Successful cities understand spatial economics
Basing a course of action on limited knowledge risks establishing a practice with little merit.
Economics is about the allocation of scarce resources. For cities one of the scarcest resources is space. Privately built floor spaces, public spaces, street spaces, recreational spaces, parking spaces…. This paper will discuss some concepts that underpin the allocation of city space and the difficulties the lack of application of these concepts can create for urbanisation, in New Zealand in particular.
Economists traditionally when looking at how scarce resources are allocated break the issue down to demand and supply concepts. I have followed that pattern for this paper.
The demand for city space is a fascinating topic, because it is central to understanding the economic transformation the world is undertaking. For several hundred years the world has been going through an urbanisation process — driven by long-term economic factors which pulled workers into urban environments — such as specialisation and agglomeration, and factors which pushed workers out of rural communities — such as decreasing agricultural employment. It is this urbanisation process which has been the long-term driver of demand for city space.
The opportunities and challenges of this global urbanisation process, I think is best described by the NYU Stern Urbanization Project.
In my country — New Zealand, 85% of its 5 million population are clustered into urban environments, on less than 1% of New Zealand’s landmass. The rural urban population split has stabalised at 85% for several decades now. Over 50% of New Zealand’s population live in the wider urban catchment areas of four cities — Auckland, Christchurch, Wellington, and Hamilton with populations roughly 1.5m, 0.5m, 0.4m and 0.25m respectively.
Projecting forward from 2018 to 2048, Auckland is predicted to continue to dominate population growth. Canterbury the second largest region is expected to have relatively fast growth — indicating Greater Christchurch will be the clear second largest urban area. Waikato will catch up to Wellington’s regional population — indicating Hamilton and Wellington will be similar sized cities in fourth place.
Although the demand for urban space is subject to some predictable long-term economic and demographic factors, there are also other demand factors which are hard to predict or control for.
The United Kingdom since the Second War War gives a good example of the difficulty in planning and providing for various unpredictable demand for urban space factors.
The basis for the United Kingdom’s modern planning system was set out by the 1947 Town and Country Planning Act. The Act tried to implement the garden city movement’s agenda of a central city surrounded by a greenbelt and satellite towns.
The 1940’s and 50’s was a period when the UK strongly believed in central planning and control (Economics & Land Use Planning, 2004, Evans, P. 3–7). Yet despite the full resources of the UK state being available to predict and control the urban environment, UK planners failed to make allowances for many important economic and social changes that would occur in the decades following the Second World War, such as;
- London’s population would not remain constant or perhaps fall as assumed, it in fact increased significantly due to a baby boom and an immigration boom. This led to a large and unpredicted increase in demand for housing in London.
- The economic troubles of the Great Depression would not reassert themselves post war and in fact incomes doubled and trebled. This led to an increase in demand for larger houses and motor vehicles, further increasing demand for urban space.
- Social changes such as a rising divorce rate and people living longer led to an increase in demand for residential spaces from many more smaller households.
- It was also assumed that public authorities would be the major provider of new housing, not the private sector as subsequently occurred. The disasters and loss in confidence in UK local authority housing provision was not predicted.
- It was assumed that inner-city slums would be cleared, and the population would move out past the newly established greenbelt to satellite new towns and to the regions. Further, that regional policies and manufacturing growth would take up this ‘slack’. It was not predicted that employment growth would primarily be in services that re-concentrated employment back in the city.
The final two points are important because it meant that London and other UK towns and cities experiencing employment growth, not only had constrained themselves from building out into greenbelts, they also lacked a private sector mechanism to intensify housing in existing suburbs to accommodate the re-concentrated employment growth.
The consequence of this inability for housing supply to adequately follow employment growth is that UK renters spend the highest proportion of their incomes in rent in Europe. The UK has had boom towns and cities for employment but not for housing. As described for other restrictive boom towns this labour market barrier to entry has serious consequences for productivity and inequality.
