The State of Environmental Migration 2013 a REVIEW of 2012

Home , Cyclone Val, Sissili Province

STUDY DECEMBER 2013 | CLIMATE The State of International Organization for Migration (IOM) Environmental Migration 2013 A REVIEW OF 2012

Edited by François Gemenne Pauline Brücker Dina Ionesco

Italy New York The New Towns of L’Aquila Hurricane Sandy Japan Louisiana Fukushima evacuees’ return Hurricane Issac India Bangladesh Haiti Recurrent Flooding Floods and Food Insurity Hurricane Sandy Burkina-Faso/Mali Desertification and drought Laos Dam-Induced Migration Nigeria Philippines Managing the displacement of 2012 Kenya Typhoon Bopha Drought and internal displacements Mozambique Samoa Mozambican floods Madagascar Cyclone Evan Cyclones Giovanna and Irina Australia Australia and its swift disaster

With the support of

Institut du développement durable et des relations internationales WITH SUPPORT OF 27, rue Saint-Guillaume www.iddri.org 75337 Paris cedex 07 France ©  Institute for Sustainable Development and International Relations (IDDRI) / International Organization for Migration (IOM)

D – IDDRI As a foundation of public utility, IDDRI encour- ages reproduction and communication of its copyrighted materials to the public, with proper credit (bibliographical reference and/or corresponding URL), for personal, corporate or public policy research, or educational purposes. However, IDDRI’s copyrighted materials are not for commercial use or dissemination (print or electronic). Unless expressly stated otherwise, the findings, interpretations, and conclusions expressed in the materials are those of the various authors and are not necessarily those of IDDRI’s board. www.iddri.org

D – IOM The opinions expressed in the report are those of the authors and do not necessarily reflect the views of the International Organization for Migration (IOM). The designations employed and the presentation of material throughout the report do not imply the expression of any opinion whatsoever on the part of IOM concerning the legal status of any country, territory, city or area, or of its authorities, or concerning its frontiers or boundaries. IOM is committed to the principle that humane and orderly migration benefits migrants and society. As an intergovernmental organization, IOM acts with its partners in the international community to: assist in meeting the operational challenges of migration; advance understanding of migration issues; encourage social and economic development through migration; and uphold the human dignity and well-being of migrants. www.iom.int

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise without the prior written permission of the publisher. The State of Environmental Migration 2013 A REVIEW OF 2012

Edited by François Gemenne Pauline Brücker Dina Ionesco

INTRODUCTION 5 In Focus METHODOLOGICAL CONSIDERATIONS 9 MIGRATION AND THE ENVIRONMENT: PART 1 THE CASE OF HURRICANE SANDY IN HAITI 32 FLIGHTS AND EVACUATIONS CYCLONES GIOVANNA AND IRINA HURRICANE SANDY IN NEW YORK MAKE THOUSANDS OF PEOPLE HOMELESS 47 AND NEW JERSEY: EVACUATION, DISPLACEMENT RESPONDING TO HURRICANE ISAAC: AND ADAPTATION 15 ASSESSING EVACUATIONS AND FEDERAL LEVEE TYPHOON BOPHA AND PEOPLE SEVEN YEARS AFTER KATRINA 61 DISPLACEMENTS IN THE PHILIPPINES 33 AUSTRALIA AND ITS SWIFT DISASTER CYCLONE EVAN IN SAMOA 49 RESPONSE MANAGEMENT 74 ASSAM AND THE BRAHMAPUTRA: RECURRENT MOZAMBICAN FLOODS AND RESETTLEMENT FLOODING AND INTERNAL DISPLACEMENT 63 PROCESSES 75 DISASTER RESPONSE IN NIGERIA: MANAGING PART 2 THE LARGEST DISPLACEMENT OF 2012 76 MOBILITY, RESETTLEMENT 2012 FLOODS IN BANGLADESH AND FOOD DESERTIFICATION AND DROUGHT RELATED INSECURITY 99 MIGRATIONS IN THE SAHEL – THE CASES OF MALI AND BURKINA FASO 79 THE NEW TOWNS OF L’AQUILA: A SUCCESSFUL POST-QUAKE REHOUSING DROUGHT AND INTERNAL DISPLACEMENTS PROJECT OR A WRONG LONG-TERM OF PASTORALISTS IN NORTHERN KENYA RECONSTRUCTION POLICY? 126 IN 2012: AN ASSESSMENT 101 THE FUKUSHIMA EVACUEES’ RETURN 117 DAM-INDUCED MIGRATION IN THE MEKONG REGION 127

3 with support of COST Action COST Action IS1101 on Climate change and Migration

COST IS1101 Climate change and migration: knowledge, law and policy and theory is a pan-European network of social scientists that aims to build a broad body of social science research on all aspects of climate change and migration. Our objectives are: sto enhance and improve understanding of climate change and migration; sto furnish state and non-state actors with state-of-the-art empirical, theoretical, legal and policy research on climate change and migration; sto inform national and international policy dialogue, such as the IPCC and other policy initiatives; sto expand research capacity in the area of climate change and migration; and sto establish a network of Europe-based social science researchers working on climate change and migration. Our work is subdivided into three working groups: Working Group I (WGI) - Knowl- edge; Working Group II (WGII) - Law and Policy; and Working Group III (WGIII) - Theory. The focus of WGI is empirical research, including method and data collection, quantification, remote sensing and geographical information systems, and modelling. The focus of WGII is normative research, including research leading to the development of policy and law. And the focus of WGIII is theoretical research, including political, cultural and social theory, postcolonial theory and critical race theory. In order to achieve our objectives, we fund a range of activities including workshops, knowledge exchange, and research dissemination, such as the annual publication of the State of Environmental Migration report. Our Action began in October 2011 and will run for 4 years, winding down in Octo- ber 2015. Dr Andrew Baldwin, Institute for Hazard, Risk, and Resilience, Durham University Chair, COST Action IS1101. The State of Environmental Migration 2011

INTRODUCTION FRANÇOIS GEMENNE, PAULINE BRÜCKER AND DINA IONESCO

For the third consecutive year, the State of Environ- climate change takes many forms, hence, requires mental Migration is geared at bringing visibility to wide-ranging partnerships as it is as the cross and new evidence on the migration-environment roads of migration, development, urban and rural nexus. This third edition follows the same rationale planning, disaster risk reduction, humanitarian, as the two previous ones, giving space to young environment and adaptation policies. researchers to present a variety of case-studies on migration, environment and climate change. The State of Environmental Migration is an opportunity THE YEAR 2012: MIGRATION for young researchers to explore research on the AND DISPLACEMENT DUE matter and to propose policy recommendations. As in the previous year, the publication has TO ENVIRONMENTAL AND benefitted from the support of the COST Action CLIMATIC FACTORS IS on Climate Change and Migration, a Eu- ropean-wide network of researchers working on The year  was again marred by a series of the themes explored in the different chapters that natural disasters, affecting both developing and follow. The Action puts a particular emphasis on developed countries. Though hurricane Sandy the support to young researchers, and we are very was clearly the disaster that attracted the most grateful for the support offered this year again. attention, this year’s edition of the State of Environ- The publication also owes a lot to Mark Koski, a mental Migration sheds light on a number of envi- graduate student from the University of Michigan. ronmental changes, both sudden and slow-onset, Mark did a fantastic editorial job on the different which were often little covered by the media. chapters, and we would like to use this opportu- Those include of course cyclones Bopha and nity to thank him for the energy and attention he Evan, which affected the Philippines and Samoa put into this report. respectively, but also recurrent floods in Nigeria The publication gathers a variety of cases stud- and India, for example. In , according to the ies, from drought in some of the poorest areas of Internal Displacement Monitoring Center (IDMC), Sub-Saharan Africa, to Hurricane Sandy devas- . million people were displaced because of tating the North-East United States. It documents natural disasters, a number that has doubled from sudden crises such as Typhoon Bopha, planned . The two largest disasters in  are both due development-induced migration from hydroelec- to flooding, leading to the massive displacement tric dams in China and the Lower Mekong Delta of . million people in Nigeria and . million in and slow-onset droughts displacing pastoralists in India. Both disasters are presented in this edition. Kenya. The  edition also analysis short-term As important as it may seem, this number does evacuations and people quickly rebuilding from not account for those who were displaced as a re-  floods in Australia, while also presenting the sult of slower, gradual environmental changes, or situation of people still displaced in Japan, Italy as a result of resettlement processes. And indeed, and Haiti. The structure of the volume that pres- behind the generic term of ‘environmental migra- ents first emerging catastrophes and second ongo- tion’, this report highlights a wide range of mobil- ing crises gives us an opportunity to examine cases ity patterns. Included are situations of flights and that made headlines in , and revisit cases that evacuations, but also situations of mobility as a have long been out of the news, but have remained risk-reduction strategy, as a way to adapt to envi- problematic for countries and displaced people. ronmental changes, and also situations of resettle- The variety of case studies proposed shows not ment, where the relocation is imposed by public only the many facets of environmental migration, authorities. but also the complexity of policy responses. Well What have in common those evacuated in an- planned migration can reduce vulnerabilities to ticipation of hurricane Sandy and those resettled the effects of environmental change and migration because of the building of dams in SouthEast Asia? is the part and parcel of adaptation, disaster risk reduction and development policies. Migration in the context of environmental degradation and . IDMC, 

IDDRI STUDY 06/2012 5 The State of Environmental Migration 2013

Not much, if not the fact that their displacement change negotiations while natural and social sci- was driven, in full or in part, by environmental ences seemed to find more common ground for changes that made their home no longer inhabit- discussion, within the side events that took place able. Is it a sufficient reason to label them as ‘envi- at the negotiations. However, migration remained ronmental migration’, as we did in this report? We marginal on the negotiation table in Doha. Even believe so. more disappointing, no further progress has been Environmental migration, as a concept, encom- made on the implementation of the paragraph passes a wide array of different types of mobility, (f) of the  Cancun agreement that anchors as we have tried to show again in this report. Yet migration in the climate change text. Yet one of environmental factors remain often neglected in the main outcomes of the COP is the inclusion migration studies, or are confined in the media of a decision on loss and damage which opens a to highly visible displacements related to climate pathway for the most vulnerable populations to change, and to sea-level rise in particular. Thus be provided with better protection and compen- what this report seeks to do is to highlight the sation. This loss and damage decision mentions diversity of mobility patterns related to different migration, displacement and human mobility in kinds of environmental changes, and their em- the Paragraph  (a)(vi) and invites voluntary work bedment in varied social and economic contexts. for increased understanding on “How impacts of Many have criticized the concept of environmen- climate change are affecting patterns of migration, tal migration for obliterating the responsibility displacement and human mobility”. The Group of of political and economic actors in migration and Least Developed Countries (LDCs) and the island displacement. This report adopts a different ap- states played a strong role in the decision and they proach, and aims to stress the increasingly impor- are also the states where the impacts of climate tant role of environmental changes in migration change strongly affect the mobility of population. and displacement across the world, through di- The loss and damage advances represent a good verse mobility patterns that would otherwise have entry door for states to consider human mobility very little in common. Doing so, it also stresses in the context of climate change. the importance of the role of the state in prepar- The year  was of importance to IOM in terms ing and managing these population movements, of migration crisis policy framework, a policy con- which sometimes goes as far as provoking and or- cept directly relevant to migration, environment ganizing these movements directly, as in the cases and climate change issues. According to IOM, of resettlement. In addition to the chapters that migration crisis “describes the complex and often document the population movements themselves, large-scale migration flows and mobility patterns this report also includes some boxes that highlight caused by a crisis which typically involve significant the role of disaster risk reduction policies in pre- vulnerabilities for individuals and affected commu- venting displacement. nities and generate acute and longer-term migration management challenges. A migration crisis may be sudden or slow in onset, can have natural or man- NEW POLICY DEVELOPMENTS made causes, and can take place internally or across AND EVIDENCE IN 2012 borders”(MC/). IOM’s Migration Crisis Oper- ational Framework (MCOF) was developed at the Three main policy areas offered increased space request of IOM Member States, pursuant to their for environmental migration in : the climate growing interest in the migration consequences domain, the migration domain, and the human of crisis situations. The frameworks considers two rights protection domain. scenarios of relevance to environmental migra- As far as the climate change negotiation agenda tion: on the one hand internal and cross-border is concerned, Parties met in  at the Confer- movements due to low-onset natural disaster; and ence of the Parties (COP)  of the United Nations on the other hand, internal and cross-border move- Framework Convention on Climate Change (UN- ments due to sudden-onset natural disaster. One FCCC) in Doha, Qatar. The COP ended with the noteworthy policy development to be mentioned adoption of the “Doha Climate Gateway” which is is on the human rights and climate policy strand. a package of agreements including a new commit- The eighth Session of the Human Rights Council ment period under the Kyoto Protocol. Additional- Advisory Committee in Geneva, included the or- ly Governments agreed on a timetable to negotiate ganization of a Seminar to “Address the Adverse a new legally binding international climate agree- Impacts of Climate Change on the Full Enjoyment ment by , which will cover all countries from . Social and human dimensions of climate . http://www.iom.int/files/live/sites/iom/files/About- change gained preponderance within the climate IOM/governing-bodies/en/council//MC_.pdf

