Hazards of Nature

ENHANCING DEVELOPMENT EFFECTIVENESS THROUGH EXCELLENCE AND INDEPENDENCE IN EVALUATION

The Independent Evaluation Group (IEG) reports directly to the Bank’s Board of Executive Directors. IEG assess-
es what works, and what does not; how a borrower plans to run and maintain a project; and the lasting contri-
bution of the Bank to a country’s overall development. The goals of evaluation are to learn from experience, to
provide an objective basis for assessing the results of the Bank’s work, and to provide accountability in the
achievement of its objectives. It also improves Bank work by identifying and disseminating the lessons learned
from experience and by framing recommendations drawn from evaluation findings.

INDEPENDENT EVALUATION GROUP

Study Series

2004 Annual Review of Development Effectiveness: The Bank’s Contributions to Poverty Reduction
Addressing the Challenges of Globalization: An Independent Evaluation of the World Bank’s Approach to Global Programs
Agricultural Extension: The Kenya Experience
Assisting Russia’s Transition: An Unprecedented Challenge
Bangladesh: Progress Through Partnership
Brazil: Forging a Strategic Partnership for Results—An OED Evaluation of World Bank Assistance
Bridging Troubled Waters: Assessing the World Bank Water Resources Strategy
Capacity Building in Africa: An OED Evaluation of World Bank Support
The CIGAR at 31: An Independent Meta-Evaluation of the Consultative Group on International Agricultural Research
Country Assistance Evaluation Retrospective: OED Self-Evaluation
Debt Relief for the Poorest: An OED Review of the HIPC Initiative
Developing Towns and Cities: Lessons from Brazil and the Philippines
The Drive to Partnership: Aid Coordination and the World Bank
Economies in Transition: An OED Evaluation of World Bank Assistance
The Effectiveness of World Bank Support for Community-Based and –Driven Development: An OED Evaluation
Evaluating a Decade of World Bank Gender Policy: 1990–99
Evaluation of World Bank Assistance to Pacific Member Countries, 1992–2002
Financial Sector Reform: A Review of World Bank Assistance
Financing the Global Benefits of Forests: The Bank’s GEF Portfolio and the 1991 Forest Strategy and Its Implementation
Fiscal Management in Adjustment Lending
IDA’s Partnership for Poverty Reduction
Improving the Lives of the Poor Through Investment in Cities
India: The Dairy Revolution
Information Infrastructure: The World Bank Group’s Experience
Investing in Health: Development Effectiveness in the Health, Nutrition, and Population Sector
Jordan: Supporting Stable Development in a Challenging Region
Lesotho: Development in a Challenging Environment
Mainstreaming Gender in World Bank Lending: An Update
Maintaining Momentum to 2015? An Impact Evaluation of Interventions to Improve Maternal and Child Health and Nutrition Outcomes in Bangladesh
The Next Ascent: An Evaluation of the Aga Khan Rural Support Program, Pakistan
Nongovernmental Organizations in World Bank–Supported Projects: A Review
Poland Country Assistance Review: Partnership in a Transition Economy
Poverty Reduction in the 1990s: An Evaluation of Strategy and Performance
The Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank’s Support Through 2003
Power for Development: A Review of the World Bank Group’s Experience with Private Participation in the Electricity Sector
Promoting Environmental Sustainability in Development
Putting Social Development to Work for the Poor: An OED Review of World Bank Activities
Reforming Agriculture: The World Bank Goes to Market
Sharing Knowledge: Innovations and Remaining Challenges
Social Funds: Assessing Effectiveness
Tunisia: Understanding Successful Socioeconomic Development
Uganda: Policy, Participation, People
The World Bank’s Experience with Post-Conflict Reconstruction
The World Bank’s Forest Strategy: Striking the Right Balance
Zambia Country Assistance Review: Turning an Economy Around

Evaluation Country Case Series

Bosnia and Herzegovina: Post-Conflict Reconstruction
Brazil: Forests in the Balance: Challenges of Conservation with Development
Cameroon: Forest Sector Development in a Difficult Political Economy
China: From Afforestation to Poverty Alleviation and Natural Forest Management
Costa Rica: Forest Strategy and the Evolution of Land Use
El Salvador: Post-Conflict Reconstruction
India: Alleviating Poverty through Forest Development
Indonesia: The Challenges of World Bank Involvement in Forests
The Poverty Reduction Strategy Initiative: Findings from 10 Country Case Studies of World Bank and IMF Support
Uganda: Post-Conflict Reconstruction

Proceedings

Global Public Policies and Programs: Implications for Financing and Evaluation
Lessons of Fiscal Adjustment
Lesson from Urban Transport
Evaluating the Gender Impact of World Bank Assistance
Evaluation and Development: The Institutional Dimension (Transaction Publishers)
Evaluation and Poverty Reduction
Monitoring & Evaluation Capacity Development in Africa
Public Sector Performance—The Critical Role of Evaluation

IEG PUBLICATIONS

All IEG evaluations are available, in whole or in part, in languages other than English. For our multilingual selection, please visit
http://www.worldbank.org/ieg

Hazards of Nature_cover.qxd 5/4/06 1:39 PM Page 2

© 2006 The International Bank for Reconstruction and Development / The World Bank
1818 H Street, NW
Washington, DC 20433
Telephone 202-473-1000
Internet www.worldbank.org
E-mail feedback@worldbank.org

The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the
views of the Board of Executive Directors of the World Bank or the governments they represent.

The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors,

denominations, and other information shown on any map in the work do not imply on the part of the World Bank any
judgment of the legal status of any territory or the endorsement or acceptance of such boundaries.

Rights and Permissions
The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission
may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant
permission promptly.

For permission to photocopy or reprint any part of this work, please send a request with complete information to the

Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470,
www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher,

World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org.

Photo credit: Banda Aceh, Northern Sumatra, Indonesia, January 2005. © Martin Adler, courtesy of Panos Pictures.

ISBN-10 0-8213-6650-5
ISBN-13: 978-0-8213-6650-9
e-ISBN 0-8213-6651-3
DOI: 10.1596/978-0-8213-6650-9

Library of Congress Cataloging-in-Publication Data has been applied for.

Ainsworth, Martha, 1955–

Committing to results : improving the effectiveness of HIV/AIDS assistance: an OED evaluation of the World Bank’s

assistance for HIV/AIDS control / Martha Ainsworth, Denise A. Vaillancourt, Judith Hahn Gaubatz.

p. cm. — (Operations evaluation studies)

Includes bibliographical references.

ISBN-13: 978-0-8213-6388-1
ISBN-10: 0-8213-6388-3

1. Economic assistance—Developing countries—Evaluation. 2. AIDS (Disease)—Economic aspects—Developing

countries. 3. HIV infections—Economic aspects—Developing countries. 4. AIDS (Diseasse)—Developing
countries—Prevention. 5. HIV infections—Developing countries—Prevention. 6. World Bank. I. Vaillancourt, Denise.
II. Hahn Gaubatz, Judith. III. Title. IV. World Bank operations evaluation study.

HC60.A4575 2005

362.196'9792'0091726—dc22
2005052329

Printed on Recycled Paper

World Bank InfoShop

E-mail: pic@worldbank.org

Telephone: 202-458-5454

Facsimile: 202-522-1500

Independent Evaluation Group

Knowledge Programs and Evaluation Capacity

Development (IEGKE)

E-mail: eline@worldbank.org

Telephone: 202-458-4497

Facsimile: 202-522-3125

Relevance of Bank Assistance

Defining Relevance of Disaster Assistance

Delivering Quick Support for the Immediate Response

Delivering New Projects for the Recovery

Reducing Vulnerability over the Longer Term

Does the Bank Respond Quickly?

Using a Long-Term View to Select Short- and Medium-Term Actions

Social Dimensions of Disaster

Participation in Natural Disaster Response

Shelter and Housing

Relocation

Vulnerable Groups

Bank Policy: Implementation and Implications

Natural Disasters in Relation to Other Emergencies

Emergency Recovery Lending Timeframe

Relief and Consumption Expenditures

Forms of Disaster Assistance

Rebuilding Physical Assets and Restoring Economic and
Social Activities

Recurring Disaster Events

Disaster-Resistant Construction Standards

Emergency Preparedness Studies

Technical Assistance on Prevention and Mitigation Measures

Addressing Longer-Term Disaster Issues

Procurement

Donor Coordination

Institutional Development

Conclusions and Recommendations

What Works in Developed Countries May Not Work Elsewhere

When to Engage and How to Stay Engaged

The Challenges Ahead: A Review of the Evidence

Lessons

Recommendations

Appendixes

Bank Policy and Products

Study Methodology

Supplemental Data—Chapter 2

109 D:

Survey Results

129 E:

Supplemental Data—Chapter 3

133 F:

Supplemental Data—Chapter 4

147 G:

Cash Support

153 H:

Housing and Shelter

159 I:

External Advisory Panel Comments

161 J:

Management Response

171 K:

Chairman’s Summary: Committee on Development Effectivess
(CODE)

i v

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

175

Endnotes

179

References

Boxes

1.1

Why Do Natural Disasters Seem to Be Increasing in Number?

1.2

A Troubled Relief Compromises the Ability to Recover

2.1

What Are Emergency Recovery Loans?

2.2

Other Organizations Have Institutional Structures for Disaster Risk
Management

3.1

Disaster Damage Undermines Development Progress

3.2

To What Extent Do CASs Develop Disaster-Specific Strategies
Appropriate to Prevailing Hazards?

4.1

Equipment Acquisition Can Be Particularly Problematic

4.2

A Sense of Urgency Can Lead to a Wasteful Response

4.3

Reducing Vulnerability May Also Reduce Poverty

4.4

Bank Helps Finance Insurance Scheme in Turkey

4.5

Damage Sustained Is a Function of a Community’s Level
of Vulnerability

4.6

Reducing Vulnerability Takes Time

4.7

Damage Assessments Are Useful But Have Some Shortcomings

5.1

The Poor Take the Heaviest Blow

5.2

Social Funds Can Be Part of a Rapid, Locally Based Response

6.1

Contingency Financing—A Learning Process

6.2

What Happens When the Bank Does Not Stay Engaged?

Figures

1.1

The Cost of Disaster Damage Is Rising

2.1

The Number of Projects Related to Natural Disasters Has Been
Rising, with Sharp Peaks in Lending about Every Five Years

2.2

The Bank Has Most Often Provided Lending for Disasters in Africa
and in Latin America and the Caribbean

2.3

Reallocations Peaked Following Each Policy Revision

2.4

Natural Disaster Portfolio Ratings: Projects Approved and
Completed, 1984–2005

4.1

Average Implementation and Extension Times for Projects
Containing Disaster Activities: All Disaster Projects

4.2

How Long Do Projects Take?

5.1

Projects Are Increasingly Addressing Women’s Vulnerabilities

6.1

The Use of ERLs Has Increased with Each Policy Revision

6.2

ERLs by Natural Disaster Type

6.3

Focus on Mitigation Has Increased with Each Successive Project
in Turkey

6.4

The Bank Often Leads Coordination, But Borrower Role Is Growing

Tables

1.1

Key Provisions of Current Bank Policy Statement

2.1

Natural Disaster Portfolio Composition, 1984 to 2005

2.2

Concentration of Lending in the Disaster Portfolio Compared with
Overall Bank Lending

C O N T E N T S

v i i

Acknowledgments

his evaluation of the World Bank’s experience with natural disasters was
done by the Independent Evaluation Group–World Bank at the re-
quest of the Bank’s Board of Executive Directors. It comes at an ap-

propriate time. Natural disasters are affecting development in many countries,
setting back hard-won development gains.

In consequence, lending for natural disasters is
a growing business for the Bank. The lessons
produced by this evaluation are expected to
inform good practice and ensure the achieve-
ment of results in Bank activities. The evaluation
is also intended for use in an ongoing revision of
the Bank’s policy statement on emergency
assistance.

The evaluation was conducted under the

leadership of Ronald S. Parker, and this report
was written by Ronald Parker and William
Hurlbut, with inputs from Anna Amato, Mark
Emmert, Silke Heuser, and Kristin Little. Helen
Phillip provided administrative support. Caroline
McEuen edited the manuscript for publication.
Peer reviewer Alexandra Ortiz provided valuable
comments on earlier drafts of the report.

The authors gratefully acknowledge the

members of the Advisory Committee—Mary
Anderson, Ian Davis, Prema Gopalan, and Franklin
McDonald—for their early guidance and their

patient and detailed review of drafts of this report.
We also gratefully acknowledge the financial
support provided by our partner, the Swiss Agency
for Development and Cooperation (SDC), which
made possible a considerable amount of the
background research upon which this evaluation is
based. We also thank Margaret Arnold, Alcira
Kreimer, and Zoe Trohanis of the Hazard Manage-
ment Unit, and Bank task manager Francis
Ghesquiere. The study conducted several field
missions. The authors thank Y´yld´yz Aydin, Murat
Sungur Bursa, Dawn French, Ali Ihsan, Cheryl
Mathurin, Francisco Rivas, S.A.M. Rafiquzzaman,
Kuljit S. Sidhu, and Krishna S. Vatsa for their support
and valuable insights during these missions.

A database of project information created for

this evaluation brought together for the first
time all the available information on every Bank
project that had disaster-related activities. This
database will be turned over to the Hazard Risk
Management Team of the Urban Unit.

Director-General, Evaluation: Vinod Thomas

Director, Independent Evaluation Group, World Bank: Ajay Chhibber

Manager, Sector, Thematic, and Global Evaluation: Alain Barbu

Task Manager: Ronald S. Parker

i x

Foreword

atural disasters occur throughout the world, but their economic and
social impacts have been increasing and are generally much greater
in developing countries than in developed ones. Disasters can wipe

out development gains and eclipse years of development investment.

In Mozambique, Bank lending financed the
construction of 487 schools over a 20-year
period, but just one recent disaster, the floods of
2000, damaged or destroyed about 500 primary
schools as well as 7 secondary schools. The
Kashmir earthquake of October 2005 caused an
estimated $5 billion in damage in Pakistan,
roughly equivalent to the total official develop-
ment assistance for the preceding 3 years, and
equivalent to the amount the World Bank had
lent to the country over the preceding 10 years.

Until recently, disasters were treated as one-

time, random events by governments and the
agencies that helped them respond. But we
know from experience that disasters strike with
regular periodicity—and repeatedly in some
parts of the world. The potential for disaster is
foreseeable to the extent that it is possible to
predict generally where an event is likely to
occur at some time in the near future (but not
precisely when or its magnitude). Hence, low-
lying coastal areas on the Bay of Bengal will
experience more flooding, and small island
states in the Caribbean and countries along the
Gulf of Mexico will be repeatedly hit by
hurricanes.

The Bank and much of the development

community must take into account in their
strategies how frequently disasters occur, and
how often they strike the same countries. The
Bank’s lending shows that disasters follow a
clear pattern: Ten borrowers accounted for 208
of the 528 disaster-related projects (39 percent)
in the portfolio over the 1984–2005 period. Bank
lending commitments also are concentrated—
7.5 percent of projects received 32 percent of
the financing.

Disaster is still sometimes treated as an

interruption in development rather than as a
risk to development in the approaches of both
the country and the Bank. Of current assistance
strategies for countries that have received Bank
support in natural disasters, 44 percent did not
mention them. Even in the 40 countries that
have had 4 or more disaster projects, one-third
of the strategies did not mention disaster. And,
for the subset of countries that had an extensive
history with disasters (more than 8), about a
third did not mention disasters at all. Project
loan documents rarely consider natural events
as a risk, even in highly vulnerable countries,
although 176 projects were adversely affected

by disaster during implementation. We need to
find ways to integrate these risks more centrally
into development assistance to improve
effectiveness.

It should be recognized that the Bank has

demonstrated considerable flexibility in its
approach to natural disasters and has learned to
manage large and small responses well. Bank
staff have often been innovative and have
demonstrated the capacity to manage massive
reconstruction on many levels: more than 60
different kinds of activities have been
undertaken in disaster-related projects, ranging
from rubble clearance to construction of
transport infrastructure systems.

The Bank has also demonstrated its ability to

work with donors in a shared response and has
adapted policies and procedures to ensure that
assistance can be delivered expeditiously. Joint
damage assessments have become an important
mechanism for engaging with other donors and
ensuring that borrower needs are met without
overlaps.

Almost 80 percent of Bank-financed natural

disaster projects were rated satisfactory for
outcomes, compared with the Bankwide
average of 72 percent for the same period.
These ratings reflect particular effectiveness in
rebuilding physical infrastructure and in
provision of materials and equipment. In
general, though, disaster responses have tended
toward the reactive and tactical, when a
proactive and strategic approach would have
had longer-term benefits. Furthermore,
attention to the poor has been especially
difficult to accomplish in disaster projects.

When disaster strikes, funds are needed

immediately, and are often diverted from
development because no contingency funding
is available. The financial cost of responding to
the most recent events has stimulated interest

in creating global and regional funding
solutions. But even if these are eventually
established and the Bank decides to be a part of
such solutions, the Bank’s engagement with
client countries will have to ensure continued
focus on avoiding the next disaster, rather than
waiting for it to happen. Countries need to
become more proactive rather than reactive,
and Bank support to countries must do more to
encourage this shift.

The funding mechanisms used by the Bank

need to be rethought: balance of payment
lending has been relatively quick-disbursing, but
disbursement often does not take place in the
post-disaster period. The Bank has increasingly
used the Emergency Recovery Loan (ERL) in
responding to disaster, even when other instru-
ments may be more appropriate to achieving
long-term reduction of vulnerability. Several
attempts to establish insurance and contingency
financing have helped focus government
attention on the development issues of
disasters, but too few have been completed and
evaluated to judge their value. Finally, loan
reallocations are used much more frequently
than other types of Bank disaster responses in
highly vulnerable countries, putting develop-
ment goals at risk.

This report recommends several adjustments

to the way the Bank currently handles natural
disasters. First, it suggests revisions to policy to
better guide staff and enhance flexibility of Bank
responses to natural disasters. Second, it
encourages increased Bank capacity to respond
to disasters and to ensure that it can be
mobilized quickly. Finally, it recommends that
the Bank prepare a strategy or action plan for
natural disaster assistance that includes an
assessment of each country’s level of disaster
risk, and differentiation of approach on the basis
of that assessment.

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Vinod Thomas

Director-General

Evaluation

x i

Avant-propos

es catastrophes naturelles se produisent partout dans le monde, mais leurs
effets économiques et sociaux se font de plus en plus sentir et sont en
général beaucoup plus importants dans les pays en développement que

dans les pays développés. Les catastrophes peuvent anéantir les progrès en
matière de développement et éclipser des années d’investissement en faveur
du développement.

Au Mozambique, les prêts de la Banque ont fi-
nancé la construction de 487 établissements sco-
laires sur une période de 20 ans, mais une
catastrophe récente, à savoir les inondations de
2000, a à elle seule endommagé ou détruit en-
viron 500 écoles primaires ainsi que sept éta-
blissements secondaires. Au Pakistan, le
tremblement de terre du Cachemire en octobre
2005 a provoqué des dégâts estimés à 5 mil-
liards de dollars, soit à peu près l’équivalent du
total de l’aide publique au développement des
trois années précédentes, et l’équivalent du
montant des prêts accordés à ce pays par la
Banque au cours des dix années précédentes.

Jusqu’à une période récente, les catastrophes

étaient traitées comme des événements isolés et
aléatoires par les gouvernements et les orga-
nismes qui les aidaient à y faire face. L’expé-
rience nous apprend toutefois que les
catastrophes frappent certaines régions du
monde avec une périodicité régulière, et de
façon répétitive. L’éventualité d’une catastrophe
est prévisible dans la mesure où il est possible

de prédire, en général, à quel endroit un évé-
nement est susceptible de se produire à un cer-
tain moment dans un avenir proche (mais pas le
moment ou l’ampleur précis d’un tel événe-
ment). On sait ainsi que les zones côtières à
basse altitude de la Baie du Bengale subiront
davantage d’inondations, et que les petits États
insulaires des Caraïbes et les pays du Golfe du
Mexique seront frappés à maintes reprises par
des ouragans.

Dans leurs stratégies, la Banque et une grande

partie de la communauté de développement
doivent tenir compte de la fréquence des catas-
trophes et de la régularité avec laquelle elles
frappent les mêmes pays. L’analyse des opéra-
tions de prêt de la Banque révèle une tendance
claire dans l’avènement des catastrophes : dix
emprunteurs ont bénéficié de 208 (39 %) des 528
projets liés aux catastrophes du portefeuille pen-
dant la période 1984–2005. Les engagements de
prêt de la Banque sont également concentrés —
7,5 % des projets ont bénéficié de 32 % du fi-
nancement.

Dans leurs approches, le pays et la Banque

continuent de traiter parfois les catastrophes
comme une interruption du développement, et
non comme un risque pour le développement.
Parmi les stratégies d’aide actuelles en faveur des
pays ayant bénéficié d’un appui de la Banque à
la suite de catastrophes naturelles, 44 % ne men-
tionnaient pas ces dernières. Même dans le cas
des 40 pays qui ont bénéficié de quatre projets
ou plus liés aux catastrophes, celles-ci n’étaient
pas mentionnées dans un tiers des stratégies. En
outre, dans les stratégies d’environ un tiers du
sous-ensemble des pays où se produisent de
nombreuses catastrophes (plus de huit), ces
dernières n’étaient pas du tout mentionnées.
Les documents de prêt des projets considèrent
rarement les événements naturels comme un
risque, même dans les pays très vulnérables,
bien que les catastrophes aient eu un effet dé-
favorable sur 176 projets pendant leur exécution.
Nous devons trouver des moyens de placer ces
risques davantage au cœur de l’aide au déve-
loppement afin d’en améliorer l’efficacité.

Force est de reconnaître que la Banque a fait

montre d’une flexibilité considérable dans son ap-
proche des catastrophes naturelles et qu’elle a ap-
pris à gérer efficacement les grandes et les petites
interventions. Les services de la Banque ont sou-
vent fait preuve d’innovation et mis en évidence
leur capacité de gérer des travaux massifs de re-
construction à nombre de niveaux : dans le cadre
des projets liés aux catastrophes, plus de 60 dif-
férents types d’activités ont été entreprises, allant
du dégagement des débris à la construction de
réseaux d’infrastructures de transport.

La Banque a par ailleurs démontré son apti-

tude à travailler avec les bailleurs de fonds dans
le cadre d’une intervention collective, et elle a
adapté ses politiques et ses procédures pour as-
surer une prompte fourniture de l’aide. Les éva-
luations conjointes des dégâts sont devenues
un important mécanisme permettant de colla-
borer avec les autres bailleurs de fonds et de
veiller à ce que les besoins des emprunteurs
soient satisfaits sans créer de chevauchement.

Près de 80 % des projets relatifs aux catas-

trophes naturelles financés par la Banque ont été
jugés satisfaisants du point de vue de leurs ré-
sultats, par rapport à une moyenne de 72 % en-

registrée à l’échelle de la Banque pour la même
période. Ces notes témoignent de l’efficacité de
la Banque dans le domaine de la reconstruction
des infrastructures physiques et celui de la four-
niture de matériel et d’équipements. En général,
toutefois, les interventions relatives aux catas-
trophes ont eu tendance à être réactives et tac-
tiques, alors qu’une approche proactive et
stratégique aurait été avantageuse à plus long
terme. Qui plus est, dans les projets liés aux ca-
tastrophes, il a été tout particulièrement difficile
de prêter attention aux pauvres.

Lorsque la catastrophe frappe, les fonds sont

immédiatement nécessaires et ils sont souvent
réaffectés au détriment du développement, car
il n’existe pas de financement pour imprévus. Le
coût financier des interventions réalisées au titre
des plus récents événements a stimulé l’intérêt
envers la création de mécanismes de financement
mondiaux et régionaux. Mais même si l’on finit
par établir de tels mécanismes et qu’elle décide
d’y participer, la Banque devra s’attacher en per-
manence, dans le cadre de son engagement au-
près des pays clients, à éviter la prochaine
catastrophe plutôt qu’à attendre que celle-ci se
produise. Il convient que les pays deviennent
proactifs au lieu de se contenter de réagir, et l’ap-
pui accordé par la Banque aux pays doit faire da-
vantage pour encourager ce changement.

Les mécanismes de financement dont se sert

la Banque méritent d’être repensés : les prêts à
l’appui de la balance des paiements se caracté-
risent certes par un décaissement relativement
rapide, mais souvent aucun décaissement ne
s’effectue durant la période qui suit la catas-
trophe. Dans ses interventions en cas de catas-
trophe, la Banque recourt de plus en plus aux
prêts d’urgence aux pays sinistrés, même lorsque
d’autres instruments peuvent être mieux indi-
qués pour réduire la vulnérabilité à long terme.
Plusieurs tentatives de mise en place de fonds
d’assurance et de fonds pour imprévus ont
contribué à attirer l’attention des gouverne-
ments sur les problèmes de développement que
posent les catastrophes, mais trop peu de ces
fonds ont été créés et évalués pour qu’on juge
de leur valeur. Enfin, dans les pays très vulné-
rables, la Banque recourt beaucoup plus fré-
quemment à la réaffectation des prêts qu’à ses

x i i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

x v

Prefacio

os desastres naturales se producen en todo el mundo, pero su impacto
económico y social ha ido en aumento y suele ser mucho mayor en los
países en desarrollo que en los desarrollados. Estos fenómenos pueden

arrasar con el progreso logrado en materia de desarrollo y significar años de
retroceso en las inversiones en desarrollo.