Advocating for new housing intensification mechanisms to address these issues is what London YIMBY is trying to remedy with its hyperlocalism proposal (declaration of bias — one of the hyperlocal proposals is mine). Hyperlocalism is about empowering some of the smallest units in the housing market — neighbours or streets or residential blocks (my proposal) of existing property owners — with new intensification tools, so they can maximise the gains from intensification whilst minimising the externality costs.
Hyperlocalism is interesting, as it is the opposite tactic to what is often proposed, as it decentralises decision making for how zoning is applied. For instance, in California which also faces a similar housing crisis, there was a proposed intensification mechanism that the state legislature remove local government’s ability to impose certain restrictive zoning rules — regarding car parking minimums, height and density restrictions etc close to rapid transit networks. Unfortunately the proposed legislative changes were voted down. In theory, hyperlocal solutions should have more political supported because it ensures existing residents in each block are agreeable.
Given the important caveat that demand for space in a city is often hard to predict, I would like to introduce several demand and supply for urban space concepts which give a flavour for the nature of city urbanisation.
Four examples of demand for urban space
1. Agglomeration
The first demand concept which I would like to introduce is agglomeration.
Agglomeration is a process by which concentrating economic activity in one place makes businesses more productive. It works through three main channels:
1. Sharing — the ability to share inputs, supply chains and infrastructure;
2. Matching — access to a large pool of workers;
3. Learning — the ability to exchange ideas and information, known as ‘knowledge spill-overs’.
Agglomeration doesn’t only influence business demand for city space, it affects consumers and workers too. Cities with amenities, such as, more and better shops, restaurants, educational facilities, clubs, societies, culture — plays, shows, sporting events, etc have an attraction which is greater than the sum of each amenity part. For workers being able access a larger pool of employment provides more future opportunities for the individual worker, as well as for their partners and extended family members.
2. Floorspace is a normal good
The second demand concept I would like to discuss is that floorspace in an urban environment is a normal good. Meaning as incomes rise, people consume more space per person. This can be best seen in the following graphs and maps of Tokyo Prefecture. The graphs are from James Gleeson who tweets under Jim @geographyjim he describes himself as a housing policy wonk and number cruncher at Greater London Authority. Recently James collated all his graphs into an article. You can access that article here along with an article describing Japanese zoning.
Although not the case for Tokyo, sometimes income growth and the desire for more personal space can be so strong that central urban areas lose population. This can be seen in Manhattan from 1910 to 1980. Improved urban transport links and modes, such as, the first affordable automobile — the Model T Ford — which went into production in 1908, meant overcrowded tenement building residents could move to more spacious dwellings in the expanding areas of metropolitan New York.
So, under some conditions, the normal good demand for personal space can cause a dispersal of a city’s population.
3. Centrally located roads are more congested
The third demand concept is congestion — a sign of excessive demand for road space — will occur in more centrally located roads first and then radiate outwards.
This demand concept holds even if a city’s population and businesses are evenly dispersed. The transport links which are closer to the centre of the city will have a greater amount of traffic than peripheral transport links.
This makes intuitive sense, consider the example of a circular shaped city with an evenly dispersed population, if half the residents randomly chose to visit the other half, basic geometry indicates, more of those journeys would go through the centre, or near the centre of city, rather than the periphery.
Ring roads can help to divert some of this traffic.
But eventually with rising demand, the number of travellers being funnelled through fixed and narrow corridors in the central parts of towns and cities will be greater than those corridors can manage. Centrally located road space will then become congested.
A common response to congestion is to build more roads. I discuss how Auckland built motorways to its CBD from the 1970s in the article Tokyo does not subsidise its transport system!
Building more and wider roads can temporarily alleviate congestion as higher capacity is provided, but there is an effect called induced demand, which means that congestion reappears, as there is a behavioural change to utilise the new road space that has been freely provided.
4. Hotelling’s model for spatial competition
The fourth demand concept is an economic force which drives demand for more central city locations — it is called Hotelling’s model for spatial competition, which explains another reason for why competing firms tend to cluster together in central locations.