6 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

of Human Rights”, devoted to understanding in- Finally, the year  also saw the release of new ternational cooperation and respect for human evidence on migration and environment, for in- rights in all climate change related situations. stance the IPCC Special Report on Managing the The seminar contributed to raising awareness Risks of Extreme Events and Disasters to Advance and understanding of the relationship between Climate Change Adaptation (SREX) that highlights climate change and human rights and to enhance the relationships between climate change and ex- cooperation among all relevant actors. Moreover, treme weather events and strategies to manage the Special Rapporteur on Human Rights of Mi- associated risks. The report Where the Rain Falls grants, François Crépeau, submitted a report on produced by the United Nations University Insti- the “Human Rights of Migrants” (August ) tute for Environment and Human Security UNU- with a thematic section dedicated to the impacts EHS and Care explores the complexities of rainfall of climate change and some of its consequences patterns and their effects on food security and hu- for migration. The report recommends that “States man mobility in eight countries, Guatemala, Peru, must collectively recognize that migration is part of Kenya, Tanzania, India, Bangladesh, Thailand the solution to global environmental challenges” as and Vietnam. Another important study with a re- well as “identify in priority vulnerable populations gional scope was released by the Asian Develop- who are susceptible to migrate internally or interna- ment Bank (ADB) on Addressing Climate Change tionally (…)”. and Migration in Asia and the Pacific presenting Over the year , the role of environmental the actual and potential impacts of climate change factors as key drivers of displacement has been in- on migration in Asia and the Pacific. This regional creasingly recognized in the protection sphere as focus also corresponds to the prevalence of Asia well. In October was launched the Nansen Initia- case studies in this publication. The importance of tive, an intergovernmental process supported by natural events with a human mobility dimension the governments of Norway and Switzerland that has also led to the launch in  by the ADB and seeks to promote a protection agenda for those dis- IOM of a new regional knowledge platform Asia placed across borders by natural disasters. Pacific Migration and Environment Network AP- Environmental migration has also made some MEN devoted to expertise exchange on environ- progress within two other significant policy tracks, mental migration. the disaster risk reduction and the development ones, yet it remained on the side ways of these processes in . At the request of the Special ABOUT THIS EDITION Representative of the Secretary-General for Disas- ter Risk Reduction, the senior level management As in previous years, this report is based on the group for Disaster Risk Reduction and Resilience works conducted by students of the course “Envi- was called to identify and act on core issues and ronment and Migration” taught at Sciences Po gaps in disaster risk reduction in preparation of the (Paris School of International Affairs, PSIA), and  Global platform on Disaster Risk Reduction would simply not exist without them. Each chapter (DRR). This collaboration also offered ground for documents a recent case of migration related to bringing human mobility issues on the DRR policy environmental changes, through a careful review agenda. Furthermore, migration was recognized of secondary sources and interviews with actors as relevant in the context of environmental coop- who were directly involved in the protection and/ eration at the  United Nations Conference on or assistance to those displaced. They do not seek Sustainable Development (Rio+), held in Brazil to cover all cases of environmental migration in June . If migration is indeed mentioned in that took place in  – this would be an impos- these policy frameworks, many more efforts of re- sible task. Instead, they aim to provide empirical search and advocacy are necessary to make human evidence of some of the most significant cases from mobility a key feature in development and DRR , and hopefully they pave the way for future frameworks. academic research and policy developments. The report is divided in two parts. The first section addresses cases of flights and evacuations. . Concept Paper Seminar on Addressing the Adverse Impacts of Climate Change on the Full Enjoyment of . http://www.ipcc-wg.gov/SREX/ Human Rights, February , OHCR . http://wheretherainfalls.org/ . http://daccess-dds-ny.un.org/doc/UNDOC/GEN/ N///PDF/N.pdf?OpenElement . ADB,  . Outcome document of the United Nations Confer- . http://www.adb.org/publications/addressing-climate- ence on Sustainable Development (Rio+) http:// change-and-migration-asia-and-pacific www.uncsd.org/thefuturewewant.html . http://www.apmen.iom.int/en/

IDDRI STUDY 09/2013 7 The State of Environmental Migration 2013

Cases that are documented here include the drought in Northern Kenya. Finally, two chap- evacuation related to hurricane Sandy in New ters tackle some specific cases of displacement, York and New Jersey, but also those related to return and resettlement: one documents the cur- typhoon Bopha in the Philippines and cyclone rent situation of the Fukushima evacuees, more Evan in the Samoa. The displacements related to than one year after the disaster, while the final recurrent floods in Assam and Brahmaputra are chapter reviews the resettlement induced by the also documented in a chapter. These chapters building of dams in the Mekong region. In this are enriched with boxes highlighting the disas- section as well, different boxes document how ter risk reduction policies that were taken with different environmental changes were addressed regard hurricane Sandy in Haiti, cyclone Gio- by authorities: the recurrent disasters in Bangla- vanna in Madagascar, hurricane Isaac in Louisi- desh and the current situation of those resettled ana, as well as floods in Australia, Mozambique because of the earthquake of L’Aquila in . and Nigeria. This year again, the State of Environmental The second section documents different pat- Migration presents a complex picture of the mi- terns of mobility, resettlement and return. The grations and displacements induced by environ- case of pastoralists driven away from their homes mental changes. Yet, as the scientific evidence by soil degradation and droughts is addressed in mounts up, policy progress remains too slow and a comparative chapter that contrasts patterns insufficient. Hopefully this edition will show the of mobility in Burkina Faso and Mali, while an- variety of policy responses that are needed, and other chapter addresses the complex situation of will prove useful for future academic research pastoralists who were displaced because of the and policy developments.

8 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

METHODOLOGICAL CONSIDERATIONS PAULINE BRÜCKER & FRANÇOIS GEMENNE

Since its first edition, the format and the themes of Internal Migration, left at the discretion of states the State of Environmental Migration have evolved, a comprehensive review of those frameworks and reflecting upon the complexity and diversity of their applications would be valuable for expanding the phenomenon and our willingness to bridge our knowledge on the political sociology of envi- and feed growing interests amongst individuals, ronmental migration management, especially in a scholars and policy makers. This third edition comparatist perspective. reviews major events that occurred in . Our Methodological concerns for research on envi- understanding of what is and what means “environmental migration have been tremendously ex- ronmental migration” is as large as witnessed panding. A multiplicity of methods have been used events suggest it to be. The challenges faced by in current and recent research on the topic, such as the contributors in their research brings new and ecological inference based on area characteristics, interesting lights on the challenges of researching individual sample surveys, time series, multilevel on extremely recent event, especially in cases of analysis, agent-based modeling (ABM), and quali- complex phenomenon. They may feed on-going tative/ethnographic studies (Piguet, ). The debates on the methodology of environmental multiplication of training activities is symptomatic migration research. of a quest for a renewed understanding of the tools to apply on the field but also in the analysis of public policies, in their great diversity. Many challenges THE CHALLENGES OF UNDERSTANDING remain however, related for instance to the nature A COMPLEX PHENOMENON of the environmental events, especially slow-onsets events triggering diffuse migration flows often par- The difficulties of researching environmental ticularly difficult to quantify (IOM, ). migration have been widely acknowledged. An The methodological reflection gathered here in expanding body of literature addresses the notion the frame of the State of Environmental Migration of “environmental migrant” and questions both reflects particularly on the conduct of research in its multi-causality (Castles, ; Kibreab, ; the aftermath of a disaster or slow onset events. Diamond, ; Lonegan, ; Foresight, ) and its internal dimension (IOM, ; Fore- sight, ) – two main challenges for scholars to THE METHODOLOGY OF THE STATE label migrants as ‘environmental migrants’ and OF ENVIRONMENTAL MIGRATION to move from correlation to causation. Secondly, providing a clear picture of the frameworks appli- The choice to divide this edition into two main cable and assessing the efficiency of the humani- sections - “Flight and Evacuations” and “Mobility, tarian responses provided to those displaced can Resettlement and Return” – derives from the will- be extremely challenging for researchers seeking ingness to focus on the type of movements rather to establish both the norms in practice and the than on the type of environmental events trig- responsibilities of all actors at stake. The protec- gering the migration. Although the division may tion of environmental migrants has indeed been seem superficial in substance, it emphasizes the increasingly discussed among scholars but also nature of migration, i.e. who is moving, how, when amongst policy makers, resulting in the sugges- and why, and the subsequent challenges associ- tions to either create new normative frameworks, ated. It takes a step away from the classical divi- or at least, to provide comprehensive guidelines sion, yet adopted in previous edition of the State to ensure systematic responses (Zetter, ; of Environmental Migration, between sudden and McAdam, ). Yet, as for now, most of the policies slow onset induced displacement. applicable to environmental migrants are disaster As for the last editions, the choice of the cases risk reduction frameworks, and possibly, national studied was based on predefined criteria. They translations of the  Guiding Principles on needed to address events that occurred in  or

IDDRI STUDY 09/2013 9 The State of Environmental Migration 2013

that were still on-going in . They also needed share some of their thoughts. The following points to present and assess the policy and humanitar- underscore those that appeared as the greatest ob- ian responses implemented, and possibly nor- stacles. Our goal in sharing them for the first time mative changes if they were to happen, pre- or in this edition, is to bring a new voice to the litera- post-disaster. ture on environmental migration methodology. Interestingly, this edition sheds light on major m Recentness of the event: The recentness of the cases that occurred in industrialized countries, event was a main challenge in all researches. Mi- including hurricane Sandy in New York, floods in gration movements were still on- going, imped- Australia, the earthquake of L’Aquila in Italy and ing any possibility to provide quantitative data, challenges related to post-Fukushima return. This or to evaluate emergency policies implemented. edition, once again, emphasizes on the many ob- Most of the data was often found contradic- stacles to protection but also brings light on best tory. The reliability of press articles was some- practices throughout the world. We truly believe times a problem and that they tried, as much that research must be held outside of the box and as possible, to confront the different secondary beyond overrun paths, as environmental migra- sources used. More generally, the recentness of tion is a truly worldwide concern. The rationale the event interrogates on whether it is possible for our geographical coverage concurs with this to take a step back to provide an informed but orientation. cautious picture of its impacts, responses and The originality of this approach allows this vol- consequences. ume to cover recent events that, precisely because m Managing the great variety of sources: a conse- of their temporal proximity, are little addressed by quence of the recentness of the analysis is the research scholars. Some would argue that those necessity to resort to a great diversity of sources. cases do not offer the appropriate hindsight to pro- All chapters of this volume are mainly based on vide a comprehensive assessment of the situation. secondary sources, especially newspaper arti- To this, we argue that there is on the contrary a cle and grey literature. Though this seemed the need to gather the first data released on such situ- most practical way, it raises questions on the ations, confront the sources in their great variety, possible treatment of media and institutional but also underscore what type of data may be miss- sources, their multiplicity, their political orien- ing in the very aftermath of a disaster, and how tation and content, at times contradictory. To this may undermine the quality of the response. overcome such obstacles, a constant care to who Accordingly, the methodology applied in all the was providing the information, how, why and following chapters try capture both quantitative when – was required from the students. They and qualitative information on each specific case. were also encouraged to raise contradiction in no field work was undertaken in this  edition, their material, and to reflect on these contradic- only one student was able to pursue local research. tions if they were pertinent with regard to the However, the joint publication with IOM has pro- subject of research. gressively facilitated students’ contacts with IOM m Accessing quantitative data: research on environ- staff members on the field, which has allowed mental migration has long focused on the ques- some of them to rely on first-hand sources. Oth- tion of numbers, as one tool to raise awareness ers have chosen to interview other experts such about a phenomenon at the beginning politically as journalists, scholars, policy makers etc. The key marginalized. Yet, providing clear-cut data on sources of information remain governmental agen- the scope of environmental migration is a main cies’ reports, international, national and non-gov- challenge raised by all contributors, especially ernmental humanitarian and emergency relief re- in such short timelines. Some students found it ports, along with a great amount of press articles. particularly surprising in countries such as the Such combination of grey literature and news ar- United States of America – this stresses how po- ticles, rather than first-hand material collected on litically charged may be any act of quantifica- the field, may be criticized by our peers. However, tion. Another problem faced was the uniqueness the goal of the State of Environmental Migration is of data: in some case, only governmental data to provide scholars and non-academics with first were available or on the contrary, international insights. Those sources allow for a comprehensive organizations’ provided by OCHA. In other cases, understanding of the diversity of the discourses they were no data on environmental migrants on environmental migration therefore analyzed in as such: aggregated data on migration at large their social context. was at times the only one accessible, highlight- This being said, students were faced with enor- ing once more the complex multi-causality of en- mous challenges in their research. We have thus vironmental migration, along with the political asked them to reflect upon their experience and to games that may hide behind numbers and labels.

