What does a good city look like, and how is it formed?


  • Professor Tony Venables (Director, IGC & University of Oxford)

  • Professor Sir Paul Collier (Director, IGC, & University of Oxford)

Prof. Venables began the session with a discussion about the features that make cities more productive and liveable. The question motivating his talk was understanding not simply the features that define successful cities, but also how good cities can engineer sustainable, pro-poor growth?

Working from the notion that cities can be high-productivity places, Prof. Venables went on to describe the dynamism and scale that defines particularly successful cities, but noted that the downside often associated with such cities was that they also tend to be much costlier places to live. Costs are driven in large part by the scarcity of land availability in densely packed urban areas.

He then went on to outline the three main types of built structures that characterise cities:

  • Commercial & Industrial structures which produce jobs and income
  • Residential structure which promote the wellbeing of households
  • Public infrastructure which ensure that public goods and services necessary for cities to function are widely and easily available.

The main constraining factor to these forces continues to be land-scarcity. This constraint can be relaxed only through effectively designed transport infrastructure. Without adequate compensation in the form of higher wages, or low-cost transport, urban environments can often produce vicious, not virtuous, circles of high costs and low wages which is considerably less efficient as an urban equilibria. The risks from higher costs can further exacerbate lower productivity and deter entry of newer firms into the market as a whole.

For most examples of successful & dense cities, the traditional urban form is most concentrated in the centre, but there are several notable examples of variation that emerge among large global cities. This is particularly true in cities developed without regard to normal efficient market principles such as Johannesburg, where the long-term impact of apartheid produced much more dispersed residential housing clusters. Brasilia & Moscow are both examples of largely government-constructed cities where residential areas were built too far out from the urban centre.

Prof. Venables went on to discuss the overall benefits that accrue in well-designed urban centres with particular emphasis on economic activities and productivity for workers. Cities are able to benefit from agglomeration effects, including knowledge spillovers and thicker labour markets that allow firms to readily access high-skilled workers. This also captures the advantages of specialising as a means of accessing higher-wage jobs. The other aspects of urbanisation that make such efficiency gains from cities possible are the features that make cities more liveable and low cost, primarily access to high-quality services such as public health and education provision, as well as well-designed, low cost transport infrastructure.

In concluding Prof. Venables urged researchers and policymakers to view cities as more holistic concepts. Moving out of vicious and into virtuous circles requires promotion of effective urbanisation processes that in turn invite greater influxes of high-productivity firms and provide financing for more effective urban infrastructure.


Following on from Prof. Venables session, Prof. Sir Paul Collier began by posing the question of how cities can move into more virtuous circles?

He went on to say that good cities act as engines of growth by creating productive jobs for ordinary people. The push to create more jobs, according to Prof. Sir Paul Collier, should form the basis of national urban policies and growth strategies. As migration brings more people into urban centres, a good development policy works to improve opportunities in both urban and rural areas.

Those who are capable, will move into newly emerging jobs in cities, leaving higher wages and more productive rural jobs for those who remain behind. This pattern, in turn, produces large urban markets into which agricultural workers can sell their produce. To this end, Prof. Sir Paul Collier argued that good urbanisation must not come at the expense of rural development.

In continuing, Prof. Sir Paul Collier identified the large young populations as a significant advantage of Sub-Saharan African economies. A large youthful workforce can better compete along dimensions that require workers to adopt and adapt to newer technologies and processes. In line with this thinking, cities are natural incubators of young, innovative technology-driven sectors, and simultaneously, they are magnets for flows of young labour market entrants. Efficient market structures will then, according to Prof. Sir Paul Collier, produce firm clusters, characterised by different kinds of labour specialisation.

In addition to the firm clusters, Prof. Sir Collier also observed that successful cities produce clusters of consumers that can serve as ready-made demand for local firms that do not produce internationally tradable goods. This again emphasises the need for efficient urban densities to trigger demands of scale and specialisation to facilitate efficient markets.

In exploring the factors that produce effective cities and ultimately drive virtuous circles, Prof. Sir Paul Collier went on to describe the central components of urban connectivity as (1) affordable transport infrastructure and (2) liveable density. These factors support and drive substantial industrial development by creating conditions for middle manufacturing firms to special their production and participate actively in global value chains.

Specialised production in particular is essential for reaping the rewards of clustering and agglomeration generated by dense urban environments. Liveability requires urban centres to produce sufficient levels of services such as health care and affordable housing. In his recommendation for African cities, Prof. Sir Paul Collier argued that the most efficient design of housing for developing cities should be apartment blocks typically built at a maximum height of five stories, 5 being the most efficient number of floors for residential housing that does not require the costs of elevator technology.

In conclusion, Prof. Sir Paul Collier proposed two questions to guide thinking on the process of urban planning and policy design. First, identifying primary decision makers that will determine the essential questions of the urban policy space? This will distribute clear responsibility for policy elements and decision-making to a specified set of decision makers. Second, even more primary is the need to determine who will appoint these decision makers in the first instance.

Summary written by Upaasna Kaul, Managing Editor & Hub Economist