The simplest analogy to explain this concept is to consider the case of two ice-cream sellers on a beach. The socially optimum location for ice-cream to be sold from the beachgoers perspective would be at the 1/4 and 3/4 positions along the length of the beach. That way beachgoers could evenly spread across the beach and only have to walk, at most, 1/4 of the length of beach to buy an ice-cream. But from the ice-cream sellers’ perspective they can get more market share the closer to the middle of the beach they locate themselves — so that is what happens.
Hotelling’s model is not an absolute law, the clustering activity in the centre of cities that is commonly seen, can become more dispersed in real world cities. Frequently cities develop polycentric centres, as land costs, congestion and other factors create opportunities for other nodes within the urban environment to be competitive.
What is not seen though, is urban areas made of discrete urban villages which have no interactions with their neighbouring villages. Alain Bertaud who has spent a lifetime working in the urban planning field, shows this in his models of the spatial distribution of jobs and populations in urban environments;
So, in summary there are factors which drive demand for both the dispersal and centralisation of urban space. But that is not the whole story, cities are not just influenced by demand factors (including many other aspects of demand which I have not discussed). Supply also plays its part.
Examples of how urban space is supplied
1. Making Room
The first supply of city space concept I would like to introduce is the Making Room paradigm, which has been developed by the NYU Stern Urbanization project. Note — Alain Bertaud, since 2012, has been a senior research scholar at the NYU Stern Urbanization Project. I met Alain and his wife Marie-Agnes at a dinner, after a talk he gave in Christchurch, New Zealand in 2014.
Making Room for urban expansion is about the division of new urban areas into public and private spaces. It is a deceptively simple concept, with many important implications, which can be grouped under two Making Room subheadings — the supply of public spaces and the supply of private spaces.
1.1 The supply of public spaces
Firstly, the basic bones of a city are planned, they are the product of public decision-making processes. The private sector does not spontaneously allocate public spaces for urbanisation. This is an uncomfortable fact for libertarians and free market ideologues. If city planning processes fail completely, the result will be sprawling slums or barrios, with little in the way of street spaces or public parks. This will take extreme measures after-the-fact to rectify.
What makes libertarians even more uncomfortable is that due to a type of market failure, to do with site assembly or contiguity, governments need to have the option of compulsory acquiring land for the network of transport right of ways and public parks. The reason for overriding private property rights is to prevent landowners holding out land which is vital for the completion of infrastructure networks and then extracting monopolistic prices for its use. The concept of contiguity is discussed further here.
The Making Room concept that for growing cities proper provision of public spaces needs to be supplied in advance of private development has at least a 2500-year pedigree. For example;
the Urban Planning Study for Piraeus (451 BC), which is considered to be a work of Hippodamus, formed the planning standards of that era and was used in many cities of the classical epoch. According to this study, neighbourhoods of around 2,400 m2 blocks were constructed where small groups of 2-floor houses were built. The houses were lined up with walls separating them while the main facets were towards the south. The same study uses polynomial formulas for the pumping infrastructure manufacture.
In modern times there is a wide variation in the allocation of street space between cities. The UN have compared 30 cities in the following report.
The relevance of street patterns and public space in urban areas — UN Habitat Working Paper April 2013. This report notes that;
cities that have around 100 crossings per km2 , on average, which allows for walking distance between crossing 100 meters apart. This is considered walkable and appropriate in many cities, in order to generate street life and for moving goods and services productively and efficiently. Also, this pattern of around 100 crossings per km2 determines the size of blocks, around 9000 m2 each which provides for good plotting within each block.
Auckland, New Zealand’s largest city, only has 18.1% of its land area allocated for street space and 72.9 intersections per square km. Both figures place Auckland in the low-end group of cities, for its supply of street spaces.
New Zealand’s central governments have long restricted local government’s ability to raise revenue to fund infrastructure, or to use off balance sheet debt instruments, such as developers who build infrastructure having access to municipal bonds which can be repaid by a targeted rate. Auckland Council in particular, due to experiencing the largest percent population growth, has had difficulty keeping up with the demand for infrastructure for new urban areas, without exceeding credit rating debt limits.
