RISK REDUCTION AND FINANCING FOR THE URBAN POOR

Bridging the gaps: stakeholder-based

strategies for risk reduction and
financing for the urban poor

CHRISTINE WAMSLER

Christine Wamsler is an A B S T R A C T This paper explores the options that can be used by aid organ-
architect and urban planner izations working in human settlement development to more effectively address
with a Master’s degree in
disaster risk management. Qualitative research was carried out in El Salvador at
International Humanitarian
Assistance. Currently, she both the household and institutional levels – to analyze the needs, capacities and
is working as a Researcher perspectives of slum dwellers and aid organizations. A clearer understanding of
at Housing Development the gaps between what households need and undertake to deal with disasters
& Management (HDM), and risk, and how organizations support them, yields important insights for the
Lund University,
Sweden, and is also a
restructuring of development aid. At the household level, the research reveals a
consultant for different huge variety of crucial but somewhat weak coping strategies. At the institutional
academic institutions level, organizational structures and mechanisms for social housing provision and
and international aid financing offer a potentially powerful platform for tackling disaster risk. However,
organizations. She current project measures are insufficient. Support for and scaling up of selected
specializes in the field of
disaster risk management
household coping strategies, combined with the expansion of social housing
for low-income settlements funding mechanisms for risk reduction and financing, are some of the options
in developing countries. proposed for targeting aid.
Address: Housing
Development & K E Y W O R D S development assistance / disaster risk management / El Salvador /
Management (HDM), Lund insurance / risk financing / risk reduction / settlement development / slum / social
University, Box 118, 22100 housing
Lund, Sweden; e-mail:
christine.wamsler@hdm.
lth.se, or wamsler_
[email protected] I. INTRODUCTION
Acknowledgement: The
author is particularly a. Background
grateful to Joanne Bayer
of the International Over the past decades, the frequency of so-called natural disasters has
Institute for Applied increased worldwide, resulting in growing human and economic losses.
Systems Analysis (IIASA),
Alfredo Stein of HDM, Lund In 2005 alone, more than 360 disasters were reported, with around 92,000
University, and Michael people killed and another 160 million suffering adverse impacts. Direct
Thompson, Musgrave material losses were about US$ 160 billion. Low- and middle-income
Institute, London and
University of Bergen for
nations bear the highest burden in terms of the human lives and pro-
their insight and the time portion of gross domestic product (GDP) lost as a result of disaster.(1)
spent commenting on draft Slum dwellers are particularly vulnerable to natural disasters. Low-
versions of this article. income human settlements are often located on marginal land near rivers
Thanks also to all who
agreed to be interviewed, or on steep slopes; housing and infrastructure are sub-standard. Among
for their time and the other problems are leaking sewage pipes from better-off settlements that
transparent manner in pass through slum areas to discharge into nearby rivers, a lack of water
which they answered the
questions, including Luis
and waste management services, limited access to information, and over-
Castillo and Jorge Gavidia crowding. Disasters make the already precarious economic, social and en-
from the cooperating vironmental conditions of slum dwellers worse, creating a vicious circle.
Environment & Urbanization Copyright © 2007 International Institute for Environment and Development (IIED). 115
Vol 19(1): 115–142. DOI: 10.1177/0956247807077029 www.sagepublications.com
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

Currently, more than one billion people worldwide live in slums and are organizations FUSAI and
UN–HABITAT–ROLAC
forced to accept inhuman and dangerous living conditions. It is estimated
(Regional Office for
that their number will double over the next 24 years.(2) Latin America and the
In recent years, increasing attention has been given to the need to Caribbean). The Swedish
reduce disaster risk within the context of development work. The stated International Development
Cooperation Agency (Sida)
aim of the Millennium Declaration to achieve a significant improvement and the German Advisory
in the lives of at least 100 million slum dwellers by 2020 alludes to this Council on Global Change
need;(3) and the Hyogo Framework for Action 2005–2015 urges governments (WBGU) provided financial
support for this research.
to address the issue of disaster risk in their sector development planning
and programmes.(4) Nevertheless, aid organizations working in human set- 1. UNISDR (2006), “Disaster
tlement development still struggle to sustainably reduce existing disaster statistics 1991–2005”, accessed
at www.unisdr.org/disaster-
risk in their everyday work. statistics/introduction.htm, 7
This paper reports on case studies carried out in El Salvador, which June 2006.
is located in one of the most disaster-prone regions in the world.(5) Its
2. UN-HABITAT (2003), The
objective is to explore and develop stakeholder-based options for aid
Challenge of Slums, Global
organizations to more effectively integrate disaster risk management (i.e. Report on human settlements,
risk reduction and financing) into their core project work. The focus is Earthscan, London.
on non-government aid organizations working, inter alia, in the field of
3. See www.
settlement development planning for the urban poor. The term “social unmillenniumproject.org.
housing organizations” will be used to describe this type of organization.
4. UNISDR (2005), “Hyogo
framework for action 2005–
2015: building the resilience of
b. Methodology and outline nations and communities to
disasters”, accessed at www.
Case studies were carried out at the household and institutional levels in unisdr.org/wcdr/intergover/
El Salvador in 2005/6 to determine the existing perceptions, needs and official-doc/L-docs/Hyogo-
framework-for-action-english.
capacities of both the urban poor and the national organizations servicing pdf, 5 June 2006.
slum communities.
The research at the household level included semi-structured inter- 5. Lavell, Allan (1994),
“Prevention and mitigation of
views with people living in 15 disaster-prone slum communities,(6) as well disasters in Central America:
as walk-through analyses, observation and a literature review. Sixty-two vulnerability to disasters at
households, comprising 331 persons, in high-risk areas were interviewed. the local level”, in Ann Varley
(editor), Disasters, Devel-
The emphasis was on analyzing: opment and Environment,
John Wiley & Sons, Chichester,
• existing disaster risk, its causes and the resulting local needs; pages 49–64.
• local capacities for risk reduction and their financial implications for
residents’ livelihoods; and 6. The communities analyzed
during research in 2005 and
• local capacities for risk financing, including formal and informal in-
2006 were: La Chacra, Llanos
surance mechanisms. de la Charcra, Quiñones
Privado, Quiñones Municipal,
At the institutional level the research included a text review, work- San Martín Privado, San Martín
shops, group discussions and semi-structured interviews with a total Municipal, Casitas del Coro,
of 22 representatives of social housing organizations and other non- Coro Nuevo, San Luis Portales,
Bolívar, Granjero II and Nueva
government organizations (NGOs), housing finance institutions or Esperanza (forming the slum
departments, government housing bodies and insurance companies.(7) area called Los Manantiales,
The focus was on analyzing the provision of social housing projects and situated in San Salvador); José
Cecilio del Valle and Divina
the related mechanisms for risk reduction and financing. A range of differ- Providencia (also situated in
ent projects was reviewed, and in-depth evaluations were also carried out San Salvador); and Refugio
of four of the projects that were implemented in the above-mentioned (situated in and made up of
people from the slums of the
15 slum communities. Bálsamo region).
The challenges and gaps identified among the perspectives, needs
and capacities at the household and institutional level were used as the 7. The organizations
interviewed were: ACSA
basis for exploring, together with the stakeholders, options for assisting (Asociación Salvadoreña
in targeting aid. A literature review was carried out to complement and de Empresas de Seguros);

116
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

CEPRODE (Centro de validate different options. For the data analysis, a combination of grounded
Protección para Desastres);
CHF (Cooperative Housing
theory,(8) systems analysis(9) and cultural theory(10) was applied.
Foundation International); Analyses of the current situation in El Salvador are now presented,
Sistema Cooperativo Financiero providing two “snapshots”, one from the household level and one from
FEDECACES; FEDECREDITO the institutional level. The gaps, challenges and potential solutions are dis-
(Federación de Cajas de
Crédito); FONAVIPO (Fondo cussed on this basis, and the main outcomes are summarized.
Nacional de Vivienda Popular);
Fundación Habitat; FUNDASAL
(Fundacion Salvadorena de
Desarrollo y Vivienda Minima);
II. FIRST “SNAPSHOT”: HOUSEHOLD-LEVEL CASE STUDIES
FUSAI (Fundación Salvadoreña
de Apoyo Integral); HFH a. Existing needs: understanding disasters
(Habitat for Humanity); IDB
(Inter-American Development In the slums analyzed, flooding and landslides, which affect many slum
Bank); INTEGRAL; national Red
Cross; Seguros Futuros; UCA
dwellers annually, and usually during the winter, were generally seen as
(University José Simeón Cañas, the main risk to lives and livelihoods. Earthquakes and windstorms ranked
Department of Architecture); next in importance. The lack of job opportunities and water provision, as
SISA (Seguros e Inversiones well as insecurity due to violent juvenile gangs (maras) were also seen as
SA); VMVDU (Vice-Ministerio de
Vivienda y Desarrollo Urbano); substantial “risks”.(11)
and different municipalities. To analyze the existing local problems and the measures needed to
All organizations were address them, slum dwellers were asked for their views on the underlying
selected through snowball and
purposeful sampling. CEPRODE,
drivers of disasters and disaster risk in slums. Interviewees reported on:
FUNDASAL and FUSAI were
operating in the 15 slum
• neighbours downhill felling trees or excavating the slopes below their
communities analyzed. houses;
• neighbours uphill building latrines close to the declivity and allowing
8. Glaser, Barney G and waste and storm water to flow onto their land;
Anselm L Strauss (1967), The
Discovery of Grounded Theory • people from outside the settlement tipping solid waste down their
– Strategies for Qualitative hills or into the nearby rivers; and
Research, Aldine Publishing • other residents not knowing how to improve their situation.
Company, Chicago.
As there is not only little sense of mutual rights and obligations in
9. Haraldsson, Hördur V (2004),
slum communities, but also a lack and unequal distribution of information
“Introduction to system
thinking and causal loop on risk reduction, the asymmetric disaster risk that inhabitants incur
diagrams”, Reports on Ecology is growing,(12) creating increased tension among neighbours. Other key
and Environmental Engineering aspects (driving disasters and disaster risk) mentioned in the interviews
2004, Vol 1, Institute of
Chemical Engineering, Lund were:
University; also Sterman, John
D (2000), Business Dynamics:
• insecure tenure resulting in slum dwellers being unwilling to invest
Systems Thinking and in reducing risk;
Modelling for a Complex World, • increases in the area of the built environment and overcrowding
Irwin/McGraw-Hill, New York. caused by growing households;
10. Thompson, Michael, Richard • inadequate housing construction and technical infrastructure;
Ellis and Aaron Wildavsky • space restrictions;
(1990), Cultural Theory, • a shortage of financial resources due to unemployment; and
Westview Press Inc, Oxford.
• little outside help.
11. The word “risk” is in
quotation marks as, in this
Furthermore, national and municipal governments were often seen by
paper, the term generally refers slum dwellers as unhelpful, and even a hindrance, to their efforts. In fact, the
to risk associated with natural actions taken by planning authorities, and the information obtained by them
disasters and/or hazards and
not to socioeconomic hazards.
with respect to the development and legalization of planned settlements,
However, as slum dwellers were viewed as contradictory and often unreliable. In some communities,
mentioned such hazards as local community cohesion and organization was affected by mistrust,
part of the risk they face, mainly related to corruption and political factionalism. In Wamsler,(13) the
some of the main related
aspects have been mentioned key variables and causal loops underlying the complex system of risk and
here. Note, however, that disaster occurrence in slum areas are described more in detail and illustrated
– in keeping with the focus in so-called causal loop diagrams, a systems analysis tool (Figure 1).(14)