En Mozambique, el Banco financió la construcción
de 487 escuelas en un período de 20 años, pero
tan sólo un desastre ocurrido recientemente —
las inundaciones del año 2000— dañó o destruyó
unas 500 escuelas primarias y siete escuelas se-
cundarias. El terremoto de Cachemira de octubre
de 2005 provocó daños por unos US$5.000 mi-
llones en Pakistán, que equivalen aproximada-
mente al total de la asistencia oficial para el
desarrollo de los tres años anteriores, y a los re-
cursos que el Banco Mundial había prestado a ese
país en los 10 años precedentes.

Hasta hace poco, los gobiernos y los organis-

mos que los ayudaban a responder ante las emer-
gencias producidas por los desastres consideraban
que éstos eran acontecimientos únicos y al azar.
Sin embargo, sabemos por experiencia que los de-
sastres naturales se producen con cierta periodi-
cidad, y que en algunas partes del mundo ocurren
repetidamente. La posibilidad de que ocurra un
desastre puede preverse en la medida en que sea
posible predecir en términos generales dónde
es probable que se produzca un acontecimiento
en el futuro cercano (aunque no exactamente
cuándo o con qué magnitud). Por lo tanto, las

zonas de litoral bajo de la bahía de Bengala ten-
drán más inundaciones, y los pequeños Estados
insulares del Caribe, así como los países del golfo
de México, sufrirán repetidamente los embates de
los huracanes.

El Banco y gran parte de la comunidad del de-

sarrollo deben tener en cuenta en sus estrate-
gias la frecuencia con que se producen los
desastres y con que éstos azotan a los mismos pa-
íses. El financiamiento concedido por el Banco
muestra que los desastres siguen un patrón defi-
nido: de los 528 proyectos relacionados con de-
sastres naturales (39%) que conformaban la cartera
en el período de 1984-2005, 208 correspondían
a 10 prestatarios. Los compromisos de préstamo
del Banco también muestran una determinada
concentración: el 7,5% de los proyectos recibie-
ron el 32% del financiamiento.

En algunos casos, en los planteamientos tanto

de los países como del Banco, los desastres se si-
guen considerando como una interrupción del
proceso de desarrollo en lugar de un riesgo para
dicho proceso. En el 44% de las actuales estrate-
gias de asistencia a los países que han recibido
apoyo del Banco en relación con desastres natu-

rales no se hace mención de tales fenómenos. In-
cluso en los 40 países en que se han realizado cua-
tro o más proyectos relacionados con desastres
naturales, un tercio de las estrategias omitió toda
mención a ellos. Asimismo, aproximadamente
un tercio del subconjunto de países que tenían
una vasta trayectoria de desastres (más de ocho)
no hizo mención alguna a estos fenómenos. En
los documentos de préstamo rara vez se considera
que los fenómenos naturales constituyen un
riesgo, ni siquiera en los países muy vulnerables,
pese a que 176 proyectos se vieron afectados ad-
versamente por un desastre durante su ejecu-
ción. Para aumentar la eficacia, debemos encontrar
la manera de integrar mejor estos riesgos en la asis-
tencia para el desarrollo.

Hay que reconocer que el Banco ha demos-

trado mucha flexibilidad en lo que respecta a los
desastres naturales y ha aprendido a proporcio-
nar respuestas adecuadas en pequeña y gran es-
cala. El personal del Banco a menudo ha recurrido
a métodos innovadores y demostrado la capaci-
dad de hacer frente a reconstrucciones en gran
escala en muchos niveles: se han emprendido
más de 60 tipos diferentes de actividades en pro-
yectos relacionados con desastres naturales, desde
la remoción de escombros hasta la construcción
de sistemas de infraestructura para el transporte.

El Banco también ha demostrado su capacidad

para trabajar con los donantes en operaciones
de respuesta conjuntas y ha adaptado sus políti-
cas y procedimientos para asegurar que la asis-
tencia llegue en forma expedita. Las evaluaciones
conjuntas de los daños se han convertido en un
mecanismo importante para entablar una rela-
ción con otros donantes y asegurar que no haya
duplicación de esfuerzos en la atención de las
necesidades de los prestatarios.

Casi el 80% de los proyectos sobre desastres

naturales financiados por el Banco fueron califi-
cados de satisfactorios en cuanto a sus efectos di-
rectos, en comparación con un promedio de 72%
en igual período a nivel de todo el Banco. Estas
calificaciones son indicativas de la particular efi-
cacia del Banco en la reconstrucción de infraes-
tructura física y el suministro de materiales y
equipo. En general, sin embargo, la respuesta
ante los desastres ha sido más bien reactiva y tác-
tica, en circunstancias en que un criterio proac-

tivo y estratégico habría producido beneficios a
más largo plazo. Además, en los proyectos rela-
cionados con desastres naturales ha sido espe-
cialmente difícil atender las necesidades de la
población pobre.

Cuando se produce un desastre, la necesidad

de recursos financieros es inmediata y a menudo
estos fondos se desvían de otros programas de de-
sarrollo porque no se dispone de financiamiento
para situaciones imprevistas. El costo financiero
que ha significado la respuesta frente a los acon-
tecimientos más recientes ha despertado el inte-
rés en encontrar soluciones para movilizar
recursos a nivel mundial y regional. Pero incluso
si se llegaran a establecer estas medidas y el Banco
decidiera formar parte de esas soluciones, en su
participación y compromiso con los países clien-
tes éste debería asegurarse de que se preste aten-
ción permanente a las actividades de prevención
de un próximo desastre, en lugar de esperar hasta
que ocurra. Los países deben ser más proactivos
que reactivos y el apoyo del Banco debe alentar
en mayor medida este cambio de actitud.

Es preciso replantear los mecanismos de fi-

nanciamiento que emplea el Banco: los desem-
bolsos de préstamos para financiar la balanza de
pagos se han producido con relativa rapidez, pero
suele suceder que no haya desembolsos durante
el período posterior a un desastre. En medida cre-
ciente, el Banco ha otorgado préstamos de emer-
gencia para recuperación en respuesta a un
desastre, incluso cuando otros instrumentos po-
drían haber sido más adecuados para reducir la vul-
nerabilidad a largo plazo. Varios intentos de
establecer sistemas de seguro y fondos para im-
previstos han ayudado a centrar la atención de los
gobiernos en los problemas de desarrollo que oca-
sionan los desastres, pero son muy pocos los que
han llegado a término y se han podido evaluar
como para determinar su utilidad. Por último, en
los países muy vulnerables se recurre con mucha
mayor frecuencia a la reasignación de préstamos
que a otros tipos de respuesta del Banco ante si-
tuaciones de desastre, con el consiguiente riesgo
de no alcanzar las metas de desarrollo.

En el presente informe se recomiendan varios

ajustes a la manera en que el Banco enfrenta ac-
tualmente los desastres naturales. En primer lugar,
se sugiere examinar las políticas a fin de orientar

x v i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

x i x

Executive Summary

he impact of natural disasters on economic well-being and human suf-
fering has increased alarmingly. In the past year alone, the earthquake
and tsunami in the Indian Ocean killed an estimated 220,000 people and

left 1.5 million people homeless, catastrophic flooding and mudslides in
Guatemala killed hundreds of people, and a massive earthquake in Kashmir
killed tens of thousands more in Pakistan and India.

The death tolls are staggering, and the costs to
the human and economic development of the
affected countries are huge and rising. Natural
disasters are becoming more costly: in constant
dollars, disaster costs between 1990 and 1999
were more than 15 times higher ($652 billion in
material losses) than they were between 1950
and 1959 ($38 billion at 1998 values). The
human cost is also high: over the 1984–2003
period, more than 4.1 billion people were
affected by natural disasters. The number
affected has grown, from 1.6 billion in the first
half of that period (1984–93) to almost 2.6
billion in the second half (1994-2003), and has
continued to increase.

Although disasters caused by natural events

occur throughout the world, losses to disaster in
developing countries are generally much greater
than in developed countries in terms of
percentage of gross domestic product (GDP) or
government revenues. The disproportionate
effect on developing countries has many
explanations. Lack of development itself
contributes to disaster impacts, both because

the quality of construction often is low and
building codes, land registration processes, and
other regulatory mechanisms are lacking, as well
as because numerous other development priori-
ties displace attention from the risks presented
by natural events.

Most natural disasters are foreseeable to the

extent that it is possible to predict generally
where an event is likely to occur at some time in
the near future (but not precisely when or its
magnitude). Small island states in the Caribbean
and states along the coast of the Gulf of Mexico
will undoubtedly be repeatedly hit by hurri-
canes; Pacific Rim states in the “ring of fire” are
highly likely to be hit by earthquakes and
volcanic eruptions; low-lying coastal areas on
the Bay of Bengal are sure to experience more
flooding; and Africa will very likely experience
more drought. Therefore, it makes sense to treat
the hazards of nature as risks to development,
especially where they occur repeatedly.

Disasters dilute hard-won development

gains. In Mozambique, for example, Bank
lending financed the construction of 487

schools. But the most recent disaster alone, the
floods of 2000, damaged or destroyed about 500
primary schools as well as 7 secondary schools.
The damage caused can outweigh years of
development assistance. The Kashmir earth-
quake of October 2005 caused an estimated $5
billion in damage in Pakistan, roughly equivalent
to the total official development assistance for
the preceding 3 years, and equivalent to the
amount the World Bank had lent to the country
over the preceding 10 years.

There is no private insurance against hazard

risks in most developing countries. While about
half of these costs of natural disasters are covered
by insurance in the United States, less than 2
percent of the costs are covered in the develop-
ing world. In addition, the cost of hedging
against natural hazard risks in developing
countries often exceeds the cost of simply paying
for damages when they arise. Further, develop-
ing countries can generally count on aid from
outside sources, a well-known moral hazard in
the disaster field. For poor households, natural
hazards are just one of the many risks they face
and are unlikely to be a high priority.

When a disaster occurs, the key concerns for

the affected country are what to do, how to do
it, and how to fund the necessary response.
Typically, funds are needed immediately, and are
often diverted from long-term development
because no contingency funding is available.
The financial cost of responding to the most
recent events has stimulated particular interest
in creating global and regional funding
solutions. A proposal has been put forward for a
regional funding mechanism in Latin America,
and another proposal would expand an existing
UN program to provide a global contingency
funding mechanism.

The World Bank has been increasingly

engaged in helping countries to recover from
the disastrous impacts of natural events and to
reduce their future vulnerability. When the
World Bank responds to a natural disaster it has
a wide array of lending and nonlending services
from which to choose. And its response spans
multiple sectors and themes, including urban,
rural, environment, infrastructure, education,
health, and social protection.

Its nonlending services can include conven-

ing of donor meetings, provision of assistance
with post-disaster assessments, study prepara-
tion, and technical assistance. Bank lending
assistance can consist of funds reallocated from
existing projects, redesigns of planned projects,
or development of new projects using a variety
of lending instruments. In addition to its
advisory and analytic services and technical
support, since 1984 the Bank has financed 528
projects that addressed natural disasters,
representing more than $26 billion in lending.

The Independent Evaluation Group examined

the Bank’s experience in disaster response over
the past 20 years to extract lessons to inform
good practice and ensure the achievement of
results in Bank-supported activities. The evalua-
tion is also intended to inform an ongoing
revision of the Bank’s policy statement on
emergency assistance.

The Bank Response

The Bank has demonstrated considerable
flexibility in its approach to natural
disaster assistance and has learned to
manage responses to events ranging from
those of very large dimensions to smaller,
more limited events.

Bank staff have often been innovative in their

response to disaster events and have demon-
strated the capacity to manage reconstruction on
a massive scale. The study identified more than
60 types of activities undertaken in disaster-
related projects, ranging from rubble clearance
and provision of emergency shelter, to construc-
tion of flood shelters and transport infrastructure,
to institutional development.

Responses to disaster have included lending

and nonlending assistance, the latter including
disaster needs assessments, advisory assistance,
and other forms of technical assistance. Among
the responses that have demonstrated the Bank’s
flexibility and innovation are the Honduras Social
Investment Fund (1999), the Maharashtra
Earthquake Project (1997), North China Earth-
quake Reconstruction (1993), Yemen Emergency
Flood Reconstruction (1989), and the drought
prevention in Niger (1988), all of which dynami-
cally adjusted to prevailing conditions.

x x

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

The Bank also has demonstrated its ability
to work with donors in a shared response
and has adapted policies and procedures
to ensure that assistance can be delivered
expeditiously.

Donor coordination was particularly strong

for Hurricane Mitch in Honduras and Nicaragua
(1999); for the Marmara earthquake in Turkey
(2000); for drought in Sudan (1989); and for
flooding in Bangladesh (1999), Mozambique
(2000), and Gujarat (2002). Joint assessments
have become an important mechanism for
engaging with other donors and ensuring that
borrower needs are met without overlaps.

Natural disaster projects financed by the
Bank have had higher ratings for outcome
and sustainability than the Bank’s portfo-
lio as a whole.

Almost 80 percent of the projects that had

natural disaster reconstruction or mitigation as a
substantial element were rated satisfactory for
outcome, compared with the Bankwide average
of 72 percent for the same period. These ratings
reflect particular effectiveness in rebuilding
physical infrastructure and provision of materi-
als and equipment.

Sustainability ratings are similarly better than

average, but institutional development ratings
are about the same as the average. The sustain-
ability rating (for what is mostly infrastructure)
reflects the likelihood that estimated net
benefits will be maintained or exceeded over a
project’s intended useful life. Experience with
the creation of disaster management capacity
has shown that it often takes more than one
project cycle to leave behind a functioning
disaster institution where none existed.

But in general, disaster responses have
tended toward the reactive and tactical,
when a proactive and strategic approach
would have had longer-term benefits.

Countries affected by disaster, as well as the

donors that try to help them, including the Bank,
have generally treated disasters as interruptions
in development rather than as a risk that is
integral to development. At the country level,
few Country Assistance Strategies (CASs) and

Poverty Reduction Strategies (PRSPs) mention
natural disaster risks, even in countries that have
experienced multiple events resulting in major
disasters. At the project level, objectives have
mainly provided for short-term fixes and rarely
addressed the root causes of the disastrous
impacts of natural hazards.

The Bank has increasingly used the
Emergency Recovery Loan (ERL), the focus
of its emergency lending policy, in
responding to disaster, even when other
instruments may be more appropriate.

The ERL offers accelerated processing and a

short implementation period of three years, and
therefore has desirable qualities valued by both
borrower and Bank staff who respond to
disasters. ERLs generally have worked well and
have high outcome ratings. But accelerated
project processing is not always desirable. For
some projects, rushed appraisal has led to long
pauses between loan approval and first disburse-
ment, poorly designed interventions, and
diminished poverty impacts.

Furthermore, by relying on a three-year

lending period, the Bank may end up emphasiz-
ing activities that are expected to have short
implementation times, while not attending to
other activities that more fully address the needs
and vulnerabilities. It often happens that activi-
ties that might contribute greatly to the recovery
effort (and to the borrower’s subsequent long-
term development) are not included in the ERL
projects because they cannot be completed in
the three years allotted—and then the project
runs long in any event.

The crucial activities for long-term
reduction of vulnerability take longer than
three years to implement and have weak
borrower demand.

Only one of the 60 activities identified in Bank-

supported projects—balance of payment
assistance—has taken less than three years to
implement, on average. The types of activities
that can have the greatest impact on reducing
vulnerability, such as building code development
or revision, development of hazard risk manage-
ment institutions, and development of insurance

E X E C U T I V E S U M M A R Y

x x i

and other mechanisms for laying off risk, are
precisely those for which borrowers are least
likely to borrow. The Bank needs to find ways to
encourage such activities.

Actions taken during the first weeks and
months after a disaster have a major
impact on the recovery process to follow,
and they need to be planned and
implemented accordingly.

Choices made immediately following a

disaster—regarding shelter, resettlement, debris
clearance, distribution of relief, and the like—
affect the later choices for longer-term solutions
and vulnerability reduction and can have severe
consequences for the ability of the poor to
recover.

Immediate post-disaster actions also need to

include the development of the capacities,
knowledge, and skills that will be required for
the recovery process. If studies are going to
produce knowledge that is critical to fully
informed project actions, they need a strong
advocate, such as the Bank. Capacity building
for procurement and preparation of bidding
documents should happen very early. Procure-
ment is among the project activities most
frequently cited in project-level evaluations as
needing improvement.

The Bank needs to be able to identify when
haste is counterproductive, lest funding
mechanisms rather than development
needs drive its response.

The funding mechanisms themselves need to

be rethought: balance of payment lending has
been a relatively quick-disbursing mechanism
but, on average, it is nowhere near as fast as it was
supposed to be, and it has only helped in very
limited circumstances. Several Bank-supported
attempts to establish mechanisms to lay off risk
(insurance and contingency financing) have
helped focus government attention on the long-
term development issues surrounding disasters,
but too few have been completed and evaluated
to make an informed judgment about their value.
Finally, loan reallocations are used much more
frequently than other types of Bank disaster
responses in highly vulnerable countries.

Recovery for the poor requires particular
attention, but is especially difficult to
accomplish in disaster projects, and poverty
impacts are generally not well documented.

When Bank projects have targeted the poor,

they have often exceeded their expected
impact: of 51 projects with documented
impacts, 41 met or exceeded the expected
impact. However, data are incomplete and
documentation of the poverty impacts is thin.

Even in the difficult circumstances of a

disaster response, beneficiary participation
during the design and implementation stages is
essential to success. The benefits of participa-
tion were demonstrated in the 1993 Argentina
Flood Rehabilitation Project, which involved
beneficiaries in all stages of the project. The
interaction between beneficiaries and the local
authorities resulted in the timely availability of
construction materials and the accommodation
of local customs in the architectural design of
new houses. Bank staff observed that this
created ownership among beneficiaries and
increased maintenance.

Experience in Turkey and Chile shows that

cash transfers and the provision of livelihood
opportunities can be especially effective for the
poor. Experience also shows that women and
other vulnerable groups need special attention
following disasters, especially in ensuring
equitable treatment.

Reconstructed housing that is built using
disaster-resistant techniques and accord-
ing to the needs of occupants reduces
vulnerability.

Building codes can improve the quality of the

built environment, but in informal neighbor-
hoods that typically do not comply with code
requirements, safer building practices need to be
disseminated in different ways. Simplicity of
message is essential to the widespread adoption
of disaster-resistant technologies, as has been
amply demonstrated in India. Because temporary
housing is sometimes occupied for long periods
of time, some projects have built temporary
shelter to slightly higher standards so that it could
become another form of housing for the poorer
once the new housing is built.

x x i i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Moreover, if shelters are built using disaster-

resistant construction techniques, not only are
they safer for the displaced living in them, but
such construction also serves as an example that
people will see, that will then potentially
influence their future construction choices.
Simple techniques can be used to ensure resist-
ance in owner- or craftsman-built small houses,
more sophisticated techniques may be used in
engineer-designed buildings such as high-rises.

Bringing Risk Management into
Development Strategy

Natural hazard risks are highly concen-
trated, so special attention needs to be
given to planning ahead for disaster and to
reducing long-term vulnerability in
countries at highest risk.

Ten countries account for 208 of the 528

disaster projects (39 percent) in the Bank’s
portfolio. Bank lending also is concentrated in
commitment terms—7.5 percent of projects
received 32 percent of the financing. Natural
hazard risks are foreseeable for many countries,
yet those risks are infrequently considered in
country programs or in project financing, even
in highly vulnerable countries.

When formulating country lending programs

and project lending, the Bank needs to elevate
the importance of natural hazards, especially for
highly vulnerable countries. To do this
efficiently, borrowing countries need to be
categorized by vulnerability level. This report
presents one way to do this, dividing borrowers
into three groups according to level of vulnera-
bility (high, medium, and low, based on the
percentage of a country’s GDP at risk from two
or more natural hazards).

The high concentration of risk also
suggests that mechanisms are needed to
finance those risks or transfer them.

Even if global or regional funds are eventually

established, they will likely address only the
short-term liquidity needs of disaster-affected
countries. The Bank needs to be seen as a part
of such regional and global solutions, but it also
needs to continue to provide the longer-term
activities directed at vulnerability reduction.

The Bank’s long-term engagement with
client countries needs to ensure continued
focus on permanent vulnerability reduction.

The Bank has supported several research

initiatives on risk hedging and private sector
involvement in reconstruction financing.
Financial approaches to mitigating loss that are
receiving attention include: reinsurance with
catastrophe bonds, national homeowner
insurance programs, disaster funds, and microfi-
nance. Additionally, 10 Bank-funded projects are
beginning to explore national insurance
schemes (5 of which are ongoing and have not
been evaluated).

Coordination Inside and Outside the Bank

The Bank has the human resources capacity
to both respond to disasters and to address
long-term country needs related to hazard
risks, but mobilizing them is cumbersome.

The Bank has a cadre of committed and

experienced staff, but it lacks an effective way to
reliably bring that staff and relevant knowledge
to its borrowers, or even to its own task teams.
Since 1999 a three-person unit has assisted Bank
task managers with natural and technological
disasters and helped provide a more strategic
and rapid response. This group is supplemented
by a thematic group comprising more than 100
staff with disaster-related experience. However,
donors and client countries do not know who to
contact when they have routine questions about
disaster and related coordination. The current
arrangement has also effectively reduced the
visibility of the natural disaster theme within the
Bank. When a disaster strikes it can be difficult
to disengage knowledgeable and experienced
staff from their ongoing tasks.

Donor coordination is especially critical to
disaster relief and recovery, in part because
of the dynamic nature of the situation, but
also because disasters typically attract the
involvement of numerous donors.

Increasingly, borrowers themselves are

providing the necessary donor coordination,
but they continue to need assistance with
coordination, especially in the early stages of
relief and recovery.

E X E C U T I V E S U M M A R Y

x x i i i

Project experience shows that the develop-

ment of a recovery strategy shared by all
requires not only an immediate Bank presence
in the disaster-affected area, but also a
prolonged presence that helps ensure that all
reconstruction needs are covered, that the plan
is appropriately designed for the available
capacity, that stakeholders’ needs are met, that
there is a reasonable distribution of labor, and
that the needs of the poor and vulnerable are
considered.

In 1989, for example, Bank negotiators in

Sudan worked with other donors to ensure that
their interests were met and that there were no
unnecessary overlaps in coverage. By keeping
the composition of the Bank’s contribution
flexible, the other donors were helped to make
adjustments in their programs. The Bank then
financed what was left.

The development community should
engage with disaster-stricken borrowers
earlier and stay engaged longer.

International experience on the impacts of

successful and unsuccessful relief management
and on the ability of key stakeholders to partici-
pate effectively in the recovery process needs to
be brought clearly to governments’ attention.
The Bank specifically needs to be present during
the emergency stage to ensure success of the
reconstruction projects it finances. Low-income
community groups need support until they
develop the capacity to manage the infrastruc-
ture that has been placed in their care.

Recommendations

Chapter 6 of the report makes a number of
specific suggestions about revisions to the
Bank’s policy for emergency lending—these are
not repeated here in their entirety.

Prepare a Strategy or Action Plan for Natural
Disaster Assistance

The Bank’s natural disaster assistance would
benefit from the development of a strategy or
action plan and related guidance that would:

• Help staff to respond to emergencies with

quick relief and well-planned reconstruction,

and to do so more effectively in a much shorter
period.

• Ensure that contingency funds (be it on a coun-

try, regional, or global scale) result in all bor-
rowing countries receiving a timely and
adequate financial response to major events.

• Help bring natural hazard risk management

to the most vulnerable countries.

The strategy or action plan needs to identify a

methodology to assess each country’s level of
disaster risk. It is suggested that countries be
divided into high-, medium-, and low-risk
groups. The action plan then needs to identify
how the Bank will assist borrowers in each
category to lower their vulnerabilities and to
build on local capacities and leadership.

In highly vulnerable countries, the action

plan needs to make provisions to give more
attention to natural hazards during the appraisal
of investment projects generally, and specifically
in the preparation of PRSPs, CASs, and other
strategic documents. Where appropriate, these
documents need to go beyond a description of
the risks, and identify monitorable mitigation
and institutional development activities.

For the most vulnerable countries, contin-

gency funding needs to be available, whether as
part of another loan, a set-aside in the CAS
lending program, or a free-standing catastrophe
fund (though these may become unnecessary if
regional or global funds are eventually
established). Another alternative worth consid-
eration is a special fund under the Bank
President’s control that can be used to fund a
quick start when disaster occurs.

Countries deemed to be at medium to high

risk need to include disaster-resilient design in
Bank-financed projects. For all countries,
disaster risks need to be considered in standard
risk assessment documents.

The strategy or action plan should be submit-

ted to the Board for discussion.

