Cities throughout history have been closely linked to armed conflict. As the locations of elites and power, they have attracted invaders and those seeking to wrest power and wealth from their incumbents. Cities have also been places of safety during times of war – think of the ramparts of ancient Middle Eastern and European cities, designed to keep invaders out and those inside safe. Cities can also concentrate conflict, leading to higher numbers of deaths than in other areas. Research by the International Committee of the Red Cross (ICRC) has found that, between March 2017 and July 2018, in four governorates of Iraq and Syria ‘Urban offensives account[ed] for eight times more conflict-related civilian fatalities …than ongoing fighting or fighting in other areas’.+ICRC, New Research Shows Urban Warfare Eight Times More Deadly for Civilians in Syria and Iraq (Geneva: ICRC, 2018) (www.icrc.org/en/document/new-research-shows-urban-warfare-eight-times-more-deadly-civilians-syria-iraq).
This section defines armed conflict, discusses conflict and critical urban infrastructure, outlines the factors and choices influencing interventions in urban armed conflict settings and identifies issues relating to working with local actors.
Defining armed conflict
Armed conflict can be defined as ‘a political conflict in which armed combat involves the armed forces of at least one state (or one or more armed factions seeking to gain control of all or part of the state), and in which at least 1,000 people have been killed by the fighting during the course of the conflict’.+Project Ploughshares, ‘Defining Armed Conflict’, undated (http://ploughshares.ca/armed-conflict/defining-armed-conflict/). International humanitarian law (IHL) recognises two types of armed conflict:+ICRC, How Is the Term ‘Armed Conflict’ Defined in International Humanitarian Law?, ICRC Opinion Paper, March 2008 (www.icrc.org/en/doc/assets/files/other/opinion-paper-armed-conflict.pdf). international armed conflict between two or more states; and non-international armed conflict between government forces and non-governmental armed groups.
Following decades of decline in the number and scale of conflicts around the world, since 2011 that trend has reversed.+M. Marshall and G. Elzinga-Marshall, Conflict, Governance, and State Fragility Global Report (Vienna, VA: Center for Systemic Peace, 2017) ( www.systemicpeace.org/vlibrary/GlobalReport2017.pdf). By 2017, there were 27 ongoing major armed conflicts; 52 new conflicts emerged between 2011 and 2017, 46 of which were in the Middle East. Conflict is increasingly affecting civilians. According to the Internal Displacement Monitoring Centre (IDMC), for example, ‘The number of new displacements associated with conflict and violence almost doubled, from 6.9 million in 2016 to 11.8 million in 2017. Syria, DRC and Iraq together accounted for more than half of the global figure’.+IDMC, Global Report on Internal Displacement (Geneva: IDMC, 2018) (www.internal-displacement.org/global-report/grid2018/). According to IDMC estimates, 40 million people were internally displaced by conflict at the end 2017.
Cities and conflict
The ICRC identifies a number of underlying vulnerabilities which make urban contexts particularly prone to conflict. They include:
As the ICRC notes, ‘The complex, interconnected systems that provide water, electrical and sanitation services essential to urban health are often among the first to fall victim to urban warfare’.+ICRC, ‘I Saw My City Die’: Voices from the Front Lines of Urban Conflict in Iraq, Syria and Yemen (Geneva: ICRC, 2017) (www.icrc.org/en/document/iraq-syria-and-yemen-five-times-more-civilians-die-city-offensives-new-report-finds), p. 18. Figure 1.3 identifies critical urban services that may be at risk during armed conflict.
The impact of conflict on urban services can be direct (such as infrastructure being targeted), indirect (where services are damaged by conflict targeted elsewhere) and cumulative (effects over time; for instance, services may worsen or degrade). Table 1.1 identifies each of these impacts in relation to critical individuals, hardware, consumables and the general public.
|Type of impact||Impact on critical people||Impact on critical hardware||Impact on critical consumables||Impact experienced by the general public|
|Direct||Casualties, restricted access due to security situation; drafting into armed forces; displacement||Destruction of or damage to instruction and/or equipment||Destruction of fuel reservoirs; destruction of stocks of chlorine; shortages due to looting||Brief interruptions in access to, reliability or quality of service; considerable public health risks|
|Indirect||“Brain drain”; retirement without replacement; no salary payments||Drops in pressure in water networks; disrepair of unused or misused equipment; negative coping mechanisms||Shortages (due to looting and/or lack of replacement); price increases on the black market||Continuous or persistent deterioration of access to, reliability or quality of service; considerable public health risks|
|Cumulative||Little to no long-term planning; loss of knowledge of system||Silting of reservoirs; leaks and increase in “non-revenue” water (unlicensed connections); mismatch of replaced items||Depletion of contingency stocks||Adaptation to poor reliability of service delivery, primarily through development of coping mechanisms; public health risks as a function of many other issues|
Source: ICRC, Urban Services during Protracted Armed Conflict.
Source: ICRC, Urban Services during Protracted Armed Conflict: A Call for a Better Approach to Assisting Affected People (Geneva: ICRC, 2015) (www.icrc.org/sites/default/files/topic/file_plus_list/4249_urban_services_during_protracted_armed_conflict.pdf). A video produced by the ICRC exploring critical urban services can be found at www.youtube.com/watch?v=eVCkzbZiYRw.
Direct impact can lead to cumulative impact. For example, in relation to public health, a relatively short disruption to a supply of drinking water (for a day or a few days) ‘can greatly increase the probability of infection from diseases already present in the environment if the quality of the service was originally reliable’.+H. L. Risebro et al., ‘Contaminated Small Drinking Water Supplies and Risk of Infectious Intestinal Disease: A Prospective Cohort Study’, PLoS One Special Issue, 2012, citing M. A. Al-Ghamdi et al., ‘Environmental Risk Factors for Diarrhoea among Male Schoolchildren in Jeddah City, Saudi Arabia’, Journal of Water and Health 7, 2009. Disease risks may increase if conflict drives more people and more disease strains into urban areas, ‘particularly when the movements of people are coupled with poor general treatment of water and wastewater, and incomplete immunization campaigns’.+Ibid.