117
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

of this research – these were

of secondary consideration
to slum dwellers compared
with the risk caused by natural
hazards.

12. Asymmetric disaster risk

is the unequal distribution
of the level of disaster risk
experienced by people living
close to and within a specific
area. Hence, the asymmetric
disaster risk of slum inhabitants
refers to the fact that the level
of disaster risk within a given
slum is not constant across the
entire community.

13. Wamsler, Christine (2006),

“Understanding disasters from
a local perspective: insights
FIGURE 1 into improving assistance for
Example of a basic causal loop diagram showing some natural social housing and settlement
key variables underlying risk and disaster occurrence as regards development”, TRIALOG
flooding and landslides (Journal for Planning and
Building in the Third World)
No 91, December, special issue
on “Building on disasters”.

b. Existing capacities: coping with disasters 14. Causal loop diagrams

portray a causal relation
“We are always trying to improve, little by little, step by step, in order to become between two variables (e.g.
A and B) by an arrow with a
more secure.” This statement by a slum dweller living in San Salvador illus- plus (+) or minus (–). A plus
trates the constant efforts that are put into coping with disasters and (+) or minus (-) indicates the
disaster risk. Key literature on disaster risk management commonly makes type of change that occurs if
variable A, at the beginning
use of the term “coping strategy”(15) – usually, however, without defining of the arrow, increases: A
it. Based on household-level research, the following definition of coping positive symbol (+) shows
strategy is proposed: “constantly changing and adapting cognitive and be- that the increase in variable
havioural efforts to manage disaster risk or disaster impacts”. These efforts A affects the increase in B.
However, a negative symbol
influence the key variables and causal loops underlying the complex (–) means that the increase
system of risk and disaster occurrence in specific slum areas. in A results in a decrease in
The research reveals three types of coping strategies employed by B. The inclusion of non-linear
relationships is one of the most
slum dwellers living at risk. These are: important advantages of causal
loop diagrams compared to
• strategies to reduce existing risk; conventional models such
• strategies to insure themselves informally or formally against possible as flow charts. Causal loop
disasters; and diagrams usually have at least
one closed loop, representing
• strategies to recover from disaster impacts. feedback. Figure 1 illustrates
how the identified key variables
Within each strategy – risk reduction, self-insurance and recovery – “rain” and “unstable soil
different thematic foci and underlying social patterns were identified. The conditions” relate to risk and
thematic foci are physical/technological, environmental, economic, social/ disaster occurrence. As can
be seen, an increase in the
cultural, organizational and institutional.(16) The social patterns, which amount of rain or in its duration
have been established by “cultural theory”,(17) are individualistic, com- could increase disaster risk
munitarian, hierarchical and fatalist. Within this research context, indi- and hence the occurrence of
flooding and landslides. In turn,
vidualistic behaviour is characterized by the use of self-help to fix things flooding and landslides can
without help from people outside one’s own household. Communitarian make unstable soil conditions
behaviour is based on the belief that everybody sinks or swims together; it worse, resulting in a further
is hence characterized by community efforts. Hierarchical patterns relate exacerbation of disaster risk.
In addition, more rainfall
to the belief in authority structures for assistance, control and organ- further destabilizes unstable
ization, including strong prescriptions. Fatalist behaviour is a non-strategy soil conditions, which again

118
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

influences the occurrence of for survival based on the idea that taking action or not taking action has
risk and disaster.
the same (negative) result.
15. For example, UNISDR
(2004), Living with Risk: A
Global Review of Disaster c. Coping strategies for risk reduction
Reduction Initiatives, Inter-
Agency Secretariat of UNISDR,
Risk reduction includes prevention (to minimize or avoid hazards), miti-
accessed at www.unisdr.
org/eng/about_isdr/bd-lwr- gation (to reduce vulnerabilities) and preparedness (to improve people’s
2004-eng.htm; also Twigg, John capacities to ensure effective response as soon as disaster strikes). Slum
(2004), Disaster Risk Reduction: dwellers use risk reduction mainly during “normal” times (i.e. pre-disaster)
Mitigation and Preparedness in
Development and Emergency so as to be less affected by future small-scale or exceptionally large-scale
Programming, Good Practice disasters. In an ideal case, risk reduction leads to an absence of disasters
Review, Number 9, March, (as hazard impact will be minimal).
Overseas Development
Household strategies to reduce risk are diverse, and include physical/
Institute (ODI), London.
technological, environmental, economic, social/cultural, organizational
16. Note that many coping and institutional measures (Tables 1–4). El Salvadoran slum dwellers, for
strategies involve elements instance, build retaining walls or embankments with old car tyres, stones,
from different categories.
However, such categories bricks or nylon bags filled with soil and cement; they plant palm trees;
are helpful for viewing and remove blockages from rivers and open water channels; take jobs outside
analyzing strategies, thereby their own settlement so as to be unaffected by local disasters; temporarily
ensuring that no household
move their families to the highest rooms in their dwellings if floods are
measures are overlooked.
forecast; and create information structures. They may also adopt more
17. See reference 10. emotionally oriented strategies, such as relying on their faith or simply
accepting their high risk. Answers such as “I just sit with my Bible and pray”
were common. However, unbearable needs push most of the dwellers to
actively adopt individualistic behaviour for survival. In fact, this research
18. See reference 15, Twigg supports Twigg,(18) who states that expressions of belief in divine power
(2004), page 139. are not incompatible with taking actions to reduce risk.
Physical or technological risk reduction was identified as including
structural and non-structural improvements of dwellings and their sur-
roundings, mostly carried out on an individual basis (Table 1). Environ-
mental risk reduction includes the use and removal of natural resources
as well as the “clean-up” of the natural environment (Table 2). These
measures are carried out individually, and to some extent in cooperation
with neighbours, the whole community and the local or national gov-
ernment. Economic and social/cultural risk reduction strategies are pre-
dominantly individualistic and were shown to include both behavioural
and cognitive measures (Table 3). Economic diversification in households
is a common strategy for reducing vulnerabilities. Household members
furthermore engage in low-risk activities or activities with differing risk
profiles. If one family member temporarily becomes jobless because, for
instance, the local tortillería or the corn mill is destroyed by a disaster,
other income sources can absorb the losses and help bridge the income
shortage. Increased household income (for vulnerability reduction) is
sometimes also achieved through the migration of family members to the
United States. In 2004, more than one million Salvadorans were resident
19. SELA (Latin American in the United States, and family remittances have become a major income
Economic System) (2005), source for El Salvador since the 1990s.(19) Organizational and institutional
Migrations and Remittances
in Latin America and the risk reduction comprises the creation of organized structures to reduce risk
Caribbean: Intra-Regional as well as strategies to access related services/assistance offered by different
Flows and Macroeconomic institutions, thereby tapping into formal or informal structures or networks.
Determinants, XXXI Regular
Meeting of the Latin American
It is often closely related to kinship networks, mutual aid and self-help, as
Council, 21–23 November 2005, well as to formal and hierarchical structures for disaster risk management.
Caracas, Venezuela. However, some strategies are also carried out individually (Table 4).

