my children to school and my wife to work and then myself to work – each leg is 30 minutes. I do the same again in the evening. If you can’t drive in the Inland Empire you’re finished.” [quoted in E. Luce, ‘Rebirth of the American city’, Financial Times, June 8 2013]

“I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description ["hard-core pornography"]; and perhaps I could never succeed in intelligibly doing so. But I know it when I see it, and the motion picture involved in this case is not that.” [Emphasis added.]

—Justice Potter Stewart, concurring opinion in Jacobellis v. Ohio 378 U.S. 184 (1964), regarding possible obscenity in The Lovers.

I’d like to re-engage a bit with the blog and continue with an occasional posting. In particular I want to pick up on where the last entry left off and spend some time over a few entries looking in more detail at the concept of car-dependent suburbanisation as a “spatial fix”. This was a term used by David Harvey to describe and analyse the post-WWII car-dependent urban growth in the US: as we saw, “[t]his arrangement was ‘spatial’ not least because it entailed a massive physical expansion of cities, and it was a ‘fix’ because it revived real growth in the national economy.” [Hackworth, 2007, p.78].

One of the points that I argued in the last post was that the Keynesian “spatial fix” to the Great Depression, manifested particularly in the US by the technologies of car-dependent suburban sprawl and by the federal-level planning of the suburban form through, for example, FHA insurance criteria, mortgage tax subsidies, and freeway construction, provided the spatial context for a new, decentralised suburban politics dominated by a middle-class, property-owning, aspirational (and largely white) electorate that was easily mobilised against explicitly redistributive policy. In the UK, a similar, though less spatially-distinct, process took place as council tenants voted for a Conservative party that enabled them to buy the houses in which they had lived as subsidised tenants.

But it might be a good idea at this point to take a step back and try to be a bit more precise about some of the terminology we are using – or at least, using the illustrated blog format, to try to give you a visual sense of what we are talking about.

What, in particular, do we mean when we talk about “car-dependent suburbanisation”?

Clearly, suburbanisation is a phenomenon that both long pre-dates the car and has been intimately connected with the evolution of transport technology. Let’s start with a couple of longish quotes from Peter Hall, one of the key thinkers in the field – the first from his superb “Cities in Civilization” (1998), summarising many of the key points that have been touched on in earlier blog entries:

“… [G]etting to work is no problem in the village or small town: a five-minute walk to field or workshop would suffice. … Since a large and growing part of all employment was found in or near the city centre, this means that the effective limit of a city’s growth was set by the ability and the inclination to walk: in practice, as a ready rule of thumb, three miles (five kilometres), equal to one hour’s travel on foot. Small wonder that as London doubled prodigiously from one to two million in the first half of the nineteenth century, people crowded ever more closely near to the centre, and human misery enormously increased; … exposure to cholera, most dreaded of Victorian epidemics, was hugely compounded by the phenomenon of population density and the near-inevitable pollution of water sources that resulted.

Transport technology provided the answer, in the form first of steam railways and horse buses and streetcars, then electric trains and subways and motor buses and electric trams, finally the private motor car and its accompanying highway system. It was urban transport technology which first allowed the city to sprout suburbs, as in late nineteenth-century London and New York, and which finally encouraged the suburbs to overwhelm the traditional city, as in mid-twentieth-century Los Angeles. But this was no easy or effortless process: it demanded not only great ingenuity in developing the technologies themselves, but also creativity in devising appropriate organizational and financial frameworks to bring the new systems into existence.”[p.612]

The next quote, focusing on the “dialectical” relationship between the evolution of transport technologies and urban form, is from his paper ‘Squaring the Circle: Can we Resolve the Clarkian Paradox?’ (1994):

“In 1957 the economist [Colin] Clark published a paper, destined to become a classic in the urban literature called ‘Transport: maker and breaker of cities’. In it he argued that – at least since the first industrial revolution, two hundred years ago – the growth of cities had been shaped by the development of their transport facilities. But these in turn were dependent on the evolution of transport technologies. For each successive development of the technology, there was a corresponding kind of city. However, the relationship was more complex than that: it was a mutual one. The transport system shaped the growth of the city, but on the other hand the previous growth of the city shaped and in particular constrained the transport alternatives that were available. So the pattern of activities and land uses in the city, and the transport system, existed in some kind of symbiotic relationship.

But, Clark stressed, the two could get out of step, and indeed very often did so. That was particularly the case because cities change more slowly than the available technologies change”.

I think this is about right: the relationship between evolving urban form and the “choice” of transport technology is highly dynamic over time, a "spatial fix" that is like all complex systems subject to multiple positive feedback loops. In order better to illustrate this point, I think it’s a good idea to use … illustrations! 

In the previous post on the subject of neoliberalism I quoted from A. Whittemore, (2013), “How the Federal Government zoned America: The Federal Housing Administration and Zoning.” He uses the examples of two residential developments in Los Angeles: Leimert Park and Westchester. The first was begun in 1928, just before the onset of the Great Depression and already well into the evolution of LA from a city of “streetcar” suburbs to one of automobile suburbs; the second was developed only a decade later, beginning in 1938 and expanding rapidly to meet the huge demand created by the aircraft manufacturing industry in LA in the early 1940s. 

What follows is taken from Wikipedia and Whittemore’s analysis.

Developed by Walter H. Leimert (for whom it is named) beginning in 1928 and designed by the Olmsted brothers (sons of New York Central Park designer Frederick Law Olmsted), Leimert Park was one of the first comprehensively planned communities in Southern California designed for upper and middle-income families, and was considered a model of urban planning for its time: automobile traffic near schools and churches was minimized, utility wires were buried or hidden from view in alleys, and densely planted trees lined its streets. Walter Leimert envisioned a self-sufficient community, with a town square, theatre and retail shopping. 

The photo shows a section of Leimert Park, still under development in the 1930s. Intimate in scale, with the commercial area in the upper centre of the photograph, multifamily construction along the radiating arterials, and single-family homes protected on interior streets. Navigable by all personal transportation modes, multiuse, and intimate in scale, it remains an excellent embodiment of the vision of its designers and to a great extent also reflects the ideals to which today’s New Urbanists aspire.

The counterpoint to Leimert Park is, in Whittemore’s view, the “perverted neighborhood unit”: Westchester in 1949, only a few miles to the northwest of Leimert Park. Bloated and dominated by vast sections of single-family use, the Los Angeles Daily News wrote: “You only have to visit Westchester to see the advantage of modern community planning over old-fashioned guesswork methods. In most old communities a hodge-podge of single-family, duplexes, apartments, and business properties are all mixed together.” 