Interventions in urban conflict
Given the nature of armed conflict, where infrastructure, people, services and humanitarian aid providers themselves may well be targets, the factors and choices influencing aid interventions are more complex than in non-conflict settings. Reflecting on humanitarian action in Syria, Francois Grünewald argues that:
To effectively support conflict-affected people in Syrian cities means working in very complex, volatile, unpredictable and dangerous environments. Humanitarian agencies have to be agile, flexible, opportunistic and risk-taking. Classic operational modalities imposed by donor procedures or guidelines, which require lengthy planning, standardised operational modalities and sophisticated accountability mechanisms, are of limited use in these highly volatile and complex urban contexts.+F. Grünewald, ‘Cities in Conflict: The Lessons of Syria’, Humanitarian Exchange 59, 2013 (https://odihpn.org/wp-content/uploads/2013/11/HE_59_web.pdf).
The ICRC recommends that, in urban conflict operations, the following factors need to be considered in determining what may be achievable in an urban armed conflict setting:
Working with local actors
Grünewald’s review of working with local actors in urban Syria underscores the need for inclusion, clarity of purpose and the ability to negotiate:
Effective humanitarian response in conflict-affected urban areas in Syria requires a capacity to engage in strategic dialogue, firmly rooted in humanitarian principles, with a wide range of actors, including the government, political/religious factions and associated armed militias and what remains of municipal institutions. Such negotiations demand language and negotiation skills, a thorough understanding of both the urban and underlying socio-political context, the networks to facilitate the necessary connections and a willingness to accept relatively high levels of risk.+Ibid.
Due to the complexity of many urban areas, exacerbated during armed conflict, the relationships implementing organisations have with key actors are highly important (see also Section 1.5 on urban actors and 2.1 on coordination). Key points include:
An initiative in conflict-affected Tamkeen in Aleppo (which at the time was in a non-government-held area) supports local actors to prioritise and manage local services. The project involves participatory activities geared towards setting up local committees that manage budgets and plans for service delivery such as health, education, food and infrastructure. A review of the project concluded that it illustrates a way to engage with local governance bodies using remote management, where priority interventions were outlined by the communities themselves. Here, most of the communities opted for basic services, infrastructure and development rather than humanitarian aid in the form of food, shelter or emergency health services.
Source: S. Dadu-Brown and A. Dadu, Strengthening Local Councils to Bridge the Aid Gap in Aleppo, IIED Briefing, January 2018 (http://pubs.iied.org/pdfs/17449IIED.pdf).
Violence can be defined as ‘the intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community, that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment, or deprivation’.+Violence Prevention Alliance, ‘Definition and Typology of Violence’ (www.who.int/violenceprevention/approach/definition/en/). Violence has traditionally not been considered a humanitarian priority, with the focus traditionally on the consequences of disasters and conflict. However, given the scale of destruction in a number of cities around the world, and given the humanitarian mandate to help people affected by acute shocks, an increasing number of humanitarian agencies, donors and think tanks are tackling urban violence.
ICRC’s authoritative 2017 report I Saw My City Die, concerning armed conflict in Iraq, Syria and Yemen, provides ten recommendations, of which the following are relevant to actors responding to emergencies:
• Parties to the conflict and the international community should refrain from displacing people and respect the rights and address the needs of those displaced within their countries.
• Authorities and the international community should protect and assist refugees from these conflicts.
• Authorities, parties to the conflict and the international community should do much more to ensure that essential service providers and humanitarian workers are protected.
• Authorities, humanitarian actors and the international community should invest more in ensuring that victims of violence have access to appropriate psychosocial and mental health support services.
• Authorities, humanitarian actors and the international community should help rebuild communities, not just infrastructure. Cities are made up of people, not just buildings.
These recommendations point to the need for humanitarian practitioners, among other things, to retain where feasible a neighbourhood-based approach (discussed in Section 3.2); prioritise protection both for affected populations and service providers (see Section 4.7); and prioritise a people-centred approach.
Source: ICRC, ‘I Saw My City Die’: Voices from the Front Lines of Urban Conflict in Iraq, Syria and Yemen (Geneva: ICRC, 2017) (www.icrc.org/en/document/iraq-syria-and-yemen-five-times-more-civilians-die-city-offensives-new-report-finds). ICRC also maintains a valuable website, ‘War in Cities’, which can be found at https://www.icrc.org/en/war-in-cities.
This section describes the scale of the issue. It presents the drivers of urban violence, including interpersonal and gender-based violence (GBV), discusses gangs and violence and explores the relationship between violence and naturally-triggered disasters. The section includes an abridged version of an article describing humanitarian action on urban violence in Central American cities.
The scale of the issue
The consequences of violence are enormous. More people die from violence (globally some 4,200 every day) than from naturally-triggered disasters or conflict. According to one estimate, violence kills around half a million people a year.+ICRC, Urban Violence and the ICRC’s Humanitarian Response (Geneva: ICRC, 2016) (www.icrc.org/en/document/urban-violence-and-icrc-humanitarian-response). Between 2007 and 2012, almost 40,000 drug-related deaths were reported in Mexico.+E. Ferris, ‘Protecting People in Cities: The Disturbing Case of Haiti’, Disasters 36(1), July (https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-7717.2012.01285.x). Violence and crime is thought to cost Latin American and Caribbean countries between 2.4% and 3.6% of their GDP. Urban violence in Latin America displaces more people than declared conflicts and war.
The impact of violence on urban life has prompted a number of agencies to consider it within humanitarian practice. The ICRC states that ‘the destructive force of urban violence on people’s lives and livelihoods – and the suffering it causes – is a major concern of the ICRC in many contexts in which it works around the world. This violence – often symptomatic of socio-economic pressures linked to rapid urbanization, soaring population growth and large population movements – will be one of the defining features, and key challenges, of the twenty-first century’.+ICRC, Urban Violence and the ICRC’s Humanitarian Response (Geneva: ICRC, 2016) (https://www.icrc.org/en/document/urban-violence-and-icrc-humanitarian-response). Similarly, Médecins Sans Frontières (MSF) has expanded its work in what have been termed ‘other situations of violence’. The 2006 Geneva Declaration acknowledged that armed violence ‘prevents humanitarian assistance from reaching people in need’, and commits signatories to ‘integrate armed violence reduction programmes into humanitarian assistance, emergency and crisis management initiatives’. Some researchers have equated urban violence with an emerging form of warfare, where ‘Chronically violent cities … [comprise] a “new” kind of armed conflict with grave implications for humanitarian action and human welfare. These and other urban centres are experiencing a variation of warfare, often in densely populated slums and shantytowns’.+R. Muggah and K. Savage, ‘Urban Violence and Humanitarian Action: Engaging the Fragile City’, Journal of Humanitarian Assistance, January 2012 (https://sites.tufts.edu/jha/archives/1524).