119
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

TA B L E 1
Physical/technological risk reduction
Focus/aim Activities identified
Constructive/ • Increasing inclination of roofs (for better run-off without damaging roof constructions)
structural house • Prolonging roof projections/eaves (to protect houses and pathways from damage/erosion)
improvements • Changing direction of roof inclination (so rainwater is discharged without causing damage/
landslides)
• Installing provisional gutters as roof eaves (so rainwater is discharged without causing
damage/landslides)
• Replacing mud walls with brick walls, wooden pillars with metallic ones, and corrugated iron
with more durable materials (e.g. duralita) (to better withstand earthquakes, rain and/or
floodwater)
• Improving electricity installations by covering cables and putting electric connections
higher up, out of reach of expected flood levels
• Regularly replacing corrugated iron, wooden pillars and beams (to better withstand rain or
earthquake impacts)
• Improving roof fixing (to better withstand earthquakes and windstorms)
• Regularly covering walls and floors with (additional) cement (for better run-off without
causing damage/erosion)
• Filling of cracks with cement (for better run-off without causing damage/erosion)
• Closing holes in corrugated iron sheets using special fillings or patches on top of or under
sheets (to prevent water entering the house)
• Changing the locations of latrines and wash places (to mitigate landslides)
Non- • Blocking wastewater pipes with stones and other objects when river levels rise (to avoid
constructive/ flooding and/or related contamination)
non-structural • Putting wood or bricks on the roof (to hold it in place during windstorms)
house • Putting plastic sheets on the roof, on the inside walls or over the bed (to prevent water
improvements entering or damaging the house)
• Building water barriers in front of the house (to prevent water entering the house)
• Digging water channels in earth floors inside the house (for better run-off without causing
damage/erosion)
• Putting pots under roofs with holes (to catch water, preventing damage/erosion)
Constructive/ • Strengthening pathways by covering them with (additional) cement and filling in cracks (to
structural im- mitigate landslides and minimize damage caused by rain and earthquakes)
provement of the • Filling in former latrine holes with earth, stones and/or cement (to mitigate landslides and
surrounding living minimize damage caused by rain and earthquakes)
environment • Repairing public infrastructure that passes through the settlement, such as wastewater
pipes (to avoid flooding and related contamination)
• Building provisional water channels with corrugated iron or cement (to discharge
rainwater without causing damage/landslides)
• Building fences to hold back soil (mitigating landslides) and/or to prevent children from
falling (fences are made of corrugated iron, mattress springs, wooden pillars and wire
netting)
• Compacting soil (to mitigate landslides and minimize damage caused by rain and
earthquakes)
• Building retaining walls or embankments from: old tyres, stones and cement; old tyres and
soil; bricks and cement; stones only; nylon bags filled with soil and cement; and other
materials (to mitigate landslides and minimize damage caused by earthquakes)
Non-constructive/ • Putting plastic sheets on slopes, often during entire year (to mitigate landslides)
non-structural • Digging water channels in earth outside the house (to discharge rainwater without causing
improvement of damage/landslides)
the surrounding • Avoiding obvious flood- or landslide-prone locations for house expansion
living environment • Replacing eroded earth with new earth (to mitigate landslides and minimize damage
caused by rain and earthquakes)
• Cleaning water gutters (to mitigate flooding)

120
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

TA B L E 2
Environmental risk reduction
Focus/aim Activities identified
Use of natural resources to reduce risk • Planting to prevent landslides
• Planting to create windbreaks
• Using the natural environment as an information source to
analyze risk situations (river level, clouds)
Removal of natural resources that • Cutting down bigger branches and trees located close to
represent risk houses (to minimize the risk of them falling down and
causing damage during earthquakes and landslides)
Clean-up of natural environment • Cleaning waste from slopes (to mitigate flooding caused by
blocked water gutters)
• Clearing objects blocking the flow of rivers, such as tyres,
plastic sheets, mattresses and branches (to mitigate
flooding)

TA B L E 3
Economic and social/cultural risk reduction
Focus/aim Activities identified
Economic diversification of individuals • Carrying out different jobs simultaneously
or families to increase overall income, which • If possible, all family members working
reduces economic impacts after disasters
and/or the dependency on the income
of specific sources or family members
Taking on income activities with differing risk • Taking (additional) jobs outside own settlement
profiles and/or low-risk activities (i.e. jobs
unaffected by local disasters)
Geographic diversification of families’ income • Taking jobs located in different geographic areas within
and/or outside the country
Reduction of household expenses to increase • Cutting down firewood instead of using gas ovens
disposable income
Learning from friends, neighbours and project • Being involved in self-help measures
measures • Copying construction types and/or economic risk reduction
strategies from neighbours
Trust in outside help • Relying on labour and/or income of family members
(e.g. regular “income” through remittances)
• Relying on a hierarchical system to supply help for risk
reduction
Psychological acceptance of risk situation • Having religious beliefs
• Downplaying the existing level of risk
• Deciding not to invest too heavily in housing or infrastructure,
as losses can be replaced more cheaply and easily

121
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

TA B L E 4
Organizational and institutional risk reduction
Focus/aim Activities identified
Creation of structures/ • Going to church (priests are a source of information)
mechanisms to access • Asking neighbours
information on • Getting information from government organizations (at local, municipal
existing risk level and and national level)
weather forecasts • Listening to radio and watching television
• Monitoring disaster-related changes (e.g. level of river, clouds) and establishing
related information structures
Creation of • Establishing local committees for risk reduction
organizational • Including risk reduction activities in the work portfolio of the local
community structures executive committee, for example, monitoring of risk situation, distribution
for risk reduction of plastic sheets, information on evacuation, clearing of waste from slopes,
accessing help from government
Accessing help for risk • Asking for help (mostly for construction materials) from different
reduction organizations, political parties and/or the municipality
Taking organizational • Preparing food in advance for the children so that, if need be, they can eat
precautions to protect and then quickly be sent to neighbours or family members in more secure areas
family members and • Temporarily moving to the highest room, another house or tents
reduce damage • Regularly observing and monitoring cracks in cement surfaces (if flooding
to belongings is expected)
• Taking belongings to another location (within the house, for example,
storing them on a bunk bed, or on a higher platform outside the house
(if flooding is expected)
Organized and • Guarding empty houses and evacuated people who are asleep during
coordinated preventive evacuation
community work • Transportation of people’s belongings to higher-level streets (if flooding is expected)
• Clearing waste and other sources of risk from slopes, rivers and streets
• Cementing of streets so that children, in the case of flooding, do not sink into the mud
• Moving to refuges (neighbours’ private houses) in anticipation of a disaster

d. Coping strategies for risk financing: self-insurance

The term “risk financing” is mainly used at the institutional level and
usually describes only formal financing measures to transfer or share
risk. The broader term “self-insurance” was selected for the household
level to include formal and informal, and monetary and non-monetary
mechanisms. The literature on disaster risk management generally uses the
terms “informal insurance”, “self-insurance” or “informal self-insurance”
synonymously and – as in “coping strategy” – without definition.(20) Based 20. See reference 15,
on the research outcomes, self-insurance is defined here as “the creation
or maintenance of formal or informal security systems that help people
access financing sources or mutual social help in the event of a disaster”.
Financing sources are, for instance, informal and formal credits, donations,
additional income, the selling price of assets and monetary compensation.
Mutual help can include offering refuge, temporary custody of children,
fostering a child, labour work for reconstruction, or washing and cleaning.
In other words, to insure themselves, slum dwellers take pre-disaster action
in the hope of obtaining direct or indirect compensation if a hazard leads
to death, injury or loss of property or income. They thus ensure that they
can bounce back faster – than if they do not have self-insurance – to their
former standard of living, or an even higher one.

122
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

Self-insurance strategies were identified as including economic, social/

cultural, organizational and institutional measures (Tables 5 and 6). An ex-
ample of the former is the acquisition and maintenance of physical assets,
such as construction materials, which can easily be sold if need be. To
alleviate financial distress, one of the slum dwellers interviewed sold seven
roofing sheets of corrugated iron, and then re-roofed his home with
an old car body. Not nailing down the corrugated iron allows it to be
resold at a higher price. Further examples of self-insurance reported by
the slum dwellers were savings, the creation of reciprocally dependent re-
lationships and the encouragement of reciprocal dependents to achieve
improved economic status. Economic and social/cultural self-insurance is
mainly applied on an individual basis (Table 5).
Organizational and institutional self-insurance strategies comprise the
creation of organized structures to insure oneself, as well as related stra-
tegies to access services/assistance offered by different institutions, thereby
tapping into formal or informal structures or networks. It includes, for
instance, accessing community emergency funds, creating linkages with
government and (mostly) non-government institutions, such as religious
bodies (which offer help after disasters), and illegally accessing formal
insurance mechanisms (Table 6). In fact, while slum dwellers are not gen-
erally believed to have a culture of insurance, in two cases residents had
illegally obtained employment certificates enabling them to pay into the
social security system, even though they were not formally employed. In
addition, 26 of the 331 people interviewed had health insurance as they
work in the formal market. One-third of the interviewees were interested
in acquiring property disaster insurance (the other two-thirds had no opin-
ion on the subject).

TA B L E 5
Economic and social/cultural self-insurance
Focus/aim Activities identified

Creation of extensive reciprocal • Having many children

and dependent relationships • Creating a large and good family network
Encouragement of dependents and • Encouraging dependents to study
other family members to achieve • Encouraging dependents to migrate to the United States
improved economic status or other foreign countries
• Supporting dependents to obtain a formal job
Increase of job capacities of • Being able to change job sectors from one where demand
individuals for different decreases after a disaster (e.g. clothes vending) to one where
sources of income demand rises after a disaster (e.g. the construction sector)
Acquiring/maintaining • Taking on a formal job
assets that can serve as • Accessing legal tenure
collateral for formal credits • Owning a legal and permanent dwelling
Saving for a “rainy day” • Individually accumulating money “under the mattress”
(not regularly, for instance monthly, but based on irregular income)
Acquiring/maintaining physical • Using reusable construction materials for own shelter
assets that can be easily sold • Owning land
if need be • Owning a home

123
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

TA B L E 6
Organizational and institutional self-insurance
Focus/aim Activities identified
Participation in informal community • Paying social fees for community emergency funds
insurance schemes
Creation of cohesion, solidarity and/or • Knowing well and interacting with people from the neighbourhood
reciprocal relationships with neighbours (e.g. buying from local shops, offering labour when needed)
and/or other community members • Employing community members for small jobs
• Engaging in community matters
Creation of linkages/ relationships with • Participating in the local executive committee
institutions at different levels • Becoming a member of a political party
(government and non-government) • Becoming a member of a religious group
• Staying in contact with NGOs that carried out projects within the
settlement
• Maintaining good contact with the local government (municipality)
and national government organizations (mostly through the local
executive committee)
Creation of illegal access to formal • Paying into the public social security system through deals with
insurance mechanisms entrepreneurs who certify the employment of the respective person
Improved possibility of accessing • Renting in a high-risk area
post-disaster assistance (partly through • Renting in risk areas where infrastructure projects
intentional increase of risk) or aid projects are planned

e. Coping strategies for recovery

Directly interlinked with self-insurance approaches are coping strategies
for disaster recovery. However, not all recovery strategies are initiated in a
pre-disaster context; they can also be ad hoc. Based on the household level
“snapshot”, recovery strategies were defined as “actions taken in a post-
disaster situation to recover as fast as possible from disaster impacts, that is,
to regain the former status quo or become even better-off than before”.
Economic and social/cultural recovery aim at loss “financing”, that is,
obtaining financial or material resources for recovery. Examples are bor-
rowing money, using credit or savings, increasing income and reducing 21. Agunias, Dovelyn (2006),
expenses (Table 7). The strategies identified are mainly carried out “Remittance trends in
Central America”, Migration
individually at the household level. However, the lack of solidarity among Information Source (MFI),
members of some households occasionally erodes such efforts: complaints April 1. See
about family members with regard to financial help were more frequent www.migrationinformation.
net/pdf/Migration_Information_
than accolades. Nevertheless, the interviews and literature confirm that Source-Central_America.pdf.
remittances from abroad play an important role in recovery. According to
Agunias, remittances rise when an economy suffers a downturn or macro- 22. UNDP (2004), Reducing
Disaster Risk: A Challenge for
economic shocks due to a natural disaster.(21) Following the 2001 earth- Development, UNDP, New York.
quakes, the Central Reserve Bank of El Salvador estimated that Salvadorans
living abroad sent home US$ 1.9 billion in remittances.(22) As of 2004, 23. See reference 19.
remittances totalled US$ 2.5 billion(23) and assisted more than one-fifth 24. UNDP (2005), Informe
of all households.(24) While organizational and institutional recovery are Sobre Desarrollo Humano El
for the most part related to kinship networks, mutual aid and self-help, Salvador 2005: Una Mirada al
Nuevo Nosotros. El Impacto de
they also include assistance from NGOs and government organizations las Migraciones, San Salvador,
(Table 8). UNDP, San Salvador.