10 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

m Sensitivity of the subject: Environmental migra- BIBLIOGRAPHY tion remains a sensitive subject for many gov- Castles, S. (2002) Environmental change and forced migra- ernments. Reluctance was sometimes observed tion: making sense of the debate, UNHCR Issues in Refugee with regard to the access to information, and Research, Working Paper No. 70 some problems occurred also at times with the Diamond, J. M. (2005) Collapse: How societies choose to field staff interviewed. fail or succeed. New York: Viking Penguin Publishers. m Comparing the incomparable? Beyond the chal- Foresight: Migration and Global Environmental Change lenges faced by each contributor, the structure (2011), Final Project Report, The Government Office for itself of the SEM is a main challenge to the so- Science, London. ciology of public policies. As we said, the mul- IOM (2009), Migration, Environment and Climate Change: tiplicity of policies managing environmental Assessing The Evidence, Geneva, 441p. migration – i.e risk management, disaster risk Kibreab, G. (1997) “Environmental causes and impact of reduction, humanitarian assistance and emer- refugee movements: a critique of the current debate”, Dis- gency, migration policies, along with the multi- asters 21(1), pp. 20–38. plicity of actors behind those rules, raises many Lonergan, S. (1998) “The role of environmental degradation question with regards to the possibility of op- in population displacement”, Environmental change and se- erating any type of compared analysis. It also curity program report, Issue no. 4. Washington: Woodrow questions the pertinence of comparing cases Wilson International Centre for Scholars, pp. 5–15. that seem incomparable in appearance– or on McAdam J. (2011), “Swimming against the tide: why a cli- the contrary may underscore how our percep- mate change displacement treaty is not the answer”, Inter- national Journal of Refugee Law, tion of what may be compared is erroneous. 23:1, pp. 2-27 Accordingly, the research produced here of- Piguet Etienne. Linking climate change, environmental deg- fers food for thought and opens perspectives onto radation, and migration: a methodological overview. WIREs Clim Change 2010, 1: 517-524. doi: 10.1002/wcc.54 future researches to be undertaken by scholars, through a methodology borrowing from the soci- Zetter, R.W., (2008) “Legal and Normative Frameworks”, Forced Migration Review, 31, Special issue on Climate ology of public policies, the sociology of migration Change and Displacement, pp: 62-63. and the political sociology of the State, in a multidisciplinary perspective. It seeks to provide a comprehensive image of who moved, who could not, why, and under which conditions.

IDDRI STUDY 09/2013 11

PART 1 FLIGHTS AND EVACUATIONS

The State of Environmental Migration 2011

HURRICANE SANDY IN NEW YORK AND NEW JERSEY: EVACUATION, DISPLACEMENT AND ADAPTATION FANNIE DELAVELLE

INTRODUCTION for future projects, particularly the option of relocation as an adaptation strategy (Part III). “Maybe things like this happen in New Orleans, places like that. But never here, not in a million years.” 1. HURRICANE SANDY AND NEW V. Baccale,  YORK’S VULNERABILITIES Vinny Baccale’s family had lived on Staten Island for generations, starting with a small holiday .. Hurricane sandy, a record bungalow to building a permanent home on the breaker seaside. This area was hit particularly hard by the storm. Like thousands of New Yorkers, Baccale, his Hurricane Sandy was a record breaker in many wife and two small kids were forced to evacuate ways. On  October , its central pressure when their seaside home was partially destroyed reached  millibars, the lowest-measured by the storm. Despite their strong roots in the area, barometric level for an Atlantic hurricane. It also they are now considering relocating to a safer caused a record surge of water for New York City, neighbourhood (J. Rudolf, ). with a water level of . meters at Battery Park on Hurricane Sandy was a shock for many New October th (Drye, ). Furthermore, Sandy Yorkers who had to admit their vulnerability to was much wider than previous storms, with hurri- environmental disasters. Vinny Baccale’s remark cane-force winds spreading over a diameter of  underscores the persisting perception in many hundred kilometres around its eye. developed countries that such disasters only hap- In the afternoon of October th, Sandy brought pen in other places, to other people. Hurricane strong winds and rains from Washington D.C. Sandy, as Hurricane Katrina before it, reversed northward. It was downgraded to the post-tropical this idea, showing that developed and relatively cyclone category before it hit Southern New Jersey wealthy metropolises like New York City are at around  p.m. EDT, near Atlantic City, with winds risk, and that environmental displacement is a of km/h (Sharp, ). Sandy moved its way concern for all countries, no matter their level of up to New York, where its surges topped seawalls development. at the Battery and flooded Lower Manhattan. The This paper begins by analysing the characteris- storm’s massive size caused flooding, rain and tics of the storm itself, and by evaluating New York high winds to continue affecting New York and City’s vulnerabilities to environmental disasters New Jersey throughout the night. On October th, (Part I). It then assesses the evacuation process as Sandy had left New York, although its tail-end was well as the forced displacements, with an empha- still causing considerable damage on the North- sis on the authorities’ management of the crisis east. The storm started to weaken in the afternoon both in the short and middle term. A special sec- of October th and dissipated on October st over tion is dedicated to the impact of the storm on the Pennsylvania. Sandy caused the death of  peo- elderly and on low-income persons (Part II). Fi- ple in the United States, including  in New York nally, the report addresses the issue of New York’s City (Drye, ). adaptation to environmental disasters, evaluating The most destructive characteristic of the hur- previous efforts and putting forward proposals ricane was the storm surge, that affected parts of

IDDRI STUDY 06/2012 15 The State of Environmental Migration 2013

Map 1. Sandy’s track: September 9-17, 2012.

Source: Wunderground (2012) the Northeast, which is home to  per cent of the to natural and man-made forces like erosion that US population (Murphy, ). A combination of reduce the amount of offshore sand. By , several factors magnified the damaging effects of experts have predicted, through the approach of the floods. the Intergovernmental Panel on Climate Change, that the sea level will rise in the New York area .. Short and long term by - cm in the best case scenario and .-. exacerbating factors: the full meter in a rapid-ice-melt scenario (Khan, ). moon and sea level rise Although the range of these projections remains very broad, they point out that hurricane of the First, Sandy made landfall during the full moon, same strength will provoke more intense damage while the tide was  per cent higher than usual as the surges are worsened by the higher baseline (by about  cm) (Khan, ). Although the sea level (Avent, ). initial waves were not as high as in previous storms, the high tide caused the water levels .. The New York Bight to flow over sea walls and other protections. Second, the impact of the full moon combined The hurricane’s unusual path led its strongest with the longer-term phenomenon of sea- winds –on its front and right-- to hit the angle level rise. Although the storm itself might not between New Jersey and Long Island that forms be directly connected to climate change, the a tip called the New York Bight. This sharp curve damaging power of the floods partly was. Since on the open coastline significantly magnified the , as a consequence of the ocean’s thermal impacts of the hurricane, as the waves composed expansion and of the melting of ice caps in the of ocean water and rain accumulated in the New Arctic, the sea level in the New York region has York harbour. The high surges were boxed in, and risen of about c m–twice as fast as water levels pushed onshore into the flood zones (Murphy, of coastal regions in other states (Khan, ). ). This mechanism was amplified by the rela- New York and New Jersey are particularly vulner- tive shallowness of NYC waters that made the able to the effects of sea-level rise, principally due surge go even farther onto land.

16 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 2. Population density in New York region

Source: Centre For International Earth Science Information Network – Earth Institute, Columbia University (2012)

Sandy’s destructive power was thus amplified by .. Demographic and a combination of short-term climatic phenomena, infrastructural vulnerabilities gradual environmental evolutions and the intrinsic geographical vulnerabilities of the region. The damage Sandy provoked was further exacerbated by the demographic and infrastructural Map 3. New York Bight characteristics of the East Coast. New York City is particularly densely populated on its extensive km of coastland (Jones, ). According to the New York City Office of Emergency Management, about . million persons would be in high-risk if a Category  hurricane hit the area --three-times the combined potentially vulnerable population in New Orleans and Miami. In addition to its demographic density, New York’s infrastructural characteristics make it especially vulnerable to hurricane damage. The many suspension bridges added to New York’s wind vulnerability, forcing authorities to close all bridges to avoid their collapse. These closures reduced the Source: Murphy, OEM/City Limits (2012) number of evacuation routes (Silverman, ).

IDDRI STUDY 09/2013 17 The State of Environmental Migration 2013

Third, New York’s subway system is quite prone Figure 1. Hurricane Irene (top) and Hurricane Sandy to flooding. When it was designed over a century (bottom) ago, authorities did not anticipate the impacts of hurricanes causing subway grates to be flooded. City officials/transportation authorities made few changes since the initial construction to provide a better protection of urban subways, mainly due to inadequate funding. As Lower Manhattan was flooded, underground networks were immediately filled with water despite the sealing off procedures that had been developed.

.. Winter storm Athena aggravates Sandy’s damages

The damage caused by Hurricane Sandy was aggravated by a second storm, which hit the New York area from - November . This early winter storm brought up to  cm of snow and rain across regions that had been significantly affected by Hurricane Sandy (Hydrometeorological Predic- tion Centre, ). The storm also caused high waves of .m, flooding many coastal roads and stalling post-Sandy repairs. An additional , households lost power in New York and New Jersey, adding to the , still without power after Sandy (The Associated Press, ). Source: NASA (2012)

.. Hurricane Irene increased new york’s vulnerability 2. EVACUATION AND DISPLACEMENT: to hurricane sandy WHEN THE UNBELIEVABLE Before it made landfall, Sandy was widely compared BECOMES REALITY to Hurricane Irene that had hit the same region in late August . Irene was ranked the seventh cost- .. The multiple waves liest hurricane in the history of the United States, of evacuation with an estimated total cost of USD  billion, and caused  deaths (Fischetti, ). However, it did Over a week after Hurricane Sandy hit, New not impact the New York area as authorities and York City authorities stated that the number of the media had anticipated, because it weakened displaced persons ranged from , to ,. from a Category  storm at sea to a tropical storm The lowest figure corresponds to the number of at landfall in New Jersey, with km/h winds. The people who registered in public shelters, where an highest recorded rainfall measurements in New official count could be made. Any number above York barely reached  cm, in comparison with the this is a broad approximation, as many people did Sandy floods that were measured in meters. The not go to shelters, and as part of the population in relatively low damage it inflicted compared to the evacuation zones did not evacuate (Jones, ). predicted disaster led many to criticize the authori- Furthermore, the numerical estimation is compli- ties’ “overreaction.” Indeed, in its preparedness, cated by the multiple waves of evacuation. the city had evacuated , people and had On Sunday, October th, Mayor Bloomberg shut down subways, airports and buses  hours issued evacuation orders for low lying areas, in- before the storm was expected to hit (Barron, ). cluding sections of lower Manhattan, parts of Although the mayor argued he based his decisions Staten Island and Brooklyn, Coney Island and the on the side of caution, the criticisms partly led the Rockaways in Queens. These areas, populated by authorities to adopt a slightly less cautious atti- , people, represent the city’s evacuation tude during Sandy, which proved to be much more “Zone A” (Gupta, ). The first wave of evacua- devastating (Nye, ). tion was therefore mostly composed of the inhabitants of Zone A who followed Mayor Bloomberg’s

18 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 4. SLOSH Model of evacuation zones

Source: New York City Office of Emergency Management (2012) order.The three evacuation zones A, B and C had evacuation zones that had not yet obeyed the may- been designed by NYC’s Office of Emergency Man- or’s order, as well as households in other areas agement in the late s, on the basis of a SLOSH who were either forced out of their houses by the model that showed the areas of the city vulnerable unexpected damage of Sandy, or voluntarily left in to flooding. precaution. Further evacuations took place after Hurricane In the following days, evacuations continued as Sandy hit New York City on Monday October th in many households were forced to move out of their the evening, as New Yorkers began to feel Sandy’s houses because of the prolonged power outages. A destructive power. This second wave of evacu- week after the storm, . million people remained ations included households in the mandatory without power, and it was weeks before electricity was restored in some areas, while the already low . Zone A represents the lowest-lying areas of the city, winter temperatures continued declining (Maze- most vulnerable to surges. Zones B and C represent the lis, ). regions that are expected to be flooded by hurricanes From November th to November th, winter of higher intensity. The order to evacuate Zone A corre- sponded to the predictions that the hurricane’s intensity storm Athena prompted further evacuations as was decreasing, but might still be Category  at landfall. snow and rain accumulated in areas that had been