Drivers of urban violence
Reviewing the drivers of extreme urban violence, one investigation+J. DeBoer, R. Muggah and R. Patel, Conceptualizing City Fragility and Resilience (Tokyo: UNU, 2016) (https://cpr.unu.edu/conceptualizing-city-fragility-and-resilience.html). identifies a collection of risk factors, including income and social inequality, rapid urban growth with poor development and access to basic services, an unemployed youth bulge and weak police and justice mechanisms that allow urban violence to prosper. Other research+C. Moser and D. Rodgers, Understanding the Tipping Point of Conflict, Working Paper 7 (Manchester: University of Manchester, 2012) (http://siteresources.worldbank.org/EXTGOVACC/Resources/GlobalPolicyReport.pdf). analyses how violence in urban areas spreads and tips over into wider city-level violence. The same investigation found three key areas for stemming the spread of violence: improving low-income informal settlements, strengthening governance structures and reducing inequality. Social cohesion and building social networks, inclusion and citizen participation, alongside social protection programmes, were also key opportunities for reducing violence.
An ALNAP paper+E. Lucchi, Humanitarian Interventions in Situations of Urban Violence, ALNAP Lessons Paper (London: ALNAP, 2013) (www.alnap.org/help-library/alnap-lessons-paper-humanitarian-interventions-in-settings-of-urban-violence). on urban violence identifies the following urban-specific challenges:
A number of other research initiatives in this area are under way. The Igarapé Institute’s project on Humanitarian Action in Situations Other than War (HASOW) ‘has as its central aim the empirical examination of the dynamics of urban violence and the changing face of humanitarian action’. HASOW comprises research on a number of themes and issues, including the impacts of violence on health and organised gangs and violence. HASOW can be found at https://igarape.org.br/en/hasow/.
Box 1.3 is an abridged version of an article by Robert Muggah, co-founder of the Igarapé Institute. It discusses the scale of urban violence in Central American cities, and the humanitarian activities currently under way in response by a number of agencies. Finally, aid organisations have typically started small, built to scale and then handed over their pilots to government or local non-governmental counterparts. Notwithstanding the temptation to undertake large-scale programmes in fragile cities, relief organisations are proceeding with caution. There are meaningful ways to scale up city-based interventions, but only if these are properly aligned with formal and informal delivery providers, with stable resourcing and political investment. To be effective, aid agencies need to keep an open mind, take risks and invest heavily in partnerships from the start.
The countries and cities of the so-called Northern Triangle of Central America – El Salvador, Guatemala and Honduras – have some of the world’s highest rates of violent deaths, with murder rates exceeding those of Afghanistan or Syria. The violence in Central America is propelled by a volatile combination of transnational gangs, drug-trafficking and weak law enforcement. Rival factions run extortion rackets and assassins for hire, and recruit heavily from poorer neighbourhoods and shanty-towns throughout the region. Most gangs are involved in extortion, protection rackets and drug transhipment and retail.
Some humanitarian agencies have launched interventions in the region. For more than half a decade, the ICRC has been quietly testing new programmes to protect civilians and facilitate better access to basic services in San Salvador, Tegucigalpa in Honduras and Ciudad Juarez in Mexico, as well as Rio de Janeiro and Medellin. MSF has also initiated violence prevention and mental health-related activities and projects to address at-risk youth, including women and girls, in inner-city neighbourhoods across Central America.
The decision whether and how to deploy humanitarian assistance is not straightforward. Many agencies and donors are struggling with how best to negotiate with municipal authorities and communities and engage productively with complex and interconnected urban infrastructures. Most directors of humanitarian organisations first ask very basic questions, including in relation to the extent of their own competencies in cities under fire. What is the organisation’s added value? Will it make a real difference on the ground? Is it safe for staff? What are the legal implications?
Humanitarian agencies that run violence prevention and emergency response programmes tend to be guided by a set of basic principles. These include being clear on the aims of the intervention, being flexible and ready to adapt, adopting highly localised interventions in partnership with civic authorities, developing strong community partnerships, planning for the long term (while also having an exit strategy) and doing no intentional harm. Agencies are taking advantage of lessons learned in war zones, but also adjusting and adapting them.
Many of the priorities of humanitarian agencies remain the same in war and non-war zones. The focus continues to be on protecting civilians and civilian assets, mitigating the effects of violence on urban populations and enabling or strengthening protective factors that limit exposure to violence. This includes investing in early childhood programmes, school-based activities, initiatives for single female-headed households, projects targeting at-risk adolescents, psycho-social support services and urban improvement schemes.
Another key goal is to supplement – rather than replace – services such as water provision, waste management and health and education. Aid agencies such as the ICRC and MSF have found it imperative to work with government institutions, rather than around them, with an emphasis more on coordination than implementation. Although there is more sensitivity today to the importance of building local capacity and ownership, working with national partners and avoiding the distortion of domestic markets is difficult. For aid agencies used to rapidly delivering aid, setting up logistics systems and working around (reluctant or interfering) state agencies, habits take time to change.
An additional critical lesson emerging from the field is the importance of high-quality data collection and real-time mapping of rapidly changing conditions on the ground. Access to a wide range of high-resolution information on beneficiary populations, service delivery systems and existing organisations and actors is critical. Even in data-scarce environments there are opportunities to harvest and analyse information, including using new technologies. Humanitarian agencies are strongly advised to build this capacity in-house.
Source: R. Muggah, ‘A Humanitarian Response to Central America’s Fragile Cities’, Humanitarian Exchange 69, June 2017 (https://odihpn.org/magazine/the-humanitarian-consequences-of-violence-in-central-america/).
Gangs are particularly urban phenomena, in part because of the concentration of population – it is simply easier to form gangs in urban settings than in rural areas – but also because economic resources are concentrated in cities. Gangs may be exceptionally powerful: in Haiti they have been ‘credited with overthrowing governments, silencing the political opposition, preventing foreign and local investment, creating a nascent kidnapping industry and terrorizing entire cities’.+A. R. Kolbe, Revisiting Haiti’s Gangs and Organized Violence, HiCN Working Paper 147, 2013, cited in J. M. Hagedorn (ed.), Gangs in the Global City: Alternatives to Traditional Criminology (Chicago, IL: University of Illinois Press, 2007), cited in A. Winton, ‘Gangs in Global Perspective’, Environment and Urbanization 26(2), 2014 (https://journals.sagepub.com/doi/abs/10.1177/0956247814544572). Many also serve as a social network providing material and non-material support for youth and groups not well served by public systems. Such associations can improve the short-term welfare and resilience of their members, albeit with a long-term cost, to them and to society more widely.