Westchester was praised by the planners; described as “low-density and functionally exclusive”, the boulevards are walled off from the residential lots backing up to them from interior streets, and the business district is buffered from the residential areas by parking lots and streets.

Whittemore focuses in his article on the role played by the Federal Housing Administration, formed in the middle of the 1930s in the deepest years of the Depression, and charged with providing Federal Government insurance to residential mortgages that conformed to the building and planning codes that were established by the FHA itself. He therefore stresses the influence that the bureaucratic aspects of zoning and the political/consumer demand driven by the need to access FHA-insured mortgages as key aspects of the evolution of built form between these two neighbourhoods. One could also look at the role played by traffic engineers in insisting on wide and relatively straight roads that would maximise the efficiency of traffic flows as cars evolved rapidly in terms of speed and sheer numbers on the road.

The point I want to make however is that visually the differences in urban form between the two neighbourhoods are indeed very clear, and they occurred within a very short period of time, subject to multiple influences driven by the underlying dynamic of the rapid evolution of car technology. 

The key elements in understanding the differences, I would argue, are the concepts of “mixed use” and “walkability” which have been fundamental to all the main debates on urban form since the middle of the 20^th century. There is no simple binary distinction between, on the one hand, a car-dependent, and on the other a “car-independent” suburb. But, to use Justice Stewart’s famous phrase, “I know it when I see it.”

Neoliberalism and the city

Back after a long break ... a few summer weeks in Europe and a semester at UNSW during which I didn't post anything. However, one of the assignments for an Urban Planning course was something of a trip down memory lane ... the old 2,000-word essay that I remember from undergrad days! I thought I might kick off a new season by posting it. The lecturer gave us some leeway in selecting the topic so I chose something a bit political.

For illustration I have chosen some photos from an amazing series by Yves Marchand and Romain Meffre, 'The Ruins of Detroit'. I think they suit the topic of the post and I hope you find them as striking as I do.

TITLE:    "Ironically, the suburbanization dynamics that had been entrenched during the postwar years of Keynesianism ... would establish the conditions of existence for the rise of a distinctly suburban-secessionist mode of neoliberal governance…" [Peck, 2011, p.901]. Discuss.

“I’ve begun recently to think that neoliberalism is poised to replace globalization as the next popular metaconcept in the social sciences. The literature on neoliberalism as a concept has exploded in recent years … States, provinces, policies, eras, people, countries, and institutions have all been deemed ‘neoliberal’ or ‘neoliberalizing’ … It is used broadly to characterize the right wing; to mean the guiding light for the ‘Washington consensus’; to mean anything related to business; to mean anything related to capitalism; to mean anything related to liberals in the US. Neoliberalism is everywhere and, apparently, everything.” [Hackworth, 2007, pp. xi-xii].

The academic literature around neoliberalism with regard to urban form and the evolution of cities is large, diverse and politically charged. There is nevertheless a consensus between left and right that one can broadly distinguish between two eras in the political economy of the post-World War II developed capitalist world: a “managerial Keynesian” era from 1945 to the early 1970s and a “neoliberal” era post-1980. The eras were divided by an interregnum of “stagflationary” economic crisis, political turmoil and intensified political conflict during the 1970s. A central tenet of the Keynesian critique of neoliberal urban planning policies is that the redistributive role of city planning has been abandoned. As one critic puts it, “[t]he belief that markets could regulate the allocation of housing as the most rational means of resource distribution … has led public policymaking towards the abandonment of the conceptual meaning of housing as a social good, part of the commonalities a society agrees to share or to provide to those with fewer resources: a means to distribute wealth.” [Rolnik, 2013, p.1059].

This paper argues that this criticism, harking back as it does to the perceived success of managerial Keynesianism in the post-war era, is misconceived for a number of reasons. First, that the political preconditions for neoliberalism’s success – as a “distinctly suburban-secessionist mode of … governance” – were themselves the outcome of a “spatial fix” dominated by managerial Keynesian policymaking. Second, it poses the question of what exactly “planning” should aim for in the context of the city – in particular whether it can realistically aim for redistributive outcomes or should rather seek to provide universal benefits. Third, to what extent planning at the city level can overcome fundamental economic, technological and demographic trends at the national level: in other words, that cities as social organisms are inherently flexible, prone to shrink and decline as well as expand, and that therefore in some circumstances city planning should aim primarily at mitigating the impact of decline.

The managerial Keynesian system emerged in 1945 led by the US (and to a residual degree the UK) as an outcome of the impact on governments, elites and policy-makers of the market failures of the Great Depression and the successful experience of large-scale planning during the New Deal and WWII. It was characterised in policy terms by activist Keynesian government intervention in the economy with a commitment to maintaining consistently high aggregate demand, a cross-party political consensus that government planning was central to improving peoples’ living standards, education levels, housing, health and access to infrastructure. Industrially, advanced capitalist economies were characterised by large, vertically-integrated oligopolies and similarly large industrial trades unions and labour movements. Globally, trade barriers were slowly dismantled, but global capital flows remained limited by capital controls. The ‘Bretton Woods’ currency architecture established a system of fixed but adjustable currencies dominated by the US Dollar as a global reserve currency pegged to gold.

How did this political and policy world relate to the evolving world of urban form and cities? Writing about the US from a Marxist perspective, Harvey developed the notion of a “spatial fix” to the Great Depression, rooted in the mutual dependence of the state and real estate capital [Harvey, 1989]. He argues that the great post-war transformation of the US urban landscape from one characterised largely by high-density industrial/commercial cities into one dominated by car-dependent suburban sprawl was driven by fundamental underlying characteristics of the capitalist dynamic – namely, to avoid or postpone an inherent tendency towards “overaccumulation”. “Within the US the radical transformation of metropolitan economies (through the suburbanization of both manufacturing and residences), as well as the expansion into the South and West, absorbed vast quantities of excess capital and labour” [Harvey, 1989, p.185].

Whether or not one subscribes to Harvey’s wider Marxist critique of capitalism’s propensity to crisis, an overarching model focusing on the confluence of technological, organisational and policy factors (and the dynamic interactions between them) to throw light on the evolution of the built environment is central to understanding the postwar growth of suburbanization. In technological terms three elements are generally accepted to be key: the transformation of the car industry by the Fordist system of standardised mass production based on the production line, and the consequent virtuous cycle between high consumer demand and falling price; a similar process that was applied to industrial-scale suburban single-family house construction, pioneered by the Levitt brothers; and the adoption of the limited-access highway pioneered by Robert Moses. In policy terms, the Federal State intervened heavily and consistently in industrial policy, macroeconomic demand management through fiscal and monetary policy, in social security and education, the subsidisation through the tax system of private housing, and through public investment in infrastructure, the key choice from the perspective of suburbanization being the huge investment in the national freeway system. Overall, “[t]his arrangement was ‘spatial’ not least because it entailed a massive physical expansion of cities, and it was a ‘fix’ because it revived real growth in the national economy.” [Hackworth, 2007, p.78].