124
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

TA B L E 7
Economic and social/cultural recovery strategies
Focus/aim Activities identified
Access to credits/loans • Borrowing money from family members
• Taking bank credits (directly or through family members)
• Taking credits from employers (directly as the employee or
indirectly through family members)
• Taking credits from aid institutions (e.g. religious institutions,
social housing organizations)
• Borrowing money from informal lenders (of own profession, for
example, market lenders for vendors)
Use of savings • Use of money that was accumulated at home, “under the
mattress” (not in saving accounts)
Trade of assets • Selling construction materials and replacing these with other
objects (e.g. corrugated iron roofing being replaced by a car body)
Reduction of expenses • Less consumption where possible (e.g. food)
Increased income • Taking on an extra job (e.g. in the construction sector)
• Changing to a more profitable job
• Working longer hours
Receipt of solidarity/help from family • Receiving remittances from family members
and/or neighbours • Obtaining food, construction materials or other support, such as
taking in foster children

TA B L E 8
Organizational and institutional recovery strategies
Focus/aim Activities identified

Coordination with neighbours to recover • Mutual employment of people in the community for reconstruction
work
• Mutual guarding of damaged houses and affected people who
have to stay/sleep in insecure places
• Cleaning up disaster impacts (washing away soil, washing clothes,
cleaning streets)
Creation of access to humanitarian • Receiving humanitarian assistance from non-government institutions
and development assistance • Accessing help from government institutions
(food, clothes, construction materials, • Moving swiftly to an even more affected area, where more help
recovery projects) is available

f. Coping strategies and incremental urban housing

Slums are generally improved incrementally, usually by self-help and
mutual help. The coping strategies identified are crucial for their con-
tinuous development, to ensure an improvement in the standard of living
and security (Figure 2). Assuming that the continuous arrow shown in
Figure 2 indicates the relative and average development of a slum situated
in a disaster-prone area, the fine and dashed arrows show how coping
strategies work together to “buffer” disaster impact and reduce recovery
time.

125
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

FIGURE 2
Importance of coping strategies for the development of slums

This research indicates that some of the various coping strategies identi-
fied are weak and deficient. In fact, backsliding is frequent. Slum dwellers
reported that it can take them several years to recover from single events
and that they are mostly dependent on outside help. If a gradual slum
development process cannot keep pace with the frequency of dis-aster
impacts, then increased insecurity and “poverty traps” can result.
Although more evidence is needed, the coping strategies of urban – as com-
pared to rural – slum dwellers appear not only to be weaker (i.e. less ef-
fective) but also less deliberate and more individualistic (as opposed to
communitarian), with a stronger focus on housing construction and land
issues, and less emphasis on productive sources of livelihood.
Weak and more individualistic coping can occur for a variety of
reasons: According to Morduch, solidarity and reciprocity work best in:

• settlements where people have family members living close by;

• where there is not too great a disparity in residents’ income levels;
• where family members or other dependents are not simultaneously
affected by disaster impacts; and
• where disasters happen repeatedly, but not too frequently, and have
mostly short-term impacts.(25) 25. Morduch, Jonathan (1999),
“Between the state and the
However, within the urban slums studied, few of these conditions market: can informal insurance
patch the safety net?”, The
apply. First, the data gathered suggest that urbanization and the increasing
World Bank Research Observer
ease of mobility related to it can enable households to “default” on their Vol 14, No 2, August,
obligations to relatives and neighbours. Second, different income levels pages 187–207.
(ranging between US$ 120–750 per household and US$ 30–500 per worker)

126
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

foster individualistic behaviour, with the better-off households opting out

of mutual and hierarchical arrangements. In addition, slum dwellers have
little to sell (e.g. no livestock or agricultural products) to help themselves
or others in need. Third, slum dwellers simultaneously and persistently
experience bad conditions over a period of years, with floods and landslides
causing adverse environmental changes (runoff, poor soil), the effects of
which continue even after the weather has returned to normal. Finally,
the lost trust in both community solidarity and hierarchical structures,
as well as the fear of being hoodwinked by the authorities, further
promotes a dominant sense of individual responsibility and ownership
at the household level, as well as a determination to “fix” things without
assistance.
Incremental housing improvements – and related coping strategies
– could be seen as equivalent to vulnerability reduction, with people
replacing cardboard walls with brickwork or temporary retaining walls
of loose tyres with cement ones. However, the research – through interviews
and expert observation – revealed cases where, from a certain point on-
wards, increased household prosperity and the concomitant housing
“improvements” did not lead to decreased disaster risk. Households, for
instance, increased the height of retention walls or enlarged their houses,
which in both cases led to decreased structural integrity. In other cases,
wealthier household members withdrew from community involvement,
again increasing their vulnerability, as social cohesion usually enhances
26. IFRC/RCC (International resilience.(26) Professional specialization (as opposed to livelihood diver-
Federation of Red Cross and
Red Crescent Societies) (2004),
sification) provides another example. In fact, coping capacity can be
World Disasters Report 2004, reduced by the need to compete economically, which at present rewards
IFCR/RCC, Geneva. (productive) specialization and intensification more than diversity and
27. See reference 22.
sustainability.(27)

g. Financial capacity to cope with disasters

The research revealed that, on average, households spend 9.2 per cent
(ranging from 0 to 75 per cent) of their income on reducing disaster risk
and preparing for the following winter, that is US$ 26 out of an average
monthly household income of US$ 284. This figure is even more intriguing
28. See also Dercon, Stefan if one considers that it excludes construction materials that are obtained
(2003), “Poverty traps and
for free (such as stones and sand from riverbeds or old tyres from friends),
development: the equity
efficiency trade-off revisited”, family members’ free labour, the opportunity costs of the considerable
Agence Française de amount of time spent on risk reduction, and the negative impacts of some
Développement (AFD) and
the European Development
coping strategies (e.g. high interest paid to money lenders, or financial
Research Network (EUDN), losses due to adjustments in assets and activities due to risk exposure(28)).
Paper presented at the There are hardly any similar studies to allow comparisons or a broader
conference on Poverty,
Inequalities and Growth: What’s
view of the expenses in question. An exception is Walker and Ryan, who
at Stake for Development Aid?, state that in risk-prone areas of India, households may sacrifice as much
13–15 November 2003, Paris. as 25 per cent of their average income to reduce exposure to shocks.(29)
29. Walker, Thomas S and
Even if risk reduction incurs a large cost, there are also substantial post-
James G Ryan (1990), Village disaster expenses, amongst others, for the replacement of belongings
and Household Economies washed away during floods and landslides; recovery efforts; temporary
in India’s Semi-arid Tropics,
Johns Hopkins University Press, income losses; and the gradual loss of investments made in the incremental
Baltimore. building of housing and community infrastructure.

127
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

III. SECOND “SNAPSHOT”: INSTITUTIONAL-LEVEL CASE STUDIES

a. Existing capacities: dealing with slums

In terms of their capacity to deal with the needs of the poor, social housing
organizations use different approaches, such as new settlement devel-
opments, in situ house improvements and settlement upgrading. Based
on lessons learned during the last decades, these kinds of projects in El
Salvador have been reducing their one-sided focus on construction-related
work and are adopting, step-by-step, a broader approach, including non-
constructive risk reduction measures (see also Section III.d).(30) 30. Wamsler, Christine (2006),
“Integrating risk reduction,
While representatives from social housing organizations report that urban planning and housing:
they struggle to reach the poorest of the poor, the slum dwellers that are lessons from El Salvador”,
included in their projects experience a rapid improvement in living stand- Open House International (OHI)
Vol 31, No 1, pages 71–83;
ards and security levels (Figure 3). Better housing and infrastructure, com- also Wamsler, Christine (2007),
bined with other risk reduction measures, result in reduced vulnerability “Integrando la gestión del
and hence reduced disaster impact. This is shown in Figure 3 where the riesgo, planificación urbana
y vivienda social: lecciones
lengths of the vertical thick arrows are reduced compared with those of de El Salvador”, INVI (Revista
the vertical fine and dashed ones. Internacional del Instituto de
However, this research reveals that projects have little impact on la Vivienda de la Facultad de
Arquitectura y Urbanismo de
people’s self-insurance and recovery capacities/strategies, which are usually la Universidad de Chile), May
not targeted within project implementation. This is illustrated in Figure 3 (forthcoming).
where the inclined fine and dashed arrows show no influence (i.e. they are
of the same length and inclination as before). Gathered data also indicate
that, in the long run, projects are not always as successful as the organ-
izations expect them to be since:
• established hierarchical and/or community structures disappear over
time (such as cooperative early-warning mechanisms and risk reduc-
tion committees);
• physical mitigation work is not maintained and is therefore not func-
tional after the life of the project;
• slum dwellers continue to use deficient coping strategies; these were
not influenced by the project measures and thus remain identical to
those that existed before the project started; and
• the incremental construction process is “built on” to the help received
(e.g. the height of retention walls increased, an additional floor added
to the house), thus undermining the effects of aid.
Furthermore, support structures for risk financing, continuous risk
reduction and incremental housing improvements are absent. The result
is a slowdown in the development process, or even a decline in the security
levels already achieved (Figure 3). This phenomenon is further related to
transitions within the low-income bracket. In slum communities, there
are the relatively well-off (the “rich poor”), a large middle segment (the
“poor poor”) and the relatively poor (the “destitute”). (31) Through project 31. Please note that Steve
Rayner is the originator of the
implementation, the relatively poor (having accessed full subsidies if these idea of varying vulnerability
are available) and the large middle segment (having obtained combined of the “destitute”, “poor poor”
credits/subsidies) can become part of the relatively rich poor. However, and “rich poor” (information
obtained orally through a
these improvements refer mainly to the physical/construction conditions discussion with Michael
rather than to the economic, social, organizational and institutional Thompson).
aspects needed to assure a consistent improvement in people’s security,
such as access to credit for incremental housing and risk reduction.
128
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