The nature and changing shape of urban gangs is a complex issue and area of understanding. Research on ‘gangs in the global city’+J. M. Hagedorn (ed.), Gangs in the Global City, cited in A. Winton, ‘Gangs in Global Perspective’, Environment and Urbanization 26(2), 2014 (https://journals.sagepub.com/doi/abs/10.1177/0956247814544572). notes that, while most gangs are unsupervised teenage peer groups, many are institutionalised in ghettos, barrios and favelas across the world. In addition, ‘Gangs are “social actors” whose identities are formed by ethnic, racial and/or religious oppression [and] through participation in the underground economy’.+Winton, ‘Gangs in Global Perspective’, p. 406.
The word ‘gang’ may be a catch-all term for a large and complex array of organised violent groups, with a wide range in terms of size, power, age (youth gangs are a defined grouping) and scope. In urban areas, research into gangs has found that they are often closely associated with the neighbourhoods they operate in. Research on urban gangs in Haiti found that:
Gangs are key urban actors and are part of the governance structures of many urban areas, in particular low-income neighbourhoods and settlements. ICRC identifies three key points for engagement with gangs. First, it is important to gain acceptance from gang leaders as gangs are often organised hierarchically; second, gang leaders need to benefit from humanitarian interventions; and third, relief efforts which undermine a gang’s position may create risks.
Interpersonal and gender-based violence
Interpersonal violence includes sexual abuse, child abuse, GBV, violence against the elderly and self-harm. One report notes that ‘every day, more than 4,000 people, over 90% of them in low- and middle-income countries, die because of violence. Of those killed, approximately 2,300 die by their own hand and over 1,500 because of injuries inflicted by another person’.+A. Butchart et al., Preventing Violence and Reducing Its Impact: How Development Agencies Can Help (Geneva: WHO, 2008) (http://apps.who.int/iris/bitstream/handle/10665/43876/9789241596589_eng pdf;jsessionid=58B8188AD0AA79563256E8ACA6F9D308?sequence=1), p. 1.
Research has found that GBV in low-income settlements in low- and middle-income countries is frequent and acute.+S. Chant, ‘Cities through a “Gender Lens”: A Golden “Urban Age” for Women in the Global South?’, Environment and Urbanization, 25(1), 2013 (https://journals.sagepub.com/doi/abs/10.1177/0956247813477809). ActionAid’s Women in the City report identifies poor urban development and infrastructure as drivers of violence and insecurity for women and girls, with a lack of safe public spaces and public transport and unreliable recourse to police or the authorities. GBV in these circumstances includes physical and sexual violence from partners and violence against children. Sexual violence can also be linked to displacement: ‘disaster after disaster produces irrefutable evidence that with displacement – be it as a result of natural hazards or conflict – the risk of physical abuse to women and girls rises substantially’.+IFRC, World Disasters Report 2007: Focus on Discrimination (Geneva: IFRC, 2017) (www.ifrc.org/en/publications-and-reports/world-disasters-report/wdr2007/).
Research by the Women’s Refugee Commission+ Women’s Refugee Commission, Mean Streets: Identifying and Responding to Urban Refugees’ Risks of Gender-Based Violence (New York: Women’s Refugee Commission, 2016) (www.womensrefugeecommission.org/gbv/resources/1272-mean-streets). on GBV against refugees in four cities concluded that the risks are a result of ‘multiple and complex unmet social, medical, and economic needs, as well as intersecting oppressions based on race, ethnicity, nationality, language, class, gender, sexual orientation, and disability. Misperceptions further contribute to discrimination toward refugees, which in turn heightens their vulnerability’. Their key recommendations for tackling urban GBV are:
For further recommendations, information and discussion, see Women’s Refugee Commission, Mean Streets: Identifying and Responding to Urban Refugees’ Risks of Gender-Based Violence (New York: Women’s Refugee Commission, 2016) (www.womensrefugeecommission.org/gbv/resources/1272-mean-streets).
Violence and naturally-triggered disasters
The incidence of violence regularly increases after a naturally-triggered disaster. Research by the IFRC+Canadian Red Cross, Predictable, Preventable: Best Practices for Addressing Interpersonal and Self-Directed Violence During and After Disasters, 2012 (www.rcrc-resilience-southeastasia.org/document/predictable-preventable-bestpractices-for-addressing-interpersonal-and-self-directed-violence-during-and-after-disasters/). highlights that post-disaster violence can increase due to a combination of factors, including increased stress, loss of belongings, a failure of services, lack of money, resort to self-medication (alcohol) and drug use and living in temporary shelters, often in cramped and poor-quality conditions. The IFRC highlights that ‘people with pre-existing vulnerabilities to violence’ are especially at risk. This may include children, women, older people and people with mental and/or physical disabilities.
The IFRC recommends that the threat of violence should be treated in the same way as a public health emergency, with ‘the same urgency, attention and resources as other preventable public health emergencies such as diarrheal disease, respiratory illnesses, malaria, measles and malnutrition’.+DeBoer, Muggah and Patel, Conceptualizing City Fragility and Resilience. The same research suggests a number of actions that can be taken and describes good practice in preventing interpersonal violence, including violence-prevention education during the disaster risk reduction phase; prioritising the prevention of violence during the response phase; rapid response to cases of violence; enhanced data collection; and support for community-based social support systems.
Access to affected neighbourhoods presents particular challenges. ICRC’s Safer Access Framework ‘proposes a structured approach to meeting the challenges of operating in sensitive and insecure contexts’. See www.icrc.org/en/what-we-do/cooperating-national-societies/safer-access-all-national-societies.
Articles on urban violence in the IDMC’s annual Global Report on Internal Displacement (GRID): see www.internal-displacement.org/global-report.
The Humanitarian Action in Situations Other than War (HASOW) project, part of the Igarapé Institute: see https://igarape.org.br/en/hasow/.
