Key message 2 – Contagion, crime, and congestion are important challenges, yet they are often poorly managed and understood

This brief highlights three downsides of density which affect all global cities – contagion, crime, and congestion. These sections will identify options for policymakers while outlining some of the biggest outstanding research questions.


Water and sewers are the most basic functions of urban government. While urbanisation without clean water and amidst refuse has, and does, occur, the health consequences are evident. The standard view is that the low life expectancy associated with city living prior to 1900 is largely a reflection of poor sanitation. The OECD has estimated that approximately 10% of the global burden of disease worldwide could be prevented with improvements to water, sanitation and hygiene, and better water resource management worldwide (OECD 2001). The larger issue is how to provide water and sewerage systems, and how much to spend. Even the most ardent believer in sewers might balk at spending $2,000 per household on waterworks in a country where per capita incomes are below $400 per year. The need for evaluation of spending and innovative implementation explains why economists, not just engineers and public health experts, need to play a role in researching sanitation investments.

Research agenda

• Infrastructure substitutes: The easiest and cheapest substitute for water infrastructure is the provision of chlorine tablets at the point of water consumption. Kremer et al (2009) have found remarkable results with low cost interventions in rural Kenya. It isn’t entirely obvious that this translates perfectly into urban areas, or that it is a substitute for sewers. Developing country cities would benefit from high quality research on randomised water related interventions that do not require expensive infrastructure.

Impact on health and the economy: If countries and aid agencies are going to make sensible decisions about how to invest in water, it seems crucial to quantify the size of the impacts on health and the wider economy. Does investing in water mains make sense if chlorine tablets are available, or should funding be targeted towards early childhood education? Without better estimates of the costs of poor water quality, we cannot answer that question.

The provision and funding of infrastructure: Should water infrastructure be built by a public entity or a public-private partnership? Today, about 7 percent of the urban population in the developing world is served by private operators, a model backed by the World Bank (Marin 2009). Yet doubts remain about the merits of this approach. No model is likely to be perfect in countries where institutional quality is weak and varies in form, but understanding the trade-offs between different institutional models is important.

Funding: In developed countries, water companies were assured of profits because they enjoyed a monopoly on water provision. Alternatives such as bore holes and wells were outlawed in many high density areas. Many developing countries have not addressed traditional methods of dealing with water and sewage. How does this change the appropriate institutional design for water provision?

Sewers and water pipes may not be the most glamorous aspects of urban life, but they are among the most important parts of a city. Economists have under-invested in sanitation research for decades, because these topics did not seem relevant to developed cities. As developing country cities grow, the need to revisit the battle for clean water becomes more urgent.


Crime has the capacity to yield direct welfare losses, but also to negatively impact investment activity, tourism and urban street life. Yet any serious analysis of crime in the developing world has to start with the recognition that city policing is far from reliable. The knee jerk reaction – hire more police and lock more people up in prison – may not be successful if the police are corrupt or ineffective.

An important consideration is that urban public safety has historically relied on private, not public, crime prevention, described by Jane Jacobs as the “eyes on the street” (Jacobs 1961). It is arguable that much of the differences in crime rates across the developing world have little to do with formal policing and much to do with neighbourhood crime prevention and the incentives to engage in criminal activity. Mumbai’s slums are much safer than Rio’s favelas, but not because the Mumbai constabulary is more effective than the Rio police. Some combination of neighbourhood cohesion and the returns to violence in the drug trade seems the likelier explanation.

The community’s role should be complemented with a public presence. Effective police forces often excel in leveraging the assets of the neighbourhood and getting residents to be their informants. For example, the Community Liaison Officer (CLO) system in Uganda seems to have strengthened links between the police and the communities they serve and encouraged NGOs to work on community safety and decreased domestic violence (Tindifa and Kiguli 1997). Given these potential outcomes, community policing and other police initiatives deserve more attention, as discussed below.

Research agenda

Data: In developed countries, researchers are typically comfortable using reported crimes, although they understand that under-reporting is a real issue. In the developing world, under-reporting is typically thought to be so enormous that victimisation surveys are crucial, unless the researchers focus only on the rare, but well-reported, crime of murder.

Corruption: Many countries attempt to reduce crime by merely increasing spending on policing. Yet the pervasiveness of corruption can frustrate any attempts to exact change. However, there are opportunities for other interventions. For example, improving the incentives for police to actually fight crime (or reduce the level of extortion by police reported in victimisation surveys) might have significant impacts on safety. In Rajasthan, Banerjee et al (2012) found that the promise of a transfer out of a less favoured assignment based on good performance more than doubled the number of drink drivers brought to court.