Within this large and complex area of study, I will make a general point before focusing on three interconnected areas to try to understand the nature of the transition to a neoliberalized polity in the 1980s. The general point is that “spatial fixes” can be seen as systemic to human and urban history, and driven by the same interaction of technological (particularly transport) breakthroughs, organisational change (particularly the division of labour and trade) and central government policy. Therefore in the industrial era the massive expansion of cities in the ‘first wave’ of suburbanisation in Europe and America along steam railway lines, and later using electric tram networks, as well as the mid-19^th century transformation of the American economy via the canal- and steam train-driven settlement of the mid-West, can be seen as earlier transformative templates for the mass car-dependent suburbanization of the mid-20^th century. So while the impact on physical landscape of car-dependent suburbanization since 1945 is indeed profound, perhaps in a historical perspective it is merely one (and perhaps not the most significant) of the transformations that have, and will continue, to shape the urban environment.

The first aspect to consider is the degree to which the characteristic urban form of car-dependent sprawl in the US, detached single family homes (on plots and in neighbourhoods defined by zoning restrictions) was itself the intended result of federal government planning, developed with the rapid expansion of direct federal government involvement in the US residential housing and mortgage markets during the New Deal and massively expanded post-1945. Specifically, this planning took the form both of broad subsidies for “middle class” home ownership (notably via the tax-deductibility of mortgage interest payments) and specific and detailed design parameters embedded in the criteria by which the Federal Housing Administration (FHA) judged mortgage applications. The FHA was “in essence, a government created and controlled insurance company functioning to insure certain kinds of loans made by private lenders, which are deemed to be socially or economically desirable.” [quoted in Whittemore, 2013, p.622].  It “codified … assumptions [on desirable residential urban forms] through standards for the mortgages on which they would provide insurance. Developers wanted to meet these standards so they could get the FHA insurance, essentially a subsidy as it removed much risk from lending. Local planners wanted FHA-insured development because it represented socially desirable and fiscally beneficial growth, and local legislators quickly saw their case, adjusting zoning accordingly. This had long-term impacts beyond development directly backed by the FHA because the FHA set the tone for safe practices throughout the private sector and because FHA-inspired zoning guided non-FHA-financed developments as well.” [ibid. p.621]. The FHA criteria for mortgages suitable for federal insurance went beyond zoning by use to require a specific urban format of generally single-family, low-density housing by mandating minimum lot sizes and widths, lot coverage, and front and side setbacks. It “effectively banned construction commonly associated with traditional urbanism: it eliminated the row-house building type, set buildings back from the street, and spread out development by requiring open space to surround buildings on all sides.” [ibid. p.630]

A second factor was that, once suburbs had developed as coherent entities with their own localised social identities and political interests, the impulses towards segregation (by race, income and building use), initially driven by road and car technology as well as FHA regulations, were increasingly reinforced by local legal and fiscal activism. As opposed to the “top-down” federally-imposed criteria of the FHA, there was also a significant “bottom-up” process in US politics and legislation which also acted to reinforce the single-family house, exclusionary-zoning version of urban morphology. For example, the framework for zoning established by the Supreme Court in the 1926 Village of Euclid vs. Amber Realty decision [Batchis, 2009] explicitly formalised the principle of zoning as a vehicle for social exclusivity, extending as it did to neighbourhoods the right to protect local property values by excluding particular housing forms and densities rather than merely specific industrial and commercial activities. Moreover, local autonomy in property and sales tax collection incentivised local suburban governments to maximise tax policies to attract investment and employment in competition with both the central cities and other suburbs.

The third, and perhaps decisive factor, was the degree to which expanding economic opportunities, growing prosperity and the evolving political consciousness of an increasingly middle class electorate, in no small part the outcome of decades of successful Keynesian intervention in infrastructure and housing markets, itself fed back into a backlash against the perceived failures and inadequacies of Keynesian planning. In her work comparing those developed countries in which radical neoliberal political movements were electorally successful (the US and UK) with those in which it remained a marginal influence (Germany and France), Prasad concludes that “the main story of Thatcherism, as of Reaganism, is of politicians mobilizing an increasingly prosperous majority that had moved across the class divide of … adversarial policies” [Prasad, 2006, p.100]. There were of course substantial tactical differences between the US and UK in terms of mobilization of electoral support for neoliberal policies (tax reduction and anti-welfare slogans in the US; opposition to the perceived excesses of trade union power and the privatisation of council housing and nationalised industries in the UK) as well as very divergent spatial and political-economic structures (decentralized, suburbanized in the US; centralized municipal-socialist in the UK). She concludes nevertheless that “[f]ree market, or neoliberal, policies did not result from any pragmatic or rational analysis showing that they were the best way to manage the crisis … Rather, they arose where the political-economic structure was adversarial. States in which the political-economic structure defined labour and capital as adversaries and the middle class and the poor in opposition to one another … provided the potential to ally the majority of voters with market-friendly policies” [p.38] … “When the majority of citizens become taxpayers and move out of the working class … support for redistribution to the poor … is fragile. If politics is particularly competitive, politicians will take advantage of this fragility in their quest for power.” [p.39]

CONCLUSION: in many respects, the redistributive aims of the Keynesian managerial era of urban planning in the US and UK (admirable though they were) contained the seeds of their own demise. The Keynesian “spatial fix” to the Great Depression, manifested particularly in the US by the technologies of car-dependent suburban sprawl and by the federal-level planning of the suburban form through, for example, FHA insurance criteria, mortgage tax subsidies, and freeway construction, provided the spatial context for a new, decentralised suburban politics dominated by a middle-class, property-owning, aspirational (and largely white) electorate that was easily mobilised against explicitly redistributive policy. In the UK, a similar, though less spatially-distinct, process took place as council tenants voted for a Conservative party that enabled them to buy the houses in which they had lived as subsidised tenants.

It is arguable therefore, that explicitly redistributive goals are fundamentally inimical to successful long-term planning in cities. Cities are “merely” the physical manifestation of the agglomeration of individual citizens who choose close connectivity for the economic and social benefits that connectivity brings, plus the technical infrastructure that makes such agglomeration possible. Even more than nations, cities need to competitively attract, retain and develop citizens capable of maintaining a flexible, resilient and high-value economy, and willing to pay the taxes necessary to build and maintain the city’s infrastructure. Individuals in a democratic society can opt out of a city, either to another city or to the suburbs, whether the underlying transport technology is cars, streetcars, railways or canals. Typically, a downward spiral of decline was driven by the collapse of concentrated nodes of heavy industry (cars in Detroit, shipbuilding on the Clyde), loss of employment and tax base, collapse of housing values, middle-class flight to the suburbs, population decline and further inability to fund infrastructure. In many cases, it is hard to imagine how cities could have overcome such profound economic shocks as the decline of the Detroit car industry or Scottish shipbuilding; or massive demographic waves such as the movement of population to the US South and West; or the post-WWII migration of rural African-Americans to northern industrial inner cities. In many cases, therefore, long-term population decline is probably inevitable and “planning” should be geared to accepting that fact, and mitigating the impact.