Revise Policy to Better Guide Staff
and Enhance Flexibility of Bank Responses
to Natural Disasters

Emergencies are of many sorts and, although
there is some overlap, most differ from the

x x i v

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

disasters created by natural events in critical
ways. Bank policy needs to reflect these differ-
ences by treating conflict and epidemic diseases
separately, with provisions that apply only to the
relevant topic. There are two ways in which this
can be done: natural disasters can either be the
subject of a separate Operational Policy (as
called for in the 1998 IEG evaluation of the
Bank’s experience with post-conflict reconstruc-
tion); or Operational Policy 8.50 could include
specific provisions for natural disasters, for post-
conflict situations, and for health and other
emergencies, so that each topic is dealt with
separately. In whatever form it takes, Bank policy
needs to focus more on disaster prevention and
vulnerability reduction in all natural disaster
operations. Policy prohibitions on relief and the
financing of recurring events need to be relaxed.

Accelerated processing and provisions for

quick disbursement for ERLs have partially
addressed the need for speed in launching
short-term activities, though they could be
fruitfully complemented by a new mechanism,
such as a special central fund managed by the
President’s office (akin to the one in place in the
Inter-American Development Bank) to fund the
most urgent needs in the early days of a disaster
response.

But the use of ERLs is less appropriate for

longer-term activities, such as mitigation,
reconstruction, and institution building, which
require a longer preparation and appraisal time
and need not be exempted from due diligence
standards and safeguard compliance.

Similarly, attention to social issues during

preparation and implementation generally

requires a longer period than has been available
under ERLs. Such activities are more suited to
standard investment lending but have often been
short-changed because of the ERL’s three-year
implementation time and the loss of borrower
interest in a second loan following the ERL.

Increase Bank Capacity to Respond
to Disasters and Ensure That It Can
Be Mobilized Quickly

Whether or not there is a designated unit to deal
with natural disasters and hazard risks, the Bank
needs the capacity to quickly gather and dissem-
inate international experience to borrowers in an
emergency. In addition, task teams need support
while conducting post-disaster assessments and
designing emergency interventions tailored to
the needs and capacities of each borrower.

Responding to disasters requires multisec-

toral expertise. Including disaster-knowledge-
able people on Bank missions following major
crises can be crucial. Being selective in staffing
identification for missions in post-disaster
settings avoids the problems of design and scale
of response that can occur when people are sent
who are not used to seeing destruction on a
massive scale or who lack country knowledge.
The Bank has very few such people, and it
currently has no consistent mechanism for
mobilizing them to respond to natural disasters.
Pulling members of the Hazard Management
Thematic Group away from their ongoing
responsibilities inevitably has a negative impact
on their normal activities. And there are so few
knowledgeable staff that the same people tend
to be called upon repeatedly.

E X E C U T I V E S U M M A R Y

x x v

x x v i i

Résumé analytique

’incidence des catastrophes naturelles sur le bien-être économique et sur
la souffrance humaine s’est accrue de manière inquiétante. Rien qu’au cours
de la dernière année, le tremblement de terre et le tsunami de l’océan In-

dien ont fait environ 220 000 morts et laissé 1,5 million de personnes sans abri,
des inondations et des coulées de boue catastrophiques ont tué des centaines
de personnes au Guatemala, et un violent tremblement de terre au Cachemire
a provoqué des dizaines de milliers d’autres morts au Pakistan et en Inde.

Le bilan est saisissant, et le prix payé par les pays
touchés, du point de vue du développement
humain et économique, est énorme et en crois-
sance. Les catastrophes naturelles deviennent
plus coûteuses : en dollars constants, les coûts
liés aux catastrophes entre 1990 et 1999 (652 mil-
liards de dollars de pertes matérielles) étaient de
plus de 15 fois supérieurs à ceux de la période
1950-1959 (38 milliards de dollars en valeur de
1998). Le coût humain est également élevé : au
cours de la période 1984–2003, plus de 4,1 mil-
liards de personnes ont été touchées par des ca-
tastrophes naturelles. Le nombre des sinistrés
s’est accru, passant de 1,6 milliard de personnes
pendant la première moitié de cette période
(1984–1993) à près de 2,6 milliards durant la
deuxième moitié (1994–2003), et il continue
d’augmenter.

Même si les catastrophes causées par des évé-

nements naturels se produisent partout dans le
monde, les pertes occasionnées par les catas-
trophes sont en général beaucoup plus grandes

dans les pays en développement que dans les pays
développés, en termes de pourcentage du pro-
duit intérieur brut (PIB) ou des recettes pu-
bliques. Un grand nombre de facteurs expliquent
cette incidence disproportionnée sur les pays
en développement. L’absence de développement
en elle-même contribue aux effets des catas-
trophes, parce que les constructions sont souvent
de piètre qualité, et en raison de l’inexistence de
codes de construction, de processus d’enregis-
trement des biens fonciers et d’autres méca-
nismes réglementaires, ainsi qu’à cause du fait que
de nombreuses autres priorités de développe-
ment détournent l’attention des risques pré-
sentés par les évènements naturels.

La plupart des catastrophes naturelles sont

prévisibles, dans la mesure où il est possible de
prédire, en général, l’endroit où un événement
est susceptible de se produire à un certain mo-
ment dans un avenir proche (mais pas le moment
ou l’ampleur précis d’un tel événement). Les
petits États insulaires des Caraïbes et les pays du

Golfe du Mexique seront sans nul doute frappés à
maintes reprises par des ouragans ; les pays rive-
rains du Pacifique situés dans le « cercle de feu »
sont très susceptibles d’être frappés par des
tremblements de terre et des éruptions volca-
niques ; les zones côtières à basse altitude de la
Baie du Bengale subiront assurément davantage
d’inondations ; et l’Afrique connaîtra très pro-
bablement plus d’épisodes de sécheresse. Il est
par conséquent raisonnable de traiter les dangers
de la nature comme des risques pour le déve-
loppement, surtout dans les régions où ils sur-
viennent de façon répétitive.

Les catastrophes annihilent une bonne partie

des laborieux progrès accomplis dans le do-
maine du développement. A titre illustratif, au
Mozambique les prêts de la Banque ont financé
la construction de 487 établissements scolaires ;
mais une catastrophe récente, à savoir les inon-
dations de 2000, a endommagé ou détruit à elle
seule environ 500 écoles primaires ainsi que
sept établissements secondaires. Les dégâts cau-
sés peuvent anéantir plusieurs années d’aide au
développement. Au Pakistan, le tremblement de
terre du Cachemire en octobre 2005 a provoqué
des dégâts estimés à 5 milliards de dollars, soit
à peu près l’équivalent du total de l’aide pu-
blique au développement des trois années pré-
cédentes, et l’équivalent du montant des prêts
accordés à ce pays par la Banque au cours des
dix années précédentes.

Dans la plupart des pays en développement,

il n’existe pas d’assurance privée contre les
risques naturels. Si environ la moitié des coûts
liés aux catastrophes naturelles sont couverts
par l’assurance aux États-Unis, moins de 2 % le
sont dans les pays en développement. En outre,
le coût de la couverture des risques inhérents
aux catastrophes naturelles dans les pays en
développement est souvent supérieur aux frais
encourus pour réparer tout simplement les dé-
gâts subis lorsque ces catastrophes se produi-
sent. Par ailleurs, les pays en développement
peuvent en général compter sur de l’aide pro-
venant de sources extérieures, ce qui repré-
sente un risque moral bien connu dans le
domaine des interventions en matière de ca-
tastrophe. En ce qui concerne les ménages
pauvres, les catastrophes naturelles ne repré-

sentent que l’un des nombreux risques qu’ils
courent, et il est peu probable qu’ils y accordent
une grande priorité.

Lorsqu’une catastrophe se produit, les pays

touchés se préoccupent principalement de ce
qu’il faut faire, de la manière de procéder, et
des moyens de financer les interventions né-
cessaires. En général, les fonds sont immédia-
tement nécessaires et ils sont souvent réaffectés
au détriment du développement à long terme,
car il n’existe pas de financement pour imprévus.
Le coût financier des interventions réalisées dans
le cadre des plus récents événements a stimulé
un intérêt tout particulier envers la création de
mécanismes de financement mondiaux et ré-
gionaux. Il a été tour à tour proposé d’établir un
mécanisme de financement régional en Amé-
rique latine, et d’étendre un programme existant
de l’ONU afin de créer un mécanisme mondial
de financement pour imprévus.

La Banque mondiale s’engage de plus en plus

à aider les pays à se remettre des effets désastreux
des évènements naturels et à réduire leur vul-
nérabilité future. Lorsque la Banque mondiale in-
tervient à la suite d’une catastrophe naturelle, elle
a le choix entre un large éventail d’activités de
prêt et de services hors prêt. Son intervention
porte sur plusieurs domaines et thèmes, no-
tamment le secteur urbain, le secteur rural, l’en-
vironnement, l’infrastructure, l’éducation, la
santé et la protection sociale.

Ses services hors prêt peuvent comprendre

l’organisation des réunions des bailleurs de
fonds, l’octroi d’aide pour évaluer les dégâts
après la catastrophe, la préparation des études
et l’assistance technique. Les activités de prêt de
la Banque peuvent consister à redéployer les
fonds de projets existants, à modifier la concep-
tion de projets prévus ou à élaborer de nou-
veaux projets en utilisant une variété
d’instruments de prêt. Outre ses services d’ana-
lyse et de conseil et son appui technique, la
Banque a, depuis 1984, financé 528 projets por-
tant sur les catastrophes naturelles et représen-
tant plus de 26 milliards de dollars de prêt.

Le Groupe indépendant d’évaluation a exa-

miné l’expérience de la Banque dans le domaine
des interventions liées aux catastrophes au cours
des 20 dernières années, afin d’en tirer des en-

x x v i i i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

seignements pour guider les bonnes pratiques
et assurer le succès des activités appuyées par la
Banque. L’évaluation vise par ailleurs à guider la
révision en cours de l’énoncé de politique de la
Banque sur l’aide d’urgence.

L’intervention de la Banque

La Banque a fait montre d’une considé-
rable flexibilité dans son approche de l’aide
en cas de catastrophe naturelle et elle a ap-
pris à gérer les interventions relatives aux
événements, des plus grands aux plus pe-
tits qui sont plus limités.

Les services de la Banque ont souvent fait

preuve d’innovation dans leur réponse face aux
événements liés aux catastrophes, et ils ont mis
en évidence leur capacité de gérer des travaux
de reconstruction à très grande échelle. L’étude
a identifié plus de 60 types d’activités entre-
prises dans le cadre de projets liés aux catas-
trophes, et qui vont du dégagement des débris
à la fourniture d’abris de secours, en passant
par la construction d’abris contre les inonda-
tions et d’infrastructures de transport, et le dé-
veloppement institutionnel.

Les interventions relatives aux catastrophes

ont consisté notamment en des activités de prêt
et des services hors prêt, ces derniers compre-
nant l’évaluation des besoins liés aux catas-
trophes, l’aide sous forme de conseils, et d’autres
formes d’assistance technique. Parmi les inter-
ventions qui ont mis en évidence la flexibilité et
l’esprit novateur de la Banque, on peut citer
celles ayant trait à la création du Fonds d’inves-
tissement social du Honduras (1999), à la ré-
ponse face au tremblement de terre de
Maharastra (1997), à la reconstruction après le
tremblement de terre du Nord de la Chine
(1993), à la reconstruction d’urgence à la suite
des inondations au Yémen (1989), et à la pré-
vention de la sécheresse au Niger (1988), les-
quelles ont toutes été ajustées de façon
dynamique aux conditions de l’époque de leur
mise en œuvre.

La Banque a par ailleurs démontré son ap-
titude à travailler avec les bailleurs de
fonds dans le cadre d’une intervention col-
lective, et elle a adapté ses politiques et ses

procédures pour assurer une prompte four-
niture de l’aide.

La coordination des bailleurs de fonds a été

tout particulièrement étroite dans le cadre des
interventions relatives à l’ouragan Mitch au Hon-
duras et au Nicaragua (1999), au tremblement de
terre de Marmara en Turquie (2000), à la séche-
resse au Soudan (1989), ainsi qu’aux inonda-
tions au Bangladesh (1999), au Mozambique
(2000) et au Gujarat (2002). Les évaluations
conjointes des dégâts sont devenues un impor-
tant mécanisme permettant de collaborer avec
les autres bailleurs de fonds et de veiller à ce que
les besoins des emprunteurs soient satisfaits
sans créer de chevauchement.

Les projets liés aux catastrophes naturelles
financés par la Banque ont obtenu des
notes plus élevées, du point de vue des ré-
sultats et de la viabilité, que celles du por-
tefeuille de la Banque dans l’ensemble.

Près de 80 % des projets dont un élément

important avait trait à la reconstruction après les
catastrophes naturelles ou à l’atténuation des
risques de catastrophe naturelle ont été jugés sa-
tisfaisants en ce qui concerne les résultats, par
rapport à une moyenne de 72 % enregistrée à
l’échelle de la Banque pour la même période. Ces
notes témoignent de l’efficacité particulière de
la Banque dans le domaine de la reconstruction
des infrastructures physiques, et celui de la four-
niture de matériel et d’équipements.

Les notes de viabilité sont également supé-

rieures à la moyenne, mais celles relatives au
développement institutionnel sont approxima-
tivement similaires à la moyenne. La note de
durabilité (des infrastructures, pour l’essentiel)
indique la probabilité que les avantages estima-
tifs nets seront maintenus ou dépassés pendant
la durée prévue de vie utile d’un projet. L’expé-
rience du renforcement des capacités en ma-
tière de gestion des catastrophes a montré qu’il
faut souvent plus d’un cycle de projet pour im-
planter de façon durable, là où il n’en existait pas,
une institution —créée après la catastrophe —
qui fonctionne.

Mais en général, les interventions liées aux
catastrophes ont eu tendance à être réac-

R É S U M É A N A LY T I Q U E

x x i x

tives et tactiques, alors qu’une approche
proactive et stratégique aurait été avanta-
geuse à plus long terme.

Les pays touchés par la catastrophe ainsi que

les bailleurs de fonds qui essayent de leur venir
en aide, y compris la Banque, ont en général
traité les catastrophes comme une interruption
du développement et non comme un risque in-
hérent au développement. Au niveau national,
rares sont les stratégies d’aide-pays et les stra-
tégies de réduction de la pauvreté qui men-
tionnent les risques liés aux catastrophes
naturelles, même dans les pays ayant connu plu-
sieurs événements qui ont entraîné de grandes
catastrophes. Les projets quant à eux ont prin-
cipalement visé à apporter des solutions à court
terme, et ils se sont rarement attaqués aux causes
profondes des impacts désastreux des catas-
trophes naturelles.

Dans ses interventions liées aux catas-
trophes, la Banque recourt de plus en plus
aux prêts d’urgence aux pays sinistrés, les-
quels sont au cœur de sa politique de prêt
d’urgence, même lorsque d’autres instru-
ments peuvent être mieux indiqués.

Le prêt d’urgence aux pays sinistrés offre un

traitement accéléré et une courte période d’exé-
cution de trois ans, et il comporte par consé-
quent d’intéressantes qualités qui sont appréciées
à la fois de l’emprunteur et des services de la
Banque chargés des interventions en cas de ca-
tastrophe. Les prêts d’urgence aux pays sinistrés
ont en général été efficaces et ont obtenu des
notes élevées en ce qui concerne les résultats.
Mais l’accélération du traitement du projet n’est
pas toujours souhaitable. Dans le cas de certains
projets, une évaluation hâtive a mené à de
longues pauses entre l’approbation du prêt et le
premier décaissement, à des interventions mal
conçues et à des impacts réduits sur la pauvreté.

En outre, en misant sur une période de prêt

de trois ans, la Banque peut finir par mettre l’ac-
cent sur les activités qui sont censées compor-
ter une courte période d’exécution, en négligeant
d’autres activités qui satisfont plus pleinement
les besoins des sinistrés et permettent de ré-
soudre les problèmes de vulnérabilité. Il arrive
souvent que les activités susceptibles de contri-

buer considérablement à l’effort de reconstruc-
tion (et au développement subséquent à long
terme de l’emprunteur) ne soient pas couvertes
dans les projets financés au moyen des prêts
d’urgence aux pays sinistrés, parce qu’elles ne
peuvent pas être achevées dans le délai imparti
de trois ans—et le projet finit de toute façon par
aller au-delà de ce délai.

Les activités cruciales visant à réduire la vul-
nérabilité à long terme nécessitent plus
de trois ans pour leur mise en œuvre, et
elles font l’objet d’une faible demande de
la part des emprunteurs.

Parmi les 60 activités identifiées dans les pro-

jets financés par la Banque —au titre de l’appui
à la balance des paiements — une seule a été exé-
cutée pendant moins de trois ans, en moyenne.
Les types d’activités qui peuvent avoir le plus
grand impact sur la réduction de la vulnérabilité,
telles que l’élaboration ou la révision du code de
construction, la mise en place d’institutions de
gestion des risques de catastrophe, l’établisse-
ment de l’assurance et d’autres mécanismes de
déplacement de risque, sont précisément celles
pour lesquelles les emprunteurs sont le moins
susceptibles de contracter des prêts. La Banque
doit trouver des moyens d’encourager de telles
activités.

Les mesures prises pendant les premières
semaines ou les premiers mois après une
catastrophe ont un impact majeur sur le
processus de reconstruction qui suivra, et
il convient de les planifier et de les mettre
en œuvre en conséquence.

Les choix opérés immédiatement après une

catastrophe — et portant sur les abris, la réins-
tallation, le dégagement des débris, la distribu-
tion de l’aide, etc. — influent sur les choix
subséquents en matière de solution à plus long
terme et de réduction de la vulnérabilité, et ils
peuvent avoir de graves conséquences pour l’ap-
titude des pauvres à se remettre des effets de la
catastrophe.

Les mesures prises immédiatement après la

catastrophe doivent en outre prévoir le renfor-
cement des capacités ainsi que l’acquisition des
connaissances et des compétences qui seront né-

x x x

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

cessaires pour le processus de reconstruction.
Si des études doivent produire des connais-
sances qui sont critiques pour éclairer pleine-
ment la prise de mesures liées au projet, il
convient qu’elles soient préconisées par un pro-
moteur puissant, tel que la Banque. Le renfor-
cement des capacités de passation des marchés
et d’élaboration des dossiers d’appel d’offres
doit s’effectuer très tôt. Dans les évaluations
réalisées au niveau du projet, la passation des
marchés est l’une des activités des projets les plus
fréquemment citées comme ayant besoin d’amé-
lioration.

La Banque doit être capable d’identifier
les situations où la hâte est plus néfaste que
bénéfique, si elle tient à ce que son inter-
vention soit déterminée par les besoins de
développement et non par les mécanismes
de financement.

Les mécanismes de financement ont eux-

mêmes besoin d’être repensés : le prêt à l’appui
de la balance des paiements a certes été un mé-
canisme de décaissement relativement rapide,
mais en moyenne il est bien loin d’être aussi ra-
pide qu’on ne s’y attendait, et il n’a été utile que
dans un nombre très limité de situations. Plu-
sieurs tentatives appuyées par la Banque pour
établir des mécanismes de déplacement de
risque (assurance et financement pour impré-
vus) ont contribué à attirer l’attention des gou-
vernements sur les problèmes de développement
à long terme que posent les catastrophes, mais
trop peu de ces mécanismes ont été créés et éva-
lués pour qu’on se prononce de façon éclairée sur
leur valeur. Enfin, dans les pays très vulnérables,
la Banque recourt beaucoup plus fréquemment
à la réaffectation des prêts qu’à ses autres types
d’interventions liées aux catastrophes.

Il convient d’accorder une attention spé-
ciale aux activités de reconstruction ci-
blant les pauvres, mais ces activités sont
tout particulièrement difficiles à réaliser
dans les projets relatifs aux catastrophes,
et les effets de la pauvreté sont en général
mal établis.

Lorsque les projets de la Banque ont ciblé

les pauvres, ils ont souvent dépassé l’impact

prévu : des 51 projets dont les effets sont attes-
tés, 41 ont eu un impact comparable ou supé-
rieur aux attentes. Toutefois, les données sont
incomplètes et les preuves montrant les impacts
sur la pauvreté sont minces.

Même dans les circonstances difficiles qui en-

tourent une intervention liée aux catastrophes,
la participation des bénéficiaires aux processus
de conception et d’exécution est indispensable
pour en assurer le succès. Les avantages de la par-
ticipation ont été mis en évidence dans le cadre
du Projet de reconstruction à la suite des inon-
dations survenues en Argentine en 1993. A cette
occasion, les bénéficiaires avaient été associés à
toutes les étapes du projet. L’interaction entre
les bénéficiaires et les autorités locales a permis
d’assurer la disponibilité en temps voulu des
matériaux de construction, et la prise en compte
des coutumes locales dans la conception archi-
tecturale des nouvelles habitations. Les services
de la Banque ont relevé que cette situation a sus-
cité l’adhésion des bénéficiaires et amélioré l’en-
tretien des ouvrages construits.

Il ressort de l’expérience de la Turquie et du

Chili que les transferts d’espèces et la fourniture
de moyens de subsistance peuvent se révéler tout
particulièrement efficaces pour les pauvres. L’ex-
périence montre par ailleurs qu’il convient d’ac-
corder une attention spéciale aux femmes et
aux autres groupes vulnérables, en veillant sur-
tout à l’équité du traitement des sinistrés.

Les logements reconstruits de façon à ré-
sister aux inondations et à répondre aux be-
soins des occupants réduisent la
vulnérabilité.

Les codes de construction peuvent améliorer

la qualité de l’environnement bâti, mais dans
les quartiers à habitat spontané où en général les
exigences relatives aux immeubles ne sont pas
satisfaites, il importe de propager selon divers
moyens des pratiques de construction plus sûres.
La simplicité du message est indispensable pour
l’adoption à grande échelle de technologies qui
permettent de construire des ouvrages résis-
tant aux catastrophes, comme il l’a été ample-
ment démontré en Inde. Étant donné que les
logements provisoires sont parfois occupés pen-
dant de longues périodes, certains projets ont

R É S U M É A N A LY T I Q U E

x x x i

bâti des abris temporaires selon des normes lé-
gèrement plus élevées, afin qu’ils puissent au be-
soin devenir une autre forme d’habitation pour
les pauvres, une fois les nouveaux logements
construits.

En outre, si les abris sont construits de façon

à résister aux catastrophes, ils seront plus sûrs
pour les personnes déplacées qui y vivront, tout
comme ils serviront d’exemple que les popula-
tions pourront voir et qui pourrait influer sur
leurs futurs choix en matière de construction.
Des techniques simples peuvent être utilisées
pour assurer la résistance de petits logements
bâtis par les propriétaires ou des ouvriers qua-
lifiés, et on peut recourir à des techniques plus
complexes dans le cas des immeubles conçus par
des ingénieurs, tels que les tours d’habitation.

Intégrer la gestion des risques à la
stratégie de développement

Les risques naturels sont très concentrés,
aussi convient-il de s’attacher tout parti-
culièrement à prendre les dispositions né-
cessaires avant que ne se produise la
catastrophe et à réduire la vulnérabilité à
long terme des pays à risque élevé.

Dix pays bénéficient de 208 (39 %) des 528

projets du portefeuille de la Banque. Les opé-
rations de prêt de la Banque sont également
concentrées du point de vue des engagements
—7,5 % des projets ont reçu 32 % du finance-
ment. Les risques naturels sont prévisibles dans
nombre de pays, et pourtant ces risques sont peu
souvent pris en compte dans les programmes na-
tionaux ou dans le financement des projets,
même dans les pays très vulnérables.

Lors de l’élaboration des programmes de prêt

aux pays et des opérations de prêt en faveur
des projets, la Banque doit amplifier l’impor-
tance accordée aux risques naturels, surtout
dans le cas des pays très vulnérables. Pour y par-
venir de façon efficace, il convient de classer les
pays emprunteurs par niveau de vulnérabilité. Le
présent rapport propose un moyen de le faire,
qui consiste à diviser les emprunteurs en trois
groupes selon le niveau de vulnérabilité de cha-
cun (élevé, moyen et faible, en fonction du pour-
centage du PIB du pays qui est exposé à deux ou
plusieurs risques naturels).

La forte concentration des risques donne
par ailleurs à penser qu’il est nécessaire de
trouver des mécanismes pour financer ces
risques ou pour les transférer.

Même si l’on finit par réunir des fonds mon-

diaux ou régionaux, ils ne serviront probable-
ment qu’à satisfaire les besoins en matière de
liquidité à court terme des pays touchés par les
catastrophes. Si la Banque doit montrer qu’elle
contribue à de telles solutions régionales et
mondiales, il importe également qu’elle continue
de réaliser des activités à plus long terme visant
à réduire la vulnérabilité.

Dans le cadre de son engagement à long
terme auprès des pays clients, la Banque
doit veiller à ce qu’une attention soute-
nue soit accordée à la réduction perma-
nente de la vulnérabilité.

La Banque a appuyé plusieurs initiatives de re-

cherche sur la couverture des risques et la par-
ticipation du secteur privé au financement de la
reconstruction. Les approches financières de la
mitigation des pertes qui suscitent de l’intérêt
sont notamment : le déplacement de risque au
moyen des obligations catastrophes, les
programmes nationaux d’assurance des pro-
priétaires, les fonds de secours et le microfi-
nancement. En outre, dix projets financés par la
Banque (dont cinq sont en cours et n’ont pas été
évalués) commencent à étudier les régimes d’as-
surance nationaux.

Coordination au sein et à l’extérieur
de la Banque

La Banque possède la capacité en ressources
humaines nécessaires pour intervenir en cas
de catastrophe et pour satisfaire les be-
soins — liés aux risques naturels — à long
terme des pays, mais la mobilisation de
ces ressources est difficile.