IDDRI STUDY 09/2013 19 The State of Environmental Migration 2013

previously hit by Sandy. The winter storm caused with the zones’ boundaries, many residents were further power outages, forcing individuals out of still unsure which zone they lived in, revealing a their houses to find shelters with heating and elec- lack of awareness of the risks posed by hurricanes. tricity, and amplified Sandy’s damage, as snow ac- The city’s evacuation and early warning proce- cumulated on broken roofs and caused the fall of dures should be improved, particularly by enhanc- further trees. ing the dissemination of information about the These differences in the time and date of evac- endured risks. uations were partly caused by the lack of obedi- ence to Mayor Bloomberg’s evacuation on Sunday .. From short-term th. One explanation for this refusal to follow evacuation to displacement the mayor’s orders is the scepticism of the population towards the authorities’ warnings after The shelter challenge their “overreaction” after Hurricane Irene in  The various evacuations waves can therefore be (Preston, ). This scepticism was amplified by distinguished according to their time of occur- the National Hurricane Centre’s downgrade of the rence, but also, and most importantly, according hurricane to a post-tropical storm before it made to their length. While some evacuees were able landfall, decreasing the population’s risk percep- to return to their homes just a few days after the tion. Furthermore, the authorities added to the hurricane hit, other households were still displaced confusion by announcing that nursing homes in in May . The length of displacement is largely evacuation zones would not be evacuated. Many to be correlated to the damage to the evacuee’s people deduced that the storm would not be much home. Whereas some buildings were only lightly more severe than Hurricane Irene, and decided damaged, many structures were either severely to ride it out (MacDonald, ). The dispersion affected or completely destroyed. In such cases, in the dates of evacuation was also caused by the evacuees turned into displaced persons, as they authorities’ two-fold response. Mayor Bloomberg were unable to return home. It is estimated that first declared that evacuation would not be neces- about , persons became homeless because sary, as the authorities were expecting a slow ac- of Sandy, in New York City alone (Mazelis, ). cumulation of water rather than a sudden surge, In the first few days after Sandy’s landfall, of- as they were comparing Sandy to Irene. However, ficials provided emergency shelters for evacuees, on Sunday th, the day before Sandy hit New usually in public schools (Bernstein N., ). York, the tone completely changed as he ordered While a few successful examples can be highlight- a mandatory evacuation of Zone A, the low-lying ed, such as the opening of recently closed Fort Mon- areas of the city (Preston, ). The late order left mouth for  displaced families (Bonamo, ), little time for people to evacuate, causing many provisions for shelters and transitional housing New Yorkers to ignore the order, both because were not adequate. In the weeks following Sandy, they were not psychologically and physically pre- the evacuees had to be relocated several times, as pared to evacuate, and because they considered schools were reopened or as they decided to move that such a late order must be an overreaction. to another shelter in the hope that it would be less Those who did follow the order had so little time overcrowded. Many were moved to drill floors and that they often took few personal belongings with armouries, but the conditions were often precari- them (assuming they would be back after a few ous (Bernstein N., ). Sandy highlighted the days) whereas they were actually often displaced lack of space in New York City, in comparison to its for extended periods of time (Barron, ). dense population, as well as the lack of emergen- Another interesting reason for the refusal to cy preparedness of local authorities. The shelters evacuate is that most New Yorkers had too lit- were usually filled to twice their capacity, and the tle experience responding to hurricanes to have city struggled to find additional spaces to welcome built a response culture. This human factor plays new comers. In response to this situation, the Fed- an important role during evacuation processes, as eral Emergency Management Agency (FEMA) –in people often measure the risks against similar past coordination with NYC officials- reserved hotel experiences (Monitz, ). As the situation was rooms, sometimes luxury ones, across Manhattan, new for most New Yorkers, many made misjudge- Brooklyn and Queens, as a transitional solution. ments on the ways to respond, for instance, by However, as evacuation centres were progressive- disregarding evacuation orders. In addition to this ly closed to rely only on hotels, officials faced the psychological element, a great obstruction to the new wave of displaced people from houses lack- efficiency of evacuations turned out to be the lack ing heat and hot water, in the cold winter, forcing of knowledge on the evacuation zones. Although them to re-open some shelters and book additional Hurricane Irene had strengthened the familiarity hotel rooms (Bernstein N., ).

20 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

When evacuation turns into displacement assistance, but actually used that money to pay for After the urge to find temporary shelter, authori- gas to drive their children to school in their previ- ties were faced with the necessity to provide longer- ous neighbourhoods, to drive to their jobs, to pay term structures for households whose homes the mortgage on their damaged house, and to buy were destroyed or inhabitable, and to launch a food. They, therefore, did not have any funds left recovery programme for damaged houses. In early to find proper lodging or to repair their homes, November, over , households were approved and had lower hopes to return to a stable existence for housing assistance to help them seek a tempo- in the short term. Furthermore, hotel stays were rary living place, or repair their houses, for a also very costly for the displaced families, as they total cost of over USD  million (Barron, ). did not have kitchens to cook in and had to order However, the management of longer-term displace- take-out or go to restaurants every day (Associated ments was as debatable as the city’s emergency Press, ). preparedness. Once power was re-established, and The number of displaced people gradually de- individuals who could return to their homes had creased from December to March as many house- left shelters, the authorities started relying almost holds were able to return to their homes after exclusively on FEMA hotel room system as a tran- they were repaired, or as they decided to leave sitional solution for displaced families. However, if FEMA hotels for a more stable situation, such as this option was adequate for transitional lodging, a rented apartment (FEMA, ). A small num- it was not designed as a long-term option. As ber of families whose houses were damaged but weeks and months went by, the authorities discov- had land available opted for FEMA’s “temporary ered that the ‘transitional’ situation turned out to housing units,” which soon became negatively re- be ‘quasi-permanent.’ Three months after hurri- ferred to as “FEMA trailers” (Barron, ). The cane Sandy hit New York and New Jersey, at least fact that some individuals preferred to move to , families remained displaced in hotels, often trailers rather than remain in hotels underscores having to change rooms or hotels as reservations how inadequate FEMA’s hotel housing programme for conferences and parties came up (Zezima, was in the long term. It should also be mentioned ). Hotels close to affected areas were taken that while some households left for other lodging over, such as the Clarion Hotel in Toms River, options, many persons that had not previously where about  per cent of the guests were part applied for FEMA financial or lodging assistance of the FEMA programme in January. The expected were forced to request help as their personal cir- cost for FEMA has amounted to tens of millions cumstances evolved. Indeed, in many cases, indi- of dollars, with an estimated rate of USD  per viduals received help from relatives and friends room per night on average. for some time, but realized after a few months A positive aspect of the hotel programme is that that this situation could not last any longer (Jones, it enabled most households to remain close to their ). In January, FEMA received about  new former living places, thus preventing the need to applications every day for financial and/or lodg- change schools for children, and long commutes ing assistance (FEMA, ). This point underlines to the workplace for adults. However, the families that the situation was constantly evolving, as dis- lived in constant fear regarding their future, as placed people that had not been accounted for yet the hotel stays expired every two weeks (Zezima, stepped out to ask for assistance. Overall, FEMA ). This lack of guaranteed stability was a great has registered over , households, and given source of worry, as the FEMA’s decision to renew over USD  million in assistance (FEMA, ). the hotel stays was made on a case-by-case basis However, it is important to highlight that although (Piore, ). Households constantly feared they the displaced households in hotels were the most would lose their right to hotel assistance, for invisible, the majority of displaced individuals were stance if the FEMA decided that their homes were “invisible,” staying with relatives or friends, rent- safe to come back to, that they had a sound insur- ing apartments, or remaining homeless. As such ance settlement or had received assistance to rent individuals have not been officially registered, it another place (Zezima, ). Although such rea- is difficult to evaluate the number of persons in- sons for the removal of hotel assistance seem rea- cluded in each category (Jones, ). sonable, they added to the psychological plight of displaced households, and were occasionally applied in an arbitral manner, without taking some . FEMA trailers are widely associated with New Orleans’ households’ characteristics into account. This was Katrina victims, as over , FEMA trailers and sometimes the case for low- and middle-income mobile homes were used after the hurricane hit the city in August , becoming part of the city’s landscape. households, who were shut out of the hotel assis- The last FEMA trailer was removed in , over  years tance programme after receiving FEMA household after the disaster. (Burdeau, )

IDDRI STUDY 09/2013 21 The State of Environmental Migration 2013

.. Sandy’s long-term displaced Sandy, New York City had the largest population of homeless people, ,, and . per cent of the Five months after Sandy made landfall, over  city’s overall population lived in public housing people remained displaced and living in hotels in structures (Huang, ). These buildings were New York, through FEMA’s temporary housing usually quite old and badly maintained, with pre- programme. Most of the remaining displaced house- existing mould problems that were aggravated by holds in hotels were either too poor to pay for their the floods. Evacuation orders were often ignored home repairs or were renters whose landlords had by the people with the least means, because of the not dedicated the necessary funds to repair their high price associated with it. Sandy would have apartments (O’Connor, ). In order to provide required evacuating relatively far away, and some- for this group, on March th, FEMA launched a times moving from shelter to shelter. These factors “Disaster Housing Assistance Programme,” in coop- led many low-income people to try and ride out the eration with the Federal Department of Housing storm. Public transportation had been stopped on and Urban Development, modelled on the one the th in the morning, before the hurricane hit, that was launched after Hurricane Katrina. This removing this cheaper option (Mathis, ). As the programme was designed to help displaced people subway and train systems remained closed or with who were still living in motels find long-term solu- running with reduced frequency after Sandy, many tions, by temporarily paying their rent. It aimed at households were unable to evacuate from their introducing more stability to their lives, by acting as disaster-struck homes, and had to survive without “a bridge from displacement to a permanent relo- power or basic food and health facilities. Another cation in their former houses” as Senator Chuck reason that led lower-income persons to stay in Schumer of New York declared. It was particularly their homes is that the jobs they occupy are usually designed for displaced families in hotel rooms who quite unstable. Missing a work day meant missing could not afford to rent an apartment. Indeed, a day’s salary, which is something that most public after Sandy hit, many real estate firms such as housing residents cannot afford (O’Connor, ). Ashore helped by finding families homes in winter The number of people who filed for first-time unem- rentals. In December, some landlords accepted to ployment insurance benefits after Sandy highlights prioritize displaced people for the rental of apart- the disastrous impact that Sandy had on the poorer ments at market price. Since these rentals were groups: , applications were made in the managed privately, it is not possible to evaluate how period between November th and November th many of those displaced found acceptable housing. only (some of those applications may also have been However, few people (or families) displaced by filed as businesses closed due to damage). Further- Sandy could afford rents at market price. Estimates more, although the most publicized media reports show that  per cent of the households that applied after the storm showed the destroyed houses in had annual income inferior to the estimated USD coastal neighbourhood, we have to think about the , needed to qualify (FEMA,  ). Although losses proportionally to the person’s assets. Home- FEMA’s Disaster Housing Assistance Programme owners who lost their house surely suffered, but represents a considerable improvement for the life usually had savings or resources to draw upon, in of displaced households, families (or people) still contrast with poorer renters who sometimes lost live in the fear of leases ending, particularly with the everything they owned (O’Connor, ). Almost beginning of the tourist season, when rent prices half of all FEMA registrants indicated their annual usually peak. Indeed, although FEMA will provide income was below USD , (Enterprise Commu- financial support by paying rent temporarily, land- nity Partners, ). Almost  per cent of them lords usually want to be assured that they will be were renters, who usually did not have insurance able to keep paying once the assistance is removed for their possessions. Furthermore, in cases when (McKelvey, ). As this programme was only set in the renter’s home was damaged but not completely motion five months after the hurricane, the condi- destroyed, they depended on their landlords’ will tions of displacement were often precarious both and financial capacity for repairs. As landowners in physical, financial and psychological terms for a often lost other large assets during Sandy, these developed city such as New York (Dawsey, ). necessary repairs were not their priority, forcing many poorer households to live in precarious condi- .. Displacement and social tions in terms of health, with increasing mould, inequality - the diverse among other issues (O’Connor, ). outcomes of Hurricane Sandy . As gasoline is expensive, and many of them did not have Low-income people constituted a particularly cars (access to cars in New York is lower than in any other vulnerable group to Sandy’s impact. Before Hurricane city in the United States).