The under supply of infrastructure and street space could explain why Auckland is particularly prone to traffic congestion and why it struggles to receive the agglomeration benefits its size indicates it should achieve. Although, as will become apparent from this general discussion on spatial economics, New Zealand struggles to efficiently allocate all kinds space for its urban areas. So, it is probably the general difficulty of managing urbanisation that is the better explanation for Auckland’s recent poor performance (from a high base) compared to the rest of New Zealand.
1.2 The supply of private spaces
The second implication of the Making Room concept, is that once public and private spaces have been allocated, it is best if private spaces are left as much as possible to spontaneously order themselves. Spontaneous order, or a tolerance of messiness within a structured framework, is something I tried to describe in my article, What is the secret to Tokyo’s affordable housing?
This tolerance for spontaneity is the issue that Marie-Agnes Bertaud Roy -marital and academic partner of Alain Bertaud — wanted to highlight when commenting on my Tokyo housing article.
Alain and I were in Tokyo and Toyama in November. We had long walk in suburban areas. We were stuck by the tolerance for large variations in lot sizes, type of buildings and commercial mix. This results in a kind of messy look, which is not unpleasant when moving around. Planners and city managers seems to concentrate on what they should always do: superb maintenance of road surface, signage and cleanness, but benign neglect for everything within private plot boundaries! What a good lesson for our New York City planners who restrict the type of zone where an «umbrella repair shop» or a « supply for artist retail shop» (no kidding, this is real) should be located!
Thus, private spaces within a city, in particular built floorspace should be allocated as much as possible by market forces, without interference by costly restrictions.
If land-use restrictions are too tight. If a city is restricted from growing up and out, then this turns competitively derived land prices to monopolistic pricing. Land prices are the biggest influence over the price of different types of built floorspace — such as housing. A point well made in an article titled — Rocket scientists do not forget about gravity so why do planners forget about economic rent?
Not even direct government intervention, by compulsory purchasing the monopolistically priced land, can convert land prices back to competitively derived pricing. As construction company — Fletchers and the New Zealand government are finding out with their joint inner-city Christchurch property development project, initiated in response to Christchurch’s 2010 and 2011 earthquakes. Seven years post-earthquakes and they still have not been able to build and sell any of their proposed $1 billion Christchurch CBD residential development onto the market.
There are strongly held norms about fairly compensating landowners when governments acquire land, which prevent governments from using their monopoly buying position to bargain down land prices. Of course, if governments acquired land which had not been affected by monopolistic pricing, such as rural land on the periphery of urban areas and then provided good transport connections to the greater urban area, this could provide some competitively derived land pricing.
This is the approach New Zealand governments took in the 1920s with the Hutt Valley Lands Settlement Act that “provided the legislative framework for the Petone — Waterloo branch line to be paid for by the betterment of the land it served. The Act set in motion the process that culminated in the construction of a double track, electrified commuter railway through the Hutt Valley”. The First Labour government of the 1930s used this integrated land and transport process to build state housing suburbs. These new urban areas were successful state housing areas compared to earlier failed attempts because fast and efficient transport links were established to the main urban area of Wellington — in particular by a new commuter rail service.
In the subsequent decades in New Zealand, which were dominated by National (Conservative) government’s, private sector developments with roads and motorway transport connections were the preferred choice for city growth.
Planning historian Chris Harris describes these public decisions as being the reason for the differing nature between Auckland, Christchurch and Wellington. Auckland being the most automobile dependent and sprawling. Although, in the last decade it has been Christchurch which has sprawled into road only connected satellite towns like Rolleston, while Auckland is finally investing in a rapid transit network.
In recent years the public in New Zealand have gradually come to terms with the fact that land-use restrictions increase the cost of new housing.
Note, land-use restrictions are not just zoning which prevents the outward expansion of a city, it is also restrictions, such as, height, setback, shade planes, heritage, view shafts, minimum car parking requirements etc, which limit the upward expansion of a city.