Robert Muggah’s analysis of urban violence in Central America is part of a wider set of articles on urban violence in a special issue of Humanitarian Exchange on ‘The Humanitarian Consequences of Violence in Central America’ (69, June 2017: https://odihpn.org/magazine/the-humanitarian-consequences-of-violence-in-central-america/).
For further discussion of assessments, partnerships and sector-specific lessons see E. Lucchi, Humanitarian Interventions in Situations of Urban Violence, ALNAP Lessons Paper (London: ALNAP/ODI, 2013).
Hugo Slim, ‘Remember the Millions of People Living in Urban Violence’, Humanitarian Law and Poverty, July 2016 (http://blogs.icrc.org/law-and-policy/2016/07/04/icrc-on-habitat-3-urban-violence/).
Marion Harroff-Tavel, ‘Violence and Humanitarian Action in Urban Areas: New Challenges, New Approaches’, International Review of the Red Cross, 92(878), June 2010 (www.icrc.org/en/international-review/article/violence-and-humanitarian-action-urban-areas-new-challenges-new).
It is by now well accepted that so-called ‘natural’ disasters are not natural. The natural phenomenon – windstorm, flood, earthquake, landslide, volcanic eruption or tsunami – needs to coincide with some form of vulnerability to produce a disaster. Until a disaster occurs, natural phenomena remain hazards, but are not disasters.
Cities are home to vulnerabilities that can coincide with a hazard to create a disaster. These include poorly built buildings, which may collapse in an earthquake (for a variety of reasons, not least corruption in the construction and regulatory process, which is discussed in Section 2.2). Another example is poorly located neighbourhoods, such as low-lying land that floods, or housing on steep slopes prone to landslides. With high population densities and concentrations of infrastructure and services, cities are prone to high levels of vulnerability.
Naturally-triggered disasters are almost certainly set to increase given the rapid pace of urbanisation, leading to more people being concentrated in often dangerous areas, usually poorer people living on lower-quality land. At the same time, the effects of climate change will cause stronger windstorms, increased flooding and rising temperatures. Indeed, some phenomena may be so overwhelming that even the best-laid plans struggle to cope, as happened in 2011 with the massive earthquake and ensuing tsunami which destroyed a number of urban areas along Japan’s east coast and damaged the Fukushima nuclear power plant.
This section discusses urban disasters. It looks at displacement from disasters and everyday urban disaster risks. It links very closely to a number of other sections, in particular Section 1.2.4 on climate change, Section 2.3 on resilience, Section 1.4 on vulnerability and Section 2.2 on corruption. This section (like most of the others) seeks to distil key relevant points related to an enormous issue. Further reading is provided on a number of points discussed.
The scale of urban disasters
The risk of naturally-triggered disasters in urban areas has increased in recent years, especially among those living in low-income settlements.+D. Dodman et al., Understanding the Nature and Scale of Urban Risk in Low- and Middle-income Countries and its Implications for Humanitarian Preparedness, Planning and Response, Human Settlements Discussion Paper Series, Climate Change and Cities (London: IIED, 2013) (http://pubs.iied.org/pdfs/10624IIED.pdf). Between 2008 and 2017, the most prevalent recorded disaster types were floods (1,522), followed by storms (1,001) and other types (622). Floods are thought to have affected 730 million people. Some two billion people have been affected by the consequences of natural hazards over the last ten years, 95% of which were weather-related (such as floods and windstorms).+According to the EM-DAT database, recorded in IFRC, World Disasters Report: Leaving No One Behind (Geneva: IFRC, 2018) (https://media.ifrc.org/ifrc/world-disaster-report-2018/). Urban disasters are also expensive: in the built environment, global expected average annual losses associated with earthquakes, floods, tsunamis, storm surges and wind from tropical cyclones is estimated at $314 billion.+UNISDR, Global Assessment Report on Disaster Risk Reduction (Geneva: UNISDR, 2015) (www.unisdr.org/we/inform/publications/42809).
At an individual, neighbourhood and city level, vulnerability comprises three elements – exposure to risk (the threat of a disaster), susceptibility to harm and ability to reduce or remove the risk. Given this definition, it is unsurprising that the most vulnerable are also the poorest or most marginalised in society). Underlying vulnerability is almost always the chief determinant of the impact of disasters. For example, the devastating 7.0 magnitude earthquake that struck Haiti in 2010 may have killed anywhere between 50,000 and 220,000 people, while the much more powerful 8.8 magnitude earthquake+It should be noted that the epicentre was at a greater depth in Chile than in Haiti, meaning that the earth absorbed more of the earthquake’s impact. that hit Chile six weeks later killed under 1,000 – still a large death toll, but substantially lower thanks to better preparation and stronger buildings and infrastructure. In Haiti, poor-quality governance and public service provision led to a lack of investment over decades and widespread poverty. At the time of the earthquake, Haiti ranked 149th on the Human Development Index, with Chile a full 100 places higher, at 49. The links between corruption and disaster are discussed in Section 2.2.
Different hazard types produce different kinds of disaster. For instance, a flood may be expensive and cause widespread damage, but the death toll may be lower when compared to a rapid-onset disaster such as an earthquake. Location also matters: where cities are relates to the degree of risk (the exposure) an area might have to a natural hazard. Globally, around two-thirds of cities with populations over five million+G. McGranahan, D. Balk and B. Anderson, ‘Risk of Climate Change for Urban Settlements in Low Elevation Coastal Areas’ in G. Martine et al. (eds), The New Global Frontier: Urbanization, Poverty and Environment in the 21st Century (London: 2007), cited in Dodman et al., Understanding the Nature and Scale of Urban Risk. are located in low-lying coastal areas or along rivers at risk of flooding. In Africa, around 50% of urban residents live in arid areas with low rainfall; in India, the equivalent figure is two-thirds.+Dodman et al., Understanding the Nature and Scale of Urban Risk. Cities located along the Pacific coast and in the Himalayas are at high risk of earthquakes and other geological phenomena. According to the Intergovernmental Panel on Climate Change (IPCC), ‘Inland cities face different hazards to coastal cities, including risks arising from floods, heat islands, desiccation, desertification, reduced fresh water supply, low food security, and the impact of diseases’.+C. E. Hanson et al. (eds), Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge: Cambridge University Press, 2007), quoted in Dodman et al., Understanding the Nature and Scale of Urban Risk, p. 20.