La Banque a un corps de cadres dévoués et ex-

périmentés, mais elle manque de moyens effi-
caces de mettre de manière fiable ce personnel
et le savoir pertinent au service de ses emprun-
teurs ou même de ses propres équipes de pro-
jet. Depuis 1999, une unité composée de trois
personnes aide les chefs de projet de la Banque
à gérer les catastrophes naturelles et technolo-

x x x i i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

giques, et elle aide à réaliser des interventions plus
stratégiques et plus rapides. À cette unité s’ajoute
un groupe thématique comprenant plus de 100
employés qui possèdent une expérience dans le
domaine des catastrophes. Cependant, les
bailleurs de fonds et les pays clients ignorent à
qui adresser leurs questions courantes au sujet
des catastrophes et de la coordination y affé-
rente. Le système actuel a en outre réduit effec-
tivement la visibilité du thème des catastrophes
naturelles au sein de la Banque. Lorsqu’une ca-
tastrophe se produit, il peut s’avérer difficile de
faire en sorte que les employés compétents et ex-
périmentés abandonnent leurs tâches du mo-
ment pour s’occuper de l’intervention.

La coordination des bailleurs de fonds est
tout particulièrement cruciale pour le se-
cours en cas de catastrophe et la recons-
truction, en raison en partie de la nature
dynamique de cette situation, mais aussi
parce que les catastrophes donnent en gé-
néral lieu à l’intervention d’un grand
nombre de bailleurs de fonds.

De plus en plus, les emprunteurs eux-mêmes

assurent la nécessaire coordination des bailleurs
de fonds, mais ils n’en continuent pas moins
d’avoir besoin d’aide en matière de coordination,
surtout au début des opérations de secours et
de la reconstruction.

L’expérience acquise dans les projets montre

que l’élaboration d’une stratégie de recons-
truction à adopter par tous exige non seulement
que la Banque soit immédiatement présente
dans la zone sinistrée, mais aussi qu’elle y reste
pendant une longue période afin de s’assurer que
tous les besoins en matière de reconstruction
sont couverts, que le plan est conçu de façon ap-
propriée pour la capacité disponible, que les
besoins des parties prenantes sont satisfaits, que
la répartition de la main-d’œuvre est raisonnable
et que les besoins des pauvres et des groupes vul-
nérables sont pris en considération.

En 1989, par exemple, les négociateurs de la

Banque au Soudan ont travaillé avec les bailleurs
de fonds pour veiller à ce que leurs besoins
soient satisfaits et qu’il n’y ait pas d’inutiles che-
vauchements dans la couverture. En maintenant
flexible le contenu de sa contribution, la Banque

a aidé les autres bailleurs de fonds à opérer des
ajustements au niveau de leurs programmes. La
Banque a ensuite financé la partie non couverte
des interventions.

La communauté du développement doit
commencer à travailler plus tôt avec les em-
prunteurs sinistrés, et maintenir cette col-
laboration pendant plus longtemps.

Il convient clairement d’attirer l’attention des

gouvernements sur l’expérience internationale re-
lative aux incidences de la gestion efficace et in-
efficace du secours et à l’aptitude des principales
parties prenantes à participer efficacement au
processus de reconstruction. Plus précisément,
la Banque doit être présente pendant la phase
d’urgence pour assurer le succès des projets de
reconstruction qu’elle finance. Les groupes de
proximité à faible revenu ont besoin d’aide jus-
qu’à ce que soit renforcée leur capacité de gérer
l’infrastructure qui a été confiée à leurs soins.

Recommandations

Le chapitre 6 du rapport formule un nombre de
propositions précises par rapport aux révisions
à apporter à la politique de prêt d’urgence de la
Banque. Ces recommandations ne sont pas
reprises en intégralité ici.

Élaborer une stratégie ou un plan d’action pour
l’aide en cas de catastrophe naturelle.

L’aide en cas de catastrophe naturelle de la
Banque bénéficierait de l’élaboration d’une
stratégie ou d’un plan d’action, ainsi que de
directives connexes, qui permettraient :

• au personnel d’intervenir en situation d’ur-

gence en fournissant un secours rapide et en
mettant en œuvre une opération de recons-
truction bien planifiée, et ce de façon plus ef-
ficace et en un temps beaucoup plus court ;

• de veiller à ce que la création de fonds pour

imprévus (à l’échelle nationale, régionale ou
mondiale) ait pour effet que tous les pays
emprunteurs confrontés à des événements
majeurs bénéficient d’une intervention fi-
nancière suffisante et à temps.

• d’initier les pays les plus vulnérables à la ges-

tion des risques naturels.

R É S U M É A N A LY T I Q U E

x x x i i i

La stratégie ou le plan d’action doit définir une

méthodologie pour évaluer le niveau de risque
de catastrophe de chaque pays. Il est proposé
que les pays soient divisés en groupes à risque
élevé, moyen et faible. Le plan d’action doit en-
suite indiquer la manière dont la Banque aidera
les emprunteurs de chaque catégorie à réduire
leur vulnérabilité et à faire fond sur les capaci-
tés et la volonté politique locales.

Dans les pays très vulnérables, le plan d’action

doit prévoir des mesures permettant de prêter
davantage attention aux risques naturels pen-
dant l’évaluation des projets d’investissement
en général, et lors de l’élaboration des Docu-
ments de stratégie pour la réduction de la pau-
vreté, des Stratégies d’aide-pays (CAS) et d’autres
documents stratégiques en particulier. Le cas
échéant, ces documents doivent, au-delà de la
description des risques, indiquer des activités
contrôlables de réduction de la vulnérabilité et
de développement institutionnel.

En ce qui concerne les pays les plus vulné-

rables, le financement pour imprévus doit être
disponible, qu’il fasse partie d’un prêt ou qu’il
revête la forme d’un fonds mis de côté dans un
programme de prêt de la CAS, voire d’un fonds
autonome de catastrophe (même si ces fonds
peuvent devenir inutiles si l’on établit en fin de
compte des fonds régionaux ou mondiaux). Une
autre solution digne d’intérêt est la création
d’un fonds spécial placé sous le contrôle du pré-
sident et pouvant servir à financer un lancement
rapide des opérations de secours lorsqu’une ca-
tastrophe se produit.

Les pays dont le niveau de risque est moyen ou

élevé doivent prévoir, dans les projets financés par
la Banque, la conception d’ouvrages résistant aux
catastrophes. Pour tous les pays, il convient de
tenir compte des risques de catastrophe dans les
documents classiques d’évaluation des risques.

La stratégie ou le plan d’action doit être sou-

mis au Conseil pour examen.

Revoir la politique pour mieux guider le
personnel et rendre plus flexibles les
interventions de la Banque liées aux
catastrophes naturelles.

Il existe un grand nombre de types de situations
d’urgence, et malgré quelques chevauchements,

la plupart d’entre elles diffèrent de façon fonda-
mentale des catastrophes causées par des évé-
nements naturels. La politique de la Banque doit
tenir compte de ces différences en traitant dif-
féremment les conflits et les épidémies, au moyen
de dispositions qui ne s’appliquent qu’à chaque
cas pertinent. Pour ce faire, il est possible de
procéder de deux manières : les catastrophes
naturelles peuvent faire l’objet d’une politique
opérationnelle distincte (tel que l’a recommandé
le Groupe indépendant d’évaluation en 1998
dans son évaluation de l’expérience de la Banque
en matière de reconstruction à la suite d’un
conflit) ; ou alors, l’politique opérationnelle 8.50
pourrait prévoir des clauses précises pour les
catastrophes naturelles, pour les situations sui-
vant un conflit, ainsi que pour la santé et d’autres
cas d’urgence, afin que chaque sujet soit traité sé-
parément. Quelle que soit la forme qu’elle prend,
la politique de la Banque doit se concentrer da-
vantage sur la prévention des catastrophes et la
réduction de la vulnérabilité dans toutes les opé-
rations portant sur les catastrophes naturelles. Il
convient d’alléger les restrictions auxquelles la po-
litique assujettit l’aide et le financement en faveur
des événements périodiques.

L’accélération du traitement des prêts d’ur-

gence aux pays sinistrés et de la prise de dispo-
sitions pour leur décaissement rapide a satisfait
en partie le besoin de célérité dans le lance-
ment des activités à court terme, même si ces me-
sures pourraient être avantageusement
complétées par un nouveau mécanisme, tel
qu’un fonds central spécial géré par le cabinet
du président (et similaire à celui qui est en place
à la Banque interaméricaine de développement)
pour financer les besoins les plus urgents pen-
dant les premiers jours d’une intervention en cas
de catastrophe.

Le recours aux prêts d’urgence aux pays sinis-

trés est toutefois moins indiqué dans le cas des ac-
tivités à plus long terme tels que la réduction de
la vulnérabilité, la reconstruction, et le renforce-
ment des institutions, qui nécessitent un plus long
délai de préparation et d’évaluation et n’ont pas
besoin d’être soustraites aux normes de diligence
raisonnable et aux mesures de sauvegarde.

De même, pour qu’une attention soit accor-

dée aux problèmes sociaux lors de la préparation

x x x i v

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

et de l’exécution, il faut en général un délai plus
long qu’on ne peut ménager dans le cadre des
prêts d’urgence aux pays sinistrés. De telles ac-
tivités conviennent mieux aux opérations clas-
siques d’investissement, mais elles y ont souvent
perdu au change en raison du délai d’exécution
de trois ans des prêts d’urgence aux pays sinis-
trés, et de la perte de l’intérêt de l’emprunteur
pour un deuxième prêt après l’obtention du
prêt d’urgence aux pays sinistrés.

Renforcer la capacité de la Banque
d’intervenir en cas de catastrophe et d’assurer
une mobilisation rapide de ses ressources.

Qu’il existe ou non une unité désignée pour
s’occuper des catastrophes et des risques natu-
rels, il est nécessaire qu’en situation d’urgence,
la Banque ait la capacité de réunir rapidement
les expériences internationales et de les diffuser
auprès des emprunteurs. En outre, les équipes
de projet doivent être appuyées lors de l’éva-
luation des dégâts après la catastrophe et de la
conception d’interventions d’urgence adaptées
aux besoins et aux capacités de chaque em-
prunteur.

L’intervention en cas de catastrophe exige

une expertise multisectorielle. Il peut s’avérer
crucial de faire en sorte que des personnes ex-
périmentées en matière de catastrophe partici-
pent aux missions organisées par la Banque à la
suite des crises majeures. En désignant sur une
base sélective les cadres devant participer aux
missions dans les régions sinistrées, on évite les
problèmes liés à la conception et à l’ampleur de
l’intervention qui peuvent survenir lorsque les
membres de la mission n’ont pas l’habitude de
voir des destructions à une vaste échelle ou
qu’ils ne possèdent pas de connaissances sur le
pays. Peu de cadres ayant le profil idéal sont en
service à la Banque, et à l’heure actuelle cette
dernière ne dispose d’aucun mécanisme cohé-
rent pour les mobiliser en vue d’une interven-
tion en cas de catastrophe naturelle. Le fait
d’éloigner de leurs responsabilités courantes les
membres du Groupe thématique sur la gestion
des risques a inévitablement une incidence né-
gative sur leurs activités normales. En outre, le
nombre d’employés expérimentés est si petit
qu’on tend à solliciter les mêmes personnes à
maintes reprises.

R É S U M É A N A LY T I Q U E

x x x v

x x x v i i

Resumen

l impacto de los desastres naturales en el bienestar económico y en el
sufrimiento de las personas ha aumentado en forma alarmante. Tan sólo
el año pasado, el terremoto y tsunami del océano Índico causó la

muerte de unas 220.000 personas y dejó sin hogar a un millón y medio de ha-
bitantes; las catastróficas inundaciones y aludes de lodo en Guatemala mata-
ron a centenares de personas, y un devastador terremoto en Cachemira causó
la muerte de decenas de miles de personas más en Pakistán e India.

Las cifras de muertos son asombrosas y los cos-
tos en términos de desarrollo humano y econó-
mico de los países afectados son enormes y siguen
aumentando. Los desastres naturales resultan
cada vez más onerosos: en dólares constantes, el
costo de los desastres naturales entre 1990 y 1999
fue 15 veces más elevado (US$652.000 millones
en pérdidas materiales) que entre 1950 y 1959
(US$38.000 millones en valores de 1998). El costo
en términos humanos también es elevado: en el
período comprendido entre 1984 y 2003, más de
4.100 millones de personas se vieron afectadas por
desastres naturales. Esta cifra ha aumentado de
1.600 millones en la primera mitad de dicho pe-
ríodo (1984-93) a casi 2.600 millones en la se-
gunda mitad (1994-2003), y continúa subiendo.

En todo el mundo se producen desastres pro-

vocados por fenómenos naturales, pero las pér-
didas en términos del porcentaje del producto
interno bruto (PIB) o de ingresos públicos sue-
len ser mucho mayores en los países en desa-
rrollo que en los desarrollados. Esta

desproporción en los países en desarrollo tiene
muchas explicaciones. La falta de desarrollo por
sí misma agrava el impacto de estos desastres, de-
bido tanto a la mala calidad de las construcciones
como a la ausencia de códigos de construcción,
procedimientos para el registro de tierras y otros
mecanismos de regulación, y a que existen mu-
chas otras prioridades de desarrollo que desvían
la atención de los riesgos que plantean los fenó-
menos naturales.

La mayoría de los desastres naturales son pre-

visibles en la medida en que sea posible prede-
cir en términos generales dónde es probable que
se produzca un acontecimiento en el futuro cer-
cano (aunque no exactamente cuándo o con qué
magnitud). Indudablemente, los pequeños Esta-
dos insulares del Caribe y los países a lo largo de
la costa del Golfo de México se verán azotados re-
petidamente por los huracanes; los Estados de la
cuenca del Pacífico situados en el denominado
“círculo de fuego” tienen altas probabilidades de
sufrir terremotos y erupciones volcánicas; con

x x x v i i i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

seguridad, las zonas de litoral bajo de la Bahía de
Bengala tendrán más inundaciones, y es muy pro-
bable que en África se produzcan más sequías. Por
lo tanto, es razonable considerar que los peligros
de la naturaleza constituyen un riesgo para el de-
sarrollo, especialmente en los lugares donde éstos
se repiten una y otra vez.

Los desastres echan por tierra los progresos en

materia de desarrollo logrados con tanto esfuerzo.
En Mozambique, por ejemplo, el Banco financió
la construcción de 487 escuelas, pero el último de-
sastre —las inundaciones del año 2000— bastó
para dañar o destruir unas 500 escuelas primarias
y siete escuelas secundarias. Los daños provoca-
dos pueden contrarrestar con creces lo logrado en
años de asistencia para el desarrollo. El terremoto
de Cachemira de octubre de 2005 causó daños por
unos US$5.000 millones en Pakistán, que equiva-
len aproximadamente al total de la asistencia ofi-
cial para el desarrollo de los tres años anteriores,
y a los recursos que el Banco Mundial había pres-
tado a ese país en los 10 años precedentes.

En la mayoría de los países en desarrollo no exis-

ten seguros privados contra este tipo de riesgo. En
los Estados Unidos, aproximadamente la mitad de
los costos relacionados con los desastres natura-
les están cubiertos por el seguro, pero en el mundo
en desarrollo esa proporción no alcanza al 2%.
Además, el costo de la cobertura contra dichos
riesgos en los países en desarrollo a menudo es su-
perior al costo que significaría reparar los daños pro-
ducidos. Por otra parte, normalmente esos países
pueden contar con que recibirán ayuda externa, un
conocido riesgo moral en el ámbito de los desas-
tres. En el caso de las familias pobres, los riesgos
naturales son sólo uno de los numerosos tipos de
riesgos a los que se exponen y es poco probable
que revistan alta prioridad.

Cuando se produce un desastre, la principal pre-

ocupación del país afectado es qué hacer, cómo
hacerlo y cómo financiar las medidas de respuesta
apropiadas. La necesidad de recursos financieros
suele ser inmediata y a menudo estos fondos se
desvían de programas de desarrollo a largo plazo
porque no se dispone de financiamiento para si-
tuaciones imprevistas. El costo financiero que ha
significado la respuesta ante los acontecimientos
más recientes ha despertado especial interés en
encontrar soluciones para movilizar recursos a

nivel mundial y regional. Se ha presentado una pro-
puesta para establecer un mecanismo de finan-
ciamiento a nivel regional en América Latina;
también se ha propuesto ampliar un programa de
las Naciones Unidas ya existente con el fin de
crear un mecanismo de financiamiento para si-
tuaciones imprevistas de alcance mundial.

En medida creciente, el Banco Mundial ayuda

a los países a recuperarse de los efectos desas-
trosos de los fenómenos naturales y a reducir su
nivel vulnerabilidad en el futuro. Cuando res-
ponde a un desastre natural, el Banco Mundial re-
curre a una amplia gama de servicios financieros
y no financieros. Además, su respuesta abarca di-
versos sectores y temas, entre ellos, urbanismo,
zonas rurales, medio ambiente, infraestructura,
educación, salud y protección social.

Los servicios no financieros pueden incluir la

convocatoria de reuniones de donantes, el sumi-
nistro de asistencia para las evaluaciones que se
llevan a cabo en cuanto se produce un desastre,
la preparación de estudios, y asistencia técnica. La
asistencia financiera del Banco puede consistir en
la reasignación de fondos de proyectos en curso,
la modificación del diseño de proyectos que se pla-
nea llevar a cabo, o la formulación de nuevos pro-
yectos mediante el uso de diversos instrumentos
de financiamiento. Además de los servicios analí-
ticos y de asesoría, y del apoyo técnico, desde
1984 el Banco ha financiado 528 proyectos rela-
cionados con desastres naturales, que represen-
tan préstamos por más de US$26.000 millones.

El Grupo de Evaluación Independiente exa-

minó la experiencia del Banco relativa a sus res-
puestas en casos de desastres en los últimos 20 años
con el objeto de extraer lecciones que permitan for-
mular prácticas recomendadas y asegurar que las
actividades apoyadas por el Banco consigan los
resultados previstos. La evaluación también tiene
por objeto servir de base para la revisión que está
llevando a cabo el Banco de su declaración de po-
lítica sobre asistencia en casos de emergencia.

La respuesta del Banco

El Banco ha demostrado mucha flexibili-
dad en su planteamiento con respecto a la
asistencia en caso de desastres naturales y
ha aprendido a proporcionar respuestas
adecuadas ante acontecimientos de enormes

dimensiones y otros de menor gravedad o
de alcance limitado.

En muchos casos el personal del Banco ha re-

currido a respuestas innovadoras ante situaciones
de desastre y demostrado la capacidad de hacer
frente a reconstrucciones en gran escala. En el es-
tudio se identificaron más de 60 tipos de activi-
dades en el marco de proyectos relacionados con
desastres naturales, tales como remoción de es-
combros, suministro de refugio de emergencia,
construcción de refugios para protegerse de las
inundaciones, infraestructura de transporte y de-
sarrollo institucional.

Las medidas de respuesta han comprendido

asistencia financiera y no financiera; esta última
ha consistido en asistencia técnica en forma de
evaluaciones de las necesidades provocadas por
los desastres, asesoría, etc. Como ejemplo de las
respuestas que han demostrado la flexibilidad y
capacidad de innovación del Banco cabe señalar
el Fondo de Inversión Social de Honduras (1999),
la asistencia tras el terremoto de Maharashtra
(1997), la reconstrucción tras el terremoto en el
norte de China (1993), la reconstrucción de emer-
gencia tras las inundaciones en Yemen (1989), y
las actividades para la prevención de sequías en
Níger (1988). En todos estos casos se hicieron en
forma dinámica los ajustes necesarios teniendo en
cuenta las condiciones imperantes.

El Banco también ha demostrado capaci-
dad para responder conjuntamente con los
donantes y ha adaptado sus políticas y pro-
cedimientos para asegurar que la asistencia
llegue en forma expedita.

La coordinación de los donantes fue particu-

larmente estrecha tras el huracán Mitch que azotó
Honduras y Nicaragua (1999), el terremoto de
Marmara, en Turquía (2000), la sequía en Sudán
(1989) y las inundaciones en Bangladesh (1999),
Mozambique (2000) y Gujarat (2002). Las eva-
luaciones conjuntas se han convertido en un me-
canismo importante para trabajar con otros
donantes y asegurar que no se produzca dupli-
cación de esfuerzos para atender a las necesida-
des de los prestatarios.

En comparación con la totalidad de la cartera,
los proyectos financiados por el Banco rela-

cionados con desastres naturales han recibido
mejores calificaciones en lo que respecta a
los efectos directos y la sostenibilidad.

Casi el 80% de los proyectos que tenían un

componente importante de mitigación o re-
construcción tras un desastre natural fueron ca-
lificados de satisfactorios en cuanto a sus efectos
directos, en comparación con un promedio de
72% en igual período a nivel de todo el Banco.
Estas calificaciones son indicativas de la particu-
lar eficacia del Banco en la reconstrucción de in-
fraestructura física y el suministro de materiales
y equipo.

Las calificaciones con respecto a la sostenibi-

lidad también son superiores al promedio, pero
las correspondientes al desarrollo institucional
son aproximadamente iguales a este último. La ca-
lificación de la sostenibilidad (principalmente in-
fraestructura) indica la probabilidad de que a lo
largo de la vida útil prevista del proyecto se man-
tengan o superen los beneficios netos estima-
dos. La experiencia acerca de la creación de
capacidad de gestión en casos de desastre ha
mostrado que a menudo se requiere más de un
ciclo completo de proyectos para establecer por
primera vez una institución encargada de res-
ponder en tales situaciones que funcione debi-
damente.

En general, sin embargo, la respuesta en
casos de desastre ha sido más bien reactiva
y táctica, en circunstancias en que un criterio
proactivo y estratégico podría haber pro-
ducido beneficios a más largo plazo.

Por lo general, los países afectados por desas-

tres, así como los donantes que tratan de brin-
darles ayuda, incluido el Banco, han considerado
los desastres como una interrupción del proceso
de desarrollo, y no como un riesgo que forma
parte integral de dicho proceso. En el plano na-
cional, son contadas las estrategias de asistencia
a los países y las estrategias de lucha contra la po-
breza en las que se mencionan los riesgos que
plantean los desastres naturales, incluso en paí-
ses donde se han producido varios aconteci-
mientos que han dado por resultado grandes
desastres. Al nivel de los proyectos, en los obje-
tivos señalados se han dispuesto principalmente
medidas de corto plazo y rara vez se han abordado

R É S U M É N

x x x i x

las causas básicas de los desastrosos efectos de los
riesgos naturales.

En medida creciente, el Banco ha otorgado
préstamos de emergencia para recupera-
ción —el elemento central de su política
de financiamiento para situaciones de emer-
gencia— para brindar asistencia en casos de
desastre, incluso cuando otros instrumen-
tos podrían haber sido más adecuados.

Estos préstamos se tramitan en forma acelerada

y el período de implementación es de apenas
tres años; por lo tanto, sus características tienen
muy buena acogida por parte de los prestatarios
y el personal del Banco que responde ante tales
situaciones. En términos generales, los présta-
mos de emergencia para recuperación han dado
buen resultado y gozan de altas calificaciones en
lo que respecta a sus efectos directos. Sin em-
bargo, la tramitación acelerada de un proyecto no
siempre es lo más conveniente. En algunos pro-
yectos, la apresurada evaluación inicial ha dado
lugar a largas pausas entre la aprobación del prés-
tamo y el primer desembolso, el diseño inade-
cuado de las intervenciones y un menor impacto
en la reducción de la pobreza.

Además, con un período de financiamiento

de tres años, el Banco puede acabar poniendo én-
fasis en actividades que suelen tener breves pe-
ríodos de ejecución, dejando de lado otras que
aborden más plenamente las necesidades y vul-
nerabilidades del prestatario. A menudo, en los
proyectos financiados con préstamos de emer-
gencia para recuperación no se incluyen activi-
dades que podrían contribuir considerablemente
a los esfuerzos de recuperación (y, posterior-
mente, al proceso de desarrollo a largo plazo del
prestatario) porque éstas no se pueden terminar
en los tres años establecidos, con la consiguiente
prolongación del proyecto.

Para llevar a cabo las actividades funda-
mentales que permiten reducir la vulnera-
bilidad a largo plazo se requieren más de tres
años y la demanda de tales actividades por
los prestatarios es escasa.

Solamente una de las 60 actividades identifica-

das en proyectos financiados por el Banco —a
saber, la asistencia para financiar la balanza de

pagos— ha requerido, en promedio, menos de
tres años. Los tipos de actividades que pueden pro-
ducir el mayor impacto en reducir la vulnerabilidad,
como la elaboración o revisión de los códigos de
construcción, y el establecimiento de instituciones
responsables de la gestión de riesgos, así como de
mecanismos de seguro y de otro tipo para mitigar
los riesgos, son precisamente aquellos para los cua-
les los prestatarios tienen menos probabilidades de
solicitar financiamiento. El Banco debe encontrar
maneras de promover ese tipo de actividades.

Las medidas adoptadas durante las prime-
ras semanas y meses después de ocurrido
un desastre producen en un gran impacto
en el ulterior proceso de recuperación, y
deben planificarse y llevarse a cabo en de-
bida forma.