The economist Edward Glaeser, who is the author of the book — Triumph of the City — has been a champion for the easing of land-use restrictions. Wikipedia describing it as;
During the 2000s, Glaeser’s empirical research has offered a distinctive explanation for the increase in housing prices in many parts of the United States over the past several decades. Unlike many pundits and commentators, who attribute skyrocketing housing prices to a housing bubble created by Alan Greenspan’s monetary policies, Glaeser pointed out that the increase in housing prices was not uniform throughout the country (Glaeser and Gyourko 2002).
Glaeser and Gyourko (2002) argued that while the price of housing was significantly higher than construction costs in Boston, Massachusetts and San Francisco and California, in most of the United States, the price of housing remained “close to the marginal, physical costs of new construction.” They argued that dramatic differences in price of housing versus construction costs occurred in places where permits for new buildings had become difficult to obtain (since the 1970s). Compounded with strict zoning laws the supply of new housing in these cities was seriously disrupted. Real estate markets were thus unable to accommodate increases in demand, and housing prices skyrocketed. Glaeser also points to the experience of states such as Arizona and Texas, which experienced tremendous growth in demand for real estate during the same period but, because of looser regulations and the comparative ease of obtaining new building permits, did not witness abnormal increases in housing prices.
A series of events in 2016 helped educate the public of the consequences of restricting the building of new housing;
- Auckland’s housing affordability became significantly worse, with the city climbing from the world’s ninth most expensive city to fifth in a year.
- In May of 2016, in response to frustration over years of slow progress in improving New Zealand’s housing supply, the opposition (now government) housing spokesman Phil Twyford requested the Government to use its upcoming National Policy Statement on housing and urban land-use to abolish Auckland’s Rural Urban Growth Boundary (RUB) and to free up density rules to accelerate Auckland’s ability to grow housing supply both out and up. This signalled the end to the political/economic acceptance of the regulatory status quo of New Zealand’s urban planning system.
- The public were shocked by news stories of families living in cars in parts of Auckland.
Housing was an election issue in New Zealand’s 2017 general election. The previous National party (conservative) led government taking a lot of criticism for not acting strongly enough to address the problems of unaffordable housing and homelessness.
What is interesting, is underneath these political disputes, in New Zealand a political and economic consensus is forming that agrees with Edward Glaeser, in attributing the cause of house price bubbles, as largely being about land-use restrictions. There is much agreement around the contention that constrained housing markets respond to positive demand shocks, by increasing house prices.
Note demand shocks come in many forms, many of which, as I have discussed, are unpredictable. Examples are — fast income growth, like New York experienced a century ago, easing of credit conditions, like the US experienced during the Greenspan years and fast population growth from immigration, such as New Zealand is currently experiencing.
In the future New Zealand cities could experience demand growth for particular urban areas due to sea level rises or natural disasters like earthquakes damaging existing urban areas. New Zealand’s emigrant diaspora (estimated to be over half a million) could return due to changes in relative economic opportunities. New Zealand’s cities could become more affordable thereby increasing their attractiveness.
Given these ongoing yet uncertain demand shock issues, constrained cities like Auckland need to increase their resilience by improving their housing supply response.
The previous Finance Minister of New Zealand — Steven Joyce commissioned a report, called — Quantifying the impact of land-use regulation: Evidence from New Zealand -July 2017, which almost exclusively used Edward Glaeser’s techniques to show that the cost of new housing in many urban areas of New Zealand, had deviated above its marginal cost due to land-use restrictions.
Auckland’s house building restrictions were assessed as being the worst of any urban centre in New Zealand.
2. The allocation of car parking spaces
Donald Shoup is considered the economist with the most expertise on how a city should allocate car parking spaces. According to Donald 30% of the land area in US cities is car parking.
Donald Shoup describes how this situation came about. In the 1920’s the sides of city streets were congested with parked cars. Cities came up with two policy responses — parking meters and requiring private landowners to create off street parking — also known as mandatory parking minimums. Local government liked the second option because it created a huge supply of parking at no cost to them and it answered the question of “how did you let such and such building be built if there is not enough parking”.