22 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

THE ELDERLY IN THE FACE OF SANDY: elderly people, which is a 48-hour process. Nursing homes were then ordered to increase staffing and to transfer the residents A TALE OF DISPROPORTIONATE who required ventilators to hospitals. VULNERABILITIES In addition to this lack of caution by NYC authorities, the effects of the hurricane on nursing home residents were wors- The intrinsic vulnerabilities of the elderly to environmental ened by the default in emergency preparedness in the facilities. disasters First, emergency backup generators are not legally required to A specific social group, the elderly, was particularly affected by be able to withstand flooding, as health inspectors only check the hurricane’s damage. Approximately 37 per cent of those who that they are functional, but not whether they are protected died in New York City because of Hurricane Sandy were 60 or against flooding. When Sandy hit, the waters rose so quickly older, whereas only 17.2 per cent of the city’s population belongs that the workers in the facilities did not have time to elevate to this age group (Synergy Home Care, 2012). Many elderly peo- them. In Queens and Brooklyn alone, over 23 nursing homes ple drowned within their houses, while others suffered from were severely flooded, leading to heat and power to go outin fatal storm related injuries and hypothermia (Parry, 2013). many cases (Preston, 2012). Elderly people therefore had to The elderly are extremely vulnerable to the effects of storms ride out the storm within the nursing homes, and were later for two main reasons. First, elderly people are often socially on forced to evacuate after the surge, due to the absence of isolated, making them more vulnerable in the case of a natural electricity. In the days following the storm, 6,300 patients were disaster. Although this isolation is sometimes unwelcomed, it evacuated from 47 facilities according to the New York State most of the time stems from a choice: elderly people are more Health Department (Durkin, 2013). These examples highlight reluctant to ask for services or seek their family’s help, because the importance of emergency preparedness of nursing homes of a will to maintain their independence, and out of concern themselves, as facilities where generators were protected from that they will bother others (Parry, 2013). This was often the surges, like Ocean Promenade in Queens, did not have to evacu- case during Hurricane Sandy, when many elderly decided to ate after the storm. Evacuees were taken to other facilities, but “ride out” the storm instead of calling their families or res- also to high schools converted to shelters, often quite far away- cue teams for help. Second, most of the disaster preparedness -even as far as Albany. They were sometimes sent without their procedures are designed for healthy individuals, sometimes medical records or medications, to places that were not special- requiring a walk of a mile to the closest evacuation centre, or ized for nursing home care. Furthermore, the evacuation flows to take crowded buses. Elderly people have a small capacity caused residents from facilities dealing with specific illnesses of adaptation because of their limited physical capability to to be mixed with residents from “normal” nursing homes, caus- leave their home independently (Solecki, 2012). Most elderly ing a great deal of confusion and psychological damage for cannot drive anymore, and find it physically or psychologically many (Preston, 2012). Some of the 160 residents evacuated difficult to take public transports (Parry, 2013). This left them from Bell Harbor in Queens had, for instance, been diagnosed no choice but to stay home or call for help. Many were also liv- with only slight psychiatric illnesses, but during the evacuation ing on the highest floors of buildings. When the power was cut they were mixed with residents who had more severe disorders, off, the elevators often stopped working for days, sometimes and were brought to the Milestone Residence where they were weeks, forcing elderly people to stay home in often precarious obliged to adopt a much more institutional lifestyle (The Associ- conditions. ated Press, 2012). In December, two months later, hundreds of Evacuating nursing homes in the face of Sandy: a tale of mis- nursing home patients were still displaced, living in temporary, management and lack of emergency preparedness overcrowded quarters, sometimes without a regular change of Elderly people living in nursing and care facilities were not clothes. The Bishop Henry B. Hucles Nursing Centre in Brooklyn spared from the hurricane’s damage. A few days before hur- was for instance full at nearly twice its licensed capacity (The ricane Sandy made landfall, Mayor Bloomberg advised 40 Associated Press, 2012). nursing homes located in the mandatory evacuation areas to The confusion was magnified by the fact that the floods cancel plans to move their 3,000 residents to safety (Preston, cut off phone lines, disabling facilities from fulfilling a major 2012). This decision was influenced by the past experience requirement: notifying relatives. Communication continued to of Hurricane Irene in 2011. At that time, officials had ordered be chaotic up to a week after Sandy hit New York, as many fami- evacuation of nursing and adult homes in specific zones. lies still were not aware of where their elderly family members While the hurricane finally caused minor damage, the cost of had been evacuated, or of their health status. No clear indi- evacuation in terms of transportation, health care, and hous- cations had been given as to which phone number should be ing amounted to millions of dollars, leading to broad criticism. called to obtain information, and many relatives were directed Some nursing homes had still not been reimbursed for Irene’s to the wrong shelters (Hallman, 2012). evacuations at the time of Sandy’s landfall. Furthermore, the The disproportionate impact of the storm on the elderly was evacuation of nursing and care homes had proved difficult and caused by this group’s intrinsic vulnerabilities, but also by the dangerous to the physical and psychological health of the resi- insufficient preparedness at the state, local and facility level. dents. However, this time around, the decision not to evacuate The fact that structures designed for this particularly vulner- had disastrous consequences. By the time the intensity of the able group were allowed to be located in low lying areas illus- storm became clear, and the surge projection rose from 1.8 to trate a sense of invulnerability of New Yorkers and authorities 3.4 meters (Wunderground, 2012), it was too late to evacuate to environmental disasters.

IDDRI STUDY 09/2013 23 The State of Environmental Migration 2013

Figure 2. Surge barrier projects

Sources: New York City Office of Emergency Management (2012)

The authority’s reliance on hotels highlights 3. REBUILDING AND that the emergency disaster management and re- ADAPTING: A POLICY PATH covery mechanisms did not take into account the radical impact of income differences on displace- FOR SANDY’S AFTERMATH ment. New York authorities failed to immediately differentiate the needs of the displaced accord- .. Action at various levels ing to their income level, causing a “tale of two Sandys”, with lower-income displaced people Climate change adaptation needs to happen at remaining in precarious living conditions for a different scales. At the household level, it could much longer length of time than households of mean thinking in a vertical manner, a certain higher means. flood requirement level and flood-proofing all the lower floors, with all essential utilities at higher floors (Solecki, ). Building defensive systems is building resilience into public structures. For nearly  years, officials have tried to compensate the erosion of shorelines due to natural causes by building hard structures such as seawalls and

24 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Figure 3. Flap-type barrier model

Source: Parsons Brinckerhoff, PBS Newshour (2012). bulkheads. Experts use the term “New Jerseyza- local marine life, while being high enough to act as tion” to refer to such structures now present on  a buffer against storm surges of a Category  hurri- per cent of the shoreline of New Jersey (Rudolf, cane (Jacobson, ). An advantage of this design ). Although sea walls are particularly well- is that it could be built off site before being laid known for their use in the Netherlands, they down in the river, thereby reducing the costs and can actually be found much closer to New York: economic and wildlife disruptions. However, such in Stamford, a city of , on the East Coast plans would force local areas to build draining sys- (Navarro, ). A five-meter high barrier was tems to deal with water flooding over the banks on built in , protecting an area of about . the sides of the gates (Jacobson, ), amounting square kilometres. The Army Corps of Engineers to an added cost of around USD  billion to USD estimated that it helped prevent USD  million in  billion to shore up the areas on the sides of the damage to houses and businesses (Navarro, ). barriers (Navarro, ). Such successful examples have attracted the In addition to being quite costly, such sea wall attention of engineers and authorities, which have plans also reflect a somewhat short-term adapta- considered applying similar systems to protect tion view. Indeed, even after the debate is won New York. Three main projects were presented in and politicians line up behind the project –Con-  at the American Society of Civil Engineers gressional approval will be needed--funding, conference, to protect New York against Category acquiring permits and conducting environmen-  hurricanes. One of the designs represented tal studies would be a lengthy process. The con- a rolling gates system, north of the Verrazano- struction would very likely not begin in the next Narrows Bridge connecting Long Island and New two decades, leaving New York vulnerable in the Jersey. The USD . billion project envisaged a meantime (Peltz, ). Furthermore, it would complex system of various gates which could be only provide protection for  to  years, as closed during a storm to prevent flood surges to sea levels continue rising (Parry, ). Sea wall affect the City and parts of New Jersey, and could construction projects also assume that current sustain a surge of nine meters –twice as high as climate change, sea-level rise and hurricane Sandy- (Jacobson, ). In normal situations, size predictions are accurate. These models are it would still allow the water to pass, in order to debatable because of the global scale and com- limit the disturbance on the ocean’s tides and on plexity of climate change. Indeed, meteorological the heavy commercial boat traffic. estimations of Sandy’s development as it moved A second design proposed, in addition to the towards the East coast have proved to be very previously mentioned sea wall, the creation of a hesitant at best, highlighting that there is still flap-type barrier at the upper East River. The sea- much to improve in this area. One can thereby wall would include a number of panels that would wonder whether sea walls built on current pro- rest flat on the ground below the water, and could jections will still be able to withstand floods and be raised in the case of a surge. This would limit hurricanes that will have evolved in size and the disturbance for boat traffic as well as to the scope in the coming decades.

IDDRI STUDY 09/2013 25 The State of Environmental Migration 2013

The barriers’ limitations are also evident in the participated in the programme received minimal shorter term: an important question is whether ratings from federal agents, which means that they would be socially fair, as they would cause local authorities had implemented the lowest water levels on the other side to rise approximate- level of adaptation measures to prevent flood ly one foot ( cm), possibly worsening flooding damage (Rudolf, ). Floods due to hurricanes in these areas –the water level could be around  like Sandy are not just “natural disasters,” they are per cent higher on the other side of the barrier (Na- failures to adopt and implement adequate land use varro, ). Furthermore, barrier plans are likely and building code requirements. to raise objections from industries arguing against Flood-proofing subway and road tunnel systems the large disruption in the city’s waterways, as in coastal areas (Jonkman, ), while adopt- well as from residents objecting to the destruction ing more nature-friendly barrier systems such as of the natural landscapes. Barriers would also re- rebuilding the harbour’s salt marshes and oyster duce the exchanges of New York’s estuarine waters beds, that once acted as softener against storms with the sea, worsening water quality and modi- (Jacobson,  and Jonkman, ) could go a fying temperatures and salinity, with potentially long way to increase the city’s resilience. As Mayor negative impacts on rebounding ecosystems and Bloomberg declared after hurricane Sandy: “What local fisheries (Jonkman, ) that benefit the we have to do is learn, and it would be great if you area both economically and in terms of recreation. didn’t put your generators in the basement… The Such policies could create a false sense of security fact that we live close to the water shouldn’t be a that blinds New Yorkers from the increasing risk, surprise to everybody.” (Chaban, ). and thereby prevent authorities from searching for smaller, more flexible and more efficient solutions .. Relocation, the key (Gessner, ). to adaptation?

.. Learning from the past A longer-term adaptation method would be to while preparing for the future encourage the relocation from low lying areas, to higher regions in Queens and Staten Island that are It can be argued that a series of minor interventions currently scarcely populated. In addition to miti- will lead, over time, to efficient and more natural gating the impacts of future storms, such measures systems to deal with climate change and sea-level would be more cost efficient than to repair struc- rise (Peltz, ). In recent years, the armouring tures after every storm. This could be achieved of the coast against erosion has for instance been through economic incentives, with higher taxes supplemented by beach replenishment systems for businesses and residences in vulnerable areas, (Wald, ). An obvious policy would be to and tax incentives for those who move elsewhere disaster-proof infrastructure, to increase their (Young, ). In the past, many coastal towns such resiliency against natural disasters. A large part of as Long Branch in New Jersey actually granted tax Sandy’s damage could have been avoided by adap- abatements for real estate developers to relocate tive measures such as burying power lines in the there. Hurricane Sandy has sent the clear message suburbs and designing closing doors for subways that this type of policy needs to be reversed. Many (Lind, ). This point highlights that recovering residents in Staten Island for instance have been after Sandy is not enough, the city needs to rebuild seeking government buyouts to be able to relo- in a way that decreases the risks of future damage. cate to higher areas. People whose homes were Populations need to return to a new normalcy, destroyed or badly damaged in flood-prone areas where they are actually in a better situation than are not inclined to stay and rebuild, either out of before. Authorities cannot only learn from the past fear that the increasingly frequent storms will take anymore, they have to take into account climate their houses every time or because they have been change related projections, including those of too emotionally affected (The Associated Press, sea-level rise (Solecki, ). Although efforts to ). This is the case in Oakwood Beach, an area increase resilience have been launched in the past, that has been rapidly developed in the last century, they were usually conducted with a short-term and was particularly hit by the storm. In February, view. The National Flood Insurance Programme Governor Mario Cuomo launched a programme Community Rating System, for instance, was through which homeowners in the area could sell introduced in  to encourage local authorities their houses at their pre-Sandy market value. To to upgrade buildings’ structures to decrease the date, In the Fox Beach section of Oakwood Beach, risk from floods, in exchange for deductions on  of  households have signed up to take a residents’ flood insurance premiums. The large buyout if one becomes available (Kaplan, ). majority of the , communities that voluntarily Nevertheless, as the focus has been on rebuilding

26 STUDY 09/2013 IDDRI The State of Environmental Migration 2011

rather than retreating, residents will have to pres- ). Officials in localities that lacked funds have sure authorities to obtain additional funding for tended to prioritize short-term economic benefits, buy-outs, for instance, through the Federal Emer- especially since many are elected for two-years gency Management Agency’s Hazard Mitigation terms. This lack of action on the part of authori- Grant Programme, that was used after Hurricane ties highlights the tension between the objectives Irene. It remains to be seen whether the will to of development and environmental protection for relocate permanently is reflected in the majority of many localities (Bernstein, ). coastal populations, or if it was solely a short-term post-storm reaction from a small share of coastal residents. CONCLUSION Another method to induce relocation from low- lying areas would be to increase insurance costs. Hurricane Sandy was a partly political disaster, Before Sandy, coastal areas benefitted from low caused by the lack of adaptation and mitigation insurance costs, which gave homeowners a false measures. Authorities had received many warn- sense of security. Increasing the individual costs of ings about the high risks of hurricanes, as early living in low-lying areas would enable to transfer as  when environmentalist Suzanne Mattei more of the risk burden to property owners, im- analysed the worst-case scenario of a “combined proving popular risk perception. In , an offi- sea-level rise/storm surge event” (Rudolf, ). cial report by the city’s Panel on Climate Change However, the federal government, the State of highlighted that New York authorities should New York as well as the City generally did not “reduce incentives that increase or perpetuate de- conduct proper adaptation programmes. velopment in high-risk locations.” (Rudolf, ). Hurricane Sandy has shown that economic de- However, this proposal was not adopted by the velopment and environmental protection are not city, that feared that in would lead to a policy of at odds and can be improved together. Political disinvestment and would encourage relocation will needs to be stirred up now, so New Yorkers do away from current urban regions, resulting in not have to live through multiple Sandys before disastrous impacts on the economy. Authorities officials act. Taking projects beyond the research/ underlined that the FEMA’s zone with  per cent development phase would allow them to be fully risk of flooding in NYC was populated with over funded and implemented through cooperation be- , persons, and provided for almost , tween federal, state and local authorities, as well jobs (Rudolf, ). as the private sector. The environment cannot be The example of Ocean Country is quite reveal- considered a stable variable in New York’s urban ing. It was one of New Jersey’s most rapidly devel- planning anymore (Funkhouser, ), and migra- oping areas in terms of population density, with tion from low-lying areas would be a particularly a total population growth of over  per cent be- efficient adaptation strategy to increasing climate tween  and  (J. Rudolf, ). Landscapes risks in New York. Resilience, mitigation and ad- that used to act as natural buffers against storms aptation can be combined to protect New Yorkers were destroyed and populated. This county, which from the impact of future natural disasters. In some issued the highest number of coastal residential cases, the cost of staying in low lying areas surpass- building permits in , was also one of the most es the cost of relocating, as households face higher devastated by the flood. Real estate representa- insurance premiums (Holborn, ), and have to tives have a large influence in the politics of New raise their houses higher above sea-level, at an ap- Jersey and New York, as they have been major proximate cost of USD , to USD , for donors to governors and local leaders, ranking the entire house (Harris, ). Although there is among the top contributors to New Jersey Gover- no legal obligation to lift houses in New York City, nor Christie (Rudolf, ). Before Sandy, there it is recommended to elevate them as a protection was no provision in these states laws requiring in- from future storms. Hurricane Sandy has changed depth review of developments on coastlines, nor cost-benefit calculations, and this should be an op- for buildings to meet reasonable storm and flood portunity for authorities as well as communities to resistance standards. Small developments of fewer adapt in a sustainable way. than  units in vulnerable areas did not require Although resilience and adaptation are im- any form of state approval, leaving the decision portant objectives, New York authorities and in- to local officials. This was specifically the type of habitants could also improve their preparedness projects that were designed throughout the s, to natural disasters. Responses to all phases of s and s, as local officials were more eager evacuation and displacement can be enhanced, to develop their localities in the short term than in order to avoid the confusion and lack of or- to adopt a sustainable long-term view (Bernstein, ganization that followed Sandy. The authorities’