Las decisiones que se adoptan inmediatamente

después de un desastre —refugio de personas, re-
asentamiento, remoción de escombros, distri-
bución de ayuda, etc.— influyen en las decisiones
que se adoptan posteriormente para encontrar so-
luciones a más largo plazo y reducir el nivel de vul-
nerabilidad, y pueden tener graves consecuencias
en la capacidad de recuperación de las personas
pobres.

En las medidas inmediatas luego de producido

un desastre también se debe incluir el desarrollo
de las capacidades, los conocimientos y las des-
trezas que serán necesarias durante el proceso
de recuperación. Para que los estudios generen los
conocimientos fundamentales que permitan adop-
tar medidas con pleno conocimiento de causa en
el marco de los proyectos, hace falta un buen pro-
motor, como el Banco. El fortalecimiento de la ca-
pacidad en materia de adquisiciones y preparación
de los documentos de licitación debe ser una de
las primeras medidas. La mejora de los procesos
de adquisiciones se cuenta entre las actividades que
se mencionan con mayor frecuencia en las eva-
luaciones a nivel de los proyectos.

El Banco debe ser capaz de determinar en
qué casos puede ser contraproducente apre-
surarse en las decisiones, para que no sean
los mecanismos de financiamiento sino las
necesidades de desarrollo las que impul-
sen su respuesta.

x l

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Es necesario replantearse los mecanismos de

financiamiento; los préstamos para financiar la
balanza de pagos han sido un mecanismo de de-
sembolso relativamente rápido, pero, en general,
no son ni cercanamente tan rápidos como debe-
rían serlo, y sólo han sido de ayuda en circuns-
tancias muy limitadas. Varios intentos apoyados
por el Banco encaminados a establecer mecanis-
mos para mitigar el riesgo (seguros y financia-
miento para imprevistos) han ayudado a centrar
la atención de los gobiernos en los problemas de
desarrollo a largo plazo que plantean los desas-
tres, pero son muy pocos los que han llegado a
término y se han podido evaluar como para emi-
tir un juicio informado sobre su utilidad. Por úl-
timo, en los países muy vulnerables se recurre con
mucha mayor frecuencia a la reasignación de
préstamos que a otros tipos de respuesta del
Banco ante situaciones de desastre.

Las actividades de recuperación dirigidas a
la población pobre requieren atención es-
pecial, pero son particularmente difíciles de
llevar a cabo en el marco de proyectos sobre
desastres naturales, y el impacto en la pobreza
no suele documentarse en debida forma.

Cuando los proyectos del Banco han estado di-

rigidos a los pobres, a menudo han sobrepasado
el impacto previsto: 41 de los 51 proyectos cuyo
impacto ha sido documentado alcanzaron o su-
peraron el impacto previsto. Sin embargo, no se
dispone de datos completos y la documentación
sobre el impacto en la pobreza es escasa.

Incluso en las difíciles circunstancias en que se

responde a un desastre, la participación de los be-
neficiarios durante las etapas de diseño y ejecu-
ción son fundamentales para obtener buenos
resultados. Las ventajas de dicha participación
quedaron demostradas en el proyecto de reha-
bilitación tras las inundaciones de Argentina, en
1993, ocasión en que los beneficiarios participa-
ron en todas las etapas del proyecto. La interac-
ción entre los beneficiarios y las autoridades
locales hizo posible disponer oportunamente de
materiales de construcción y tener en cuenta las
costumbres locales en el diseño arquitectónico de
las nuevas viviendas. El personal del Banco pudo
observar que todo esto permitió que los benefi-
ciarios se identificaran con el proyecto y que me-

jorara el mantenimiento de las obras.

La experiencia en los casos de Turquía y Chile

muestra que las transferencias de efectivo y las
oportunidades para ganarse la vida pueden ser es-
pecialmente eficaces para los pobres. También
indica que las mujeres y los grupos vulnerables ne-
cesitan atención especial después de un desastre;
concretamente, se debe asegurar que reciban un
trato equitativo.

La reconstrucción de viviendas con técnicas
de construcción resistentes a los desastres
y conforme a las necesidades de los mora-
dores reduce el grado de vulnerabilidad.

Cuando existen códigos de construcción se

puede mejorar la calidad de las edificaciones,
pero en los barrios informales en los que no se
suelen cumplir dichas normas es necesario pro-
mover por distintos conductos la utilización mé-
todos de construcción más seguros. Es
fundamental que la información se divulgue en un
lenguaje sencillo para lograr la adopción genera-
lizada de técnicas de construcción resistente a los
desastres, como se ha demostrado ampliamente
en India. Dado que a veces las viviendas tempo-
rales se ocupan durante períodos prolongados, en
el marco de algunos proyectos se han construido
refugios temporales conforme a normas ligera-
mente más estrictas, de modo que pudieran con-
vertirse en otra forma de vivienda para personas
más pobres una vez construidas las nuevas vi-
viendas.

Además, si los albergues se construyen con

técnicas resistentes a los desastres, no sólo son
más seguros para sus moradores desplazados,
sino que también sirven de modelos que la gente
puede apreciar, con la posibilidad de que lleguen
a influir en sus decisiones de construcción en el
futuro. Se pueden utilizar técnicas de construc-
ción sencillas para asegurar la resistencia de pe-
queñas viviendas construidas por sus propietarios
o por artesanos; en edificaciones diseñadas por
profesionales, como edificios de altura, se pueden
utilizar técnicas más complejas.

Integración de la gestión de riesgos en
las estrategias de desarrollo

Los riesgos naturales están muy concen-
trados, por lo que es necesario prestar aten-

R É S U M É N

x l i

ción especial a la planificación antes de que
se produzca un desastre, así como a la re-
ducción del grado de vulnerabilidad a largo
plazo en los países que corren mayor riesgo.

De los 528 proyectos relacionados con desas-

tres naturales (39%) que conforman la cartera
del Banco, 208 corresponden a 10 países. Los
compromisos de préstamo del Banco también
se encuentran concentrados: el 7,5% de los pro-
yectos recibió el 32% del financiamiento. En el caso
de muchos países, los riesgos naturales son pre-
visibles, pero pocas veces se tienen en cuenta en
los programas para los países o en el financia-
miento de los proyectos, incluso cuando se trata
de países muy vulnerables.

Al formular los programas de financiamiento

para los países y el financiamiento para proyectos,
el Banco debe dar mayor importancia a los ries-
gos naturales, especialmente en el caso de los pa-
íses muy vulnerables. Para proceder de manera
eficiente, los países prestatarios se deben clasifi-
car por nivel de vulnerabilidad. En el presente in-
forme se señala una manera de hacerlo, dividiendo
los prestatarios en tres grupos según su grado de
vulnerabilidad (alto, mediano y bajo, según el
porcentaje del PIB del país en cuestión expuesto
al riesgo en caso de que se produzcan dos o más
riesgos naturales).

La alta concentración del riesgo también
indica que se necesitan mecanismos para fi-
nanciar o transferir dichos riesgos.

Incluso si se llegan a establecer fondos mun-

diales o regionales, éstos probablemente sólo aten-
derán las necesidades de liquidez a corto plazo de
los países afectados por un desastre. El Banco
debe considerarse parte de tales soluciones re-
gionales y mundiales, pero también debe seguir par-
ticipando en actividades a más largo plazo dirigidas
a reducir el nivel de vulnerabilidad de los países.

En su participación a largo plazo para aten-
der las necesidades de los países clientes, el
Banco debe asegurar que se preste aten-
ción continuada a la permanente reducción
de la vulnerabilidad.

El Banco ha apoyado varias iniciativas de in-

vestigación relativas a la protección contra riesgos
y la participación del sector privado en el finan-

ciamiento de actividades de reconstrucción. Entre
las soluciones financieras para mitigar pérdidas en
que se está centrando la atención cabe mencio-
nar las siguientes: reaseguro con bonos de ca-
tástrofe, programas nacionales de seguro para
los propietarios de viviendas, fondos para situa-
ciones de desastre y microfinanciamiento. Ade-
más, en el marco de 10 proyectos financiados
por el Banco se están empezando a estudiar pla-
nes nacionales de seguro (cinco de esos planes
están en marcha pero no han sido evaluados).

Coordinación dentro y fuera del Banco

El Banco cuenta con los recursos humanos
necesarios para responder ante situaciones
de desastre y atender las necesidades a largo
plazo de los países relacionadas con los
riesgos naturales, pero la movilización de di-
chos recursos es complicada.

El Banco tiene un cuerpo de funcionarios de-

dicados y con vasta experiencia, pero carece de
una forma eficaz para hacer llegar de manera con-
fiable ese personal y los conocimientos perti-
nentes a sus prestatarios, e incluso a sus propios
equipos. Desde 1999, una unidad integrada por
tres personas asiste a los jefes de proyectos del
Banco en materia de desastres naturales y tec-
nológicos, y ayuda a entregar una respuesta más
estratégica y expedita. Este grupo se comple-
menta con un grupo temático integrado por más
de 100 funcionarios con experiencia en casos de
desastre. No obstante, los donantes y los países
clientes no saben con quién ponerse en contacto
cuando tienen preguntas de rutina sobre situa-
ciones de desastre y las actividades de coordina-
ción conexas. Con los mecanismos que existen
actualmente, de hecho también se ha reducido la
visibilidad del tema de los desastres naturales en
el Banco. Cuando ocurre un desastre, puede ser
difícil sacar de sus tareas habituales al personal con
conocimientos y experiencia en la materia.

La coordinación de los donantes reviste es-
pecial importancia para las actividades de
asistencia y recuperación tras un desastre,
debido en parte al carácter dinámico de la
situación, pero también porque los desas-
tres normalmente atraen a numerosos do-
nantes que desean participar.

x l i i

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

En medida creciente, los propios prestatarios

se ocupan de la necesaria coordinación de los do-
nantes, pero siguen requiriendo asistencia para la
coordinación, sobre todo en las primeras etapas
de asistencia y recuperación.

La experiencia recogida de los proyectos mues-

tra que para elaborar una estrategia de recupe-
ración compartida por todos se requiere no sólo
la presencia inmediata del Banco en la zona afec-
tada por el desastre, sino también una presencia
prolongada que permita asegurar que se atiendan
todas las necesidades de reconstrucción, que el
diseño del plan sea adecuado a la capacidad dis-
ponible, que se satisfagan las necesidades de las
partes interesadas, que la distribución del trabajo
sea razonable, y que se tengan en cuenta las ne-
cesidades de los grupos pobres y vulnerables.

En 1989, por ejemplo, negociadores del Banco

en Sudán trabajaron con otros donantes para ase-
gurar que se consideraran los intereses de estos
últimos y que no hubiera una innecesaria dupli-
cación de esfuerzos. Al mantener la flexibilidad en
la composición de la contribución del Banco, se
ayudó a los demás donantes a introducir ajustes
a sus programas. Luego, el Banco financió el resto.

La comunidad del desarrollo debería tra-
bajar con los prestatarios afectados por un
desastre en una etapa más temprana y con-
tinuar participando por más tiempo.

A los gobiernos se les debería hacer presente

con toda claridad la experiencia internacional
sobre los impactos que produce una gestión acer-
tada o desacertada de la asistencia y sobre la ca-
pacidad de las principales partes interesadas para
participar de manera eficaz en el proceso de re-
cuperación. Concretamente, el Banco debe estar
presente durante la etapa de emergencia para
asegurar el éxito de los proyectos de recons-
trucción que financia. Los grupos de menores
ingresos de la comunidad necesitan apoyo hasta
que desarrollen la capacidad para ocuparse de la
gestión de la infraestructura que se haya puesto
su cuidado.

Recomendaciones

En el Capítulo 6 del informe se presentan varias
sugerencias concretas con respecto a la revisión
de la política del Banco relativa al financiamiento

para situaciones de emergencia; dichas sugeren-
cias no se repiten aquí en toda su extensión.

Preparar una estrategia o plan de acción para
la asistencia en casos de desastres naturales

La asistencia del Banco en casos de desastres na-
turales podría mejorar si se elaborara una estra-
tegia o plan de acción —con las correspondientes
orientaciones— que:

• ayudaría al personal a responder ante situa-

ciones de emergencia con asistencia expedita
y actividades de reconstrucción bien planifi-
cadas, y a actuar con eficacia en un período
mucho más breve;

• aseguraría que los fondos para situaciones im-

previstas (ya sea a escala nacional, regional o
mundial) representaran para todos los países
prestatarios una respuesta financiera oportuna
y adecuada ante acontecimientos graves;

• Ayudaría a gestionar los riesgos naturales en los

países más vulnerables.

En la estrategia o plan de acción se debe señalar

una metodología para determinar el nivel de
riesgo de que ocurran desastres naturales en los
distintos países. Se recomienda que los países se
dividan en tres categorías: grupos de riesgo alto,
mediano y bajo. Luego, se debe establecer la ma-
nera en que el Banco brindará asistencia a los
prestatarios de cada una de estas categorías con
el propósito de reducir su vulnerabilidad y apro-
vechar las capacidades y el liderazgo de cada uno
de ellos.

En los países muy vulnerables, el plan de ac-

ción debe disponer que se preste más atención
a los riesgos naturales durante la etapa de eva-
luación inicial de los proyectos de inversión en ge-
neral, y específicamente durante la preparación
de los documentos de estrategia de lucha contra
la pobreza, las estrategias de asistencia a los paí-
ses y otros documentos de esa naturaleza. Cuando
corresponda, estos documentos no deberán li-
mitarse a presentar una descripción de los ries-
gos, sino señalar además actividades de mitigación
y desarrollo institucional susceptibles de segui-
miento.

En el caso de los países más vulnerables, se

debe disponer de financiamiento para situaciones

R É S U M É N

x l i i i

imprevistas, ya sea como parte de otro préstamo,
fondos de destinación especial en el programa de
financiamiento de la estrategia de asistencia a los
países, o un fondo para catástrofes de carácter in-
dependiente (aunque esto podría ser innecesario
si a la larga se establecieran fondos regionales o
mundiales). Otra alternativa que valdría la pena
considerar es la creación de un fondo especial ad-
ministrado por el Presidente y que podría utilizarse
para financiar el inicio inmediato de las activida-
des en cuanto se produzca un desastre.

El diseño de los proyectos financiados por el

Banco en países que se consideren de riesgo me-
diano o alto deberá poder adaptarse en caso de
que ocurra un desastre. En los documentos ha-
bituales de evaluación de riesgos que se preparan
para todos los países se deberán considerar los ries-
gos de que se produzcan desastres naturales.

La estrategia o el plan de acción se deberá en-

viar al Directorio para su discusión.

Revisar la política del Banco para orientar
mejor al personal y aumentar la flexibilidad de
las respuestas de la institución ante los
desastres naturales.

Las emergencias son de diversa naturaleza y, pese
a que existen ciertas semejanzas, la mayoría difiere
en aspectos muy importantes de los desastres
provocados por fenómenos naturales. La política
del Banco debe reflejar esas diferencias y consi-
derar por separado los conflictos y las enferme-
dades epidémicas, con disposiciones que se
apliquen únicamente al tema pertinente. Hay dos
maneras de hacerlo: los desastres naturales pue-
den ser materia de una política operacional in-
dependiente (como se pide en la evaluación del
Grupo de Evaluación Independiente de 1998 re-
lativa a la experiencia del Banco en actividades de
reconstrucción después de un conflicto), o bien
la OP 8.50 podría incluir disposiciones específicas
en relación con los desastres naturales, las situa-
ciones posteriores a los conflictos, la salud y otras
emergencias, de manera que cada tema se trate
por separado. Cualquiera sea la forma que adopte,
la política del Banco debe centrar más la atención
en la prevención de desastres y la reducción de
los niveles de vulnerabilidad en todas las opera-
ciones relativas a estos fenómenos. Se deben fle-
xibilizar las prohibiciones establecidas en la política

con respecto a la asistencia y el financiamiento en
el caso de acontecimientos recurrentes.

La tramitación acelerada y las disposiciones

para el rápido desembolso de los préstamos de
emergencia para recuperación han atendido par-
cialmente la necesidad de iniciar de inmediato las
actividades a corto plazo, pero podrían comple-
mentarse fructíferamente con un nuevo meca-
nismo, como un fondo central especial
administrado por la oficina del Presidente (simi-
lar al que existe en el Banco Interamericano de De-
sarrollo) para financiar las necesidades más
urgentes durante los primeros días de respuesta
después de un desastre.

No obstante, los préstamos de emergencia

para recuperación son menos adecuados para
actividades a más largo plazo, como las de miti-
gación, reconstrucción y fortalecimiento institu-
cional, que requieren más tiempo para la
preparación y la evaluación inicial, y no necesitan
quedar eximidos del cumplimiento de las normas
de salvaguardia y diligencia debida.

Del mismo modo, la atención de los problemas

sociales durante la preparación y ejecución de
las actividades suele requerir más tiempo del que
se dispone en el marco de los préstamos de emer-
gencia para recuperación. Tales actividades son
más apropiadas para las operaciones habituales
de financiamiento para proyectos de inversión,
pero con frecuencia se han dejado de lado habida
cuenta del período de implementación de tres
años de los préstamos de emergencia para recu-
peración y debido a la pérdida de interés de los
prestatarios en tomar un segundo préstamo.

Aumentar la capacidad de respuesta del
Banco en casos de desastre y asegurar que
ésta se pueda movilizar sin demora.

Ya sea que exista o no una unidad designada que
se ocupe de los desastres naturales y los riesgos
que éstos plantean, el Banco debe estar en con-
diciones de reunir y diseminar rápidamente su ex-
periencia internacional a los prestatarios en caso
de una emergencia. Además, los equipos de tra-
bajo necesitan apoyo para llevar a cabo las eva-
luaciones que deben realizarse luego de un
desastre, y para diseñar intervenciones de emer-
gencia que se ajusten a las necesidades y capaci-
dades de cada prestatario.

x l i v

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Para responder ante situaciones de desastre se

requieren conocimientos y experiencia en diver-
sos sectores. La inclusión de personas con cono-
cimientos en materia de desastres en las misiones
del Banco tras una crisis grave puede resultar
fundamental. La selectividad a la hora de identi-
ficar a los funcionarios que integran las misiones
en situaciones posteriores a un desastre permite
evitar los problemas relacionados con el diseño
y la escala de la respuesta que se pueden presentar
cuando se envía a personas que no están acos-
tumbradas a ver destrucción en gran escala o que

no conocen el país. El Banco cuenta con muy
pocas personas con esas características, y actual-
mente no tiene un mecanismo congruente para
movilizar a estas personas responsables de res-
ponder ante un desastre natural. El hecho de
sacar a los integrantes del Grupo temático sobre
gestión de riesgos de sus responsabilidades pro-
duce inevitablemente un impacto negativo en
sus actividades habituales. Además, dado que el
número de funcionarios que cuentan con los co-
nocimientos necesarios es tan reducido, se suele
recurrir a las mismas personas una y otra vez.

R É S U M É N

x l v

x l v i i

ACRONYMS AND ABBREVIATIONS

ADB

Asian Development Bank

AFR

Sub-Saharan Africa Region

Bank Procedure

CAS

Country Assistance Strategy

CGIAR

Consultative Group for International Agriculture

DEC

Development Research Group (World Bank)

DFID

Department for International Development (U.K.)

DRI

Disaster Risk Index

DRU

Disaster Reduction Unit (UNDP)

EAP

East Asia and the Pacific Region

ECA

Europe and Central Asia Region

EERL

Emergency Earthquake Reconstruction Loan (Turkey)

ERL

Emergency Recovery Loan

GDP

Gross domestic product

HIPC

Highly Indebted Poor Countries Initiative

HMU

Hazard Management Unit

HRM

Hazard Risk Management (Team)

IBRD

International Bank for Reconstruction and Development (World Bank)

IDA

International Development Association

IDB

Inter-American Development Bank

IEG

Independent Evaluation Group

IEG-WB

Independent Evaluation Group–World Bank

LAC

Latin America and the Caribbean Region

MDGs

Millennium Development Goals

MFIs

Multilateral financial institutions

MNA

Middle East and North Africa Region

NGO

Nongovernmental organization

O&M

Operation and maintenance

OECD

Organisation for Economic Co-operation and Development

Operational Directive

Operational Policy

OPN Operational

Policy

Note

PIR

Policy Implementation Review

PIU

Project Implementation Unit

PRSP

Poverty Reduction Strategy Paper

PTI

Poverty Targeted Intervention

SRO

Small Rural Operation

UNDP

United Nations Development Program

UNESCO

United Nations Educational, Scientific, and Cultural Organization

GLOSSARY

Disaster

A serious disruption of the functioning of a community or a society causing
widespread human, material, economic, or environmental losses that exceed
the ability of the affected community or society to cope using its own
resources.

Disaster risk

The systematic process of using administrative decisions, organization,

management

operational skills, and capacities to implement policies, strategies, and coping
capacities of the society and communities to lessen the impacts of hazards.

Disaster risk

The conceptual framework of elements considered with the possibilities to

reduction

minimize vulnerabilities and disaster risks throughout a society, to avoid

(disaster

(prevention) or to limit (mitigation and preparedness) the adverse impacts of

reduction)

hazards.

Hazard

A potentially damaging physical event, phenomenon, or human activity that
may cause loss of life or injury, property damage, social and economic disrup-
tion, or environmental degradation.

Mitigation

Structural and nonstructural measures undertaken to limit the adverse impact
of natural hazards, environmental degradation, and technological hazards.

Preparedness

Activities and measures taken in advance to ensure effective response to the
impact of hazards, including the issuance of timely and effective early
warnings and the temporary evacuation of people and property from threat-
ened locations.

Prevention

Activities to provide avoidance of the adverse impact of hazards and means to
minimize related environmental disasters.

Recovery

Decisions and actions taken after a disaster with a view to restoring or improv-
ing the pre-disaster living conditions of the stricken community, while encour-
aging and facilitating necessary adjustments to reduce disaster risk.

Relief/response

The provision of assistance or intervention during or immediately after a
disaster to meet the needs of those affected. It is generally immediate and
short term.

Resilience/

The capacity of a system, community, or society potentially exposed to hazards

resilient

to adapt, by resisting or changing to reach and maintain an acceptable level of
structure and functioning.

Risk

The probability of harmful consequences, or expected losses (deaths, injuries,
property and livelihood loss, economic activity disrupted or environment
damaged) resulting from interactions between natural or human-induced
hazards and vulnerable conditions.

x l i x

Nature, Disaster,
and Recovery

Disasters reflect the ways societies structure themselves and allocate their resources.

n December 26, 2004, an undersea earthquake measuring 9.0 on the
Richter scale hit off the coast of Sumatra. The fast-moving tsunamis
it generated devastated the shores of countries from Indonesia to

Somalia, killing an estimated 220,000 people and leaving 1.5 million people
homeless.

Three months later, an 8.7 magnitude earthquake
with roughly the same epicenter generated
widespread panic at the prospect of another
tsunami and killed almost 2,000 people in
Indonesia. Then, in the fall of 2005, a record
number of hurricanes battered the Caribbean,
Mexico, and the U.S. Gulf Coast. At the same time,
Central America experienced a series of natural
disasters including a hurricane, flooding, and an
earthquake.

Most recently, world attention focused on the

powerful 7.6 magnitude earthquake that struck
the region bordering Pakistan and northern
India on October 8, 2005. A humanitarian
disaster of enormous proportions, the
earthquake devastated towns and villages
throughout the region, leaving tens of
thousands of dead and injured, and millions
homeless. In Pakistan, the official government
estimate of the death toll was 86,000. India did
not escape the devastating effects of the quake,
with estimates of over 1,000 deaths.

The reported number of disasters has been

increasing, growing from fewer than 100 in 1975 to
more than 400 in 2005. This increase has many

possible explanations
(box 1.1). Without doubt,
though, the cost of
disaster damages has
been exploding: the
economic costs of major disasters in constant
dollars are now estimated to be 15 times higher
than they were in the 1950s—$652 billion in
material losses in the 1990s (IMF 2003) (figure 1.1).

The human cost is also high: over the

1984–2003 period, more than 4.1 billion people
were affected by natural disasters (Guha-Sapir,
Hargitt, and Hoyois 2004, p. 85). The number of
people affected has also been rising, from 1.6
billion in the first half of that period (1984–93)
to almost 2.6 billion in the second half
(1994–2003), and has continued to increase in
the current decade (Guha-Sapir, Hargitt, and
Hoyois 2004, p. 85).

This growth in damage to built environments

and to the societies that use them is a product of
human actions. Social and economic vulnerabil-
ity to natural events is rising as the world
becomes more populated. The pursuit of
economic opportunity brings more people into

Natural disasters are
becoming increasingly
destructive.

urban settings and fragile
coastal areas where the
damage of earthquakes
and extreme weather
events is often greatest.
The more vulnerable

people are, the more disastrous a natural event
will be. Increasing environmental degradation
also contributes to the intensification of the
effects of natural events.

• In drought, problems associated with a short-

age of water are exacerbated by deforestation,
soil erosion, and inappropriate land use.

• Floods are caused by the silting up of rivers

and the loss of absorptive capacity of the soil—
both legacies of poor agricultural practices
that destroy groundcover and other natural en-
vironmental defenses.

• Human actions that contribute to the de-

structiveness need to be addressed.

• Destruction of forests and overgrazing to meet

the needs of growing population leads to
desertification.