In the Keynesian managerial view (to be a bit polemical) such fundamental economic and social change is seen as a “failure of planning”, to be combatted by even more “planning”, as if the eternal competitiveness of car or shipbuilding industries that underpinned a city’s glory days could be ensured by city (or if not city, then national) Keynesian planners. In general terms, planning in cities will perhaps be more successful if, rather than aiming for redistribution, it pursues “universalist” goals on the model of France and Germany in Prasad’s study: bipartisan commitment to decent and affordable public infrastructure such as schooling, security and transport, some degree of legislative protection for housing tenants and structures to allow new building, as well as a commitment to promoting good design and urban form: basic good city management in other words. Not difficult in theory; harder in practice.

 

REFERENCES

Batchis, W. (2009). “Enabling urban sprawl: revisiting the Supreme Court’s seminal zoning decision Euclid v. Ambler in the 21^st Century”. Virginia Journal of Social Policy & the Law 17:3 373-403

Brenner, N. and Theodore, N. (2005). “Neoliberalism and the urban condition.” City 9(1), 101-107.

Glaeser, E.L. (2011). Triumph of the City: how our greatest invention makes us richer, smarter, greener, healthier and happier. The Penguin Press.

Hackworth, J. (2007). The Neoliberal City: Governance, Ideology, and Development in American Urbanism. Cornell University Press.

Harvey, D. (1989a). The Condition of Postmodernity. Basil Blackwell.

Harvey, D. (1989b). “From managerialism to entrepreneurialism: The transformation in urban governance in late capitalism.” Geografiska Annaler 71 B(1) 3-17

Kirkpatrick, L.O. and Smith, M.P. (2011). “The Infrastructural Limits to Growth: Rethinking the Urban Growth Machine in Times of Fiscal Crisis”. International Journal of Urban and Regional Research 35(3) 477-503.

Peck, J. (2011). “Neoliberal Suburbanism: Frontier Space.” Urban Geography 32, 6 884-919.

Peck, J. (2012). “Austerity Urbanism.” City: analysis of urban trends, culture, theory, policy, action 16:6 626-655

Prasad, M. (2006). The Politics of Free Markets: The Rise of Neoliberal Economic Policies in Britain, France, Germany & the United States. The University of Chicago Press.

Rolnik, R. (2013). “Late Neoliberalism: The Financialization of Homeownership and Housing Rights.” International Journal of Urban and Regional Research 37(3) 1058-1066

Sager, T. (2011). “Neoliberal urban planning policies: A literature survey 1990-2010.” Progress in Planning 76 147-199.

Whittemore, A.H. (2013). “How the Federal Government zoned America: The Federal Housing Administration and Zoning.” Journal of Urban History 39(4) 620-642.

 

More cars and suburbia: the economics of car use and sustainability …

“No rational concert promoter would decide how big to build a stadium based on the number of people who would come to see the Grateful Dead if the tickets were free. But that is often how transportation planners decide highway capacity: they estimate how many trips would be made on an unpriced facility, then try to build a facility big enough to accommodate that number of trips.” Moore and Thorsnes, ‘Transportation/Land Use Connection’, American Planning Association p.57

“There is nothing revolutionary about providing ‘free’ transportation. It is done by every elevator in any office building or apartment. The cost of the service is collected as part of the rent from all tenants, whether they use it 100 times a day or not at all … The driver benefits when his neighbour leaves his car at home and takes the subway or bus. But if everyone who comes downtown each day by public transport drove a car, there would be an enormous traffic jam. It is true, of course, that not everyone wants to drive downtown. But, then, not everyone uses a park or playground, yet all pay for it.” Blumenfeld, ‘The Modern Metropolis’, p.166

Let’s spend this post looking in a bit more detail at the cost structures associated with the different transportation options that underpin the built environment in our two idealised cities, “Los Angeles” and “Paris”. Of the two forms, the car and underground rail, the cost structure of the former is far more complex and, I would argue, much more relevant to the question of “sustainability”; so we’ll spend most of the post looking at the costs to the individual and society of private car usage, and try to look at some recent data in order to begin to quantify the various costs.

Let’s however first summarise the cost structure and urban density implications of underground rail. The fixed costs – digging tunnels and underground stations, and equipping them with signalling and trains – are very high, and therefore the debt load necessary to build the system is also high. The variable costs (electricity, maintenance and personnel) are comparatively low. It is relatively inflexible but once built it is in place virtually indefinitely, and therefore suits large cities that are stable in spatial terms. It is relatively straightforward to build high capacity for ‘peak passenger loads’ into the system, for the rest of the time the system will generally operate with substantial spare capacity. The feeder network is pedestrian – i.e. the system operates on a “walk-train-walk” basis in order to go door-to-door, and a convenient walking distance of say half a mile to the nearest station determines the catchment area for each station. The system is therefore only cost-effective in large high-density (compact) cities. However, for those cities the capital costs are relatively easy to amortise over a long period of time because they tend to have high value-added economies, high incomes, high property values and therefore large tax bases. Apart from requiring density, the underground system is independent of the design of the urban streetscape.

With Blumenfeld’s elevator analogy, it is theoretically possible to imagine an approach that says that in a high rise building, movement via the stairwell is free, but that the additional costs of installing an elevator will be met by charging the residents of the higher floors a fare for each trip. Obviously, this would be “fair” for the inhabitants of the ground floor. In reality the costs of building and operating an underground system are impossible to cover by charging fares (they are simply too high). But just as are the high rise building and the elevator are entirely symbiotic (movement by stairwell becomes increasingly inconvenient and then virtually impossible the higher you go), so is the underground rail network with a large and compact city. In this sense the provision of transport as a free good is both necessary and probably inevitable.

Turning to the car, what is the cost structure of car ownership? These are somewhat more complex than for a public transit system: they fall under the general headings costs to the individual or household, infrastructure costs which must be borne by the city as a whole, and what Jane Jacobs would probably regard as “urban form costs” which are less quantifiable but in my view just as critical.