Displacement from disasters
Figure 1.4 indicates that many more people are displaced by weather-related hazards such as floods, compared to geophysical events such as earthquakes. Disasters also contribute to rural to urban migration. Traditional ways of life are affected, for example by land being rendered unproductive by rising sea levels and coastal erosion, resulting in climate-induced migration (see Section 1.2.4 for further discussion).+A. Gero, J. Kohlitz and J. Willetts, ‘Informal Settlements in the Pacific and Links to Sustainable Development’, Development Bulletin 78, 2017 (https://opus.lib.uts.edu.au/bitstream/10453/118057/1/Informal%20settlements%20in%20the%20Pacific%20and%20links%20to%20sustainable%20development.pdf).
The two major earthquakes that struck Nepal in 2015 caused extensive damage there and in northern areas of the Indian state of Bihar. An assessment visit shortly after the first earthquake found that many unreinforced masonry buildings as well as reinforced concrete houses had collapsed, as well as heritage structures such as temples. The study attributed the damage in part to a ‘widespread lack of preparedness even when the seismic hazard of the Himalayan region is well established’, including poor construction practices.
Source: C. Rai Durgesh et al., ‘Reconnaissance of the Effects of the M7.8 Gorkha (Nepal) Earthquake of April 25, 2015’, Geomatics: Natural Hazards and Risk 7(1).
Source: IDMC, Global Report on Internal Displacement, 2017, p. 32.
The Indian city of Chennai recorded nine large-scale floods between 1943 and 2015. Issues that increase flood risk include ‘haphazard town planning, choked drains, poor garbage management, and the rampant destruction of mangroves, forests, and pastures’. The severe flood of December 2015 was attributed to El Niño and a low pressure system, which led to extremely heavy rains.
Source: F. Rafiq et al., ‘Urban Floods in India’, International Journal of Scientific and Engineering Research 1, January 2016 (www.ijser.org/researchpaper/Urban-Floods-in-India.pdf).
Flooding can cause massive damage and displace millions of people for weeks or months – or indeed for good if they relocate elsewhere. Recent examples include the 2012 Bangkok floods and extensive flooding in Pakistan in 2010 (and subsequent years), which covered one-fifth of the country and affected some 14 million people in urban and rural areas. The example in Box 1.5 is from the 2015 floods in Chennai.
A common government response to hazards is often relocation to safer areas. While there is a strong case for this, often poorer neighbourhoods are forcibly relocated. There is a great deal of research and work concerning the unfairness – and often illegality – of forced evictions of low-income settlements by the authorities.+See for example the archive of the Centre On Housing Rights and Evictions (COHRE) at www.cohre.org/. In some instances neighbourhoods may be relocated to other equally risky locations, doing little to reduce the threat of disaster. In one Indian state, ‘state and local authorities have been building resettlement tenements on inland marsh areas using centrally sponsored schemes for affordable housing. These have been used as a “quick fix” after disasters, but without addressing communities’ underlying needs and inequalities. Their siting has also increased flood risk across the urban area, creating new risks’.+G. Jain, C. Singh and T. Malladi, Rethinking Post-disaster Relocation in Urban India, IIED Briefing, August 2017 (http://pubs.iied.org/pdfs/17430IIED.pdf). Neighbourhoods may be relocated far from jobs and existing networks, storing up problems for the future.
Everyday urban disaster risk
Flooding, which is for many a seasonal, recurring hazard, is one example of an ‘everyday’ risk that may affect urban areas. Others are captured in Table 1.2, drawn from the 2010 IFRC World Disasters Report, which focuses on urban risk. Urban poverty and vulnerability are further discussed in Section 1.4.
|Aspect of human poverty||Implications for everyday risk||Implications for disaster risk|
|Inadequate and often unstable income and thus inadequate consumption of necessities, including food and, often, safe and sufficient water. Often, problems of indebtedness, with debt repayments significantly reducing income available for necessities. Inability to pay for insurance.||Very limited capacity to pay for housing, which in urban areas means living in the worst-quality and most overcrowded homes in illegal settlements on dangerous sites lacking provision for infrastructure and services – so very high levels of environmental health risk.||In most cities and many urban centres in low- and middle-income nations, most low-cost housing is on dangerous sites, e.g. at high risk from flooding or landslides. The lack of public provision for infrastructure and services adds to such risks, particularly for flooding.|
|Inadequate, unstable or risky asset base (e.g. property, skills, savings, social networks) for individuals, households or communities.||Very limited capacity to cope with stresses or shocks in everyday life – including rising prices or falling incomes, injuries or illness.||Very limited capacity to cope with disaster events when they occur, including lacking assets that are not damaged or destroyed by the disaster and having no insurance.|
|Poor-quality and often insecure, hazardous and overcrowded housing (often rented) located on dangerous sites such as flood plains, steep slopes and soft or unstable ground.||High risk levels from physical accidents, fires, extreme weather and infectious diseases – with risks often increased by overcrowding.||Housing is often of poor quality so at risk from storms/high winds, earthquakes, landslides, floods, fires and disease transmission, which may cause epidemics.|
|Inadequate provision of ‘public’ infrastructure (piped water, sanitation, drainage, roads, footpaths, etc.), which increases the health burden and often the work burden.||High levels of risk from contaminated water, inadequate sanitation, house flooding from lack of drainage.||Lack of protective infrastructure against flooding. Lack of roads, footpaths and drains inhibiting evacuation when disaster threatens or happens.|
|Inadequate provision of basic services – day care, schools, vocational training, healthcare, emergency services, public transport, communications, policing and good information on safe building practices.||Unnecessarily high health burden from diseases and injuries because of lack of healthcare and emergency response.||Lack of healthcare and emergency services that should provide rapid response to disaster (and should have had a role in reducing disaster risk and in disaster preparedness).|
|Limited or no safety net to ensure basic consumption can be maintained when income falls; also to ensure access to housing, healthcare and other necessities when these can no longer be paid for (or fully paid for).||Very limited capacity to cope with stresses or shocks in everyday life – including rising prices or falling incomes, injuries and diseases.||Very limited capacity to recover from disaster, for instance to afford food and water, rebuild homes and livelihoods. Lack of documentation often means not getting post-disaster support.|
|Lack of influence over what government does, including what it does in post-disaster responses.||Low-income survivors often not allowed to move back to their former settlement and rebuild their homes and livelihoods.||Little external support for low-income groups and their organisations to rebuild in ways that reduce disaster risk.|
|Limited influence over external civil society actors such as international aid agencies during disaster risk reduction and response.||Lack of local input can lead to inappropriate development investments or missed opportunities to reduce risk and to build more secure local economies and livelihoods.||International humanitarian actors can overwhelm local government and civil society organisations alike. Lack of partnership inhibits good governance.|
Source: IFRC, World Disasters Report (Geneva: IFRC, 2010), p.21.
