Bigger cities are more productive but higher cost: what policy could do but doesn’t

by Professor Paul Cheshire

22 Feb 2016

Cities are founded on specialisation. They were ‘discovered’ about 14,000 years ago and are arguably humanity’s most important invention. Why did we invent the wheel?  Probably because the productivity advantage those early cities generated led to the need and ability to pay for food from ever further afield. The productive city-dwellers created not only a need – to transport food further and in larger quantities – but the resources to pay for it.

This strength of cities – specialisation – persists and except for military defence – the advantages and specialisms of cities are still today as they emerged millennia ago: specialised services, artisans and skilled manufacturing, cultural activities and industries, specialised retail and government and administration.

Cities, however, are the constructs of people and they are about people: systems to produce welfare and increase productivity. So buildings, design and infrastructure are a means to an end, not the end itself. And enhancing the built environment cannot promote prosperity or happiness in itself – but can if it improves people’s lives or productivity and is critical in helping cities deliver welfare and productivity more effectively. Yet, while cities are economic & social constructs, policy is dominated by ‘design’ & ‘engineering’ modes of thought. Architecture and urban design have been the intellectual traditions urban policy has drawn from rather than urban economics or sociology. But urban economics for certain has made great progress in the past 15 years or so and has valuable things to say relevant to urban policy.

As cities get bigger their capacity to increase productivity and deliver welfare increases. Economists call this phenomenon ‘agglomeration economies’.  Over the past 15 years or so we have succeeded in generating quite credible estimates and it seems as if doubling a city’s size produces about a 5% increase in total factor productivity (TFP), holding everything else constant. This implies that going from a city the size of Bradford or Cardiff to one the size of Leeds would increase TFP all else equal  by 5 or 6% and going from a city the size of Leeds to that of London by 18%.

There are also direct consumption benefits of city-size: bigger cities, all else equal, improve the quality of people’s lives. There is a wider range of goods and services available and more neighbourhoods of different character. One of the major sources of satisfaction is one’s relationships with neighbours and the larger a city is, the more varied are its neighbourhoods and the easier it is to find congenial neighbours. Another obvious advantage of larger cities is that many activities depend on audiences, so the larger a city is, the more specialised its venues can be. To take football as an example (but it could equally be opera, classical music, or art) in a city as large as London a football fan can easily choose to watch any one of four or five world class teams and expect to see the greatest stars in the world playing when they visit. This is only possible because of the size of the audience.   For these consumption benefits of larger cities, however, there are not yet any really reliable quantitative estimates.

The problem is, however, that costs also rise as cities get larger. Partly because of agglomeration economies more firms and people bid to get access to the higher incomes, revenues and welfare bigger cities generate, so the price of space rises. So, too, do other costs such as congestion and pollution. Until very recently people assumed that these increasing costs just ended up cancelling out the agglomeration economies and cities stopped growing; even got too big.

Recent research by a French team is the first to provide a serious quantitative estimate of how these costs increase with size and what particular features of cities drive the rise in costs. It finds that while, with a fixed land supply, costs do rise at about the same rate as agglomeration benefits in productivity, if land supply is elastic, the rate of increase of costs with size is only about 0.4 times that of the benefits.

The clear message for urban policy is to relax constraints on land supply subject to possible environmental costs. Alternatively, that the increase in price of farmland at the fringe of British cities that would result from allowing it to convert to housing is a measure of foregone agglomeration economies. Getting permission to build houses on a hectare of farmland on the northern fringes of London would increase its price from perhaps £20,000 to £12m: a clear signal of the loss of value imposed by not letting more people live where they really want to. Letting more people live where they find life most attractive and productive we could reap the benefits of bigger cities but pay far less in terms of costs. Cities can be both bigger and better; and better by being bigger.

The two other types of cost that rise with city size are those of congestion and pollution. Again urban economics has a clear policy message. We should impose a charge on congestion that reflects the extra costs any journey causes to others as accurately as possible. The London Congestion Charge is better than nothing but does not do the job as effectively as it might since it is not closely related to the congestion costs any given journey creates and, as a cordon charge, in fact generates an incentive once you have paid the charge to use your vehicle.

Pollution is, like congestion, an externality; the price any individual pays does not reflect the costs imposed on others by their actions but pollution is less obviously amenable to pricing. Instead regulation can be very effective but often needs to be at well above the city level both because the wind carries pollutants across administrative boundaries and because imposing technical changes on, for example farming methods or vehicles, requires action at the national or international level. It may seem odd to include farming in this account of urban policy but the most recent findings are that emissions from intensive agriculture – mainly of ammonia – are the single most important cause of premature death for city dwellers in Western Europe from air-borne particulates.

Paul Cheshire is Emeritus Professor of Economic Geography at the London School for Economics and Political Science. He is co-author of Urban Economics and Urban Policy: Challenging Conventional Policy Wisdom, Edward Elgar Publishing, 2014.

Sign up to our email newsletters