The costs of car ownership for the individual (or, more generally, household) are in simple terms divided into fixed costs (purchase of the car, insurance, tax and depreciation) and variable costs (overwhelmingly petrol/gasoline but also maintenance and any road tolls or parking costs). Things obviously get more complicated in reality: for example, depreciation will also increase when the car is driven a lot, cars themselves vary significantly in cost and fuel consumption, and technologies such as hybrid and electric cars change the parameters.

However, let’s look at some recent data to get a broad picture of the scale of costs for the household – this from a Canadian survey:

So we’re talking about approximately CAD$7,000 fixed costs per annum for a single medium sized car. It’s probably safe to assume that the average suburban household is at least a two-car household: therefore let’s say $14,000. As far as variable cost is concerned, it is interesting to note both the absolute amounts but also how it will escalate dependent on fuel costs – a doubling of the price of petrol will roughly double the variable cost component of the car driven 32,000 kilometres per year (i.e. about 55 miles per day) to about $8,000 p.a., leading to a total household annual cost per car of $15,500. Therefore a two-car household, with both cars being driven 32k kilometres p.a., would be facing total costs in the order of $31,000 p.a. in the event that petrol prices double from when the survey was done. 

With regard to variable costs, the important things to conclude are: first, that while the marginal cost per kilometre driven is a very small number ($1,500/12,000 or $0.125 per kilometre for this car), when an individual or household is locked into a requirement to make long commutes the costs are significant; and second, that those costs escalate in step with petrol price rises.

Let’s take a look at what has been happening to petrol prices (data from the US):

In other words, since 2008, despite the US (and most developed economies) having flat-lined at either very low growth levels or outright recession, petrol prices have risen substantially, reflecting global crude oil prices. The following charts shows two significant processes: first, that just as after the 1989 oil price spike, US drivers are in the process of significantly reducing their consumption of petrol; and second, that crude oil prices remain at far higher levels than in the 1990s and most of the early 2000s despite the ongoing recessions in many of the major developed oil-importing countries.

Without delving into the dynamics of the world oil market it is probably fair to conclude that elevated crude oil prices over a period of several years reflect a combination of two factors: on the supply side a ceiling having been reached in world production, and on the demand side the rise of a car economy in Asia (and in particular China). The relevance of these data for the sustainability of the car suburbia model of urban growth is fairly clear – namely that if a fundamental shift has taken place which brings into question the premise of cheap petrol, and that this reflects a long-term shift in supply/demand balance of the global oil market, then the question arises how far the red line of average per capita gasoline consumption can continue to decline while car dependent suburban life remains viable?

But while it is probably a truism to say that car suburbia lives or dies depending on the price of petrol in the long run, this is not to say that a suburban car-dependent lifestyle which is “sustainable” one year with crude oil at $20 a barrel will be “unsustainable” the next year with crude oil at $100 a barrel. Reality is much more complex: to go back to a Jane Jacobs quote from a previous post, just as “the ideal of the suburbanised anti-city was developed architecturally, sociologically, legislatively and financially” on the foundation of a single technology, that of affordable individual car travel, so any transition to a new urban model will similarly depend on the ability of societies to cope with the political, financial, technical, architectural and legislative aspects of that transition.

In many respects the most interesting questions revolve around the technical modalities and time-frames of re-fitting and adapting the existing car suburbia format to alternative transport structures and living arrangements. However, the history of previous transformations of urban form in the US is ominous: in particular the social, political, economic and financial implosion of many of the inner cities in the 1970s and 1980s. Just as Jacobs identified complex feedback loops and self-reinforcing dynamics as the key drivers of spatial dispersion in car suburbia, I would argue that the financial and economic dynamics that promoted suburban sprawl are subject to equivalent feedback loop phenomena, with the attendant threat that long-term “virtuous cycles” of expansion can rapidly turn into “vicious circles” of contraction. I will take a closer look at the infrastructure aspects of suburban cost structures and the financial-economic dynamics in the next post.

Cars and suburbia: urban density, the compact city, sustainability …

“One of the chief contributing factors to traffic congestion is crowds of pedestrians interrupting the flow of traffic at intersections.” Dallas planner Vincent Ponte, quoted in W. Whyte, ‘City: Rediscovering the Center’ [p.198]

“The present relationship between cities and automobiles represents … one of those jokes that history sometimes plays on progress. The interval of the automobile’s development as everyday transportation has corresponded precisely with the interval during which the ideal of the suburbanised anti-city was developed architecturally, sociologically, legislatively and financially. But automobiles are hardly inherent destroyers of cities … We went awry by replacing, in effect, each horse on the crowded city streets with half a dozen or so mechanised vehicles, instead of using each … vehicle to replace half a dozen or so horses.” [J. Jacobs, ‘The Death and Life of Great American Cities, pp.446-7]

Since photos of freeways and parking lots are not very exciting, let's illustrate with the help of a few car movies!

 

We ended the last post by quoting Hans Blumenfeld: “High density requires transit, and transit makes high density possible. Low density requires individual car driving, and universal use of the car requires low density.” Both he and Jacobs were writing in around 1960, during the early stages of the build-out of the freeway system and suburban sprawl in North America. However, while Blumenfeld clearly identified the degree to which lower residential densities erode, and finally eliminate, public transit options by making feeder networks uneconomic, he nevertheless implies that there is a more or less stable trade-off, and hence choice, between the high-density/public transit and low-density car systems. In this respect, and with the advantage of 50 years’ hindsight, Jacobs was in my opinion more prescient insofar as she very clearly recognised that a central aspect of this trade-off was that it was certainly not stable, but rather highly dynamic and subject to severe feedback loops – in other words, that once a society began the process of moving from a high-density to a low-density model of organisation, that this process would feed upon itself and therefore was unlikely to reach a stable equilibrium at any meaningful level of density. It could be argued, from a 21^st century perspective, that there may be no straightforward and stable “choice” between car-based suburban sprawl on the one hand and high-density cities on the other, but that rather the complex positive feedbacks involved in car-based low-density living mean that the urban environment will be subject to wild and unstable fluctuations (or, more accurately, decreases) in density – effectively a “chaotic” system rather than a linear system. (Many modern “behaviouralist” economists, such as Robert Shiller who has also focused on the American housing market, are moving in this direction).