• Earthquakes are most destructive in countries

with poor building code enforcement and high
seismic vulnerability of construction, and when
landslide-prone, steeply sloped land loses grass
and forest cover and becomes occupied by in-
formal housing.

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Several factors contribute to the apparent increase in the number of reported disasters.

Source: EM-DAT: The OFDA/CRED International Disaster Database-www.em-dat.net-Université Catholique de Louvain, Brussels.

•

Increases in relief and reconstruction assistance have encouraged international reporting of more disasters. This is particularly

the case for smaller events, which were previously treated as a local concern (IFRC 2005).

•

More specialized agencies are tracking natural events and their disastrous impacts. Many country governments have now de-

veloped specialized agencies for tracking and reporting on natural disasters. The increased accuracy of observation and re-

porting on the weather contributes to the increase in reported extreme weather events—a 50 percent increase each decade

from the 1950s to the 1990s (Guha-Sapir, Hargitt, and Hoyois 2004).

•

Sea temperatures have risen. A rise in tropical sea temperatures of up to 2 degrees Fahrenheit over the past century has con-

tributed to an increase in weather-related disasters, some of which may be cyclical in nature (Webster and others 2005).

Box 1.1: Why Do Natural Disasters Seem to Be Increasing in Number?

550

500

450

400

350

300

250

200

150

100

1975

1980

1985

1990

1995

2000

Number of disasters

Year

Human actions that

contribute to the

destructiveness need to be

addressed.

Developing Countries and
Natural Disasters

Although disasters caused by natural events
occur throughout the world, losses to disaster in
developing countries are generally much higher
than in developed countries in percentage of
gross domestic product (GDP) or government
revenues. For example, Maldives’ tsunami losses
amounted to 66 percent of GDP. Hurricane
Mitch caused losses equal to 41 percent of GDP
in Honduras.

Even when the national impact is relatively

small, the local impact may be catastrophic—the
2004 tsunami is estimated to have reduced
Indonesia’s GDP growth by just 0.1–0.4 percent,
but the province of Aceh suffered destruction of
its capital stock equivalent to 97 percent of its
GDP. Small island nations can lose multiples of
their GDP to natural disasters: Grenada lost 200
percent of its GDP to Hurricane Ivan.

The damage caused by large disasters can also

outweigh development assistance. The Kashmir
earthquake of October 2005 caused an estimated
$5 billion in damage in Pakistan,

roughly equiva-

lent to the total official development assistance
for the preceding three years, and equivalent to
the amount the World Bank had lent to the

country over the preced-
ing 10 years.

The impacts of natural

disaster on societies also
differ, depending on their
level of development. In December 1999,
landslides in Venezuela and storms that hit France
caused similar amounts of physical damage,
estimated at about $3 billion in both cases. But the
human costs differed greatly: there were 50,000
deaths in Venezuela compared with 123 in France.

The disproportionate effect on developing

countries has many explanations. One is simply
that areas of the world that have harsh climatic
conditions, are subject to extreme weather, or
have unstable geology are difficult to develop,
and development gains can be fragile and easily
overwhelmed by the effects of natural events.
Lack of development itself contributes to disaster
impacts, both because the quality of construction
often is low and building codes, land registration
processes, and other regulatory mechanisms are
lacking, as well as because numerous other
development priorities
displace attention from
the risks that natural
events present.

N AT U R E , D I S A S T E R , A N D R E C O V E R Y

Figure 1.1: The Cost of Disaster Damage Is Rising

100

200

300

400

500

600

700

1950–59

1960–69

1970–79

1980–89

1990–99

US$ billions, 2001 values

Source: IMF 2003.

Note: Data are for “great” disasters, in which the ability of the region to help itself is distinctly overtaxed, making interregional or international assistance necessary.

Impacts of natural
disasters are very large in
developing countries.

Lack of development
contributes to
destructiveness.

Response, Recovery, and Reconstruction

The first days following a disaster are typically
very dynamic, in part because the global and
national media sharply and swiftly focus
attention on the immediate needs of victims and
because of the risks to health and social order.
Donors and governments are compelled to act.
However, their initial actions affect all future
actions.

A Harvard University study of 30 disaster and

relief and reconstruction efforts concluded that
initial actions are never neutral—they either
support longer-term development or undermine
it (Anderson and Woodrow 1989). In the first
months after a disaster the situation remains

highly dynamic—needs
can shift very quickly and
missteps are common
and can have serious
consequences. In Bolivia,

for example, initial relief efforts created additional
difficulties for the recovery (box 1.2).

Without question, attention to natural disasters

is growing. In particular, the recent Asian
earthquakes focused the world’s attention on the
magnified effects of disaster in developing
countries and generated commitments of support
from donor nations. Shortly after the tsunami

disaster, an unprecedented outpouring of interna-
tional support provided assistance to the affected
populations. But donor promises during the first
few weeks following a disaster are usually reduced
later, or even dropped altogether, when initial
estimates of need prove too high or when the
sudden inflow of assistance exceeds the country’s
ability to manage the funds.

In any event, most of the costs of recovery

ultimately are borne by the country itself. The
Organisation for Economic Co-operation and
Development (OECD) Development Aid
Committee has reported that outside financing
and donations usually offset less than 10 percent
of a country’s disaster losses (Linnerooth-Bayer
and Amendola 2000). Hence, disasters can
represent a permanent loss of development
momentum.

Although the destructive impacts of disasters

are tightly connected with development,
disasters are typically treated as an interruption
in development rather than as risks that should
be a calculated part of development. Some
countries are in a near-permanent state of
recovery. The countries themselves also tend to
lose sight of long-term priorities related to
reducing their vulnerability to disasters as
immediate needs are met and media attention

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

The actions of the first

few days affect all future

decisions.

Following an earthquake that devastated the urban areas of

Aiquile, Totora, and Mizque on May 22, 1998, the Government of

Bolivia requested an International Development Association

(IDA) reallocation of $5 million from the El Niño Emergency As-

sistance Credit to help finance reconstruction. A Bank techni-

cal mission found that a number of troubling decisions had been

made on how temporary shelter would be provided:

•

Victims of the disaster had been obliged to abandon their

homes and possessions and move to refugee camps under

military control, where they were fed for free. This effectively

destroyed the local economy and, without commerce, few

could afford to pay for services.

•

Victims had not been allowed to recover the recyclables from their

homes (doors, windows, floor and roof tiles, kitchen and bathroom

fixtures), which represented about 70 percent of their cost.

•

Bulldozers knocked down damaged buildings without pre-

serving the boundary lines between properties, paralyzing sub-

sequent rebuilding efforts.

•

Temporary settlement camps had been set up in a manner that

led to social disintegration and abetted robbery and assault.

The Bank ultimately directed the requested reallocation to

rural housing. The way in which the early part of the process

was managed had severely compromised the recovery of the

urban areas, because the Bank funds would have filled a major

gap in the assistance needed to rebuild the cities, had they been

directed there.

Box 1.2: A Troubled Relief Compromises the Ability to Recover

Source: Field interviews and observation in 1998 and World Bank data.

turns elsewhere. Mitigation, prevention, and
disaster risk management often drop off the
development agenda and may be neglected as
attention returns to other pressing development
priorities.

Cleaning up the damage and rebuilding

structures without addressing the human
actions that turn recurring natural phenomena
into disasters only ensures that the inevitable
next event will be as disastrous as the last.
Annual flooding only regenerates agricultural
soil when human settlements are not located in
floodplains. Where environmental degradation
turns seasonal events into disaster, environmen-
tal restoration needs to be part of the solution.
In other places, increased attention to
infrastructure and settlement design is all that is
required to increase disaster resilience. Effective
activities that address root causes of vulnerabil-
ity and mitigate the potential for future damage
are crucial to reducing the steady erosion of
development gains that natural disasters
represent.

The World Bank and Natural Disasters

The World Bank has financed reconstruction
since its inception and increasingly has been
engaged in helping countries recover from
disasters and reduce their future vulnerability to
natural hazards. Since 1984, most of the Bank’s
borrowers—110 countries in all—have sought
emergency financial assistance related to
disaster.

Since the 1970s, requests for Bank financing of

post-disaster reconstruction projects, humanitar-
ian crises, and post-conflict recovery have grown
steadily. A succession of Bank policy statements
has been developed to guide this work:
Operational Policy Note (OPN) 10.07, Guidelines
for Bank Participation in Reconstruction
Projects after Disaster (1984); Operational
Directive (OD) 8.50, Emergency Recovery
Assistance (1989); and Operational Policy (OP)
8.50, Emergency Recovery Assistance (1995).
Table 1.1 outlines the key provisions of the current
policy (Annex A analyzes the evolution of policy
over its 20-year existence). The changes in policy
over time were made primarily for institutional
reasons (because all Bank policies were changed

from OPNs to ODs, and
then into OPs).

The Bank’s three

policy statements reflect
an evolution in thinking
about its response to
emergencies such as
natural disasters. The three statements differ in
how they characterize emergencies, what
emergencies are covered, and in several other
areas, but all make timeliness a key concern. This
concern led to the creation a quick-disbursing
instrument in the 1984 OPN. The Emergency
Recovery Loan (ERL), as it came to be called, has
become the instrument of choice in lending for
natural disaster emergencies. Although policy has
evolved, the changes have been made without
benefit of evaluation or knowledge of Bank
experience.

Although it has a policy

on emergency assistance
for disasters, the Bank has
never had a strategy for
that assistance. Therefore,
strategic planning for
natural disasters has been confined to country-
level analyses in Country Assistance Strategies
(CASs) and Poverty Reduction Strategy Papers
(PRSPs).

The Bank is one of a large number of institu-

tions that countries can call on and coordinate
after a disaster. Indeed, the number of institutions
that respond to disaster has been growing, making
donor coordination an increasing challenge.

The roles of the various institutions are not

fixed and have blurred over time, though they
tend to follow traditional strengths. For
instance, the Red Cross/Red Crescent Society
and other international and national nongovern-
mental organizations, along with the military,
typically participate in the immediate response.
The United Nations Development Program
(UNDP) focuses more
on the social aspects of
recovery. The World
Bank typically concen-
trates on infrastructure
and housing during the
reconstruction, given its

N AT U R E , D I S A S T E R , A N D R E C O V E R Y

Disasters are often
treated as an
interruption in
development rather than
as a risk to development.

Most Bank borrowers
have sought assistance
related to natural
disasters.

The number of
institutions responding to
disaster is growing and
coordination is becoming
more complex.

comparative advantage in that area. However,
the Bank also has considerable experience with
disaster recovery, as well as an important role in
assisting with coordination that ensures that
country needs are met with as few overlaps and
conflicts of priorities as possible.

A key partnership of the Bank in natural

disasters is the ProVention Consortium,
launched in February 2000 to reduce disaster
risk in developing countries and to make
disaster prevention and mitigation an integral
part of development efforts. The Consortium is
an international network focused on sharing
knowledge and leveraging resources to reduce
disaster risk. Though launched by and originally
housed in the Bank, it is currently under the
management of the International Federation of
Red Cross and Red Crescent Societies in
Geneva. This report does not evaluate the
partnership, but the Independent Evaluation
Group (IEG) will be evaluating it in 2006, and
therefore offers no judgments on its perform-
ance in this report. The current evaluation does
draw on ProVention analyses.

The Evaluation

Heightened global awareness, increased public
and private generosity, growing Bank investment
in disaster recovery as well as disaster prevention
and risk management, and greater ability to
anticipate some natural events make this an
appropriate time for the World Bank to review
and update its policy and upgrade practices with
respect to natural disasters. This evaluation has
been undertaken to inform that process.

No assessment has previously been done of

the Bank’s disaster-related assistance. The
evaluation assesses the relevance and effective-
ness of Bank activities related only to emergen-
cies caused by natural events. Emergencies
caused by armed conflict have been the subject
of an earlier evaluation (IEG 1998), and thus are
not covered by this report.

The study was conceived as a review of the

implementation of Bank policy and examined the
relevance and effectiveness of Bank operations, as
well as their institutional development impact, to
develop lessons from experience. The evaluative
questions addressed are detailed in Appendix B.

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Relief and consumption

The Bank does not finance relief and consumption (par. 2 and 4).

Support for damage and

The Bank provides immediate support in assessing the emergency’s impact and develops a recovery strategy

needs assessment

(par. 3).

Implementation time

ERLs are fully implemented in two to three years (par. 6).

Procurement rules

Standard Bank Operational Policies, including those on procurement, consultants, and disbursement, apply to ERLs

(par. 8).

Suitability for

Regular investment projects (not ERLs) may be preferable for recurrent disasters (floods) and slow-onset

recurrent disasters

disasters (droughts) (par. 5).

Design standards,

ERLs use disaster-resilient construction standards, emergency preparedness studies, and technical assistance

prevention, and

for prevention and mitigation. Prevention and mitigation projects carry out studies of vulnerability and risk

mitigation

assessment, reinforce vulnerable structures, adjust building and zoning codes, and acquire hazard-reduction

technology (par. 6).

Institutional and

The Bank helps countries to establish an adequate institutional and regulatory framework for prevention and

regulatory framework

mitigation (par. 10).

Donor coordination

Collaboration with the UNDP and other international agencies, local nongovernmental organizations, and donors

is helpful in designing the recovery assistance strategy under an ERL and in designing specific prevention and

mitigation programs (par. 9).

Source: Operational Policy 8.50 – Emergency Recovery Assistance, August 1995.

Note: ERL = Emergency Recovery Loan.

Table 1.1: Key Provisions of Current Bank Policy Statement

Evaluative Instruments and Methods

The study examined the Bank’s experience from
several angles. The basic approach was to avoid
sampling, and instead identify all Bank-financed
projects with natural disaster activities. For
example, in the staff survey, all task managers that
worked on at least one project with disaster activi-
ties were invited to respond. When the evaluation
looked at an activity (such as housing) or a disaster
type (such as tropical storms), all the relevant
projects were reviewed. Hence, different analyses
use different numbers of projects, but all use the
full universe of projects relevant to that issue. The
key study instruments were as follows (see
Appendix B for details and the methods used):

• Expert knowledge –

through an extensive
review of the literature
and the use of an Ad-
visory Panel

• Bank-financed proj-

ects and activities – through a review of the
portfolio of projects and analysis of a textbase
of project information

• Staff knowledge – through surveys and in-

terviews

• Detailed examination of critical issues

and countries – through issues papers and
case studies (field-based and desk studies).

N AT U R E , D I S A S T E R , A N D R E C O V E R Y

No previous assessment
has been done of Bank
assistance for natural
disasters.

1 1

The World Bank Responds

atural disasters directly affect the traditional beneficiaries of World Bank–
financed development assistance—98 percent of the 211 million peo-
ple affected by natural disasters each year from 1991 to 2000 were from

developing nations (IFRC 2001).

And although a few disasters have been devastat-
ing to the better-off (such as when cities with a
large number of poorly built and badly designed
high-rise apartments are hit by earthquakes),
unquestionably the most vulnerable to disaster
are the poor, who live in dangerous zones, on
marginal lands, and have precarious livelihoods.

The Bank has a major stake in ensuring that

country assistance following a disaster
contributes to long-term development, and that
the potential future impacts of disasters are
reduced or prevented. As the human and
economic losses from natural disasters continue
to rise, they represent a large and growing
obstacle to sustainable development. Hence,
the portfolio of projects supported by the Bank
includes activities related to their immediate
response, recovery, reconstruction (with mitiga-
tion), and prevention.

The Bank’s Natural Disaster “Portfolio”

The portfolio of projects approved by the
Executive Board of the World Bank (International
Bank for Reconstruction and Development
[IBRD] and International Development Associa-
tion [IDA]) since 1984 that have had some activity

involving natural disasters
contains projects entirely
devoted to disasters,
projects with formal
disaster components as
part of a larger project,
and projects with smaller disaster activities.

consists of 528 projects hereafter referred to as
“disaster projects” when discussing the full
universe of activities (see Appendix B for more on
how these projects and the amounts of lending
were identified). A total of $26,281 million in
activities have focused on natural disasters.

These

528 projects represent 9.4 percent of all Bank loan
commitments since 1984.

Some projects are entirely devoted to natural

disasters. These include Emergency Recovery
Loans (ERLs; including IDA credits and grants),
as well as projects using other instruments. Such
projects represent less than a third of the 528
identified projects (table 2.1). The value of these
loans and credits is $12,200 million.

The Growth of Bank Lending for Disaster

Looking at the full portfolio (all 528 projects),
increasing Bank involvement over time is

Natural disaster
assistance accounts for
9.4 percent of
commitments since 1984.

evident (figure 2.1). In
addition, as a share of
overall Bank lending,
assistance related to
natural disasters has

been growing (see Appendix C, figure C.1).
Much of this assistance is quite recent: about 43
percent of all disaster-related loans have yet to
close. Lending for disasters is noticeably cyclical,
with peaks about every five years. Even so, Bank

lending for disaster has risen gradually over the
past 20 years. If projects that are still open (those
approved since fiscal 1999) reallocate according
to the historical trend, the numbers in the most
recent years will rise. Project reallocations are
discussed later in this chapter.

Among the 528 projects, the amount of

disaster-related support ranges from a few
thousand dollars for fire detection towers in a
forestry project to a $500 million loan for post-

1 2

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Total

Completed

Ongoing

All projects with some disaster activity

528

303

225

Emergency Reconstruction Loans

Disaster projects using other instruments

Projects with at least one full disaster component

130

Projects with an identified disaster activity below component level

258

131

127

Source: World Bank data.

Note: Data based on project approval year.

Table 2.1: Natural Disaster Portfolio Composition, 1984 to 2005

Lending for disasters has

been gradually

increasing.

Figure 2.1: The Number of Projects Related to Natural Disasters Has Been Rising, with Sharp

Peaks in Lending about Every Five Years

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Approval fiscal year

Number of projects

500

1,000

1,500

2,000

2,500

3,000

3,500

Millions of US$

Number of projects

Loan commitments

Source: World Bank data. Reallocations are recorded by approval fiscal year as well.

earthquake reconstruction. In recent years the
scale of individual operations financed by the
Bank has grown. The 1999 Turkey earthquake
alone led to Bank commitments of over $1.1
billion and a Bank-coordinated reconstruction
program of $1.7 billion. Other relatively recent
events have unleashed large responses because
they have affected multiple countries (such as
the Indian Ocean tsunami, Hurricane Mitch, and
the El Niño phenomenon).

Lending by Disaster Type

The kind of disaster for which countries most
frequently requested Bank financing was
flooding. Of the 528 projects in the portfolio,
243 had a flood-related activity. Drought was the
second most common disaster type, with 107,
and fire was third, with 95 projects. Since
drought is the only type of disaster that is
entirely slow onset, about 80 percent of the
portfolio is primarily focused on rapid-onset
events.

Notably, 127 of the 528 projects

responded to more than one disaster (floods

and tropical storms, for
instance), either because
they struck together or
because another disaster
occurred during imple-
mentation and the proj-
ect had to be modified to take the later event(s)
into account.

Of the 140 completed and

ongoing projects that were completely devoted
to disaster, 21 (in 17 countries) have focused
wholly on prevention activities.

The Bank has approved 89 ERLs for natural

disasters since 1984.

Reflecting the general

trend for the whole portfolio, floods dominated
this group of projects, earthquakes were the
second most common, and tropical storms were
the third.

Lending by Region

The Sub-Saharan Africa Region had the largest
number of disaster projects, 134 (figure 2.2).
The Latin America and Caribbean Region was a
close second with 128. The Middle East and

T H E W O R L D B A N K R E S P O N D S

1 3

Flooding most frequently
gets a Bank response, but
many projects respond to
multiple disasters.

Source: World Bank data.

Figure 2.2: The Bank Has Most Often Provided Lending for Disasters in Africa and in

Latin America and the Caribbean

100

120

140

160

Africa

Latin America

and the

Caribbean

South Asia

East Asia

and the

Pacific

Europe and

Central Asia

Middle East

and

North Africa

Number of projects

Ongoing

Completed

North Africa Region had the smallest portion of
projects. This distribution generally follows the
broader trend in Bank lending. The Bank has
helped the Africa Region to confront natural
disasters for many years. It is the only Region
where drought is the most common hazard, and
it also had the largest proportion of completed
projects. Recent attention has gone to the East
Asia and Europe and Central Asia Regions, which
have a majority share of their portfolios in
ongoing projects. The evaluation also compared
Regional disaster lending with overall Regional
lending to identify Regions where disaster
impacts were disproportionate (see Appendix
C, figure C.2).

Lending by Sector

The rural sector implemented 40 percent of the
natural disaster portfolio. This is not surprising,
since many disasters, such as floods, droughts,

and fire, and especially
their mitigation and
prevention activities,
take place in projects
involving forest manage-
ment, irrigation, and

disaster-resistant crops. Other sectors that have
been highly involved are transport (13 percent),
environment (10 percent), and urban (9
percent). Among the 528 projects were 12
multisectoral projects, accounting for 2 percent
of the total.

Analytic Work

The Bank has prepared 65 publications, working
papers, articles, and reports on natural disaster
topics since 1999.

The most frequently

researched topics have been risk management
and financing mechanisms. Although much
research work on these topics is done collabora-
tively with country staff or staff in other depart-
ments with various expertise, 31 were done
under the auspices of the Hazard Management
Unit. As might be expected, most of these are
global in scope or topically oriented. However,
about a third of them (10) focused on a particu-
lar country or Region.

Regional or country offices are responsible

for an additional 17 reports or publications. The

East Asia Region has produced five reports on
topics such as risk management in the Pacific
region, with a pilot study of Vanuatu and a report
on the Philippines focusing on enhancing
poverty alleviation through disaster reduction.
Among the Latin America and the Caribbean
Region’s four reports is one on the catastrophe
insurance market in the Caribbean and another
on “Risk and Vulnerability in Guatemala.” This
seems a particularly relevant report, as
Guatemala is in the “top ten” list of borrower
countries that are at greatest economic risk for
natural disasters, and it has not borrowed from
the Bank for natural disasters in 20 years. South
Asia’s four studies have included “Financing
Rapid Onset Natural Disaster Losses in India”
and “Bangladesh: Climate Change and Sustain-
able Development.” Europe and Central Asia
have done two research reports on the topic,
most notably one in Turkey entitled “Poverty and
Coping with Crises.” One of the two done in
Africa was on “Systemic Shocks and Social
Protection.”

In total, notwithstanding authorship, 15

countries were the subject of specific analytical
work on natural disasters. All but one of the
countries are considered to be of elevated
economic vulnerability and/or mortality risk.

The Environment Department of the Bank

has devoted resources to at least six reports on
natural disasters—most on climate change. The
Development Research Group (DEC) has
produced five reports in the area in the past six
years, including “Natural Disasters and Develop-
ment” and “Can Financial Markets be Tapped to
Help Poor People Cope with Weather Risks?”
The Social Protection Department has done two
studies that have included major sections on
natural disaster risk. The Consultative Group on
International Agricultural Research (CGIAR) has
produced a study with the Bank on rebuilding
agriculture in countries affected by natural
disaster.

Concentrations in Lending Patterns

A majority of Bank member countries have
turned to the Bank for emergency financial
assistance (110 countries) following a natural
disaster. However, lending has been highly

1 4

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

The rural sector

implements the largest

share of natural disaster

projects.

concentrated: 53 countries had only one or two
loans with disaster activities, while the top 10
countries had 208 projects (39 percent) among
them (table 2.2).

In terms of commitments as well, Bank

lending has been quite concentrated: 32 percent
of the $42,552 million in commitments over the
20-year period from fiscal 1985 through fiscal
2005 went to 7.5 percent of the projects. The 10
largest loans for reconstruction and/or preven-
tion totaled $3,882 million (table 2.3).

Reallocations

After a natural disaster, when a country requests
assistance, Bank country staff first examine the
existing country portfolio and identify loans
from which funds can be reallocated for
reconstruction. Over the past 20 years, funds
from 217 projects have been reallocated (see
Appendix C, table C.1).

The importance of reallocation as an

emergency response is highlighted by compar-
ing it with emergency recovery lending: since
1984, the Bank has made more than $3,047
million available for natural disaster response
through loan reallocations

and has dedicated

$9,021 million toward disasters through ERLs.
In a typical example, following the 2001
earthquake in Gujarat, 12 projects were restruc-

tured, providing a total
of $416 million for
immediate reconstruc-
tion. Funding of $10 to
$130 million per project
was taken from the
original implementing
agencies and given to another implementing
agency handling reconstruction. Projects
changed not only their scope, but also their
components and the sectors they targeted.

In part, reallocations appear to have been

sensitive to periodic increased awareness of
disaster in the Bank related to policy develop-
ment. The number of reallocations each year
varied from fewer than 10 in the 1980s to 17 in
the 1990s (figure 2.3).

The first jump in 1984 coincided with the

introduction of OPN 10.07, which encouraged
reallocations, and a second jump appeared in
1990, perhaps in part because of the renewed
visibility of emergencies with the introduction of
OD 8.50. While in the pre-policy period (1976–83)
reallocations averaged 2 each year, they averaged
11 each year from 1984
through 1999.

After 1999 disaster-

related reallocations
decreased, presumably

T H E W O R L D B A N K R E S P O N D S

1 5

Ten countries account for
39 percent of the lending
and 41 projects account
for about a third of the
lending.

A total of $3,047 million
has been reallocated from
217 projects since 1984.