FIGURE 3
Impact of social housing project implementation on slum dwellers’
security and long-term vulnerability. RR = risk reduction

b. Provision and financing of social housing

Projects carried out by social housing organizations were described as
being divided into three main phases: preparation and community organ-
ization, construction through mutual and self-help, and credit repayment
(at times combined with further community development). The time
frame of the different phases varies considerably. Projects take 1–12 years
to complete, not including the repayment phase.
Infrastructure, social housing and housing improvements are financed
by a combination of government and non-government subsidies, micro-
credits, family savings, and mutual or self-help labour efforts (Figure 4).
The percentage of resources allocated to the different mechanisms differs
from organization to organization and from project to project. The average
cost of a new slum dwelling is around US$ 4,000–5,000, including land
and basic services.
To provide microcredits, social housing organizations often use seed
capital from donor organizations to create so-called revolving or rotating
funds; as these must maintain their original value, recovered credits are
reinvested in new loans to families in the same income bracket. Interest
rates range between 5–23 per cent and repayment periods between 4–10
years, during which time clients are carefully monitored. Detailed loan
analyses are carried out to screen the capacity of credit beneficiaries to
repay, and the clients’ ability to afford credit (i.e. the amount of credit
should be within 15–25 per cent of clients’ secure income level) and their
payment behaviour are checked regularly. Regarding the latter, potential
project beneficiaries usually have to save a specific amount of money

129
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

FIGURE 4
Combined financing mechanism for new settlement developments,
in situ housing construction and settlement upgrading

over a period of 6–18 months, and these savings are then used as a credit
down payment. Apart from housing microcredits, smaller sums of between
US$ 1,500–2,000 are loaned for housing improvements, micro-enterprise
development, land purchase and legalization.
A specialized finance department within the social housing organ-
ization generally manages the housing loan portfolios and subsidies,
or they are coordinated with, and sold to, private microfinancing insti-
tutions (MFIs), which are then responsible for credit repayment. So
far, Salvadoran social housing organizations have not worked through
financing cooperatives.
Only one of the social housing organizations analyzed, FUSAI, works
with government subsidies and credits from the National Public Housing
Fund FONAVIPO (as opposed to donations and own credit funds). FUSAI
uses its revolving fund as bridging finance until financial resources from
the national system can be accessed.
As far as completed social housing projects are concerned, none of
the organizations analyzed has a mechanism in place to offer their former
project beneficiaries microcredits for future risk reduction, housing im-
provements or housing enlargements. Furthermore, families wishing to
obtain such credits from other institutions cannot use the project houses
as collateral, as assisted housing cannot become bank property in the event
of default. This may have a negative impact on future developments, as
project beneficiaries are unable to use their assets effectively.
Specialized financing departments within social housing organ-
izations, and also cooperating MFIs, asserted that they are currently
developing or outsourcing a wider range of financial services for the poor.
Some MFIs, for instance, are developing additional savings schemes and
consumer loans based on remittances paid by family members living in
the United States. Neither type of scheme, however, is related to social
housing projects.

c. Risk and loss financing in social housing provision

It was ascertained that credits for social housing provision generally in-
clude compulsory life insurance to cover the debts. The insurance fees are

130
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

included in the monthly credit repayment, and are calculated on the basis
of the loan balance (around one per cent per year), or on the basis of the
total credit amount (around 0.005 per cent). Hence, monthly costs are
either fixed or decreasing, and generally range between US$ 0.25–0.80.
Depending on the insurance policy, the credit is either completely or
partly cancelled if the borrower dies. Three social housing organizations
have included life insurance as part of their credits only in the last three
years, after recent disasters. One of them is Habitat for Humanity, which
has adopted life insurance directly in combination with disaster property
insurance. This policy package costs around US$ 2.20 per month, a price
negotiated based on the organization’s yearly construction work in
El Salvador of around 600 housing units.
None of the other social housing organizations has yet added disaster
property insurance to their housing credit schemes. However, interest in
risk financing is slowly growing. In fact, Hurricane Stan in 2005 led to
ongoing negotiations among several social housing organizations and
national insurance companies. The insurance companies that were inter-
viewed offer disaster property insurance for social housing with premium
rates of 0.034 per cent of the house value plus administration fees, or rates
of a total of around US$ 2 per month. Damage from all types of natural
disasters is covered, as is damage related to construction errors. However,
representatives of insurance companies admit that the impact of such
an insurance policy on promoting risk reduction is poor, as mere com-
pliance with formal construction procedures is seen as a sufficient basis
for buying insurance.
After repayment, the insurance included in the housing credit scheme
is cancelled and is not replaced by any mechanism for financing future
damage. This is despite the fact that some representatives of the organ-
izations reported that around 4–5 per cent of their project houses, as well
as the organizations themselves (for instance, their private access roads),
have been affected by disasters.
Social housing organizations further provide special funds for non-
recoverable credits, by including a small percentage in the housing credit
(i.e. 0.5 per cent annually during the life of the debt, or a certain per-
centage included in the interest charges). Such funds are seldom used to
ease disaster-affected slum dwellers. An exception is the case of Habitat
for Humanity, which, after hurricane Stan, used its fund to amortize the
outstanding credits of six project beneficiaries living in a location that was
officially declared uninhabitable. None of the organizations provides a
financing mechanism that could finance the relocation of affected former
project beneficiaries.
Subsidies from national and international organizations are not
usually conditional upon an insurance policy being bundled with accom-
panying credits.
With regard to informal risk and loss financing, local mechanisms for
self-insurance and recovery are generally little supported and analyzed
within the framework of the projects.

d. Provision and financing of risk reduction measures

The social housing organizations interviewed stated that resources that
are earmarked for housing frequently cannot be used for risk reduction

131
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

measures. This legal constraint occasionally results in help being denied

to the most vulnerable slum dwellers, as their inclusion in projects would
require supplementary mitigation work. It was reported that, generally,
the resources available can only be allocated to construction and design
improvements that entail little or no extra cost. These costs, in turn, are
included in the individual credits of the project beneficiaries. Increased
costs are partly also compensated for by longer repayment periods. Existing
government funds for settlement upgrading, which were accessed by the
Salvadoran Vice-Ministry of Housing through the Inter-American Devel-
opment Bank, allow some flexibility with regard to the integration of risk
reduction, and can be obtained by participating in bidding processes.
However, according to social housing organizations, projects that include
risk reduction are difficult to get through such bidding procedures, as they
often involve higher project costs. This situation did not seem to concern
the representatives of donor and national government institutions, who
reported that risk reduction is sufficiently supported and guaranteed
through: market forces, existing formal construction procedures, project
work through local municipalities, and the participation of project
beneficiaries.
Thus, organizations willing to integrate risk reduction measures into
their project design are working increasingly with municipalities and a
range of other additional financial partners. “Soft” risk reduction measures,
such as risk analyses, the elaboration of mitigation plans and maps, or
capacity building for reducing economic, social, organizational and insti-
tutional vulnerabilities included in the project design, are generally fully
paid for by international, often post-disaster, donations. Only some large-
scale physical mitigation work is financed using subsidies from national
and municipal governments, usually through matching funds combined
with help from project beneficiaries.
Neither social housing organizations nor cooperating MFIs offer credits
for risk reduction, even though some reported a demand for these from
communities. Nationally, there is no established formal mechanism for
financing risk reduction. Funds are available only sporadically. For instance,
after Hurricane Stan in 2005, a temporary risk reduction fund was estab-
lished by FONAVIPO, offering individual subsidies of up to US$ 2,200.
Coping strategies for risk reduction are not generally supported by the
projects, (i.e. risk is generally reduced through alternative arrangements).
Only some coping efforts are fostered, and often only temporarily, for
instance through established institutional relations between the local and
national authorities. Furthermore, only a few mechanisms, such as par-
ticipatory work with beneficiaries, are in place to guarantee the long-term
maintenance or continuation of the risk reduction measures implemented.

e. The institutional landscape: the right to secure housing

versus market forces?
Widely differing organizational philosophies drive social housing pro-
jects. At one extreme are organizations, such as FUSAI, which design
housing financing mechanisms that are as close as possible to market
structures. Credits and revolving funds have to be financially sustainable
and competitive, and housing donations are only exceptionally used for
project implementation. Formal insurance, being a market instrument,

132
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

is seen as an important future and complementary mechanism. The in-

tegration of risk reduction into project design is criticized by some, as
increasing implementation periods and costs ultimately endanger organ-
izations’ financial sustainability. However, risk reduction is also identified
as an upcoming market, resulting in strategic engagement by these insti-
tutional actors.
At the other extreme are organizations, such as FUNDASAL, which see
housing first and foremost as a human right, and who work primarily with
subsidies from donor organizations and less with microcredits. Formal
insurance is seen as an inadequate mechanism for the low-income sectors
that would only increase costs and endanger the access to housing of the
poorest people. Increasingly, integral and comprehensive projects are the
aim, including risk reduction. The increased duration of projects and
the expenses involved are seen not in terms of cost but, rather, of having
the working focus right, that is, being dedicated to the urban poor.
While organizational philosophies differ strongly, focusing on social
patterns with either an individualistic or a communitarian bias, the working
approach for project implementation is not as diverse, and tends to be based
on community action and the establishment of hierarchical structures. In
fact, participation, mutual help and the establishment of community struc-
tures were identified as being among the most fundamental principles of
project implementation.