How does Jacobs make her argument? It is worth laying out the logic at some length: “Too much dependence on private automobiles and city concentration of use are incompatible. One or the other has to give. In real life, this is what happens. Depending on which pressure wins most of the victories, one of two processes occurs: erosion of cities by automobiles, or attrition of automobiles by cities … We also have to be aware that surface traffic in cities exerts pressures upon itself. Vehicles compete with each other for space and for convenience of their arrangements. They also compete with other uses for space and convenience.” [pp.454-5]. The urban planner Victor Gruen did an analysis of Fort Worth’s downtown traffic requirements in 1955: “Gruen, in order to understand the size of the problem … began by calculating the potential business that Fort Worth’s currently underdeveloped and stagnating – but traffic jammed – downtown ought to be doing by 1970, based upon its projected population and trading area. He then translated this quantity of economic activity into numbers of users [shoppers, workers etc] … [He then] translated the putative future users into numbers of vehicles [and] calculated how much street space would be required to accommodate [them]… He got an outlandish figure of roadbed needed: sixteen million square feet, not including parking [compared with five million in 1955]. But the instant Gruen had calculated his sixteen million square feet, the figure was already out of date and much too small. To obtain that much roadbed space, the downtown would have to spread out physically to an enormous extent. A given quantity of economic uses would thereby be spread relatively thin. To use its different elements, people would have to depend much less on walking and much more on driving. This would further increase the need for still more street space … Differing uses … would be so far from one another that it would become necessary to duplicate parking spaces themselves … This would mean spreading the downtown even thinner… As Gruen pointed out …, the more space that is provided cars in cities, the greater becomes the need for use of cars, and hence for still more space for them … Swiftly or slowly, greater accessibility by car is inexorably accompanied both by less convenience and efficiency of public transportation [because of uneconomic feeder networks], and by thinning-down and smearing-out of uses, and hence by more need for cars.” [pp.456-7] “When a city has become a sufficiently homogeneous and thin smear, it should have the traffic problem, at any rate, in hand. Such a state of equilibrium is the only possible solution to a positive feedback process like city erosion. This point of equilibrium has not yet been approached in any American city.” [p.461]

Can it be said in 2013 that such a point of equilibrium has been reached in any American urban environment? I leave that judgement to those better qualified, but the figures are quite striking: “between one third and one half of urban America’s land is typically dedicated to the driving and parking of vehicles. In Los Angeles, that ratio jumps to two thirds. Houston provides the equivalent of 30 asphalt parking spaces per resident.” [J. H. Kay, ‘Asphalt Nation’, p.64].

The critics of suburban sprawl, and particularly of the car suburb, are legion (books such as ‘Suburban Nation’, ‘Asphalt Nation’, ‘The Geography of Nowhere’, and much of Lewis Mumford’s writings) and there is no point trying to summarise their analyses in any detail here.

I will continue the theme of looking at the underlying structures and dynamics of transport networks at an abstract level, as well as at their underlying cost structures, in order to establish some sort of a basis for comparison and maybe some tentative conclusions about re-configuring urban environments and their transport networks. We have made the assumption that the raison d’etre of a city was physical communication – that is ease of movement for its inhabitants as a corollary of the division of labour and multiplicity of choice in work and living. In simple terms that was of course exactly the radical promise made for the car and the truck – they would liberate individuals and industry by providing the convenience and efficiency of universal door-to-door communication. This would be the case both for interurban (via the freeway network) and intraurban traffic. The complexity, redundancy, inefficiencies and costs of the multiple modes of transport that constitute public transit systems, and the need to transfer between them to achieve a door-to-door service, would be thereby be eliminated.

Let’s get back to our concept of “Los Angeles”, an abstract and simplified city of 15 million people which is fully dependent on the car and truck for personal and goods transport, as a comparator to “Paris”. At 7,500 people per sq mile “Los Angeles” will extend over 2,000 square miles; i.e. it will fit into a square with approximately 45-mile sides. In order to facilitate the efficient movement of cars the city will be served by a number of high capacity intraurban freeways, so that any door-to-door journey outside the immediate neighbourhood will involve driving to the nearest freeway, driving on the freeway network to as close to the final destination as possible, and exiting to use the “feeder network” of smaller roads to get to the final destination. The longest possible drive within this city is clearly dependent on the positioning and number of freeways, and there’s not much point in being excessively detailed, but as an indication the distance from one corner of the square city to the opposite corner will involve a drive of around 64 miles.

But as with underground rail, the freeway system needs to be built to cope not with the average daily number of car and truck journeys evenly distributed, but rather to cope as far as possible with the two daily rush hour spikes of traffic. The various engineering and human behavioural aspects of traffic management are extremely complex, but the basic reality is that it is (to say the least) much harder to build sufficient capacity into a freeway system than into an underground rail system. It may well be that there are now technologies in the pipeline (such as automatic steering for cars in convoy) that will substantially improve the efficiency of use of road space; but the historical experience to-date has been that all attempts to add capacity by building new freeways or adding lanes to existing freeways have been negated within a short period of time by increased traffic. “A recent University of California at Berkeley study covering 30 California counties between 1973 and 1990 found that, for every 10% increase in roadway capacity, traffic increased 9% within four years’ time” [quoted in Duany et. al., Suburban Nation, p.88].

This reflects a general and worldwide phenomenon, which is due to the fact that the reasons for it reflect the interaction of fundamental human behaviour and the feedback loops of car suburbia – induced traffic and latent demand. “The mechanism at work behind induced traffic [can be described by the aphorism]: ‘Trying to cure traffic congestion by adding more capacity is like trying to cure obesity by loosening your belt’. Increased traffic capacity makes longer commutes less burdensome, and as a result, people are willing to live farther and farther from their workplace. As increasing numbers of people make similar decisions, the long-distance commute grows as crowded as the inner city, commuters clamour for additional lanes, and the cycle repeats itself.” [Suburban Nation, p.88]. Latent demand refers to the fact that, “since the real constraint on driving is traffic, not cost [an issue we will look at below], people are always ready to make more trips when the traffic goes away.” [p.91] The level of traffic on any road reflects the aggregate of drivers’ individual choices: “How crowded a roadway is at any given moment represents a condition of equilibrium between people’s desire to drive and their reluctance to fight traffic. Because people are willing to suffer inordinately before seeking alternatives … the state of equilibrium of all busy roads is to have stop-and-go traffic. The question is not how many lanes must be built to ease congestion but how many lanes of congestion do you want?” [pp.91-2]

There are a couple of points that I think are worth addressing before closing this post: first, that there are ways of mitigating traffic congestion – primarily by pricing; and second, that to some extent you might find similar dynamics with regard to passenger traffic on underground rail systems in large cities. Governments can always reduce demand for a service by charging for its use – specifically by making it more expensive to drive by charging tolls either in general or focused on high congestion times of day. London for example has instituted a “congestion charge” on car commuters. This is certainly do-able when congestion becomes unbearable, but it has tended in practice to be politically difficult and is perceived as discriminatory in favour of the wealthy against the poor. With respect to the second question, many underground systems (London for example) suffer with serious under-capacity and overcrowding, and there may be elements of induced traffic and latent demand creation. However, as we saw in the previous post, while it is expensive there is no meaningful limit to the additional capacity one can add to an underground line because of the inherent efficiency of fitting people rather than cars into a given amount of space and time.