Number of

Lending to projects

Rank:

disaster

with disaster activities

all Bank projects,

all Bank commitments,

Rank Country

projects

(US$

millions)

1984–2005

India

8,257

China

4,902

Bangladesh

2,880

Brazil

2,349

Honduras

712

Turkey

3,390

Yemen

306

Madagascar

327

Mexico

2,145

Vietnam

1,232

TOTAL

208

Source: World Bank data. Each reallocation counted as a separate project.

Table 2.2: Concentration of Lending in the Disaster Portfolio Compared

with Overall Bank Lending

because, for the most
recent years, the
ongoing projects are
still new and realloca-
tions tend to occur

toward the end of the project life, though it
may also be because ERLs have been used
instead of reallocations (more than a third of

the ERLs in the portfolio are ongoing; see
table 2.1).

While reallocations can give borrowers

flexibility to react to unforeseen needs, the Bank
has been developing promising alternatives.
Additional lending to existing projects is already
in use, though it has not yet been applied in a
disaster response. In addition, a specialized

1 6

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Approval

Original loan amount

Country

Project name

fiscal year

(US$ millions)

Turkey

Marmara Earthquake Emergency Reconstruction

2000

505

India

Emergency Tsunami Reconstruction

2005

465

India

Gujarat Emergency Earthquake Reconstruction

2002

443

Mexico

Mexico Natural Disaster Management Project

2001

404

Turkey

Istanbul Seismic Risk Mitigation

2005

400

Mexico

Earthquake Rehabilitation & Reconstruction

1986

400

Turkey

Emergency Flood Recovery

1999

369

India

Drought Assistance

1988

350

Turkey

Earthquake Rehabilitation & Reconstruction

1993

285

India

Maharashtra Emergency Earthquake Rehab. Project

1994

261

TOTAL

3,882

Source: World Bank data.

Table 2.3: Ten Largest Loans for Disasters

The level of reallocations

has responded to policy

changes.

1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Number of reallocations

OP 10.07

OD 8.50

OP 8.50

Figure 2.3: Reallocations Peaked Following Each Policy Revision

Source: World Bank data.

Note: Based on project approval years.

form of development policy lending (the
Contingent Hazard Recovery and Management
Loan) is in development.

These new alternatives may further help to

avoid the diversion of funds from their original
purposes, as in reallocations. In 18 cases reallo-
cations have undermined achievement of the
original objectives of the project. For example, in
January 1999, as a result of the earthquake in the
coffee growing region, the Government of
Colombia decided to use 55 percent of the
Secondary Education Project (a $90 million loan)
together with funds from three other projects to
support reconstruction. Instead of supporting
improvements in education, the proceeds went
into the Reconstruction Fund for the coffee
production zone. This decision was made
despite the concern of the Ministry of National
Education that the municipalities left out had no
other possibility of accessing supplementary
resources for educational improvement.

Project Performance

Bank-financed projects with at least one natural
disaster reconstruction and mitigation com-
ponent have had outcome and sustainability
ratings that are higher than the Bankwide

portfolio (figure 2.4).
Almost 80 percent of the
projects were rated
satisfactory for outcome
compared with the Bank
average of 72 percent for
the 1984–2004 period. A
sustainability rating of likely was attained by 63
percent of the disaster portfolio, seven points
better than the Bankwide average for the
1984–2004 period.

Where Bank-Financed Projects Perform Best

Disaster Type

Project outcome ratings
for most disaster types
were well above the Bank
average, with landslide
and earthquake projects
rated satisfactory on
outcome more frequently
than other disaster types.
The story is similar for
sustainability. Earthquake projects were most
frequently rated likely for sustainability, at a rate 21
points above the Bank average of 56 percent.

T H E W O R L D B A N K R E S P O N D S

1 7

Reallocations give
borrowers flexibility but
can prevent achievement
of planned development
goals.

Disaster projects perform
better than average on
outcome and
sustainability, and about
average on institutional
development.

Figure 2.4: Natural Disaster Portfolio Ratings: Projects Approved and Completed, 1984–2005

100

Outcome satisfactory

Sustainability likely

Institutional development

impact substantial

Percent satisfactory

Natural disaster portfolio

ERLs

Disaster only non-ERLs

One disaster component

Bankwide

Source: IEG data.

Institutional development impact performance

was closer to the Bank average. Only drought and
tsunami projects were rated above the average.
One reason why drought projects received higher
ratings than other disaster types was that such
slow-onset disasters provided a bit more time to
get the institutional framework right. These
projects tended to take longer to implement as
well, averaging almost seven years from approval
to completion (see also Chapter 4), so there was
more time for institutional capacity development
(see Appendix C, figures C.6–C.8).

Emergency Recovery Loans

ERLs (described in box 2.1) were rated substan-
tially higher than either the disaster portfolio or
the overall Bank portfolio (figure 2.4). Only 4 of
59 completed ERLs were rated unsatisfactory on
outcome, accounting for a 93 percent satisfac-
tory outcome rate.

Comprehensive Multisector Projects
and Other Subsets

Although only 12 in number, multisector
projects were the best performers, followed by
another subset of 9 social sector projects
(including health projects). Of the larger
subsets, the urban sector performed best,
although the rural and transportation sectors,
which contain the bulk of the portfolio, still
performed well above the Bank average (see
Appendix C, figure C.9).

Regions

There was a 22 percentage point difference
between the best-performing Region (Middle
East and North Africa), which was also the

smallest portfolio, and the worst (Africa), which
was the largest. Out of 13 completed projects in
the Middle East and North Africa Region, only
one was rated unsatisfactory on outcome. All but
the Africa Region performed above the Bank
average (see Appendix C, figure C.10).

Outputs and Outcomes of Projects

The Bank’s capabilities in assisting countries
with their response to natural disasters are
apparent in the results of its projects. Not
surprisingly, Bank-financed projects were best at
restoring physical assets. In 115 completed
projects, damaged infrastructure was success-
fully restored; in 28 projects, infrastructure
reconstruction was not fully completed or not
entirely successful (Appendix C, table C.2).

In addition, 86 projects had successful mitiga-

tion activities, while such activities were
unsuccessful in 32 projects. Despite this ability to
reconstruct infrastructure and provide some
additional security to the population, 73 projects
still recorded that a subsequent disaster lessened
the project’s impact (an additional 55 projects
were not disaster responses but were also
affected by disaster during implementation).

Based on limited experience with prevention

activities, the Bank may not yet have learned
enough about what activities are effective. Of
the 21 projects wholly devoted to prevention, 8
have been completed and evaluated, with about
63 percent rated satisfactory. Of the eight, three
were rated unsatisfactory on outcome and one
was rated highly unsatisfactory. This is a very
small sample from which to make judgments,
but it seems to indicate an area that may require
more Bank research.

Some of the negative results indicate the

difficulty of addressing social issues in disaster
projects (discussed in Chapter 5). Thirty-five
projects successfully restored economic assets,
and nine successfully restored social assets.
However, in six projects, stakeholders and
vulnerable groups were neglected, and target
groups were missed in two projects.

Lessons Learned

The lessons from project-level evaluations of 303
completed disaster projects exhibit some

1 8

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

An Emergency Recovery Loan (ERL) is a three-year lending instru-

ment. It has several advantages over other Bank lending instruments:

it allows for expedited processing from project initiation through Board

approval, quick disbursement through a positive list of imports, and delay

in meeting some safeguard and fiduciary requirements. It requires an

ad-hoc advisory committee headed by the country director. The ERL also

limits the use of conditionality.

Box 2.1: What Are Emergency Recovery Loans?

persistent themes—the top 12 are listed in table
2.4. Because certain lessons keep coming up, it
suggests that they are not being learned (IEG
2005c).

Hazard Management in the Bank

The growth in disaster-related lending in the late
1990s prompted the creation of a separate Bank
unit to guide staff. Between fiscal years 1999 and
2005 the Disaster Management Facility, later the
Hazard Management Unit (HMU), assisted Bank
task managers with natural and technological
disasters and helped provide a more strategic
and rapid response. The status of this unit
changed in 2005, as detailed below.

Bank management has recently adopted a

distributed, decentralized approach to hazard
risk management in the institution rather than
retaining a specialized central unit. The
current Hazard Risk Management Team of the
Urban Unit serves as the anchor for the Hazard
Risk Management Thematic Group, which
consists of more than 100 Bank staff in the
various organizational units with a particular
interest in hazard risk management.

Other

international organizations have found it

useful to centralize the
hazard risk manage-
ment function (see box
2.2 and Appendix C,
table C.3).

This arrangement handles emergencies un-

evenly, however. When a disaster strikes, the
country teams that are the Bank’s interface with
the borrowers are unlikely to have the expertise
needed and must call on others, either in the
Hazard Risk Management Team or the Hazard
Management Thematic Group, for technical
assistance. The quality of the result is partly a
function of who is around to answer the calls.

With the recent change, an important reserve

capacity has been lost. Soon after it was founded,
what was then the Disaster Management Facility
became the secretariat for
the ProVention program—
making the team far more
visible internationally. But
when that program left
the Bank, the team lost
staff that could help in
emergencies. Three people are too few to spread
across the natural disasters that occur every year,

T H E W O R L D B A N K R E S P O N D S

1 9

Projects performed best
when they restored
physical assets.

The lessons from disaster
projects do not seem to
have been learned—the
same ones are recorded
repeatedly.

Category

Times in database

Disaster management, preparedness, and mitigation need to be addressed

Simple and flexible procurement is fundamental to expeditious implementation

Lessons regarding Project Coordination Units (PCU) and/or working

with existing agencies (pros and cons)

Maintenance is critical for sustainability

Simple project design is more important when activities to be implemented are urgent

Community participation produces several identifiable benefits

Trade-off between careful project preparation and quick action

Emergency projects need experienced staff during project preparation

Assure borrower ownership by involving the highest levels of government

Donor coordination: cofinancing is preferable to parallel financing

Reconstruction often requires careful assessments and long-term efforts that

extend beyond the three-year implementation period for emergency operations

Studies need to be prepared before project approval

Source: IEG project database.

Table 2.4: A Dozen Lessons Learned from Natural Disaster Projects

and they are too few to be
both the face of the Bank
to the donor community
and to serve the needs
of countries affected
by disaster, while also
ensuring attention to
long-term reduction of

hazard risks in client countries and lending
programs.

Overall, the perceptions of the HMU were

positive among staff surveyed about the unit’s
relevance and effectiveness.

Eight respondents

(22 percent) stated that they had used the HMU in
their projects (see Appendix D for details of the
survey results). This number reflects the period
under study: only 6 of the 20 years reviewed
overlapped with the period in which the unit was
in existence. Among the respondents who had
used the HMU, the most helpful assistance was

seen as providing advice (7 respondents), provid-
ing project documentation or institutional
memory (4 respondents), and maintaining a
consultant database (2 respondents).

The task managers that used the HMU cited

additional services the HMU could provide,
including seed funds for supervision and
increased staff availability for missions. Task
managers were also asked to provide sugges-
tions of how the services provided by the unit
could be improved. Their requests were for
more assistance in designing prevention
policies, more training, the organization of a
more active thematic group, the promotion of
adjustments to the Bank’s ERL guidelines, the
inclusion of mitigation as a safeguard and the
mainstreaming of mitigation in regular lending
activities, and the promotion of the Bank’s
operational experience in conferences around
the world.

2 0

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

The Asian Development Bank (ADB), the U.K. Department for In-

ternational Development (DFID), Inter-American Development

Bank (IDB), and UNDP have specialized units and/or decen-

tralized disaster specialists that deal with disaster prevention,

mitigation, and management.

For instance, UNDP’s Disaster Reduction Unit (DRU) helps

country offices set up and provide more effective responses for

natural disaster reduction. The DRU is made up of 8 Geneva-

based professionals, 4 Regional Disaster Reduction Advisors (lo-

cated in Bangkok, Nairobi, New Delhi, and Panama), and 20–24

National Disaster Reduction Advisors in highly disaster-prone

countries. These senior officials act as permanent government

counterparts, identify disaster risk management opportunities,

and assist with relevant initiatives and oversee linkages with

UNDP programs.

Similarly, the IDB has 36 disaster risk management focal

points—26 individuals in the country offices, and 10 in key de-

partments in headquarters.

Box 2.2: Other Organizations Have Institutional Structures for Disaster Risk Management

The Bank has adopted a

less-centralized approach

to hazard risk

management than other

international

organizations.

2 3

Disasters and
Bank Strategy

Vulnerability to disasters is “largely dependent on development practices that

do not take into account susceptibility to natural hazards.”

ata available to the Bank on natural disasters have historically been
vague and fragmented, constrained by a shortage of reliable sources
in some countries, a relatively short history of data collection, and in-

consistent methodologies.

Reports typically presented a static view of
disasters, focusing on the number of people
killed and affected and on estimated disaster
damage. Disaster was rarely considered an
ongoing development challenge. The lack of
information, together with the perception that
disasters are random and unpredictable, limited
the Bank’s strategic planning for them. Hence,
the Bank had no overall strategy for disasters.
But if disasters are predictable, then planning
for them should be a normal part of develop-
ment work.

Predictability of Disasters

The preceding chapter showed that some Bank
borrowers frequently confront disasters brought
on by natural events. Two recently completed
studies on natural disaster risks confirm this
pattern and dispel much of the uncertainty and
unpredictability surrounding such events. An
understanding of the main messages of these
reports can broaden and deepen the
understanding of the Bank and borrowers, and
together with the results of this review,
culminate in a significant shift in strategic
thinking regarding recurrent natural disasters.

In February 2004, the

United Nations Develop-
ment Program (UNDP)
report Reducing Disaster
Risk: A Challenge for
Development described
the global trends in exposure, risk, and vulnera-
bility to natural disasters. From an international
development perspective, the report was signifi-
cant for two reasons. First, it featured a disaster
risk index (DRI), which measured and compared
physical exposure levels to four natural hazard
types,

vulnerability, and risk among some 200

countries. Second, the report identified signs of
vulnerability associated with development activi-
ties under way that could lead to higher disaster
risk.

The statistical approach of the report

allowed the UNDP to draw comparisons between
a particular country’s vulnerabilities and the
different natural hazards.

Natural Disaster Hotspots: A Global Risk

Analysis (World Bank 2005) identified countries
prone to experiencing a high frequency of
natural disasters accord-
ing to single or multiple
disaster variables. The

The exposure, risk, and
vulnerability of countries
to natural disasters are
known.

Disasters frequently recur
in the same countries.

ProVention study on which it was based
presented a set of data on the risks of mortality
and economic losses associated with six major
natural disaster types

and determined the

prevalence of natural disasters using a common
geospatial unit of reference in all countries. In
addition, the report ranked countries in terms
of highest risk potential in order to influence
risk mitigation investments and to better inform
the Bank on how to manage its future
emergency lending. The remainder of this
chapter incorporates the analysis detailed in the
Natural Disaster Hotspots report.

Both studies identify areas likely to be

affected by severe events and then determine
where disastrous impacts are likely to occur
because of the risks attached to the density of
human occupation. At some point in any analysis
of vulnerability, the event needs to be
uncoupled from human actions, at least until
risks are understood as distinct from being
inherent in the event itself. Such thinking should
be the foundation for any strategic approach to
disaster assistance.

Planning Implications for the World Bank

A significant number of the Bank’s disaster loans
and credits can be characterized as ad hoc
responses to what all involved parties consid-
ered unforeseeable acts of nature. Yet it is only
necessary to look at which countries have
borrowed the most for disasters in the past
(table 2.2) to know with considerable certainty
which ones will borrow the most in the future.

Most natural disasters are foreseeable to the

extent that it is possible to predict generally
where an event is likely to occur at some time in
the near future (but not precisely when or its
magnitude). It is also possible to know the
fragility of the built environment and the likeli-
hood that the siting of a given human settlement
will expose it to potentially destructive natural
events. Therefore, disasters should be anticipated
as more predicable events, with human and
financial risks calculated in advance, and Bank

policy and practices need
to provide a supportive
framework for such an
approach.

In terms of strategic thinking and policy

formulation, the Bank can go beyond acknowl-
edging the general existence of natural disasters
and identify with relative precision the
geographic “hotspots”—the countries most
vulnerable to natural disasters—anticipate the
foreseeable human and economic risks, and
then encourage borrowing targeted at reducing
risks, in line with these calculations, ahead of
the disaster event.

Based on the list of hotspot countries in

Natural Disaster Hotspots, 50 of the Bank’s
borrowers are at relatively high risk from two or
more hazards; 47 of these actually borrowed
during the period analyzed. Though these
countries received 46 percent of all Bank
lending projects, they accounted for 56 percent
of the natural disaster projects and 62 percent of
the reallocated project loans.

The countries that experience extreme

events with the greatest frequency, therefore,
also experience the most interruptions to non-
disaster lending, which can increase the impact
of disaster and impede their overall develop-
ment (box 3.1). Because many reallocations
occur in these countries, it also suggests that
neither the Bank nor its borrowers are planning
sufficiently for potential disasters in the places
they are most likely to occur. This is borne out
by analysis of disaster planning in the two main
strategy documents used by the Bank and its
borrowers: Country Assistance Strategies (CASs)
and Poverty Reduction Strategy Papers (PRSPs).

Disaster Planning in Poverty Reduction
Strategies

The PRSP is a product of borrowing country
governments that is intended to improve the
poverty impact of external partner lending and
the effectiveness of technical advice. Given the
effect major and recurrent disasters can have on
the life of the poor, disaster mitigation and
prevention might be expected to be featured in
these documents, especially in highly vulnerable
countries.

Instead, of the 59 PRSPs

prepared to date,

only 9 have incorporated aspects of hazard risk
management.

Of those 9, only 3—Honduras,

Nicaragua, and Vietnam—are highly vulnerable

2 4

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Reallocations are highly

concentrated in the most

vulnerable countries.

countries (see Appendix E, table E.2a). This
suggests that not only is hazard risk manage-
ment rarely addressed in PRSPs, but that the
PRSPs that do address it tend not to be in
countries with a relatively high economic risk
from multiple hazards.

Disaster Planning in Country Assistance
Strategies

Since the CAS is a planning document, evidence
that the country and Bank have given some
thought to disaster prevention might also be
expected in that strategy. Often, however,
natural disasters get no attention in the CASs. Of
current CASs for countries that have already
received Bank support for work related to
natural disasters, 44 percent did not mention
natural disasters (table 3.1). Even in the 40
countries that have had 4 or more disaster
projects, one-third of the CASs did not mention
disaster. And, for the subset of countries that

had an extensive history
with disasters (more
than 8), about a third did
not mention disasters at all. Moreover, CASs for
countries that are prone to repeated disasters of
the same type do not include those disasters in
their planning (box 3.2) (IEG 2005d).

All CAS documents contain a one-page matrix

titled “CAS Summary of Development Priorities”
that lists country and Bank priorities. Priorities are
rated as high, moderate, or low, and the main issues
are identified. The CASs for three countries—
Bangladesh, India, and Mozambique—flag natural
disasters in the Development Priorities Matrix. For
these three countries, both the Bank and the
country rated it as a high priority. In addition,
the Turkey CAS includes
natural disasters among its
major issues, though it
does not flag it separately
in the priorities matrix.

D I S A S T E R S A N D B A N K S T R AT E G Y

2 5

Two examples illustrate the enormous brake that disasters place

on economic development and how easily hard-won gains can

be lost.

The World Bank has been helping Honduras to build its high-

way system since 1955, when it was estimated that it had 2,500 kilo-

meters of roads. In the course of seven transportation projects

(totaling $120 million in commitments) between 1958 and the mid-

1990s, the Bank financed construction of 1,270 kilometers of high-

ways and feeder roads. By the time Hurricane Mitch hit in 1998, there

were approximately 10,000 kilometers of roads in the country. The

hurricane destroyed 6,000 kilometers of the better roads—almost

five times what the Bank had helped to build. In addition, more than

163 bridges were damaged or destroyed. Estimates of the damage

to roads from the hurricane were on the order of $454 million.

Disasters have also taken a development toll in Mozambique.

Bank lending financed the construction of 487 schools. Just the

most recent disaster, the floods of 2000, damaged or destroyed

about 500 primary schools as well as 7 secondary schools.

Box 3.1: Disaster Damage Undermines Development Progress

Few countries anticipate
disasters in their PRSPs.

Few CASs consider the
possibility of disruptions
from natural disasters.

Number of their

Number of disaster

Number of countries

CASs with no discussion

projects in a country

with this count

of disasters

Percent

More than 8 disaster projects

4 to 7 disaster projects

2 to 3 disaster projects

1 disaster project

Total

Source: IEG data.

Table 3.1: Many CASs That Should Discuss Natural Disasters Do Not

Bank support for

disaster work clearly
reflects the importance
of including this topic in
strategy documents. IEG
research also found that

including disaster in the CAS affected its inclusion
in projects prepared under the CAS. The average
number of Bank-supported projects for highly

vulnerable countries that
mention disaster in their
CAS is 7.9. In contrast,
the average number for
the highly vulnerable
countries that do not
mention disasters in their

CASs is only 2.4. Not taking disaster into account
in the CAS has an opportunity cost in that it may
lead to significant under-investment in risk
management and prevention.

Categorizing Borrowers According to Disaster
Risk—An Illustrative Approach

When formulating country lending programs,
the Bank needs to elevate the importance of
disasters, especially for highly vulnerable coun-
tries. To do this efficiently, borrowing countries
would have to be divided into categories accord-
ing to their disaster risk levels.

Using the list of hotspot countries in the

Natural Disaster Hotspots study as a starting
point, 35 countries have a high vulnerability,
because 50 percent or more of their GDP is
classed as being at risk from natural disasters in
the report. Fifteen countries have a medium
vulnerability to natural disasters because natural
disasters could place between 30 and 50 percent
of their GDP at risk. The remaining borrowing
countries have been classed as having a low
vulnerability level, because natural disasters are
a relatively small risk.

Based on a working hypothesis of a country’s

level of vulnerability—and subsequent events can
and will change understanding and improve the
accuracy of any categorization scheme—the Bank
needs to develop and adopt specific plans of
action. For example, countries of all vulnerability
levels would consider disaster risks systematically
along with their consideration of macroeconomic
and other threats in the risks identification
section of relevant Bank documents.

For countries with medium and high vulnera-

bility levels, both disaster-related and regular
lending for infrastructure, technical assistance,

2 6

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Of the 43 countries that have received Bank financing for flooding,

only 4 CASs mentioned either the development of an early warning sys-

tem or land use planning. Only three mentioned development of a dis-

aster-appropriate legal framework.

Of the 13 countries with earthquake projects, only 3 CASs men-

tioned seismic strengthening of critical facilities such as hospitals,

schools, and the like. In El Salvador, for example, seismic-resistant de-

sign and the reconstruction of 594 destroyed or severely damaged

schools was envisaged in the CAS.

Of the 27 countries affected by drought, 8 of the CASs mentioned

food security or removing drought-related impediments to growth in agri-

culture. Strengthening the safety net was discussed in seven. Targeting

of interventions for orphans and vulnerable persons was addressed in four

CASs, and one CAS mentioned creation of off-farm income opportunities.

Box 3.2: To What Extent Do CASs Develop

Disaster-Specific Strategies Appropriate to

Prevailing Hazards?

Source: IEG data.

Borrowers can and

should be classified by

their disaster

vulnerability.

Specific, differentiated

plans of action can be

developed within projects

for the more vulnerable

groups and regions.

Disaster-contingent line

Bank loans to incorporate

Bank documents

of credit in CAS

disaster-resilient designs

to consider

Vulnerability level

lending program

and/or environmental restoration

disaster risks

High (>50% of GDP)

Medium (30–50% of GDP)

Low (<30% of GDP)

Table 3.2: Natural Disaster Risk Can Be Mainstreamed in the Bank’s Lending

2 9

Relevance of
Bank Assistance

here does the World Bank fit in the scheme of things as borrowers
respond to natural disasters and prepare for recovery and recon-
struction? The external Web site states its role succinctly: “The

World Bank is the largest funder of disaster recovery and reconstruction in the
world.”

What this means for borrowers facing a severe
crisis is that if the Bank does not finance the
solutions to the bigger reconstruction
problems, they will often be addressed only
piecemeal, if at all. Not returning to full function-
ality has a multitude of small impacts, and it has
been argued that much of the longer-term GDP
cost of disaster is the result of incomplete
reconstruction (Linnerooth-Bayer and Mechler
2005).

When the Bank responds to a disaster, it often

uses several of its funding mechanisms. For
example, in Honduras, following Hurricane
Mitch, the Bank’s primary response was balance
of payment support through an Emergency
Recovery Credit. This was supplemented by
reallocations from seven existing projects and
redirection of a social fund to assist in the
recovery and reconstruction. A follow-on loan is
being implemented to reduce vulnerability
through strengthening municipal institutions
for disaster management. The Bank has
responded to other disasters—the Turkey
earthquake, flooding in Bangladesh—with a
similar mix of financing.

Flexibility and innovation are essential to

success with a natural disaster response. Among
the Bank-financed responses that have
demonstrated such flexibility and innovation are
the Maharashtra Earthquake Project (1997),
which attempted to address the needs of the very
poor and nomadic; North China Earthquake Re-
construction (1993), which integrated cultural
heritage in the response; Yemen Emergency
Flood Reconstruction (1989), which had particu-
larly effective donor coordination; drought
prevention in Niger (1988), which made highly
efficient use of limited resources; and the
Honduras Social Investment Fund (1999).