Pacific Island nations are undergoing rapid urbanisation, much of it unplanned and informal. Tropical Cyclone Winston, which struck Fiji in 2016, caused widespread damage in the capital, Suva. Key priorities in the immediate recovery concerned water, sanitation and emergency shelter. Following extensive flooding in Honiara, the capital of the Solomon Islands, a review+A. Trundle and D. McEvoy, ‘Honiara Urban Resilience and Climate Action Plan (HURCAP), Honiara City Council and the Solomon Islands Government’, 2016. of low-income settlements and disaster risk identified a number of challenges, including underlying chronic problems that exacerbated disaster risk, such as poor sanitation, lack of drainage, poorly built buildings and unsuitable siting of settlements in areas prone to landslides.
Source: A. Gero, J. Kohlitz and J. Willetts, ‘Informal Settlements in the Pacific and Links to Sustainable Development’, Development Bulletin 78, 2017 (https://opus.lib.uts.edu.au/bitstream/10453/118057/1/Informal%20settlements%20 in%20the%20Pacific%20and%20links%20to%20sustainable%20development.pdf).
Climate change exacerbates the risk of disasters and increases migration. The IPCC warns that ‘global warming is likely to reach 1.5°C+A. Trundle and D. McEvoy, ‘Honiara Urban Resilience and Climate Action Plan (HURCAP), Honiara City Council and the Solomon Islands Government’, 2016.between 2030 and 2052 if it continues to increase at the current rate’.+IPCC, Global Warming of 1.5 Degrees: Summary for Policymakers, 2018 (http://report.ipcc.ch/sr15/pdf/sr15_spm_final.pdf), p. 6. As a result, cities could experience more heatwaves and flooding (resulting from sea-level rises and changing rainfall patterns) and stronger windstorms.
This section summarises the impacts of climate change on cities. It discusses urban climate change consequences, the impact of climate change on informal settlements and activities associated with climate change adaptation. This section is very closely related to Section 1.2.3 on naturally-triggered disasters, where there is further discussion on and links to sources of information concerning disasters and hazards. Other closely aligned sections include Section 2.3 on resilience (concerning adaptive capacity) and Section 1.4 on vulnerability.
Climate change and cities
Cities both contribute to, and are impacted by, climate change. Globally, cities produce up to 80% of greenhouse gas emissions and 70% of the world’s waste.+World Bank, Natural Hazards, Unnatural Disasters: The Economics of Effective Prevention (Washington DC: World Bank, 2010) (www.worldbank.org/en/news/feature/2010/11/15/natural-hazards-unnatural-disasters-the-economics-of-effective-prevention). A 1.5°C rise in temperature by 2050, as forecast by the IPCC, could expose more than 350 million people across a number of large cities (including Lagos and Shanghai) to deadly levels of heat.
Climate change is impacting cities in different ways, creating complex challenges. Table 1.3 summarises the direct and indirect impacts of climate change on urban areas.
|Change in climate||Direct impacts on urban areas||Indirect impacts on urban areas|
|Tropical cyclones, storm surge||High winds Storm-surge induced-flood Heavy rainfall||Disruption of livelihoods and city economics Damage to infrastructure, including homes and businesses Loss of life and assests|
|Extreme rainfall||More intense flooding Higher risk of landslides||As for tropical cyclones, storm surge and precipitation|
|Drought||Water shortages||Higher water and food prices Food insecurity Disruption of hydo-electricity Distress migration from rural areas|
|Extreme temperature events||Heatwaves Coldwaves||Short-term increase in energy demands for cooling/heating Effects on human health|
|Abrupt climate change||Possible extreme sea-level rise Extreme temperature change||As for sea-level rise As for extreme temperature events|
|Change in means|
|Temperature||Fewer cold days and nights Warmer and more frequent hot days and nights||Increased energy demands for cooling Reduced energy demands for heating Worsening air quality intensified by urban heat islands Creation of vector habitats in new areas|
|Increased risk of flooding Increased risk of landslides||Distress migration from rural areas Interruption of food supply networks Increased transmission of malaria Increased spread of cholera|
|Coastal flooding||Reduced income from agriculture and tourism Salinisation of water resources Damage to coastal infrastructure Displacement of urban populations|
Source: D. Dodman et al., Understanding the Nature and Scale of Urban Risk in Low- and Middle-income Countries and its Implications for Humanitarian Preparedness, Planning and Response (London: IIED, 2013) (http://pubs.iied.org/pdfs/10624IIED.pdf).
|Vulnerability||Time period||Population||City estimate|
|Extreme heat||Present day||Over 200 million people||Over 350 cities|
|2050s||Over 1.6 billion people||Over 970 cities|
|Extreme heat and poverty||Present day||Over 26 million people||Over 230 cities|
|2050s||Nearly 215 million people||Over 490 cities|
|Water availability||2050s||Over 650 million people||Over 500 cities|
|Food security||2050s||Over 2.5 billion people||Over 1,600 cities|
|Sea-level rise||2050s||Over 800 million people||Over 570 cities|
|Sea-level rise and power plants||2050s||Over 450 million people||Over 230 cities|
Source: The Future We Don’t Want: How Climate Change Could Impact the World’s Greatest Cities, UCCRN Technical Report (https://c40-production-images.s3.amazonaws.com/other_uploads/images/1789_Future_We_Don’t_Want_Report_1.4_hi-res_120618.original.pdf).
Table 1.4 indicates the impact of the effects of climate change in terms of populations and numbers of cities affected.
As the IPCC report cited above makes clear, heatwaves are a major threat to urban areas. According to recent research,+The Future We Don’t Want: How Climate Change Could Impact the World’s Greatest Cities, UCCRN Technical Report (https://c40-production-images.s3.amazonaws.com/other_uploads/images/1789_Future_We_Don’t_Want_Report_1.4_hi-res_120618.original.pdf). the number of cities experiencing temperatures of over 35?C in a three-month average is expected to reach 970 by 2050, an increase from 354. Many of these cities will be in Africa; the report found that ‘by 2050, many of the most at risk cities with large urban populations in poverty will be in West Africa, as well as Sudan and Egypt’.