But the basic point to bear in mind is that these sorts of questions are really aimed at dealing with symptoms rather than the underlying disease. The disease is the one Jane Jacobs identified, the inexorable and self-reinforcing dynamic between car use and urban form over time. To repeat and emphasise an earlier quote: “the more space that is provided cars in cities, the greater becomes the need for use of cars, and hence for still more space for them … Swiftly or slowly, greater accessibility by car is inexorably accompanied both by less convenience and efficiency of public transportation [because of uneconomic feeder networks], and by thinning-down and smearing-out of uses, and hence by more need for cars.”

One for transport nerds: urban density, the compact city, sustainability …

“Does anybody know what the ‘right’ density is? I do. It is 12,000 to 60,000 persons per square mile of residential area (20 to 100 persons per acre). In other words, acceptable conditions can be created within a wide range of densities … but there are upper and lower limits beyond which serious disadvantages appear.” [H. Blumenfeld, “The Modern Metropolis: Its Origins, Growth, Characteristics, and Planning”, p.172]

“If sustainable development is so dependent on higher densities, then the question is higher than what …?” [M. Jenks and N. Dempsey, ‘The Language and Meaning of Density’, in Jenks & Dempsey, “Future Forms and Design for Sustainable Cities”, p.287]

 

I’d like to spend a couple of posts driving around the veritable “Spaghetti Junction” (http://en.wikipedia.org/wiki/Spaghetti_Junction) of issues related to the subjects of urban density, the compact city, and sustainability. These complex and interrelated topics have provided planners, scholars and urban historians with material for robust debate since at least the early days of the commuter suburb in the late 19^th century. In recent decades, as technological options have multiplied, the range of densities at which people live has widened considerably, and as the question of the relationship (if any) between urban density and sustainability has become globalised with rapid worldwide urbanisation, the scope for complexity and confusion seems to have expanded with them.

What I would like to do as a contribution to unscrambling some of the complexity is to make some of my definitions and assumptions explicit, and by simplifying a couple of categories at least to pick out a couple of useful signposts along the route.

The first point of simplification is that I will focus on the impact of urban transport technologies on how the built environment has been (and can be) shaped, and how that can potentially affect the issue of sustainability. Although one can view the subject of urban sustainability through many lenses (such as energy generation and consumption, food supply, water and sanitation, even no doubt cultural and psychic) I believe that understanding the evolution and dynamics of transport technology is the starting point for any meaningful discussion. In this, I will draw heavily on Hans Blumenfeld’s wide-ranging series of essays, written between 1940 and 1965 and published as “The Modern Metropolis”.

The second point of simplification will be to distinguish a number of abstract, “ideal type” city structures in order to illustrate the symbiotic and dynamic interplay between a given transport technology and population density. My aim in starting with a very abstract categorisation of city types is twofold: first, to give a mental image of the kind of city we are talking about; and second, to try to focus on the essentials before reintroducing some levels of real world complexity.

Let’s start with a long quote from Blumenfeld which helps illustrate, using the example of 19^th and early 20^th century technical progress, why the evolution of transport technology is so critical to understanding the urban landscape:

“The 19^th century gridiron plan of our cities was designed without any mental image of the body of the city, and the buildings were constructed without relation to any design of the city as a whole. No one could guess from a map of Manhattan from which field on this checkerboard rises the fantastic silhouette of the city’s skyscrapers.

Yet it is possible to discern a definite pattern of the modern city that has gradually superimposed itself on the ubiquitous gridiron. This pattern is essentially a product of the growth of transportation, which at different stages developed centralizing and decentralizing tendencies.

In the first stage interurban traffic was revolutionized by steamships and railroads. Where they met, and only there, could modern industry assemble the masses of coal – its driving power – and of raw materials and food that it needed. And only from these points could it easily ship its products to distant markets. Factories attracted workers, and the presence of many workers of various skills created favourable conditions for more factories.

But while steamships and railroads carried huge masses of goods and passengers to and from the far corners of the earth, traffic within the city moved, as of old, on foot or by horse and buggy; and while the telegraph carried news around the world within a few seconds, communications within the city were still carried by messengers. So factories and offices and dwellings all tried to be close to the centre of the city, crowding each other.

Only after several decades did the technical revolution reach the interior communications of the city. Suburban railroads, streetcars, elevated trains, subways, buses, automobiles, and the telephone overcame the distances within the urban area – as steamships, railroads, and telegraph had already succeeded in overcoming distances between cities. While interurban traffic continues to act as a centralizing force, concentrating business and population in metropolitan areas, intraurban traffic acts as a decentralizing force within the limits of these areas. The densely crowded agglomeration of the 19^th century with its concomitant, the fantastic skyrocketing of urban land values, turns out to have been a short-lived passing phenomenon necessitated by the time lag between the transformation of interurban and intraurban traffic, respectively; it was bound to disperse once this lag was overcome.

Though it would disperse, it would not dissolve. Sources of power, raw materials, and markets may be equally accessible outside the metropolitan area, but it is only here that employees and employers have a wide range of mutual choice, as skills become ever more varied and specialized. The modern metropolitan area is primarily a labour market; it extends only as far as people can commute daily to and from work.” [Blumenfeld, pp.31-2].

I think it will be useful to look at three “ideal type” city formats, and the transport technologies that underlie each of them and which are necessary to make them work. I hope that, although the characterisation of the cities will be highly simplified and abstract, it will nevertheless be useful in order to draw out some of the basic constraints and interrelationships involved. The first city type has a density of 75,000 people per square mile and is served by a public transport network of subway lines – let’s for sake of argument call it “Paris”. The second type is a city composed entirely of low density “automobile suburb” development (7,500 people per sq m) and has no public transport network, being solely dependent on its road network – let’s call it “Los Angeles”. The third type is a hybrid with a density of 25,000 people per sq m: characteristic of, for example, an outer suburb of a large city like London or a smaller European city (let’s say Nottingham, Nantes or Freiburg). This city is served by a mixed transport network of light rail (trams) and cars.

However, before turning to look at each of the cities in turn, let’s first address a couple of the definitional complexities and confusions that plague the debates around urban density in order to clarify our terms a little. The first confusion is “between high densities and overcrowding … The Garden City planners and their disciples looked at slums which had both many dwelling units on the land (high densities) and too many people within individual dwellings (overcrowding), and failed to make any distinction between the fact of overcrowded rooms and the entirely different fact of densely built up land. They hated them both equally … and coupled them like ham and eggs, so that to this day housers and planners pop out the phrase as if it were one word, ‘highdensityandovercrowding’” [J. Jacobs, p.268]. Clearly, one can find population densities of 75k-100k people per sq m in either the central districts of Paris or Hong Kong, or in a third world slum or in the slums of 19^th century London or Chicago. The differences are pretty obvious, and come down essentially to the amount of space available to each person, so we don’t need to be confused.