IV. GAPS AND CHALLENGES: HOUSEHOLD-LEVEL REALITY

VERSUS INSTITUTIONAL INTERVENTIONS
An understanding of the gaps between what households actually undertake
to reduce and finance their disaster risk, and the endeavours supported
by social housing organizations, yields important insights for reforming
social housing assistance to address disaster risk more effectively. The
challenges arising at the household and institutional levels, and the gaps
between them, can be summarized as follows:
• People cope with disasters and risk through risk reduction, self-
insurance and recovery strategies, while social housing organizations
mainly look at how to reduce risk, with a focus on physical risk and,
increasingly, also other types of risk. In this context, organizations
rarely analyze the key variables and causal loops underlying the
complex system of risk and disaster occurrence in a particular project
area, or consider existing local risk reduction strategies. Hence,
after project implementation, people usually continue to cope – as
before – without having obtained better structures for carrying out
and financing their own efforts for risk reduction, self-insurance or
recovery (Sections II.a–f and III.a–e).
• While people’s strategies for coping with disasters are hetero-
geneous, continuous and based mainly on individualistic behaviour,
organizations focus more on providing uniform, short-term and
community-based measures to tackle housing deficits and disaster
risk, while struggling to make the projects accessible to the poorest
(Sections II.b–f and III.a–e).
• Local coping strategies are diverse and crucial for the incremental
development of slum areas; they influence the key variables and
causal loops underlying the complex system of risk and disaster

133
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

occurrence in slum areas. Nevertheless, the strategies are insufficient

to keep pace with disaster impacts (Section II.a–g).
• There is an indication that incremental slum development does not
always correlate with reduced vulnerability, especially once people are
in transition to being relatively well-off (becoming part of the “rich
poor”). This is related to local processes and – in areas where pro-
jects have been carried out – to the nature of project interventions
(Sections II.f and III.a).
• Risk and loss financing is usually not integrated into housing finance
mechanisms (i.e. government and non-government subsidies, micro-
credits, family savings, and mutual or self-help) (Section III.c).
• Risk reduction measures are increasingly included in project design;
however, they are not financially sustainable and depend strongly on
donations, which are rarely available during “normal” times (Section
III.a and d).

While the list of gaps and challenges is long, there are also oppor-
tunities. Existing institutions and structures of housing provision and
financing provide a promising platform for supporting disaster risk man-
agement. There is ample scope for potentially beneficial interventions
that transcend purely constructive measures, leading to more sustainable
housing provision.

V. BRIDGING THE GAPS

The research revealed a demand for improvements from both the house-
hold and institutional levels in current approaches to risk reduction
and related practices. To address the challenges and gaps identified in
El Salvador, possible stakeholder-based options for social housing organ-
izations were developed. These reflect the analyses of the perceptions,
needs and capacities of the different stakeholder groups. As far as possible,
they were discussed with both the implementing organizations and the
slum dwellers. A review of practical experiences from other countries was
carried out to analyze, validate and complement the options identified
for El Salvador. The following sections provide an overview of practical
measures.

a. Framework for analyzing and supporting coping strategies

Since coping strategies were identified as crucial, although deficient, it is
essential within development efforts to consider encouraging and scaling
up selected strategies, as well as offering better alternative strategies where
needed. As shown in Figure 5, this could help achieve reduced disaster
impact through:
• improved risk reduction; and
• better “bouncing back” (in time and level) through adequate self-
insurance and recovery mechanisms.
To support coping strategies, the organizations interviewed indicated that
they would first need an adequate framework for viewing and analyzing
them. In Section II.c–e, a framework for analyzing coping strategies was
elaborated that provides a first indication as to how project measures could

134
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

FIGURE 5
Improved project implementation through improved risk reduction,
the integration of risk financing and the establishment
of recovery mechanisms. RR = risk reduction

be designed to constructively match up with the efforts of slum dwellers.

Coping strategies were, first, divided – based on their respective objectives –
into risk reduction, self-insurance and recovery; and second, grouped –
related to their thematic foci – into physical/technological, environmental,
economic, social/cultural, organizational and institutional aspects. Then,
within each of the thematic foci, they were classified into sub-groups that
express the more specific aims of the different measures (see left-hand
columns of Tables 1–8). Finally, based on the social patterns underlying
the different coping strategies, they were categorized as individualistic,
communitarian, hierarchical or fatalist.
Naturally, not all commonplace household measures can or should be
supported. Careful attention should be given to the cost-effectiveness and
sustainability of assistance. With regard to the latter, for operational pur-
poses, additional typologies are needed. Based on the research outcomes,
it is proposed to divide coping strategies into:
• those that can increase the capacity of slum communities to manage
urban disasters and disaster risk in both the short and long terms;
• those that increase capacities in the short term but decrease them in
the long term; and
• those that decrease capacities in both the short and long terms.
The focus of this division is not on individual, but on community,
gains. In fact, local communities or regions may be stuck in “poverty traps”
linked to individualistic social patterns and rational coping behaviour,
such as overexploitation of natural and other resources (e.g. excavating
slopes; cutting down trees; and removing stones from river beds to build

135
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

retention walls), or high population growth. However, such actions may

harm each person individually in the long run and/or contribute to the
greater impoverishment of the whole community. Identifying which
coping strategies are sustainable or do not work well, and why, is a major
challenge. Causal loop diagrams of different levels of detail can help in
this process, providing an understanding of the local context, local needs,
and people’s interrelated coping efforts (Section II.a).
Examples of coping strategies that, in the slum areas analyzed, proved
to be effective in the short term but might be ineffective in the long term
are: borrowing from money lenders at high interest rates; selling assets off
cheaply during the post-disaster period; spending money on temporary
arrangements (e.g. short-lived water barriers and channels); cutting down
trees to prevent them falling on nearby houses or to use as firewood to
save money; covering slopes with plastic sheets that pollute the environ-
ment, blow into rivers and block them up; and the “production” of
many children to provide informal social security. Parents do not always
take into account child-related costs (e.g. for education) or, where over-
crowding occurs, related negative externalities, such as congestion and
envionmental degradation.(32) 32. See reference 25. Note
that “short term” refers here
Examples of coping strategies that might be ineffective in both the to a longer period compared
short and long terms are, for instance: passive behaviour (owing to lost to the other listed examples,
trust in planning authorities or to the belief in divine forces); dysfunctional since even in communities
arrangements such as using corrugated iron as retention walls, or roofing where children’s labour is
accepted and relied on, it
houses with loose corrugated iron weighted with heavy objects that en- takes several years before the
danger neighbours during windstorms; and full dependence on family initial “investment” of time and
members’ labour or outside help. Examples of coping strategies that might expense eventually pays off.
be effective in both the short and long terms are: encouraging children
to study; cooperating with neighbours and the local executive committee
(e.g. for savings, mutual help or early warning); learning from friends and
others; accumulating assets for use as collateral or for sale in post-disaster
times without making a loss; reduction of unnecessary expenses; safe and
convenient saving arrangements; and accessing loans with favourable
conditions.

b. Scaling up (and down) of coping strategies

Based on the framework presented, social housing organizations and
cooperating MFIs can support or even scale up effective coping strategies
through formal and informal structures. Possible examples could be the
assistance for, or insurance of, local saving schemes; the establishment of 33. See reference 25; also see
reference 15, Twigg (2004).
linkages between formal and informal early warning systems to ensure
that the clients are informed in a timely fashion about potential disasters; 34. Note that the result of
and professional training for slum dwellers in low-budget structural risk new structures should be that
people are better off than
reduction.(33) Being able to use project housing as collateral when applying under the current situation. If
for credit would be another means of reducing barriers to coping among they only replace what is there,
slum dwellers. it is not a great help. Some
In addition to support for effective coping strategies, ineffective stra- of the most telling evidence
on crowding out comes from
tegies could be scaled down and alternatives offered.(34) In this context, South Africa (1993), when the
the creation of alternative formal or informal self-insurance mechanisms government extended basic
is of prime importance. Remittance-transfer schemes could be offered for pension benefits to black South
Africans, replacing informal
financing new housing and housing improvements, including disaster means of coping with aging
insurance for property. A borrower living in the United States could pay and economic downturns. See
an intermediary agent located in the United States who, in turn, would reference 25.