IDDRI STUDY 06/2012 27 The State of Environmental Migration 2013

preparedness and emergency response system was will support the final decision. Everyone having not efficient enough to deal with the immediate a say in the debate makes improves stakeholder human displacement consequences of Sandy, as sense of ownership Third, promoting conversa- shown by the low level of pre-storm evacuations, tions and debates on the issue increases the popu- as well as the precariousness of the shelter system, lation’s awareness of the risks, making them more highlighting the dire lack of space for New York likely to make sound, long-term decisions rather City to welcome displaced persons. As hurricanes than short-term ones based on pre-conceived no- become stronger and more frequent, authorities tions or political inclinations. could also improve their policies targeted at par- Migration–—and more precisely displacements ticularly vulnerable groups such as the elderly and relocations—has been one of the main, and and the low-income households. Services such as yet least-reported, consequence of Hurricane special phone lines for the elderly would be use- Sandy. It has been essential to the short-term re- ful, to enable them to call for help if they do not sponse and remains crucial to the long-term ad- have any family or friends help them to safety. Past aptation process. Technology, high income levels censuses could be particularly used to identify the and stable institutions are likely to mitigate the areas where many vulnerable people live, enabling consequences of disasters. However, the central- authorities to increase their activity and presence ity of migration during and after Hurricane Sandy in such places in the case of a natural disaster – highlights that environmentally-induced displace- particularly before the disaster strikes. ments are not restricted to developing countries or Responses to long-term displacements have yet to lower-income cities like New Orleans, as natural to be created to deal with displaced families whose disasters strike regardless of the level of develop- houses have been destroyed. Many programs were ment. The impacts of Hurricane Sandy further- created after Sandy hit without proper coordina- more underscore that relocation is not necessarily tion or prior reflexion, leading to uneven and of- a consequence of disasters, but should also be con- ten poor responses. The FEMA’s hotel program is a sidered as an opportunity to adapt to a changing perfect illustration. world, through relocation to safer, less vulnerable Finally, the efficiency of preparedness, resilience areas. and adaptation measures relies on the authorities’ communication strategy. In terms of preparedness, Picture 1. After hurricane Sandy enhancing the communication with communities and households would allow for easier and more efficient management of evacuations, through a better awareness of evacuation zones, of the risks involved, and of available means of transportation. Working with local associations, NGOs and religious groups, among others, could prove extremely useful –many households are more likely to be persuaded by their local pastors than by official messages on T.V. Improved communication with local communities is also fundamental in the adaptation process, for three main reasons. First, locals can prove very useful to develop efficient adaptation measures, as they usually know their locality and its geographical and demographical characteristics very well. Second, involving communities in the process and taking their concerns into consideration increases the chances that they Source: Coastal America, 2013.

28 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Monitz, Gary, 2013. “Lessons from Bangladesh for a post- BIBLIOGRAPHY Sandy New York”, International Institute for Environment and Development, January 22nd 2013, http://www.iied.org/ B  R lessons-bangladesh-for-post-sandy-new-york. Murphy, Jarrett, 2012, “Why NYC Is So Vulnerable to Hur- Holborn, 2012. “Superstorm Sandy”, Holborn Perspectives, ricanes”, CityLimits, November 1st 2012, http://www.cityl- th December 12 2012, http://www.holborn.com/Data/ imits.org/blog/blog/224/why-nyc-is-so-vulnerable-to-hur- Sites/1/pdfcontent/2012_SuperStormSandy.pdf ricanes#.UV8CCY7aOC4m Hydrometeorological Prediction Center (the), 2012. “Storm Piore, Adam, 2012. “Mailman School Disaster Expert Ir- Summary Message”, National Centers For Environmental win Redlener on the Lessons of Hurricane Sandy”, Colum- Predictions, November 8th 2012, http://www.webcitation. bia University of the City of New York, November 28th 2012, org/6C1o51YlE http://news.columbia.edu/research/2992 Orton, Phillip, 2012. “Updated Forecasts from SSWS and Solecki, William, 2012. “Sandy and urban systems, infra- ET-Surge”, Sea and sky NY, October 29th 2012, http:// structure and vulnerabilities”, Department of Geography & seaandskyny.com/page/2/ CUNY Institute for Sustainable Cities - Hunter College, http:// www.geo.hunter.cuny.edu/news/Sandy%20and%20 urban%20systems.pdf A Young, Rob, 2013. “Hurricane Sandy Relief Bill Fails to Face Bernstein, Fred A., 2012. “It Happened Here: Hurricane Coastal Realities”, Yale Environment 360, January 7th 2013, Sandy’s Impact,” Architectural Record, December 2012, http://e360.yale.edu/feature/hurricane_sandy_relief_ Vol. 200 Issue 12, p19. bill_fails_to_face_coastal_realities/2606/ Cameron, Charley, 2013. “Sociologist Eric Klinenberg Dis- cusses How We Can Better Adapt for Extreme Weather Dis- P A asters in This Week’s New Yorker”, Inhabitat New York City, January 8th 2013, http://inhabitat.com/nyc/sociologist- Associated Press, The, 2012. “Nor’easter bears down on eric-klinenberg-discusses-how-we-can-better-adapt-for- Sandy-battered NYC and New Jersey”, Fox News, November extreme-weather-disasters-in-this-weeks-new-yorker/ 7th 2012, http://www.foxnews.com/weather/2012/11/07/ Center for International Earth Science Information Net- sandy-battered-nyc-and-nj-prepare-for-noreaster/. work, 2012. “Sandy’s Surge Affected More Than 1.4 Million Associated Press, The, 2012. “Hundreds of elderly and disa- in 11 States”, Blogs from the Earth Institute, November 12th bled New Yorkers evacuated during Hurricane Sandy are 2012, http://blogs.ei.columbia.edu/2012/11/12/hurri- still homeless”, Daily News, December 26th 2012, http:// cane-sandys-storm-surge-affected-more-than-1-4-million- www.nydailynews.com/new-york/elderly-disabled-sandy- coastal-residents-in-11-states/ victims-homeless-article-1.1227517 Cooper, Matthew, 2005. Future sea level rise and the new Associate Press, The, 2013. “Superstorm Sandy renews Jersey Coast, Science, Technology and Environmental Policy debate of whether to rebuild or relocate”, NOLA, Janu- Program, Woodrow Wilson School of Public and Internation- ary 21st 2013, http://www.nola.com/hurricane/index. al Affairs, Princeton University, November 2005, http:// ssf/2013/01/rebuild_or_relocate_superstorm.html www.princeton.edu/step/people/faculty/michael-oppen- Avent, Ryan, 2012. “Costs to come”, The Economist, October heimer/recent-publications/Future-Sea-Level-Rise-and- 31st 2012. the-New-Jersey-Coast-Assessing-Potential-Impacts-and- Barron, James, 2011. “With Hurricane Irene Near, 370,000 Opportunities.pdf in New York City Get Evacuation Order”, The New York Times, Funkhouser, David, 2012. “‘This is a wake-up call – don’t August 26th 2011, http://www.nytimes.com/2011/08/27/ hit the snooze button”, The Earth Institute of Columbia nyregion/new-york-city-begins-evacuations-before-hurri- University, November 6th 2012, http://blogs.ei.columbia. cane.html?pagewanted=all edu/2012/11/06/this-is-a-wake-up-call-dont-hit-the- Barron, James, 2012. “With $200 Million in U.S. Housing snooze-button/ Aid, Officials Begin Relocating the Displaced”, The New Huang, Al, 2012. “Hurricane Sandy’s Disproportionate York Times, November 5th 2012, http://www.nytimes. Impact on NYC’s Most Vulnerable Communities”, Natural com/2012/11/06/nyregion/housing-relocation-begins- Resources Defense Council Staff Blog, November 15th 2012, after-hurricane-sandy.html?_r=0 http://switchboard.nrdc.org/blogs/ahuang/hurricane_ Bernstein, Nina, 2012. “Storm Bared a Lack of Options for sandys_disproportion.html the Homeless in New York”, The New York Times, November Kahn, Brian, 2012. “Superstorm Sandy and Sea Level 20th 2012, http://www.nytimes.com/2012/11/21/nyre- Rise”, National Oceanic and Atmospheric Administration, gion/storm-worsens-shortage-of-nyc-housing-for-homeless. November 5th, 2012, http://www.climatewatch.noaa.gov/ html?pagewanted=all&_r=0 article/2012/superstorm-sandy-and-sea-level-rise. Bonamo, Mark J., 2012. “Decision to move families dis- Mazelis, Fred, 2012.” Humanitarian Crisis in the Wake placed post-Sandy into Fort Monmouth -a godsend”, New of Hurricane Sandy”, Centre for Research and globaliza- Jersey News, November 12th 2012http://www.nj.com/mon- tion, November 6th 2012, http://www.globalresearch. mouth/index.ssf/2012/11/decision_to_move_families_ ca/humanitarian-crisis-in-the-wake-of-hurricane- displaced_post-sandy_into_fort_monmouth_a_godsend. sandy/5310820 html

