If agglomeration economies require urban densification why are Houston, Paris and Atlanta so successful?

When we look at strategies followed by the UK’s metropolitan mayors, we see plans that take the polycentric urban areasand seek to impose a single centre model, the opposite of the successful model

Agglomeration theory has always bugged me. Ever since I first encountered it doing my masters degree, the idea that economic and technical advancement is determined, in large part, by the density of urban environments has seemed like the urbanist version of the labour theory of value - superficially appealing but in the end not even close to the whole story.

There’s a very good essay written by John Myers, Ben Southwood and Sam Bowman called ‘The Housing Theory of Everything’ which posits that the failure to build enough housing exacerbates all the other malaises of western economies:

“Try listing every problem the Western world has at the moment. Along with Covid, you might include slow growth, climate change, poor health, financial instability, economic inequality, and falling fertility. These longer-term trends contribute to a sense of malaise that many of us feel about our societies. They may seem loosely related, but there is one big thing that makes them all worse. That thing is a shortage of housing: too few homes being built where people want to live.”

As readers will know, I have a load of sympathy with the idea that many of the economic and social problems facing western societies would be mitigated if not eliminated by having genuinely affordable housing (without the need for price fixing or public subsidy). But I don’t see overpriced housing being, in and of itself, the cause of a parallel problem, poor productivity. And it is productivity that is the principle (perhaps the only) gain from urban agglomeration - we get from proximity what are called agglomeration economies. These come from three sources usually called pooling, matching and learning: labour and supply is pooled, that labour and supply is better matched to opportunity, and the concentration of people results in organised and serendipitous learning. Some add a fourth source of agglomeration economy - greater amenity (or the ‘consumer city’ as we might call it).

All of this makes sense until we spot that outcomes are not consistent with this theory since the economic performance of places does not map simply onto levels of urban density. The Federal Reserve Bank of New York commissioned some research looking at the extent to which productivity is linked to the density of human capital - one of the fundamental assumptions of agglomeration theory. The research found that there was no correlation between density and productivity, at least in aggregate.

When they dug down into the data, it became clear that, while for lots of industries there were no obvious agglomeration economies, for some the gains were very large. And these industries - finance, arts & entertainment, professional services and information - are the industries we most associate with successful dense urban places.

Agglomeration, it seems, isn’t quite the silver bullet that city boosters argue, indeed there’s a school of thought which says the only thing that matters is the gross size of the metropolitan area rather than the intensity of human capital within that urban area:

“The shape of the city in space, including for example its residential density, matter much less than (and are mostly accounted for by) population size in predicting indicators of urban performance. Said more explicitly, whether a city looks more like New York or Boston or instead like Los Angeles or Atlanta has a vanishing effect in predicting its socio-economic performance.”

We know that places that have allowed cities to expand (Houston, Paris, Atlanta) have lower housing costs than places where expansion has been constrained by public policy (Sydney, London, San Francisco). We also know that there is no obvious connection between encouraging density through policy - densification has, in effect, been UK government policy since 1997 - and improvements in productivity. Indeed, the use of policy to densify human capital results instead in an increased competition between potential land uses within the city with the effect of driving up not only residential rents but also rents for commercial, industrial and leisure uses.

Given that urban density in and of itself isn’t the main reason for agglomeration economies and the associated evidence that constraining land use at the urban margin results in higher housing costs, we should be cautious in seeing policy ideas such as street votes as any sort of panacea for unaffordable housing. Nor should we focus more on new public transport infrastructure than we do on roads and active travel. For cities with an established and scaled mass transit system (London, New York, Tokyo) further intensification of public transport makes sense, but for large cities without these legacy systems the construction of new fixed rail introduces a measure of inflexibility and is extremely expensive. Moreover the nature of large urban areas without legacy metro networks is very polycentric.

A large urban metropolis like Los Angeles, developed largely during the age of the car and, therefore, focused on freeways and personal vehicles, has evolved with a multitude of urban centres. One study of LA concluded that:

“The interplay of agglomeration at different geographic levels suggests a highly complex and connected space economy. Large metropolitan areas have the advantage of scale, offering a highly diverse and specialized labor force, dense networks of intermediate suppliers, and extensive transport and communication systems. The presence of multiple centers likely lowers the land, wage, and congestion costs of agglomeration by geographically spreading economic activity while still preserving agglomeration benefits.”