Factors that influence crime: There is a need to understand all of the factors which can influence crime, especially those that have nothing to do with policing. For example, there is long standing debate about the connection between unemployment and crime. Education, especially early childhood education, has also been found to reduce crime (Yoshikawa 1995). Even the introduction of debit cards as cash substitutes have been found to reduce the amount of criminal activity (Wright et al 2014).

Impact of crime: Finally, it is important to measure the benefits of spending public money to reduce the level of crime. What does crime do to economic activity in developing world cities? There are sophisticated studies of the link between crime and property values in wealthier cities, for example, Gibbons (2003) shows that a 10 percent reduction in recorded crime adds around 1.7 percent to the selling price of a home in London. However there are few equivalent studies in developing countries.


While contagion and crime affect most cities to some degree, congestion is a problem everywhere. London and New York have traffic jams that can be just as severe as those in Delhi and Lagos. Indeed, without proper management, traffic problems can easily get worse as incomes rise because a greater fraction of the population is able to afford motor vehicles.

The public sector always plays two significant roles in transportation, and usually takes on a third: 1) managing, or directly providing, transportation-related infrastructure; 2) implicitly or explicitly setting the rules of using that infrastructure, including any prices that are charged; and 3) usually but not always, providing public transit services that use that infrastructure.

Transportation infrastructure normally begins with a street plan, and the setting aside of some space as public thoroughfares. Such a plan costs little and has strong benefits. Paul Romer has argued that developing country cities should adopt grids and protect streets from private development even if the city is unprepared to invest in the infrastructure that surrounds the streets. Yet protecting a grid will be extremely difficult in many developing country cities. If the movement towards street planning becomes stronger, there will be a role for economists to play in evaluating different plans for protecting public spaces against private encroachment.

Infrastructure that goes beyond the grid becomes more debatable. The most basic step is paving for automotive and pedestrian paths. But how much space should be allocated for pedestrian use relative to cars? Developing country cities differ significantly in the pedestrian space that they have made available. For example, walking around the streets of Lusaka is quite easy, while walking around much of Mumbai is considerably more challenging. It may be tempting to argue that all dense cities need decent sidewalks, but since many do not have them, even research on such a basic question would have some value.

Beyond paving comes highways and then elevated or submerged transport links, including subways and bridges. These investments can be very expensive and it is important that we understand the benefits as well as the costs. Traditional analysis has focused on the direct benefits to riders. While still important, a new and exciting literature is looking at the ways in which transportation is shaping local economies. These papers typically compare cities with rail or highway linkages to cities without similar infrastructure. The big challenge is how to prove that transportation investments were not correlated with other factors that influence local economic success.

The impact of transport links on economic success

Two of the best papers that use metropolitan area differences in highway building are Baum-Snow (2007) and Duranton et al (2013). Baum-Snow (2007) convincingly shows that post-World War II highway construction in the United States led to increased suburbanisation and depopulation of the central city. This paper reminds us that there can be geographic losers from transportation investment. Valparaiso, Chile, for example, famously shrank hugely after the opening of the Panama Canal rerouted ocean traffic.

Duranton et al (2013) also makes an important point: vehicle miles travelled increase almost one-for-one with highway miles built. They are essentially confirming a behavioural response to new infrastructure that had long been suspected. If a country builds more highways, they will get more drivers. If the effect is literally one-for-one, then new infrastructure has done nothing to reduce congestion, but merely managed to put more drivers on the road.

Research agenda

Congestion pricing: With car-related transportation infrastructure, the point is to reduce congestion by providing more space. Yet the world’s traffic jams show that building more roads merely encourages drivers to drive more, given they aren’t being charged for the social costs of their actions. In response, the economist William Vickrey implemented congestion pricing, an intervention that may be easier when fewer people are used to driving for free. This suggests that it might be appropriate to begin doing experiments with congestion pricing in developing country cities. Naturally this will involve a host of issues surrounding enforcement which may be quite difficult in areas with weak institutions.

Buses: They are cheap and flexible, which is important in places where future development is unpredictable. Yet the economics literature on buses is quite limited. In many developing country cities, there is a debate between privately run jitneys versus public buses. There is also a question of appropriate subsidies.

Trains: Trains are the biggest and most expensive form of transportation. They can become a lifeline for a city, as they can provide the one means of getting around when streets become hyper-crowded. Yet Meyer et al (1965) argued that there wasn’t much that you could do with a train that you couldn’t do with a bus on a dedicated lane or tunnel, and the bus would cost much less.

Metro: Many developing world cities have invested significantly in metro systems. Yet rarely have the results been properly evaluated. Given the vast sums involved, this is also an area that could use more research.

Linking transportation infrastructure to wider objectives: What is the economic impact of investment in subways or highways? What is the health impact? What do different transportation systems do to the shape of urban society? While there is an abundance of hypotheses, there are few hard facts.