136
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

35. In contrast to related transfer the money to a social housing organization or cooperating MFI in
suggestions by Woo, the
El Salvador.(35) It must be emphasized that formal insurance represents an
responsible credit organization
would have their seat in El alternative to supporting informal self-insurance and recovery mechanisms
Salvador, not in the United only when these are very weak or expensive. To scale down related self-
States. See Woo, Gordon insurance mechanisms, formal money-transfer systems, which may be
(2001), “Risk acceptance
as a charitable donation”, more efficiently delivered than private transfers, could be offered,(36) as
Proceedings of the First well as capacity building for family planning. Further related examples
Annual IIASA-DPRI Meeting are presented in the following section.
on Integrated Disaster Risk
Management: Reducing
Socioeconomic Vulnerability,
1–4 August 2001, Laxenburg, c. Extending social housing financing mechanisms for risk
Austria. reduction and financing
36. Cox, Donald and Emmanuel The research indicates that the institutional and structural platform of
Jimenez (1997), “Coping with
apartheid: inter-household housing provision and financing can be used to foster disaster risk man-
transfers over the lifecycle in agement and overcome the gaps between the household and institutional
South Africa”, Working Paper, levels through, first, integrating risk and loss financing into existing
Boston College and World
Bank.
housing financing mechanisms (i.e. microcredit, subsidies and savings),
and second, expanding those mechanisms to finance risk reduction, (self-)
insurance and recovery for the urban poor (Figure 6). Such mechanisms
are needed not only during, but also after, project implementation, to
support incremental housing processes. Hence, it is crucial that social
housing organizations provide related regulatory and institutional frame-
works to improve households’ access to microcredits, subsidies, safe and
convenient savings opportunities, self- or micro-insurance and, where
possible, employment opportunities. These will be discussed now.
With institutional structures in place, additional microcredits for emer-
gencies, housing improvements and/or risk reduction could be offered
during and after project implementation. These could be monetary or
in the form of construction materials. Interviewees stated that adequate
credit portfolios for risk reduction need to be developed by social housing
organizations or their cooperating MFIs, not only for individuals but also
for communities, to reduce default risk. Furthermore, if social housing

FIGURE 6
Extending social housing financing mechanisms
to further support disaster risk management

137
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

organizations use revolving funds for housing credits, it was suggested 37. Note that conditionality
on of the part of international
that these could be used as bridging funds for risk reduction until gov-
donor agencies has to be
ernment subsidies were accessible. Alternatively, seed funds could be considered carefully so as
provided by donor organizations to create complementary revolving not to hinder successful
funds for risk reduction and emergency loans. With regard to subsidies, partnerships for poverty
reduction. See DFID (2005),
both government and non-government donors could make access to “Partnerships for poverty
housing subsidies conditional upon accompanying microcredits being reduction: rethinking
bundled with property disaster insurance and life insurance.(37) conditionality”, UK Policy Paper,
There is widespread positive experience with saving arrangements in DFID, UK.
Asia, for instance in Bangladesh, India and Indonesia.(38) However, the 38. For example, Rutherford,
present research supports Morduch, who indicates that in Latin America Stuart (1999), “The poor
there has also been an overhaul of savings-related programmes.(39) “Com- and their money”, Finance
and Development Research
pulsory savings” are already included within the framework of social Programme Working Paper
housing projects to evaluate potential borrowers’ payment behaviour. Series 3/1999, University
Regular deposits are made to build up collateral against loans, which of Manchester, Institute for
cannot normally be withdrawn while loan repayments are outstanding Development Policy and
Management, Manchester.
(Section III.b). These existing saving mechanisms could, interviewees
confirmed, be expanded to offer independent emergency savings schemes 39. See reference 25.
during and after project implementation. Such schemes, combined with
contingency credit facilities, could meet the immediate post-disaster 40. D’Cruz, Celine and David
Satterthwaite (2005), “Building
needs of slum dwellers in terms of replacing or repairing lost assets. So- homes, changing official
called crisis credits are, for instance, being used successfully by urban approaches: the work of
poor federations.(40) Furthermore, social housing organizations or their urban poor organizations and
their federations and their
cooperating MFIs could insure the compulsory savings. Alternatively, a contributions to meeting the
certain percentage of beneficiaries’ savings could be put into an emergency Millennium Development Goals
fund, which could be made available quickly to disaster-affected borrowers in urban areas”, Working Paper
in the form of emergency loans.(41) Such funds could also be created on Poverty Reduction in Urban
Areas Series, Number 16, July,
through housing microcredit instalments. At the institutional level, organ- IIED, London. See www.iied.
izational emergency reserves could be created, eventually backed by org/pubs/pdf/full/9547IIED.pdf.
international donors.
The institutional level “snapshot” revealed that micro-insurance can be 41. See reference 15, Twigg
(2004).
included in housing microcredits, if additional costs can be accepted by
project beneficiaries or be compensated by other means. The work through 42. In Manizales, Colombia, the
MFIs or financing cooperatives could allow the establishment of insurance city has made an agreement
with an insurance company to
solidarity schemes, through which policies for the poor are subsidized or allow any resident to purchase
fully paid from policies sold to people in higher-income groups.(42) Such insurance coverage through
schemes can even be legislated for, as in India.(43) Increasing experience the municipal tax collection
and lessons learned by social housing organizations and cooperating MFIs system. Once 30 per cent of
the insurable buildings in the
regarding microcredits can be partly translated into micro-insurance, for metropolitan area participate
instance, the selection of applicants, payment mechanisms, follow-up in the plan, the insurance
and reduction of transaction costs. Independent insurance policies could coverage extends to the
thus be offered after credit repayment, and eventually even to poor properties of the poorest
population sectors that are
slum dwellers who cannot access housing credits. In general, insurance exempt from property tax. See
companies indicated that they cover only stochastic and unpredictable The World Bank (Marianne Fey
(i.e. highly uncertain) events. Hence, community insurance policies (as editor) (2005), The Urban Poor
in Latin America, Directions
opposed to individual policies aimed only at slum dwellers living in high- in Development, Washington
risk areas) might be a solution. DC. See
In contrast with existing insurance arrangements in El Salvador, careful http://siteresources.worldbank.
attention should be given to encouraging risk reduction rather than “moral org/INTLACREGTOPURBDEV/
Home/20843636/
hazard”.(44) Insurance should not discourage people from taking steps to UrbanPoorinLA.pdf.
reduce physical risk, nor encourage them to take even greater risk. This
can be achieved by means of: 43. Mechler, Reinhard and
Joanne Bayer (2006), Disaster
• index-based insurance schemes where claims are independent of Insurance for the Poor? A
Review of Microinsurance
losses;(45)
138
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

for Natural Disaster Risks • schemes where people who have carried out constructive mitigation work
in Developing Countries,
ProVention Consortium/IIASA
pay lower premiums than those taking no measures to reduce risk; and
Study, July 2006. • insurance policies which make access conditional on risk reduction.

44. The term “moral hazard” is With regard to the latter, in Fiji a structural engineer must certify that
generally used to describe the houses have certain cyclone-resistant features before owners can access dis-
situation in which someone
aster property insurance.(46) International donors could provide support
insured against disaster risk
will ignore disaster risk, or regarding technical and administrative aspects of insurance schemes.(47)
even purposely engage in They could also offer community insurance policies or reinsurance in co-
risky behaviour, knowing that operation with national NGOs and national or international insurers.(48)
any costs incurred will be
compensated by the insurer. Even if organizations decide not to include disaster property insurance
in social housing credits, they could lobby governments or commercial
45. Linnerooth-Bayer, Joanne, insurance firms to cover at least schools, bridges and hospitals that serve
Reinhard Mechler and Georg
Pflug (2005), “Refocusing the poor.
disaster aid”, Viewpoint, Based on experience in the health sector, it is clear that, in order to
Science Vol 309, pages become successful, micro-insurance needs to be complemented by non-
1044–1046.
financial preventive measures.(49) In the housing sector, for instance,
46. See reference 15, Twigg disaster property insurance could be linked to preventive construction
(2004). programmes that involve training of community construction workers,
47. Sinha, S (2002), Strength
or the establishment of village advisory services. Ideally, social housing
in Solidarity: Insurance for organizations or housing financing MFIs would offer risk reduction meas-
Women Workers in the Informal ures to ensure that credits are paid back and no insurance claims become
Economy, Self-Employed necessary.
Women’s Association (SEWA),
Ahmedabad, India. Finally, targeted transfers could provide a kind of self-insurance for the
poor (e.g. workfare programmes and employment guarantee schemes).(50)
48. Obermayer, Bastian For instance, programmes for individuals or communities affected by
and Marc Baumann (2006),
“Lebensversicherung für 80 disasters, offered by social housing organizations and/or governments,
cent”, GDV Position Vol 48, May, could provide contingent transfers to finance labour for reconstruction.
Gesamtverband der Deutschen
Versicherungswirtschaft (GDV).

49. In India, the NGO SEWA d. Matching heterogeneity

provides street vendors with
umbrellas, based on its life This research demonstrates that coping strategies are not homogeneous
insurance data that the sun but, rather, are distinct with regard to their objectives, thematic foci,
could be increasing their risk of
illness. See reference 47. See underlying patterns of social relations, sustainability and effectiveness.
also Brown, Warren and Craig In Section V.b, it was argued that it is important and possible to take
F Churchill (2000), Insurance advantage of this heterogeneity to design projects that appeal to the
Provision in Low-income
Communities, Part II: Initial various perspectives and efforts of the related stakeholders. But this is
Lessons from Micro-insurance only the start. People do not just have strategies. As identified by “cul-
Experiments for the Poor, tural theory”,(51) under certain conditions they can move from the under-
Development Alternatives,
Microenterprise Best Practices
lying social pattern of one strategy to another pattern.(52) Projects can help
Project, USA. or hinder such transitions. Hence, in selecting project measures, devel-
opment organizations should be sensitive to the social forms of behaviour
50. See reference 28; also see
identified in a specific slum, so that the social patterns can be matched or,
reference 25.
where needed, channelled.
51. See reference 10. Introducing systems of mutual rights, accountability and community
organization to support a shift in social patterns could be an important
52. Thompson, Michael and
Aaron Wildavsky (1986), “A step in overcoming asymmetric risk (Section II.a). However, this needs to
poverty of distinction: from be followed up over time (i.e. after project implementation) and needs to
economic homogeneity to be complemented by further strategies, such as trying to ease people away
cultural heterogeneity in the
classification of poor people”,
from fatalism and offering solutions for improved individual coping. Im-
Policy Science Vol 19, proved local coping strategies are crucial for the process of incremental
pages 163–199. housing and security, particularly after project implementation. Section V.c
provided a variety of related measures for achieving this.

139
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 19 No 1 April 2007

The philosophies, which drive the social housing organizations

that service slum dwellers, differ greatly (Section III.e). However, project
measures are quite uniform and uni-dimensional, with development aid
generally being understood as assistance for mutual community help. A
greater variety of project measures is needed to also take into account
the strategizing heterogeneity of slum dwellers’ efforts. Moreover, where
predominantly commercial mechanisms, such as formal insurance, are
inappropriate, alternative informal structures, such as those described in
Section II.d, could be considered.