The problem though is that parking requirements are set arbitrarily, planners do not have any training on how to allocate parking space. In the US, parking minimums can be based on hospital beds, gallons of water in swimming pools, the square footage of restaurant dining areas…. Donald Shoup calls this a pseudoscience, like bloodletting.
Parking minimums is an expense that drivers do not have to pay. Land which is expensive for housing, commercial or other uses becomes free for parking. This expense is added to business costs which all consumers pay. Thus, non-driving consumers pay a subsidy to driving consumers.
Even worse, parking minimums mean that businesses must dedicate more space to parking than to the business itself. This pushes businesses further and further away from each other. Making it hard to walk from business to business, thus necessitating more driving and more car parking……
3. Managing the supply of existing road spaces
The choices city residents make around transport modes has a large effect on how congested a fixed quantity of road space becomes. Cities need to make strategic choices on what transport options are available and individuals need to face the full costs of those choices.
David Lupton — an economist and international transport consultant introduced me to the concept of congestion road pricing. David has long been an advocate of congestion road pricing. His interest dating back to personally meeting William Vickrey, who was an early congestion pricing economic theorist. William Vickrey had Georgist beliefs as do Harold Hotelling and Donald Shoup. Georgism dates from the 19th century writings of the economist and social reformer Henry George. The Georgist paradigm seeks solutions to social and ecological problems, based on principles of land rights and public finance which attempt to integrate economic efficiency with social justice.
Regarding congestion road pricing;
Hotelling pointed out that when local public goods like roads and trains become congested, users create an additional marginal cost of excluding others. Hotelling became an early advocate of Georgist congestion pricing and stated that the purpose of this unique type of toll fee was in no way to recoup investment costs but was instead a way of changing behavior and compensating those who are excluded.
Road pricing helps best allocate the supply of space set aside for transport corridors. Travelers have various transport options, each with a different demand for space, including taking alternative routes or delaying travel to non-peak/non-congested times.
Auckland is reaching its limits for its ability to increase road capacity for private automobiles cheaply. It needs a mechanism like road pricing to change behaviour and even up the playing field between private automobile use and other more space efficient transport modes.
In the mid-term future, new technology, such as electric vehicles, when taken up widely, will mean a decrease in fuel taxation revenue, which will encourage reform in the direction of electronic/GPS road user charges.
Frank McRae has also laid out a logical argument for why autonomous vehicles, should they become viable, will make road pricing necessary -in an article titled Driverless cars, sprawl and density.
An economist — Peter Nunns — who occasionally write for the Greater Auckland blog explains why road building costs in Auckland are escalating. He notes it is comparatively cheap to lay down highways on empty paddocks, and expensive to retrofit them into built-up urban areas or thread them through geographic pinch-points.
Going forward, Peter Nunns gives four options for new motorway corridors in Auckland:
- Dig a tunnel (eg Waterview);
- Build a bridge or viaduct (eg Reeves Road flyover);
- Reclaim some land (eg East West Connections); or
- Buy and bowl some houses.
All these options are very costly, as can be seen in the above graph showing increasing costs for additional lane-km’s in Auckland. This rising cost pressure is the reason why the allocation of road space by congestion charging needs to be considered.
David Lupton has a plan for the affordable development of cities, focusing on reforming New Zealand regulatory environment so that price signals lead to a better allocation of city space. It takes into account two of the spatial supply considerations — the supply of private space and how to better manage the supply of road space. I explain how this might work using Tokyo as example in a paper titled — Tokyo does not subsidise its transport system! New Zealand needs more of this sort of big picture strategic thinking so that it can benefit from its growing cities.
I would like to conclude my paper with the request that to better manage New Zealand’s urban areas there should be a strategic consideration of spatial economics for all the different types of urban space.
I hope this article encourages this process.
New Zealand is reaching its agricultural production limits. In the future New Zealand’s urban areas will need to do more of the country’s economic heavy lifting, by embracing an affordable diversified urban based economy.
It is my firm belief that the best way for New Zealand to successfully progress into the future is to conduct a well-managed urbanisation project.
Article significantly updated and edited 18.02.18