When we look at the strategies followed by the UK’s metropolitan mayors, we see plans that take the polycentric urban areas of the West Midlands, Greater Manchester and West Yorkshire and seek to impose a single centre model, the complete opposite of the successful model evolving in more productive places such as Los Angeles, Atlanta and Houston. Instead groups like Centre for Cities select very dense European cities like Milan or Barcelona as models resulting in a focus on very expensive public transport systems rather than less expensive private road travel.

The reasoning behind this rejection of car-permissive urban strategies relates to the idea that private transport is inefficient and polluting whereas public transport (or active travel options such as cycling) is efficient and clean. The problem here is that the motor car is such a dominant part of our transport system that no substitute has the capacity to replace the car. We can (and should) encourage alternatives, especially in denser urban areas, and ideas like 15-minute cities have merit as a design principle rather than as an urban management strategy. Where there is no fixed rail mass transit system, installing one is disruptive, expensive and, if Manchester is a guide, not especially effective. Such systems seem more a badge of urban success than a creator of that urban success in the first place. Lots of French cities have trams but this isn’t necessarily a reason for those cities’ success (or indeed, as a place like Roubaix shows, lack of economic success).

Returning to familiar themes, Sam Bowman in his latest essay argues (rightly) that Britain needs to ends the vetocracy that is preventing economic growth and development - everything from too few homes in London through the inability to build a bridge over the Thames estuary to the chronic and crippling lack of laboratory space in and around Cambridge. I worry, however, that the focus on agglomeration leads to misplaced urban strategy:

“The big four, as I see them, are housing, childcare, transport and energy. The first three of these relate to “agglomeration” – the aggregate increases in productivity and innovation we get from letting people live and work near each other. That’s the root principle behind the “housing theory of everything” – a city is greater than the sum of its residents, because being close together lets them specialize and collaborate more deeply, and generate more useful ideas together, than if they were working in smaller towns, or alone.”

It isn’t the agglomeration economies per se but rather the assumption, as we see in Bowman’s essay, that the way to realise these economies is accelerated urban densification. This is despite Bowman telling us about Paris:

“Housing supply there is freer: the overall geographic extent of Paris’s metropolitan area roughly tripled between 1945 and today, whereas London’s has grown only a few percent”

The success of Paris, at least up to recent times, has been in its physical expansion - it is far more like Houston or Atlanta than London or San Francisco. And, while Britain is geographically smaller than France or Texas, this doesn’t mean our only option to meet housing needs and address declining productivity is to increase levels of urban density.

Bowman is right that the best way to deal with NIMBYs is to buy them off but, for all that I support its introduction, I don’t see the street votes idea working at scale. I do see a neighbourhood plan model working where, as Pierre Poilievre, Canada’s opposition leader proposes, national government funding for any community is predicated on the delivery of new housing. Communities can stop housing if they want but they lose government grants for their communities as a result. Just as important, this approach takes away the ‘no loss’ benefit from anti-housing campaigners. Somebody might get elected opposing new housing but in doing so they are also going to stop funding for the new sports centre, health facility, road link or bus station.

Urban agglomeration theory, especially the Jane Jacobs version where random encounters in the street or at a coffee shop create economic growth, is not a function of densification but a function of  urban scale and the ability of residents to move freely around that metropolis. Transport and housing policies should focus on this ability to move from home to work, leisure and education rather than get trapped in prescriptive models based on assumptions about the impact of densification. We should, I feel, think more about enabling development rather than about controlling, managing or directing development. The fastest growing US cities, in economic terms as well as in terms of population, are not the dense urban legacy cities like San Francisco and New York but rather the sprawling new places in the Sun Belt and Texas. Maybe, instead of looking at Milan, Barcelona or Tokyo, UK economic planners should turn their eyes to Houston, Nashville and Atlanta?