Heatwaves are causing increased fatalities in India. In 2015, Andhra Pradesh and Telangana alone tallied nearly 2,000 heat-related deaths.56 In the Gujarati city of Ahmedabad, low-income settlements suffer heat extremes both night and day, with limited ventilation. According to one report, the adoption of a ‘heat action plan’ has reduced fatalities from extreme heat from over 1,000 in 2010 to just seven in 2016. The plan combines public awareness, building medical capacity and hospital readiness and having an effective early warning system.
Source: ‘PROJECT: Deepening and Expanding Heat Health Action in India’ (https://cdkn.org/project/deepeningand-expanding-heat-health-action-in-india/?loclang=en_gb).
The consequences of urban climate change
In urban areas, the consequences of climate change could include:
The IPCC predicts that Uganda’s temperature will increase by 1.5 degrees by 2030, and by 4.3?C by 2080. As a result the capital, Kampala, may see a 20% increase in rain, accompanied by stronger and more frequent storms. Both are likely to lead to increased flooding. Sanitation conditions are poor – pit latrines are widely used. Increased flooding combines poorly with pit latrines, leading to overflows of effluent and the contamination of water sources (on which the majority of the population relies). This in turn leads to more outbreaks of disease: Kampala suffered five recorded outbreaks of cholera between 1997 and 2008, which were linked to flooding.
Source: R. Wilson and T. Smith, Urban Resilience to Climate Change Challenges in Africa, Policy Research Project on Urban Resilience to Climate Change Challenges in Africa (Austin, TX: University of Texas, 2015) (https://repositories.lib.utexas.edu/handle/2152/31021).
Impacts on informal settlements
Inevitably, people living in low-income settlements are least equipped to deal with climate change. The effects of heat, for example, are amplified substantially by poorly ventilated buildings; corrugated iron roofing also increases temperatures. One recent report+The Future We Don’t Want, p. 18. predicts that, by 2050, some 215 million people will be living in poverty in 495 cities that have a regular three-month average temperature of over 35?C – which is eight times the current number. Settlements that lack adequate drainage or sanitation facilities tend to fare worse when flooding occurs than better-maintained neighbourhoods. For further discussion on the most vulnerable in cities see Section 1.4.
Climate change adaptation
Climate change adaptation (CCA) can be defined as ‘the process of preparing for, and adjusting proactively to, climate change – both negative impacts as well as potential opportunities’.+World Bank, Guide to Climate Change Adaptation in Cities (Washington DC: World Bank, 2011) (http://siteresources.worldbank.org/INTURBANDEVELOPMENT/Resources/336387-1318995974398/GuideClimChangeAdaptCities.pdf), p. 4. The World Bank’s 2011 report Guide to Climate Change Adaptation in Cities notes that key actions for implementing CCA include urban design and planning, improved urban transport, more efficient use of buildings and better waste management. The guide provides a roadmap for city decision-makers to develop and implement CCA plans, including city adaptation plans, policies and actions. It also discusses the central importance of policies and practices around building resilience, in particular in relation to adaptive capacity (discussed below).
The concept of resilience is discussed in Section 2.3.
The extent to which CCA can be enacted depends on adaptive capacity, which can be defined as ‘The ability of systems, institutions, humans and other organisms to adjust to potential damage, to take advantage of opportunities, or to respond to consequences’.+IPCC, ‘Glossary’, undated, p. 118. This in turn depends on cost (financial, social and political), the policy environment and a willingness to change behaviour. This poses a challenge for a number of cities that are at high risk. For example, those cities in Asia most vulnerable to climate risk (namely Manila, Dhaka and Jakarta) also have the lowest adaptive capacity.+WWF, Mega-stress for Mega-cities: A Climate Vulnerability Ranking of Major Coastal Cities in Asia (Gland: WWF International, 2009), cited in Dodman et al., Understanding the Nature and Scale of Urban Risk, p. 17.
Given the potentially enormous costs of CCA, funds that arrive in low-income neighbourhoods post-disaster need to be used, not only to meet present needs, but also invested to reduce the impact of future shocks. One example is geographically-focused initiatives such as area-based approaches (ABAs, discussed in Section 3.2). This can be coupled with a systems-based perspective on the city, so that strategic investments can be made that meet immediate needs, but which also contribute to mitigating long-term impacts. Another opportunity lies in engaging in post-disaster and post-conflict reconstruction planning that uses morphological approaches, for example working with existing topography to minimise flooding, or adopting low-carbon approaches in post-crisis (disaster and conflict) reconstruction. A recent systematic review of energy efficiency in buildings, low-carbon transport and sustainable waste management concluded that ‘low-carbon measures can help to achieve a range of development priorities, such as job creation, improved public health, social inclusion, and improved accessibility’.+A. Gouldson et al., The Economic and Social Benefits of Low-Carbon Cities: A Systematic Review of the Evidence (London and Washington DC: Coalition for Urban Transitions, 2018) (https://newclimateeconomy.report/workingpapers/wp-content/uploads/sites/5/2018/06/The-Economic-and-Social-Benefits-of-Low-Carbon-Cities-A-systematic-review-of-the-evidence.pdf), p. 1.
In conclusion, more frequent and worse floods and windstorms, combined with higher temperatures and more severe drought, will lead to increasing demand for humanitarian response, which in turn will need to be much more strategic in its actions (not only because of climate change, but also because of the other issues described in this review around the scale and challenges of urban humanitarian action). Strategic decisions need to consider a changing future, not just focus on the present.
The website of C40, ‘a network of the world’s megacities committed to addressing climate change’, comprises extensive information, guidance and research on a variety of themes, including heat extremes, climate change and poverty, water availability and food security. The website is https://www.c40.org/.
D. Broekhoff, G. Piggott and P. Erickson, Building Thriving, Low-Carbon Cities: An Overview of Policy Options for National Governments (Stockholm: Stockholm Environment Institute,2018) (https://newclimateeconomy.report/workingpapers/workingpaper/buildingthriving- low-carbon-cities-an-overview-of-policy-options-for-national-governments/).
World Bank, Guide to Climate Change Adaptation in Cities (Washington DC: World Bank, 2011) (http://siteresources.worldbank.org/INTURBANDEVELOPMENT/Resources/336387-1318995974398/GuideClimChangeAdaptCities.pdf).