IDDRI STUDY 09/2013 29 The State of Environmental Migration 2013

Bronars, Stephen, 2012. “70,000 Workers Displaced by Hur- Jaffe, Eric, 2013. “Why New York’s Sandy Commission Rec- ricane Sandy in New Jersey: Unemployment Rate May Reach ommendations Matter”, Mother Jones, January 15th 2013, 11%”, Bronars Economics, November 29th 2012, http:// http://www.motherjones.com/environment/2013/01/ sbronars.wordpress.com/2012/11/29/70000-workers-dis- new-yorks-sandy-commission-recommendations-matter- placed-by-hurricane-sandy-in-new-jersey-unemployment- climate-change rate-may-reach-11/ Jones, Malcolm, 2012. “What to Do With the Thousands Burdeau, Cain, 2012. “Last FEMA trailer leaves 6 years after Kat- Displaced by Hurricane Sandy?”, The Daily Beast, No- rina”, The Washington Times, February 15th 2012, http://www. vember 8th 2012, http://www.thedailybeast.com/arti- washingtontimes.com/news/2012/feb/15/last-fema-trailer- cles/2012/11/08/what-to-do-with-the-thousands-dis- leaves-6-years-after-katrina/ placed-by-hurricane-sandy.html Chaban, Matt, 2012. “When It Comes to Protecting New York Jonkman, Bas, 2012. “A Wake-Up Call to Consider the Op- From the Next Hurricane, Mayor Bloomberg Suggests You tions”, The New York times, November 4th 2012, http:// Fend for Yourself”, The Observer, November 1st, 2012, http:// www.nytimes.com/roomfordebate/2012/11/01/should- observer.com/2012/11/when-it-comes-to-protecting-new- new-york-build-sea-gates/hurricane-sandy-could-force- york-from-the-next-hurricane-mayor-bloomberg-suggests- new-york-to-consider-flood-prevention you-fend-for-yourself/ Kaplan, Thomas, 2013. “Cuomo Seeking Home Buy- Coastal America, 2013. “NOAA, USGS: Climate change impacts outs in Flood Zones”, The New York Times, February 3rd, to U.S. coasts threaten public health, safety and economy”, Coast- 2013, http://www.nytimes.com/2013/02/04/nyre- al America A partnership for Action, January 29th 2013, http:// gion/cuomo-seeking-home-buyouts-in-flood-zones. coastalamerica.wordpress.com/tag/hurricane-sandy/ html?pagewanted=all Dawsey, Josh, 2013. “Sandy Victims Still in Hotels By Thou- Lind, Michael, 2012. “Hurricane Sandy and the Case for Ad- sands”, The Wall Street Journal, February 5th 2013, http://on- aptation to Climate Change”, The Breakthrough, November line.wsj.com/article/SB100014241278873249002045782864 2nd 2012, http://thebreakthrough.org/index.php/voices/ 82091427450.html. michael-lind/hurricane-sandy-and-the-case-for-adapta- Drye, Willie, 2012, “A Timeline of Hurricane Sandy’s Path of tion-to-climate-change/ Destruction”, National Geographic, November 2nd 2012, http:// Mathis, Sommer, 2012. “Hurricane Sandy Provokes the newswatch.nationalgeographic.com/2012/11/02/a-timeline- Largest Preemptive Public Transportation Shutdown in of-hurricane-sandys-path-of-destruction/ U.S. History”, The Atlantic Cities, October 29th 2012, www. Durkin, Erin, 2013. “City officials defend decision not to evacu- theatlanticcities.com/commute/2012/10/hurricane- ate hospitals and nursing homes before Hurricane Sandy”, sandy-provokes-largest-preemptive-public-transportation- Daily News, January 24th 2013, http://www.nydailynews.com/ shutdown-us-history/3727/ new-york/officials-defend-move-evacuate-hospitals-sandy-arti- Navarro, Mireya, 2012. “Weighting sea barriers as protec- cle-1.1247397 tion for New York”, The New York Times, November 7th 2012, Fischetti, Mark, 2012. “Sandy versus Katrina, and Irene: Mon- http://www.nytimes.com/2012/11/08/nyregion/after- ster Hurricanes by the Numbers”, Scientific American, Octo- hurricane-sandy-debating-costly-sea-barriers-in-new-york- ber 29th 2012, http://www.scientificamerican.com/article. area.html?pagewanted=all&_r=2&. cfm?id=sandy-vs-katrina-and-irene MacDonald, Jeffrey G., 2012. “Defying Hurricane Sandy Gessner, David, 2012. “Sandy recovery: Walls won’t stop su- evacuation orders prompts moral debate”, NOLA, Novem- th perstorms“, Salon, November 6th 2012, http://www.salon. ber 15 2012, http://www.nola.com/hurricane/index. com/2012/11/06/sandy_recovery_walls_wont_stop_super- ssf/2012/11/defying_hurricane_sandy_evacua.html storms/ McKelvey, Wallace, 2013. “Families displaced by Hurricane Gupta, Prachi, 2012. “Mayor Bloomberg orders mandatory evac- Sandy face competition for housing as summer visitors ar- th uation for 375,000 New York City residents”, Salon, October 28th rive”, Press of Atlantic City, March 16 , 2013. 2012, http://www.salon.com/2012/10/28/mayor_bloomb- Nye, James; Stebner, Beth, 2012, “Here comes Sandy! erg_orders_mandatory_evacuation_for_375000_new_york_ State of emergency declared across New York as Halloween city_residents/ ‘Frankenstorm’ roars towards East Coast”, The Mail Online, th Hallman, Ben, 2012. “After Sandy, Communication Break- October 26 2012, http://www.dailymail.co.uk/news/ down Hampered Efforts to Find Evacuated Seniors”, article-2223047/Hurricane-Sandy-Forecasters-warn-pre- Huffington Post, November 16th 2012, http://www.huff- Halloween-perfect-storm-strike-East-Coast.html ingtonpost.com/2012/11/16/sandy-communication-evacuat- O’Connor, Julie, 2012. “After Hurricane Sandy, understand- ed-seniors_n_2141699.html ing a disaster’s worst impacts: Q&A”, New Jersey News, th Harris, Elizabeth A., 2013. “Going Up a Few Feet, and Hoping to November 11 2012, http://blog.nj.com/njv_editorial_ Avoid a Storm’s Path”, The New York Times, April 15th 2013, http:// page/2012/11/after_hurricane_sandy_understa.html www.nytimes.com/2013/04/16/nyregion/after-hurricane- Parry, Wynne, 2013. “Future Disasters: 10 Lessons from sandy-homeowners-elevate-property.html?pagewanted=all Superstorm Sandy”, LiveScience, January 28th 2013, http:// Jacobson, Rebecca, 2012. “Engineers Consider Barriers to www.livescience.com/26640-future-disasters-lessons-su- Protect New York From Another Sandy”, PBS Newshour, No- perstorm-sandy.html vember 20th 2012, http://www.pbs.org/newshour/run- Parry, Wynne, 2013. “Why Disasters Like Sandy Hit the El- down/2012/11/engineers-draw-barriers-to-protect-new-york- derly Hard”, LiveScience, March 8th 2013, http://www.lives- from-another-sandy.html. cience.com/27752-natural-disasters-hit-elderly-hard.html

30 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Peltz, Jennifer, 2012. “NYC Sea Barrier Could Have Stopped Traub, James, 2012. “Adapt or Die,” Foreign Policy, No- Surge During Hurricane Sandy, Say Experts”, The Huffing- vember 2nd 2012, http://www.foreignpolicy.com/arti- ton Post, November 1st 2012, http://www.huffingtonpost. cles/2012/11/02/adapt_or_die?page=0,0 com/2012/11/01/nyc-sea-barrier-could-have-stopped- Wald, Matthew, 2013. “Storm Panel Recommends Major surge_n_2057981.html. Changes in New York”, The New York Times, January 6th Preston, Jennifer, 2012. “Behind a Call That Kept Nursing 2013, http://www.nytimes.com/2013/01/07/nyregion/ Home Patients in Storm’s Path”, The New York Times, De- new-york-state-storm-panel-recommends-major-changes. cember 2nd 2012, http://www.nytimes.com/2012/12/03/ html?_r=1& nyregion/call-that-kept-nursing-home-patients-in-sandys- Zezima, Katie, 2013. “For thousands displaced by Hur- path.html?pagewanted=all&_r=0 ricane Sandy, anxious lives in hotels and motels”, Press of Rosenthal, Sandy, 2012. “Should Flood Survivors of Hur- Atlantic City, January 25th 2013, http://www.pressofatlan- ricane Sandy Relocate?”, Huffington Post, November 2nd ticcity.com/news/breaking/for-thousands-displaced-by- 2012, http://www.huffingtonpost.com/sandy-rosenthal/ hurricane-sandy-anxious-lives-in-hotels/article_550a1ece- should-flood-survivors-of_b_2067613.html 6745-11e2-9b70-001a4bcf887a.html Rudolf, John, 2012. “Hurricane Sandy Damage Ampli- fied By Breakneck Development Of Coast”, Huffington W Post, November 12th 2012, http://www.huffingtonpost. com/2012/11/12/hurricane-sandy-damage_n_2114525. FEMA. http://www.fema.gov/ (Consulted in December html?ncid=txtlnkushpmg00000040 2012, February 2013 and April 2013) Sharp, Tim, 2012. “Superstorm Sandy: Facts About the National Oceanic and Atmospheric Administration, http:// th Frankenstorm”, Livescience, November 27 2012, http:// www.noaa.gov/ (Consulted in December 2012, January www.livescience.com/24380-hurricane-sandy-status-data. 2013 and April 2013). html NASA, http://www.nasa.gov/ (Consulted in January, Feb- Silverman, Rena, 2012. “Why New York City Is the Worst ruary and April 2013). Place for a Hurricane”, National Geographic, Novem- ber 1st 2012, http://news.nationalgeographic.com/ New York City Office of Emergency Management, http:// news/2012/11/121101-new-york-city-sandy-hurricane- www.nyc.gov/html/oem/html/home/home.shtml, (Con- bight-science-environment-nation/ sulted in November, December 2012, January, March 2013) Synergy Home Care, 2012. “Hurricane Sandy Hurts Sen- Weather Undergound, http://www.wunderground.com/ iors”, Synergy Home Care Official Blog, http://www.syn- hurricane/at19034.asp (consulted on April 17th 2013) ergyhomecare.com/blog/index.php/2012/11/hurricane- sandy-highlights-vulnerability-of-senior-population/ Tollefson, Jeff, 2012. “Hurricane sweeps US into climate- adaptation debate”, Nature, November 6th 2012, http:// www.nature.com/news/hurricane-sweeps-us-into-climate- adaptation-debate-1.11753

IDDRI STUDY 09/2013 31 The State of Environmental Migration 2013

IN FOCUS Migration and the Environment: The case The housing damage resulted in internal displacement, as over 33,760 families, including 7,450 families already in IDP camps, lost their homes. of Hurricane Sandy in Haiti Most of these people were accommodated by friends, family or neighbors. Others stayed in evacuation shelters. Over 2,000 of these people could not Juliette Cody return home as their houses and lands had been destroyed and did not have access to alternative shelter solutions (FAO 2012:7). They stayed with fami- From 23 to 26 October 2012, the western and southern regions of Haiti were lies, friends or in temporary shelters at the site of their destroyed houses heavily affected by Hurricane Sandy. After a drought and Tropical Storm for several months, until they were able to rebuild or repair their houses Isaac, Hurricane Sandy was the third natural disaster to strike the country (Mellicker 20132). in 2012. Haiti was still recovering from the lingering impacts of the 2010 There is a general consensus that, on the long run, Hurricane Sandy’s earthquake. Over 347,000 people remained in 450 Internally Displaced Per- most important effect on migration relates to food security (Mellicker sons (IDP) sites – i.e. camps and more ad-hoc facilities hosting evacuees - 20133). In order to cope with the severe damage inflicted on agricultural at the end of 2012 (IOM 2013:1). Cholera, gender-based violence, and food output, crops and cattle, some people used migration as a positive adapta- insecurity, among other issues, remained prevalent. tion strategy, moving to less affected rural areas or to urban areas. It is Preparedness and preventive evacuation also assumed that migration towards the Dominican Republic increased Prior to and during Hurricane Sandy, the Government of Haiti issued suc- significantly (Gitau 2013). No mechanisms so far have tracked these popu- cessive Orange and Red Alerts and activated the National Risk Manage- lation movements and it is therefore hard to evaluate the numbers of people ment and Disaster Plan (PNGRD) (FAO 2012:3). The Haitian Government, concerned. led by the Civil Protection Directorate and with the support of the Haitian Conclusion Red Cross, Haitian Civil Society and international partners, worked to put The success of the strategy implemented prior to and during Hurricane into practice a preparedness strategy, involving awareness-raising among Sandy in Haiti demonstrated the importance of preparedness and pre- vulnerable populations to natural disasters and a “newly-developed sys- ventive evacuation to decrease vulnerability. It underlined the necessity tem of short-term, preventive evacuation” for the Port-au-Prince metro- to consolidate such strategies by improving disaster preparadedness- politan area, building on the success of the first preventive evacuation mechanisms such as tracking mechanisms, monitoring, and alert raising organized in response to Tropical Storm Isaac (Mellicker 2013:1). Imple- systems. The damage resulting from the storm also emphasized the need menting such a strategy was initially challenged by the reluctance of a few for mitigation measures, as well as the necessity to limit environmental international actors and vulnerable people to evacuate (Gitau 2013). There degradation in Haiti. were concerns that it would lead to long-term displacement or increase insecurity and theft. None of these scenarios took place and the preventive evacuation was considered to be successful, thus earning support both from the evacuees and the international partners. The Haitian National References Agencies maintained leadership and received assistance from their inter- Email correspondence with Bradley Mellicker, IOM Programme Manager - Disaster Risk Reduction / Emergency Preparedness and Response, 25 April – 3 May 2013, 23-27 August national partners, which were extensively involved in funding and building 2013 the shelters used for preventive evacuation. The international community Interview with Rosalia Gitau, ex-IOM Officer Liaison Officer & Associate Legal Officer (Land also supported evacuations and provided the evacuees with humanitarian and Property), conducted on 3 May 2013 assistance to meet their basic needs. Over 20,000 people would have been FAO, Haiti: Emergency Appeal Hurricane Sandy 2012, 2 November 2012 (Consulted April- evacuated to 136 shelters (IOM 2012, p1). Some vulnerable families also August 2013) http://www.fao.org/fileadmin/templates/tc/tce/pdf/Emergency_Appeal_ evacuated to other types of shelter, for example, with host families. Most Hurricane_Sandy_Haiti.pdf evacuees left the shelters in the days after the passing of Hurricane Sandy International Organization for Migration, Displacement Tracking Matrix Report January (Mellicker 20131). 2013 http://iomhaitidataportal.info/dtm/index2.aspx (Consulted March-August2013) Migration as a response to damaged homes and livelihoods International Organisation for Migration Displacement, Crise Alimentaire (presentation HTC), March 13 2013 Despite evacuations, Hurricane Sandy heavily impacted the Haitian popu- Mellicker Bradley, “Showing Haiti Skeptics the Error of Their Ways.” In Migration the Big lation. Severe flooding and landslides resulted from the violent winds and Issue, 19 February 2013 http://weblog.iom.int/blog/2013/02/19/showing-haiti-skeptics- heavy rain and caused immediate and significant damage to houses, error-their-ways (consulted on April-August 2013) public infrastructure and crops. The combination of Hurricane Sandy, the Système National de Gestion des risques et des dégâts et EShelter & CCCM Cluster, Oura- drought and Tropical Storm Isaac led over 3.44 million people to be affected gan Sandy, Rapport de Situation #9, November 10 2012 (Consulted April-August 2013) by food insecurity (IOM 2013:2).