Defining Relevance of
Disaster Assistance

The relevance of the actions taken following a
natural disaster depends on the extent to
which they are timely, appropriate to country
needs, and reduce vulnerability. However, the
needs of the country and of those affected by
disaster change dramatically day by day and
month by month following a disaster. This
makes it difficult to remain relevant without

building a measure of flexibility into planned
actions.

In the dynamic circumstances following a

natural disaster it can be difficult to do activities
in order of priority. Several government officials
interviewed during field visits made the point
that readiness to implement too often
outweighed the priority of the activity. This
meant that some immediately relevant critical
activities that required a plan (such as rehabilita-
tion of public markets or housing for the poor)
were put off until one could be prepared, while
other activities not requiring a plan (such as
repaving urban streets) were started sooner
than necessary, diverting attention from the
more critical activities. With a longer-term view
of disaster risks, countries might recognize that
getting the priorities right would require a level
of advance planning by government ministries,
as well as through institutions specifically
focused on disaster risk management.

Over the past 20 years, Bank financing has

supported 60 distinct types of activities in

response to disasters,
exhibiting a high level of
innovation and flexibility.
Projects contained any-
where from one activity

type to 22, but generally projects have been
designed to provide solutions specific to the
unique situation presented by each event, taking

into account geographic, cultural, and social
factors, as well as hazard risks (see Appendix F,
figure F.1). Table 4.1, which lists the 10 most
frequent activities, shows that many projects
have pursued activities related to reduction of
vulnerability (such as those related to disaster
management and to public awareness).

Timeliness is also extremely important—

when an activity is done and how long it takes to
complete are at least as important as what activi-
ties are undertaken. In the case of an
earthquake, food is needed immediately post-
disaster and not later. If it arrives later, it can
actually detract from the sales of food harvested
(and/or salvaged) in the interim (Jackson 1982).
As droughts become increasingly serious, there
is a time when the poor consume seed stocks
and sell off domestic animals. Considering this,
interventions that lead to speedy recovery need
to begin before the next year’s crops are
consumed and draught animals are gone.

The same principle applies to Bank activities.

For example, if balance of payment finance is to
do any good, it must be disbursed quickly
enough in the year following the event to cover
the cost of imports needed for reconstruction,
and not those unrelated to reconstruction.
India, Madagascar, Mozambique, and Turkey
used balance of payment lending for petroleum
imports. In the following year, however,
petroleum imports declined sharply, which

suggests that the balance of payment loan may
have led to an over-purchase of those products.

The activity most frequently pursued—

provision of supplies and equipment—may be
of dubious relevance. Specialized equipment

purchased through a loan needs to be used
effectively and efficiently. Accomplishing this
may require additional investment by the
borrower (see box 4.1)—investment that may
not be forthcoming or may lose out to other
development priorities once the most pressing
needs have been met.

For emergency shelter to do any good at all, it

should be erected only after victims have been
assisted as far as possible to provide themselves
with shelter close to where their homes used to
stand (or, as often happens, have permanently
resettled elsewhere). Of course, relevance also

3 0

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

The Bank has responded

flexibly with a wide
variety of activities.

Activity

Number of projects

Provision of supplies and equipment

Rehabilitation of road infrastructure

Rehabilitation of flood control infrastructure

Studies and research on disaster management

Institutional development for disaster management

Fire prevention activities

Early warning and public awareness campaigns

(Re)Development of education facilities

Planning for disaster management

Rehabilitation of irrigation and drainage

Source: IEG data.

Table 4.1: The 10 Most Frequently Pursued

Activities

requires that they only be constructed in areas
that experience severe winters, and early
enough to be useful—otherwise it represents a
waste of scarce resources.

The importance of timeliness is reflected in

project performance. Of completed projects for
which the time between approval and effective-
ness was below the median (half the portfolio),
86 percent had satisfactory outcome ratings. For
those above the median only 67 percent were
rated satisfactory. Therefore, for projects experi-
encing difficulty in meeting effectiveness
requirements, the performance was lower
compared with those able to make a speedy first
disbursement.

The funding mechanism used and the

approach chosen depend on whether the intent
is to finance an immediate response to urgent
need, a medium-term response to assist
recovery and reconstruction, or a long-term
response for reconstruction and mitigation. As
will be seen, however, this is not always the case.
There is a general pattern, but there are
exceptions. Each circumstance is worth consid-
ering separately.

Delivering Quick Support for the
Immediate Response

Countries affected by natural disasters often
request quick assistance to replace lost capital
and to prevent cascading negative economic
effects from growing and multiplying. To help
with this, the Inter-American Development Bank
has established an Emergency Reconstruction

Facility that permits the
commitment of up to
$20 million “in the first
hours after the disaster
takes place.”

The World Bank has nothing similar. Instead,

it has relied on reallocations to fill this expressed
need. Unfortunately, the documentation for
reallocations is so incomplete that it has not
been possible to determine the activities and
uses of reallocated
funds. Based on experi-
ence, however, the Bank
has sometimes been
pressed into ill-consid-
ered responses during
the early part of a natural
disaster response (box 4.2).

Reallocations are highly relevant where the

relevance of the original project is reduced by
disaster. Furthermore, those funds often keep their
broad sector dedication. For example, funds
originally intended for school improvement have
been reallocated to school reconstruction after a
hurricane or volcanic eruption had destroyed
schools (the 1995 Honduras Basic Education Project,
the 1995 Nicaragua Basic Education Project, and the
1993 Papua New Guinea Education Development
Project). In other cases,
projects have had slow-
disbursing components
that, in an emergency, can
be formally reallocated to
reconstruction purposes

R E L E V A N C E O F B A N K A S S I S T A N C E

3 1

Timeliness matters as
much or more than the
activities undertaken.

The quickest Bank
response is reallocation,
but the use of reallocated
funds is not transparent.

Activities requiring
supplemental effort by the
country may remain at
risk.

Equipment acquisition requires careful planning to ensure its pro-

ductive use. Technical assistance provided under emergency proj-

ects may persuade public officials to invest in modern equipment for

early warning, improved communication, and emergency supplies

and equipment for police, firefighters, and other first responders.

However, project experience has demonstrated a strong ten-

dency for the devices to be bought but not installed. In other cases,

equipment was installed or supplies warehoused, but no budget was

made available for the people who were supposed to use it.

In some cases, large databases of geographic information

have been established to provide information on different types of

hazards, areas especially vulnerable to disasters, and available re-

lief material, but they have not been regularly updated. In the Ma-

harashtra Emergency Earthquake Project, data on disaster

vulnerability was collected, but the government then classified it

top secret, and it was not shared with the officials in charge of dis-

aster management.

Box 4.1: Equipment Acquisition Can Be Particularly Problematic

without regard to sector
(such as Bolivia 1998,
Vietnam 1995, and Mexico
1993).

For some countries,

however, reallocation is
not an option. Small

island states and small countries with fragile
economic and political systems

often have few

ongoing loans to reallocate, and even if they do,
they are small relative to the assistance required.
These borrowers are limited to requesting new
lending from the Bank, which takes time to
process and adds to their overall debt.

An alternative to reallocation in some cases

has been the redirection of a social fund, which
by its nature is multisectoral and intended to
respond to community needs. In Honduras and
Nicaragua, following Hurricane Mitch, social
funds were an important part of the Bank’s early
response, transforming themselves overnight
from centralized social investment funds into
nimble rehabilitation and reconstruction
agencies. Such redirections can provide critical
support to the poor in affected communities,
but they ultimately do little to address the
problems faced by the government. More recent

alternatives to realloca-
tion show some promise
for meeting immediate
needs, but their value
remains to be seen.

Delivering New Projects for the Recovery

In the short to medium term, countries typically
request assistance such as budget support,
technical assistance for reconstruction planning,
rapid reconstruction of transport infrastructure
essential to international trade, and the provision
of equipment and supplies. The Bank now
customarily provides such support through ERLs.

Balance of Payment Support

Balance of payment support is intended to be a
quick-disbursing activity that meets the most
pressing financial needs of affected countries.
Designed to provide quick inputs to stabilize
macroeconomic conditions and facilitate
recovery following a calamity, this kind of
support is not very common; only 15 loans have
been made for balance of payment support
following natural disasters.

Despite its emphasis on the rapid disburse-

ment of funds, balance of payment support took
an average of about 7 months (214 days) to reach
effectiveness and 2.4 years (860 days) to reach
closing. It thus did not meet institutional
intentions that it be an effective means of provid-
ing quick transfer of resources to affected
countries.

As one task manager who imple-

mented this type of support noted, “I am very
skeptical about disaster relief/emergency projects
going through the budget in most countries. In
[reference to country deleted] we were supposed
to finance reconstruction through the budget,

3 2

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Zimbabwe experienced its worst drought of the century during

the 1991/92 growing season. The Bank responded with an ERL

supplemented by reallocations from three ongoing loans—a

total of $37 million.

The project was to be implemented by established units in

several ministries. However, their capacity was insufficient to

handle the extra volume of work. Procurement planning, for ex-

ample, should have started at the earliest opportunity, but only

began after credit effectiveness was declared.

The results were mixed. Relief goods such as foodstuffs were

quickly imported, but the recovery was marred by delays in the

emergency water program and underachievement of targets for

agricultural recovery. Only limited institutional capacity was cre-

ated to assist with future drought management, and proposed

policy and organizational improvements were not carried out. In

the end, more foreign exchange was made available than could

be readily absorbed—$23.5 million of the IDA credit was can-

celed and hardly any reallocated funds were used.

Box 4.2: A Sense of Urgency Can Lead to a Wasteful Response

Source: IEG project database.

Reallocations can be

highly relevant, but

reallocation is not

always an option.

The ERL is typically used

for short- to medium-term

assistance.

which never happened and, in the end, the
emergency project had neither a positive balance
of payment nor reconstruction impact.”

Support for Social and Economic Recovery

Ideally, the Bank tries to separate activities that
benefit from a nearly immediate start-up and
those where there is a need to plan more slowly
for medium-term recovery and mitigation. OP
8.50 and Bank Procedure (BP) 8.50 call for
consolidating project preparation activities,
expediting processing procedures, and disburs-
ing resources rapidly. Disaster projects are
allowed accelerated processing and are expected
to have a short implementation period. OP 8.50
and BP 8.50 establish a timing framework for
emergency recovery lending that limits the
elapsed completion time to three years.

Although Bank policy for emergency lending

emphasizes a quick response and compressed
implementation time, experience with the
earthquake, drought, and tropical storm
projects suggests that a quick response does not
always lead to the most relevant response. The
completion time of disaster projects appears to
be determined by the time needed to complete
each activity, an interval that is predictable.
Responding effectively to a specific disaster
requires picking and choosing from the full
menu of 60 activities, and not concentrating on
those that can be quickly accomplished.
Therefore, a quick response is necessary and
possible only in cases where the required activi-
ties can be carried out in a short time.

Having a disaster policy that emphasizes time

limits for the intermediate term may lead Bank staff
to rush certain activities unnecessarily. For example,
in the St. Lucia Watershed and Environment
Management Project, pressure to start reconstruc-
tion led to inadequately analyzed designs and
implementation that did not reduce vulnerability to
the next storm. Specifically, silt was removed from
waterways, but the denuded hillsides in the
surrounding watershed were not stabilized, and the
waterways quickly refilled with silt.

Another risk inherent in a policy with time

limits is that projects may exclude some activi-
ties that take longer but achieve long-term
development goals. The analysis of activity

frequency specifically for
ERLs shows that of the
most commonly imple-
mented activities, the
top 10 are dominated by
rehabilitation (of roads,
schools, water systems,
irrigation systems, health facilities, flood control
infrastructure) and the purchase of equipment.

Among the 10 least frequently pursued ERL

activities are land and water resource manage-
ment, community-based disaster prevention,
fire prevention, and forest management. The
degree to which rehabilitation has been allowed
to overshadow prevention, the low relevance of
some activities, and the absence of follow-on
projects indicate that greater attention to
country needs and flexibility with regard to
implementation time, coupled with a more
varied assortment of
lending instruments,
may be called for.

The disaster activities

carried out during re-
covery and reconstruc-
tion exhibited significant
variance in implementation time, ranging from
two to seven years. Figure 4.1 shows how long
projects containing each disaster activity took,
on average. Only one of the activities was carried
out in projects completed within three years, as
stipulated for ERL projects by OP 8.50.

Almost all of the most crucial disaster response

activities required more than three years, in part
because they were conducted without pre-
disaster planning. For instance, projects involving
land acquisition activities took an average of 7.5
years, and for those with infrastructure activities,
6.5–7 years was the average, depending on the
type of infrastructure.

In planning for projects related to natural

disaster recovery, too little time was generally
budgeted for implementation. This is high-
lighted by the revised implementation time for
most disaster activities
(shown in figure 4.1). For
example, on average, the
completion of projects
with irrigation and drain-

R E L E V A N C E O F B A N K A S S I S T A N C E

3 3

Balance of payment
support has not been an
effective means of
providing quick transfer
of resources.

The time required to
complete particular types
of activities affects
project completion time.

Only one activity type has
been completed in less
than three years.

3 4

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

Figure 4.1: Average Implementation and Extension Times for Projects Containing

Disaster Activities: All Disaster Projects

2.0

3.3

4.4

4.0

3.7

3.5

3.4

4.0

4.4

4.0

3.1

3.7

3.9

4.1

4.3

4.6

4.8

4.4

3.7

4.0

3.9

4.3

4.1

4.5

4.6

3.2

4.4

4.7

4.3

5.0

4.5

4.0

4.2

5.0

4.6

4.7

4.6

4.8

5.0

4.7

4.5

5.3

4.9

5.0

5.6

5.5

4.9

5.6

5.8

5.5

5.8

5.9

6.1

6.0

6.1

0.4

0.5

0.4

0.7

1.1

1.4

0.8

1.0

0.6

1.0

2.0

1.4

1.2

1.1

1.0

0.6

0.5

0.9

1.7

1.4

1.5

1.2

1.3

1.5

1.1

2.5

1.3

1.0

1.5

0.9

1.4

1.9

1.8

1.0

1.4

1.5

1.4

1.2

1.5

1.9

1.0

1.5

1.1

1.2

1.8

1.3

1.1

1.5

1.2

1.1

1.3

1.4

1.1

0.0

BoP/imports (FIN)

PIU support (PJT MGT)

Contingency financing (FIN)

Assistance subsidies (FIN)

Rehab of water/sanitation (INFR)

Search/rescue & medicine (DIS MGT)

Consulting (PJT MGT)

Community-driven prevention (PREV)

Policy reform (DIS MGT)

Disaster insurance (FIN)

Rehab of urban water (INFR)

Leveraging funding (DONOR)

Interfacing with govt/donor (DONOR)

Rehab of electr/ener/comm (INFR)

Health facilities (ECO STRGTH)

Institutional development (DIS MGT)

Design/supervision (PJT MGT)

New constr of shelters (INFR)

Pest control activities (AGR)

Rehab of irrig/drainage (AGR)

New constr of elec/ener/comm (INFR)

Retrofitting/strengthening (PREV)

Transport facilities (ECO STRGTH)

Early warning (DIS MGT)

Rehab of housing (HSG)

Education facilities (ECO STRGTH)

Rehab of rural water (INFR)

Loans to beneficiaries (FIN)

Procurement (PJT MGT)

New constr of housing (HSG)

Project training (PJT MGT)

Public facilities (ECO STRGTH)

Water resource management (AGR)

Supplies/equipment (AGR)

New constr of water/sanitation (INFR)

Commercial facilities (ECO STRGTH)

Research (DIS MGT)

Tech assist: Engineering (DIS MGT)

Tech assist: Non-eng (PJT MGT)

Planning (DIS MGT)

Studies (PJT MGT)

Rehabilitation of roads (INFR)

Rehab of flood control (INFR)

Cash transfer (FIN)

Resettlement (RESTTLMT)

New constr of urban water (INFR)

New constr of roads (INFR)

New constr of irrig/drainage (AGR)

Training (DIS MGT)

Flood control activities (PREV)

Demolition/rubble collection (INFR)

New constr of rural water (INFR)

Resp to slow-onset event (PREV)

Forest mgmt (AGR)

Operations/maintenance (O&M)

New constr of flood control (INFR)

Livestock/fisheries mgmt (AGR)

Fire prevention activities (PREV)

Land acquisition (AGR)

Land management (AGR)

Disaster activities

Number of years

Average project time

Average extension time

Source: IEG data.

Note: Each bar represents the average actual implementation time of all projects with a given activity.

age infrastructure took 1.5 years longer than
planned, and demolition took 1.8 years longer.
Overall, of 303 completed projects, 28 percent
were completed when they were expected. The
rest ranged between 18 days and five-and-a-half
years late. Of ERLs, 42 percent were completed
in the timeframe originally anticipated, though
estimated times often exceeded the policy
requirement.

Reducing Vulnerability over the
Longer Term

The Bank’s standard lending instruments gen-
erally are well suited to efforts that reduce vul-
nerability through new or rehabilitated
infrastructure (shelters, early warning systems,
flood control works) or through developing the
capacity to manage disaster risks. As figure 4.1
shows, such activities take longer than three years
on average, and the time needed to implement
them usually is underestimated by a year or more.
Some ERLs have overreached by trying to cover
such activities. Yet if such activities are left out of
the ERL with the expectation that additional
borrowing will follow, those activities may never
be undertaken. ERLs have been followed up with
normal disaster investment projects only 17
percent of the time. Of 59 completed ERL
projects, only 10 have had follow-on projects
that took place within three years after they
closed (with activities related to either
reconstruction or prevention).

For most low-income

countries that have not
faced a disaster recently,
reducing vulnerability to
disasters is just another of
their many development priorities, and for those
that have experienced disasters recently, vulnera-
bility reduction quickly falls off the development
agenda as governments turn their attention
elsewhere. As seen in Chapter 3, disaster risks do
not make it into the CAS or PRSP as often as
country exposure to such risks would seem to
warrant. When a CAS does discuss natural
disasters, it is likely to discuss activities related to
vulnerability reduction (such as strengthening
disaster management, long-term planning, early
warning systems; see Appendix E, table E.1).

When vulnerability is addressed, it can take a

long time, but it can have a lasting impact on
poverty (box 4.3). It can also represent signifi-
cant savings.

Insurance that covers

disaster damage is an
important part of vulner-
ability reduction in most
developed countries,
but it is rare in develop-
ing countries. While
about half of the costs of
natural disasters are covered by insurance in the
United States, less than 2 percent of the costs
are covered in the developing world. The study

R E L E V A N C E O F B A N K A S S I S T A N C E

3 5

Too little time is generally
budgeted for
implementation.

Investment lending
instruments can fill
disaster needs, but ERLs
are rarely followed up
with such lending.

The Small Rural Operations (SROs) Project in Niger adopted a

development approach to reducing drought vulnerability based

on transfer of responsibility to beneficiaries. By intensifying off-

season crop production through widespread use of existing

simple, low-cost technologies, the project generated rural income

and helped mitigate food shortages.

The project took 11 years to implement (1988–98). It contributed

to the financing of 88 SROs in off-season production of horticul-

tural products and fruit trees, soil conservation, smallholder fish

production, stock-raising, and animal husbandry. About 35,000

rural farmers benefited directly from these income-generating

activities. A strong impact on the incomes of the farm families arose

from higher cropping intensities, cultivation of higher-value crops,

and diversification toward non-crop activities.

The project’s impact on food security and poverty reduction has

been positive in much of the project area, which has a chronic food

deficit and where other donors are noticeably absent. The impact

on the welfare of women beneficiaries, who dominate the project’s

horticultural marketing and food processing activities, has been

significant.

Box 4.3: Reducing Vulnerability May Also Reduce Poverty

Source: IEG project database.

database shows that the
Bank has supported
activities focused on
laying off risk in 10
projects (see Appendix
F, table F.3). These have
included catastrophe

bonds, index-based insurance, catastrophe
models to facilitate coverage, work through
microfinance institutions, contingency lines of
credit, and the creation of a catastrophe
insurance pool.

The limited experience has seen some

success and encountered a number of difficul-
ties. Perhaps the most successful of these experi-
ments was the catastrophe insurance pool
developed for Turkey following the Marmara
earthquake, but the long-term success of that
experiment is not yet assured (box 4.4). Among

the difficulties that need to be
faced is getting people to under-
stand how insurance works and
the benefits of paying into it.
There also are good reasons that

insurance coverage may not be available. In many
developing countries, for example, the cost of
hedging against risks exceeds the cost of simply
paying for damages when they arise. This is an
issue the Bank will need to address in future
attempts to put insurance schemes in place.

Does the Bank Respond Quickly?

Disaster projects are prepared and implemented
under difficult working conditions. They may have
to overcome shortages of critical materials,
severed transportation links, and weak capacity or
capacity loss due to disaster damage. Under such
circumstances a quick response is difficult, but
essential, to achieve. Based on analysis of the 528
projects examined, the Bank does not appear to
have increased its response time by using ERLs.

The average amount of time that elapsed

between an emergency event and the presenta-
tion of a loan or credit to the Bank’s Board of
Executive Directors (figure 4.2) was 6.7 months
(201 days). When disaggregated, the difference
between ERLs and disaster projects that use
other instruments was small. The non-ERL

3 6

H A Z A R D S O F N AT U R E , R I S K S T O D E V E L O P M E N T

In order to extend liquidity to homeowners, reduce government

liability, and lessen dependence on foreign donors in the event

of future disasters, Turkey launched a disaster insurance scheme

in September 2000. The scheme, which will cover $1 billion in

damages in the event of a disaster, was launched in a timely fash-

ion, and soon had more than tripled the level of insurance pen-

etration for earthquake coverage compared with that previously

achieved by the private insurance market.

The scheme also demonstrated its ability to pay claims quickly in

a number of small- and medium-size earthquakes that have occurred

since its initiation. A successful public awareness campaign, com-

bined with an affordable average annual premium of $20, brought the

scheme significant initial penetration (the average is currently $46),

and the level has held steady at around 2 million policies over the past

three-and-a-half years. Increased penetration is awaiting passage of

the draft Earthquake Insurance Law. Meanwhile, the Bank is inves-

tigating several measures to increase the number of policyholders.

The scheme faces some difficulties in not only sustaining but

increasing the numbers of insured in the country, however. Two

newly enacted laws pertaining to state-owned enterprises, the Pub-

lic Financing and Supervision Law and the Procurement Law for

State-Owned Enterprises, may terminate the insurance program’s

current exemption from all state regulations applying to govern-

ment-owned enterprises. (It was created as a special, non-

government entity, as it does not have government employees, and

the government only intervenes in the event of a disaster calling

for over $1 billion in coverage.)

Whether these laws will apply to the catastrophe insurance pool

is currently being clarified. The current Disaster Law #7269 (which

provides easy compensation to earthquake victims), along with the

lack of a strong insurance culture in Turkey, adds to the chal-

lenges faced by the scheme.

Box 4.4: Bank Helps Finance Insurance Scheme in Turkey

Source: Project Performance Assessment Report: Turkey (IEG 2005b).

The Bank averages about

seven months to get a

disaster project to the

Board.

Limited experience with

insurance shows

difficulties in

implementation and

sustainability.

disaster projects were found to have an average
of 7 months (209 days) pass before they received
Board approval. For ERLs, the preparation time
averaged 6.6 months (198 days)—just 11 days
less than projects that used other instruments.

The projects in the Bank’s disaster portfolio

required an average of seven months (208 days) to
reach the effectiveness date, the first day on which
a borrower is eligible to withdraw against an
approved loan. Of these projects, however, only 27
percent

were able to reach effectiveness within

four months of Board approval. If the time from
event to Board approval is added to the time from
approval to effectiveness, it is clear that having the
current emergency lending mechanisms already in
place does not significantly accelerate project
processing and it does little to expedite the
disbursement of disaster funds.

The ongoing projects, on average, have also

taken longer to prepare and to reach effective-
ness than those that have closed, though in
some of the largest responses, such as those for
the Indian Ocean tsunami, the Marmara and
Hebei earthquakes, and Hurricane Mitch, have
been very quick. As the definition of emergen-

cies has broadened and
more projects have at-
tained emergency lend-
ing status, project
preparation and effec-
tiveness times have
grown increasingly similar for ERLs and non-
ERLs (figure 4.2). In other words, the more
recent projects are, on average, slower than
those of a few years ago. Considering that even
emergency projects were prepared and reached
effectiveness roughly one year after the disaster
event, which was similar to non-ERL disaster
projects, it appears that the Bank has usually
reacted slowly to emergencies, and this trend is
becoming more pronounced.

Disaster projects as a whole took longer to

close than regular investment lending. From
Board approval to the final project closing date,
the average implementation time of disaster
projects

was 6.6 years (2,396 days), with a

median of 7 years (2,541 days). In contrast, the
overall Bank portfolio took less time, with all
projects

having an average implementation

time of 6.1 years (2,214 days) and a median of

R E L E V A N C E O F B A N K A S S I S T A N C E

3 7

ERLs have not moved
significantly faster than
other lending
instruments.

4.5

6.6

7.0

2.6

3.9

7.6

81.0 (6.6 years)

46.8 (3.9 years)

28.8 (2.4 years)