A second area of confusion and complexity is of terminology and measurement. In their survey of the British historical experience of recognising the significance of, and attempting to legislate for, urban density criteria, Jenks and Dempsey identify the “wide range of different measurements [which] have been used including: persons per hectare; dwellings per hectare; habitable rooms per hectare; bed spaces per hectare; and floorspace per hectare.” [p.293] Moreover, there is the question of whether density is measured in gross (the whole area including all uses) or net (solely concerned with residential square footage) terms. “[D]ensity ‘can be life-threatening when in the wrong hands’. Using net residential density alone fails to take into account wider issues of land capacity, mixed uses, and gives no guide for assessing aspects such as ‘walkability’”. [p.293] Let’s react by acknowledging these complexities, but assume that issues such as mixed use are a given in the interests of thinking more abstractly: think “Paris” or “L.A.” in terms of the general format of the city, or I will put in some pictures to illustrate the sort of urban places I’m thinking of.

“PARIS”: what, at least for the purposes of this analysis, is the point of a city? In Blumenfeld’s view, the “basic raison d’etre of the modern metropolis is the need for cooperation and communication resulting from the division of labour… Primarily the metropolis is a labour market, a place for making a living. This also sets its limits. It is a commuting area, extending as far as daily commuting is possible and no farther.” [p.123]. Let’s go with this definition, maybe broadened a bit to say that the point is ease of physical communication of all kinds and therefore mobility – the ability of the individual to move around the city reasonably fast and reliably, at a reasonable cost, for whatever reason.

Starting with a residential density of 75,000 let’s assume a mega-city of 15 million. Purely abstractly this is a city of 200 square miles, or one that therefore fits into a square with sides of just over 14 miles. Let’s further assume that we are going to serve the city with a public transport network consisting solely of a heavy rail underground system. (There is clearly always going to be a requirement for the road network to deal with some private car usage, transport for tradesmen and the delivery of goods, but let’s make the reasonable assumption that with the great majority of intraurban trips made by underground the road network is both adequate for high quality pedestrian usage and occasional car/delivery van usage.)

What does moving around the city look like? In order to go door-to-door anyone making a trip within the city will obviously need to walk from the starting point to the closest subway station, take the subway to the nearest subway stop to his/her destination, and walk to the final destination. Pretty straightforward. Making the standard assumption of an average 3 miles per hour walking speed, a 10-minute walk is half a mile. Therefore if we are going to say that every inhabitant is a maximum approximately 10-minute walk away from the nearest subway station, a subway station needs to be placed at the centre of every square mile block of the city – i.e. 200 stations which are connected by 14 13-mile west-to-east subway lines and one (higher capacity) 13-mile north-to-south line. So the longest trip anyone will have to make in the 15 million inhabitant city is: 10 minutes’ walk to the nearest subway station; a subway trip of 28 stops with 2 changes of train line; and 10 minutes’ walk to the final destination.

What are the advantages of a modern subway system for moving people around a city? Obviously, it is independent of the surface shape of the city, in particular the design of its street network. This means that there is no trade-off between the design choices that can be made to enhance pedestrian accessibility and liveability on the one hand, and the design choices necessary to ensure car mobility on the other. The second point is really a subset of the first: that its capacity (the number of people that can be moved per hour) is for practical purposes unlimited. This is significant because intraurban trips are not evenly spread throughout each 24 hour period but rather are heavily concentrated in the 2-hour morning and evening commuter rush hour periods; therefore in terms of capacity any transport system has to deal not just with the total number of trips made during the day but with the 2 daily “spikes” of demand on working days – therefore 20 hours of the average working week in comparison to the total 168 hours of a week.

If we want to get a little bit nerdy, there are a number of ways the carrying capacity of a subway system can be increased through the design of trains and stations: trains can be enlarged – widened, lengthened or made double-decker; headways (the time between each train) can to some extent be decreased (although there are technical constraints that mean that minimum headways are not lower than around 90 seconds or 2 minutes); moving passengers on and off the trains can be made faster by having platforms on both sides of the train, one for passengers entering and the other for exiting. If necessary, more tunnels can be dug. There are some technical limits to train speed stemming from the fact that a train has to accelerate and then decelerate within (in our example) one mile; however, even this can be radically increased by having two sets of trains and tunnels on the highest-capacity routes, one set for a smaller “local” service stopping at each station and one set for an “express” service stopping every 4-5 stations. Anyway, the net result is that there are no meaningful limits on the design capacity of underground rail to move very large numbers of people, with high concentrations at rush hours, efficiently and reliably.

The main problem with underground rail systems is capital expense – primarily digging the tunnels and the stations, and equipping them with signalling, control equipment and trains. Once built, the operating costs are relatively low and stable – electricity, maintenance and salaries for the relatively small number of people necessary to run the system. A subway system is also by definition inflexible – it can’t cheaply be adjusted if part of the city suddenly becomes relatively depopulated and/or the city expands rapidly in terms of area. However, the debt incurred through the high capital costs of a subway system is relatively straightforward to amortise over a long period of time when you have a large and densely populated city which is relatively stable in land area and population over extended periods of time, and which will tend to have relatively a high and stable tax base (in terms of individual incomes, business productivity and property values.) Such cities go together with subway systems, as Jane Jacobs would put it, like “ham and eggs”.

Having looked at an idealised high-density city, let’s begin to look at the dynamics involved in the transportation options available to a low-density city, “Los Angeles”. Blumenfeld points out that “[i]n every big metropolis a battle royal is raging between the advocates of a rapid-transit system and the partisans of a freeway system … But neither rapid-transit lines nor freeways constitute a system. Neither can carry persons or goods from door to door. Given the money, anyone can build a rapid-transit line or a freeway … The problem is: how do you get to it from where you are, and from it to where you want to go? A transit system is no better than the feeder system that brings people to and from the stations, and a road system is no better than the interchanges, streets, parking and loading spaces to which the cars and trucks must get from the freeway.

If one thinks in terms of a transit system and a road system, it is not difficult to define the role that each can play best. If you go back to your home in one suburb from a party in another suburb at two in the morning, you cannot expect a bus or train but have to drive your car. But when a thousand people want to go from point A to point B in the same five minutes, it is obviously more sensible to carry them in one train than to have a thousand cars compete for street and parking space. High density requires transit, and transit makes high density possible. Low density requires individual car driving, and universal use of the car requires low density.” [p.141]