VI. CONCLUSIONS
The research shows how dwellers living in 15 slum communities in
El Salvador cope with disasters and disaster risk, and reveals the variety
of strategies, tactics and mechanisms they have elaborated to reduce risk,
to insure themselves and to recover fast if disasters occur. The strategies
are based on different patterns of social behaviour, with a strong focus
on individualistic behaviour for survival. Coping strategies are crucial for
the incremental, step-by-step development of slums, and influence the
key variables and causal loops underlying the complex system of risk and
disaster occurrence. However, although they entail considerable costs for
the poor – on average 9.2 per cent of their income – they are not always
sufficient to keep pace with the frequency of disasters.
Until recently, social housing organizations paid little attention to
disasters and disaster risk. This is unfortunate as “pro-poor” housing and
settlement development is not necessarily synonymous with vulnerability
reduction. Greater wealth and better living standards may not themselves
reduce risk. While increasing efforts are made to mainstream risk reduc-
tion in both project design and implementation, there are hardly any
mechanisms for financing risk, loss, and risk reduction. The lack of such
mechanisms is unfortunate as there is increasing evidence that risk and
disaster shocks are a major cause of lower growth, reduce poor people’s
income growth, and possibly cause “poverty traps”.(53) 53. See reference 28; also
Mechler, Reinhard (2004),
This research indicates that social housing organizations and co- Natural Disaster Risk
operating MFIs have the potential to provide a powerful platform to Management and Financing
support disaster risk management. For improved project design and im- Disaster Losses in Developing
plementation, first, risk reduction needs to be improved to reduce disaster Countries, Verlag für Versicher-
ungswissenschaft, Karlsruhe.
impact, and second, adequate (self-) insurance and recovery mechanisms
need to be integrated to improve people’s chances of “bouncing back”
quickly and to a reasonable level after a disaster. These objectives can be
achieved by:
• gaining a better understanding of what urban dwellers perceive as
disasters and disaster risk as well as their underlying drivers;(54) 54. Related studies can be
supported by systems analysis
• encouraging and scaling up effective (i.e. sustainable) coping and its causal loop diagrams.
strategies;(55) See references 13 and 14.
• crowding out of unsustainable coping strategies by offering alternative
55. Their identification is
formal or informal mechanisms; possible through the presented
• reducing barriers to coping; analysis framework.
• integrating risk and loss financing into existing social housing financ-
ing mechanisms (i.e. microcredits, government and non-government
subsidies, and family savings);

140
 RISK REDUCTION AND FINANCING FOR THE URBAN POOR

• expanding existing social housing financing mechanisms to finance

risk reduction, (self-) insurance and recovery for the urban poor;
• offering multi-dimensional measures to match and, where needed,
channel the patterns of social behaviour that underlie people’s ways
of coping;
• given incremental development processes in slums, offering mech-
anisms that work or come into effect after project implementation;
and
• improving trust and relations between national, municipal and
local authorities and the slum dwellers (e.g. through improved com-
munication structures and the creation of community rights and
obligations).
In the El Salvador case studies, the emphasis was on developing a
grounded theory on the existing situation, and options as to how it could
be improved. The generalization of the outcomes will be validated in a
subsequent study.

REFERENCES

Agunias, Dovelyn (2006), “Remittance trends in Ecology and Environmental Engineering 2004,
Central America”, Migration Information Source Vol 1, Institute of Chemical Engineering, Lund
(MFI), April 1. See www.migrationinformation. University.
net/pdf/Migration_Information_Source-Central_ IFRC/RCC (International Federation of Red Cross and
America.pdf. Red Crescent Societies) (2004), World Disasters
Report 2004, IFCR/RCC, Geneva.
Brown, Warren and Craig F Churchill (2000), Insurance
Lavell, Allan (1994), “Prevention and mitigation of
Provision in Low-income Communities, Part II: Initial
disasters in Central America: vulnerability to dis-
lessons from Micro-insurance Experiments for the
asters at the local level”, in Ann Varley (editor),
Poor, Development Alternatives, Microenterprise
Disasters, Development and Environment, John
Best Practices Project, USA.
Wiley & Sons, Chichester, pages 49–64.
Cox, Donald and Emmanuel Jimenez (1997), “Coping
Linnerooth-Bayer, Joanne, Reinhard Mechler and
with apartheid: inter-household transfers over the Georg Pflug (2005), “Refocusing disaster aid”,
lifecycle in South Africa”, Working Paper, Boston Viewpoint, Science Vol 309, pages 1044–1046.
College and World Bank. Mechler, Reinhard (2004), Natural Disaster Risk
D’Cruz, Celine and David Satterthwaite (2005), Management and Financing Disaster Losses in
“Building homes, changing official approaches: Developing Countries, Verlag für Versicherungswis
the work of urban poor organizations and their senschaft, Karlsruhe.
federations and their contributions to meeting the Mechler, Reinhard and Joanne Bayer (2006), Disaster
Millennium Development Goals in urban areas”, Insurance for the Poor? A Review of Microinsurance
Working Paper on Poverty Reduction in Urban for Natural Disaster Risks in Developing Countries,
Areas Series, Number 16, July, IIED, London. See ProVention Consortium/IIASA Study, July 2006.
www.iied.org/pubs/pdf/full/9547IIED.pdf. Morduch, Jonathan, (1999), “Between the state and the
Dercon, Stefan (2003), “Poverty traps and development: market: can informal insurance patch the safety
the equity efficiency trade-off revisited”, Agence net?”, The World Bank Research Observer Vol 14,
Française de Développement (AFD) and the No 2, August, pages 187–207.
European Development Research Network (EUDN), Obermayer, Bastian and Marc Baumann (2006),
Paper presented at the conference on Poverty, “Lebensversicherung für 80 cent”, GDV Position
Inequalities and Growth: What’s at Stake for Vol 48, May, Gesamtverband der Deutschen
Development Aid?, 13–15 November 2003, Paris. Versicherungswirtschaft (GDV).
DFID (2005), “Partnerships for poverty reduction: rethink- Rutherford, Stuart (1999), The Poor and Their Money,
ing conditionality”, UK Policy Paper, DFID, UK. Finance and Development Research Programme
Glaser, Barney G and Anselm L Strauss (1967), The Working Paper Series 3/1999, University of Man-
Discovery of Grounded Theory – Strategies for Quali-tative chester, Institute for Development Policy and
Research, Aldine Publishing Company, Chicago. Management, Manchester.
Haraldsson, Hördur V (2004), “Introduction to system SELA (Latin American Economic System) (2005),
thinking and causal loop diagrams”, Reports on Migrations and Remittances in Latin America and the

141
E N V I R O N M E N T & U R B A N I Z AT I O N Vol 18 No 2 October 2006

Caribbean: Intra-Regional Flows and Macroeconomic UNISDR (2006), “Disaster statistics 1991–2005”,
Determinants, XXXI Regular Meeting of the Latin accessible at www.unisdr.org/disaster-statistics/
American Council, Caracas, Venezuela, 21–23 introduction.htm.
November 2005. UN–HABITAT (2003), The Challenge of Slums, Global
Sinha, S (2002), Strength in Solidarity: Insurance for Report on human settlements, Earthscan,
Women Workers in the Informal Economy, Self- London.
Employed Women’s Association, Ahmedabad, Walker, Thomas S and James G Ryan (1990), Village and
India. Household Economies in India’s Semi-arid Tropics,
Sterman, John D (2000), Business Dynamics: Systems Johns Hopkins University Press, Baltimore.
Thinking and Modelling for a Complex World, Irwin/ Wamsler, Christine (2006), “Understanding disasters
McGraw-Hill. from a local perspective: insights into improving
Thompson, Michael and Aaron Wildavsky (1986), “A assistance for social housing and settlement devel-
poverty of distinction: from economic homo- opment”, TRIALOG (Journal for Planning and
geneity to cultural heterogeneity in the classi- Building in the Third World) No 91, December,
fication of poor people”, Policy Science Vol 19, special issue on “Building on disasters”.
pages 163–199. Wamsler, Christine (2006), “Integrating risk reduction,
Thompson, Michael, Richard Ellis and Aaron Wildavsky urban planning and housing: lessons from El
(1990), Cultural Theory, Westview Press, Inc, Oxford. Salvador”, Open House International (OHI) Vol 31,
Twigg, John (2004), Disaster Risk Reduction: Mitigation No 1, pages 71–83.
and Preparedness in Development and Emergency Wamsler, Christine (2007), “Integrando la gestión del
Programming, Good Practice Review, Number 9, riesgo, planificación urbana y vivienda social:
March, Overseas Development Institute (ODI), lecciones de El Salvador”, INVI (Revista Inter-
London. nacional del Instituto de la Vivienda de la Facultad
UNDP (2004), Reducing Disaster Risk: A Challenge for de Arquitectura y Urbanismo de la Universidad de
Development, UNDP, New York. Chile), May (forthcoming).
UNDP (2005), Informe Sobre Desarrollo Humano El Woo, Gordon (2001), “Risk acceptance as a charitable
Salvador 2005: Una Mirada al Nuevo Nosotros. El donation”, Proceedings of the First Annual
Impacto de las Migraciones, San Salvador, UNDP, IIASA-DPRI Meeting on Integrated Disaster Risk
San Salvador. Management: Reducing Socioeconomic Vul-
UNISDR (2004), Living with Risk: A Global Review of nerability, 1-4 August 2001, Laxenburg, Austria.
Disaster Reduction Initiatives, Inter-Agency Secre- World Bank (Marianne Fey editor) (2005), The Urban
tariat of UNISDR, accessible at www.unisdr.org/ Poor in Latin America, Directions in Develop-
eng/about_isdr/bd-lwr-2004-eng.htm. ment, Washington DC. See http://siteresources.
UNISDR (2005), “Hyogo framework for action 2005– worldbank.org/INTLACREGTOPURBDEV/
2015: building the resilience of nations and com- Home/20843636/UrbanPoorinLA.pdf.
munities to disasters”, accessible at www.unisdr. www unmillenniumproject.org.
org/wcdr/intergover/official-doc/L-docs/Hyogo-
framework-for-action-english.pdf

142