2. Correspondence 1. Correspondence 3. Correspondence

32 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

TYPHOON BOPHA AND PEOPLE DISPLACEMENTS IN THE PHILIPPINES ULRIKE JULIA WENDT

1. INTRODUCTION will follow. It is important to note that, at the time of writing in , policy responses are still being Typhoon Bopha, locally known as Pablo, made implemented and are changing. Policy recommen- landfall in the southern part of the Philippines on dations are discussed in the conclusion.  December . It was not only the sixteenth and strongest typhoon that hit the country in , but it was also the most powerful and southern typhoon 2. THE PHILIPPINES: A COUNTRY that was recorded in the Western Pacific in the FAMILIAR WITH NATURAL past  years, a worrying indicator how climate change could affect the Philippines in severe DISASTERS AND CONSEQUENTIAL ways (IOM Philippines, ). Considering that PEOPLE DISPLACEMENTS the Philippines is a country familiar with natural disasters and consequential people displacements, The Philippines is one of the most disaster-prone this paper examines the extent to which Typhoon countries in the world. According to the World Risk Bopha has caused large-scale displacements of Report , it was ranked third out of  coun- people and significantly contributed to decreased tries, meaning the third most exposed and vulner- standards of living. In so doing, the paper will able country to disasters (Alliance Development refer to “displaced people” as those who have been Works, : ). Between  and ,  forced or obliged to leave their homes as a result natural disasters occurred in the country, which of the threats or impacts of Typhoon Bopha. This is nearly a third of the total number of disasters comprises of people who were evacuated or fled by that occurred in the South-East Asian region in themselves before or after the onset of the event. this period (GFDRR, : ). Typhoons are the It also includes situations where people lost their most common and detrimental disasters. Between homes or livelihoods and now stay in different June and December, an average of  typhoons sites (Yonetani : ). hit the country annually, usually accompanied by The first section of the paper will provide a short intense rainfall and flooding, causing the death of background about the Philippines’ long history hundreds of people and damaging the country’s of environment-induced displacements. It then economy significantly (GFDRR, : ). elaborates on Typhoon Bopha and its detrimental impacts, which forced many people to leave their .. The Philippines homes before and after the storm. The next section geographical vulnerabilities examines these displacements in greater detail. Subsequently, a critical study of the major policy Natural disasters strike the Philippines due to responses that have been implemented to address various reasons. Its geographical situation contrib- the particular needs of the displaced population utes to the high vulnerability that is further exacerbated by the risks resulting from climate change. In . The description used in this paper to identify “displaced fact, the Philippines have recorded its most detri- people” is informed by the United Nations Guiding mental, largest and deadliest typhoons in the past Principles Document that defines Internally Displaced  years (GFDRR, : ). Furthermore, envi- People (IDPs) as “persons or groups of persons who have been forced or obliged to flee or to leave their homes or ronmental degradation has increased the coun- places of habitual residence, in particular as a result of try’s risks of disasters. Poor land-use, planning and or in order to avoid the effects of armed conflict, situa- population growth contribute to the enormous tions of generalized violence, violations of human rights or natural or human-made disasters, and who have not damage of the environment and lead to the defor- crossed an internationally recognized state border”. estation of large regions (GFDRR, : ). As a This paper refers to people that were displaced due to result, landslides and flash flooding have increased Typhoon Bopha without differentiating between those and some areas that have significantly lost their who have been displaced internally or across border. The available data suggests that the majority of people dis- woodland have become more vulnerable to the placed due to Bopha have been displaced internally. damage caused by typhoons (GFDRR, : ).

IDDRI STUDY 09/2013 33 The State of Environmental Migration 2013

The high occurrence of natural disasters has , the Philippines have supported the Hyogo caused the displacement of millions of people in Framework for Action (HFA), which is a ten-year the Philippines. Between  and , more plan that aims to explain, illustrate and specify than twelve million people were displaced due the work that is needed from different fields and to natural disasters (Yonetani : ). After Ty- stakeholders to better address issues related to phoon Washi, locally known as Sendong, hit the disasters (NDRRMP, : ). In , the Filipino country in December , about  per cent of the Government signed the Philippines Disaster Risk population of the city Cagayan de Oro and about Reduction and Management Act (PDRRM) that  per cent of the population of the city Iligan in aims to reform the country’s main laws and poli- the country’s southern part were displaced (Gin- cies in line with the HFA (Ginnetti et al., : ). netti et al., : ). Clearly, these numbers help In order to specifically address issues related to explain why the Internal Displacement Monitoring Internally Displaced Persons (IDPs), the country Centre (IDMC) and the Norwegian Refugee Coun- has formulated a number of laws and bills, such as cil (NRC) have identified the Philippines as one of a bill to protect the rights of IDPs that was passed the countries with the highest number of people in February  (UNHCR, ). The PDRRM has displaced due to natural disasters. In , they since included a provision on IDPs as well (Ginnetti have ranked the country third; in , seventh; et al., : ). and in , fourth (Yonetani, : , , ; Yon- Despite the governmental attention given to etani : ). DRR and IDPs and the consequent adoption of do- A number of factors highly increase the risks of mestic policies based on international standards, the displacement of people. For example, homes Typhoon Bopha has caused large-scale people dis- and buildings are often not built to the same placements and significantly contributed to their standards as in developed countries (Ginnetti et decreased standard of living. al., : ). Consequently, the risk of them being destroyed is far greater and can lead to the people’s displacement. Poor urban governance and 3. THE TYPHOON AND ITS weak political accountability are the underlying DETRIMENTAL IMPACTS reasons for this because they often lead to unsuccessful and/or unenforced land planning and building standards (Ginnetti et al., : ). Fur- .. Formation and path thermore, Ginnetti et al. (: ) cite that a lack of of Typhoon Bopha understanding of the effects of climate change and unsatisfactory early-warning systems for natural The origins of Typhoon Bopha were first sighted disasters contribute to the country’s high risk of around  November  (Gutro, ). Bopha people’s displacement. formed in an uncommon and distinctive manner because it developed fewer than five degrees north . Policy adaptation to of the equator (Met Office, ). This is an area the high level of risks where tropical storms do not typically emerge due to the low Coriolis force that helps typhoons to Given this high risk of natural disasters and its form by causing their rotation (Haeseler, ). consequences on people’s displacement, the Phil- In the following weeks, Bopha steadily increased ippines have enacted a variety of policies and legis- in its size and intensity. On  December, it hit the lations on disaster risk reduction (DRR). In fact, southern part of the Palau archipelago (around similar efforts have a long history in the country. , km east of the Philippines) where it devas- As a colonial power starting in the th century, tated several areas, caused power outages but no Spain set up a record system to keep informa- casualties (OCHA, a). tion on environmental events. Interestingly, the On  December at . am, Bopha made landfall records of the environmental events that occurred over Bagaga in Davao Oriental, which is in eastern between  and  have informed today’s Mindanao in the southern part of the Philippines early warning systems (Ginnetti et al., : ). (NDRRMC, ). Bopha recorded wind strengths Since the end of the th century, laws on disas- of about  km/h in its center, gusts of up to  ters associated with natural and human-induced km/h and rainfall of - mm per hour within its hazards and institutions on disaster manage- diameter of  km (IOM Philippines, a). This ment have increasingly been enacted. In , was three times the wind strength and twice the President Marcos issued Presidential Decree  rainfall than Tropical Storm Washi had brought that remains the basis for the country’s disaster to the Philippines in  (OCHA, : ). In the management (Ginnetti et al., : ). Since following days, Bopha traversed the landmasses

34 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 1. The path of Typhoon Bopha

Source: OCHA 2013: iv. Map provided courtesy of the UN Office for the Coordination of Humanitarian Affairs. The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the United Nations

IDDRI STUDY 09/2013 35 The State of Environmental Migration 2013

of Mindanao and turned north while it weakened on  December, to  on  December and to , (NDRRMC, ). As noted earlier, Bopha moved on  December (IOM Philippines, ; OCHA, along the most southern course of a typhoon that b; OCHA, c; OCHA, d; OCHA, e). has been recorded in the Western Pacific in the Many factors explain this increase in numbers: it past  years. Bopha was downgraded to a tropi- took several days until Bopha completely left the cal depression when it left the Philippine area on  country and thereby killed people in the days after December (NDRRMC, ). The typhoon’s great  December. In addition, Bopha hit many remote strength and its uncommon trajectory was a ma- communities where it was challenging to obtain jor reason why it had a detrimental impact on the comprehensive data on causalities. Mindanao region and its population. Bopha affected various specifically vulnerable groups. Some reports revealed that up to  per .. Affected regions cent of the affected population lived below the poverty line (Protection Cluster, ). Poor peo- With high winds along coastal areas, mudslides ple were often dependent on income generating in highland villages and flash floods in low-lying activities, such as subsistence farming, and lived areas, Bopha damaged vast regions during its in houses built with less robust materials (IFAD, course through Mindanao (Shelter Cluster, ). : ). Given that Bopha particularly destroyed In particular, it affected the eastern part of the farmland and housing, which will be later ex- island, which is one of the poorest regions of the plained in greater detail, poor people were excep- Philippines. About . million people live in this tionally vulnerable to the typhoon. Other reports area that encompasses the Davao and Caraga using different criteria to identify the most af- region (OCHA, : ). Out of the country’s  fected people disclosed that more than  per cent regions, these regions are respectively the th ( were indigenous people (Protection Cluster, ). persons per square meter) and th ( persons Indigenous people frequently live in the isolated per square meter) most populated areas (National regions of Mindanao. Therefore, they may not Statistics Office ). Specifically, Bopha had have been informed early enough to evacuate be- detrimental effects on the provinces of Davao fore Bopha’s landfall and were less likely to obtain Oriental, Compostela Valley, Suriago del Sur and humanitarian assistance in a relatively quick man- Agusan del Sur, each of which it destroyed and ner (Protection Cluster, ). Women and chil- impacted in their particular way. For example, dren also belonged to the most affected groups. Davao Oriental especially suffered from the UNICEF () reported that about one third of all damages to its agricultural lands and Compostela affected people were children. They often experi- Valley experienced a particularly high number of enced profound stress due to the death or injury casualties. The high level of deforestation that of their loved ones, which can have significant ef- characterized Mindanao before Bopha may have fects on their well-being and future development contributed to the extent to which these effects (UNICEF, ). Furthermore, many women were were so damaging (Morrison, ). affected because they had become the main provider for their family. Their men had often left .. Human cost of Bopha after the storm to look for employment elsewhere (Protection Cluster, ). Bopha affected over . million people (OCHA, a: ). According to the Inter-Agency Standing .. Agriculture damages Committee (n.d.), this includes “all people whose lives have been affected in some way by the crisis” Bopha had devastating impacts on the agriculture beyond those in need of humanitarian assistance. of eastern Mindanao. National authorities reported More specifically, the typhoon left , people that thousands of hectares of banana and coconut dead,  missing and , injured (OCHA, plantations were destroyed, mainly in the provinces e, OCHA, a: ). These numbers have of Compostela Valley and Davao Oriental (DSWD et constantly and drastically risen since Bopha hit al., ). Small growers were particularly affected the country on  December. The reported number and had lost between  and  per cent of their of affected people increased from about , total farming areas (DSWD et al., ). The overall on  December, to  million on  December, and damage costs to agriculture were estimated to be at to . million on  December. Similarly, the about USD  million (OCHA, f). number of people killed increased from about  Before the typhoon, the majority of the population of eastern Mindanao was dependent on . Interview with IOM Communications Specialist, con- subsistence farming of basic commodities, includ- ducted on  March  via Skype. ing coconuts, bananas and rice (OCHA, : ).

36 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Picture 1. Aftermath of Typhoon Pablo (Bopha)

Credits: Cateel, Davao Oriental, Creative Commons, Photo: Sonny M. Day, 5 December 2012. Available: http://www.flickr.com/photos/89715795@N02/8289315117/in/pho- tostream/, Accessed: 27 August 2013., Accessed: 21.05.2013.

In addition to growing parts of their own food, even iron roofs had been carried through the wind households purchased food items from local mar- like ”flying machetes” (Mullen, ). Damages to kets (OCHA, : ). However, Bopha had dam- infrastructure were estimated to amount to over aging impacts on this mixed food economy. After USD  million (OCHA, ). Power and elec- its landfall, many people were no longer able to tricity were only slowly restored in the following grow their own food and buy food from the mar- months, which may have contributed to communi- kets. Not only was their agricultural land com- ties being slowly informed of the current situation pletely destroyed, but also market supply chains (OCHA, : ). had been cut off (OCHA, : ). Furthermore, the damages to agricultural lands contributed to .. Education, health a loss of income for many people. People had to and safety impacts look for other jobs elsewhere to make their living; however, these coping strategies were often not Large numbers of schools were destroyed or used sufficient to cover all food needs (OCHA, : - as evacuation centers, which made it impos- ; DSWD et al., ). These detrimental impacts sible for many children to attend school (Center contributed to food insecurity and dramatically for Disaster Philanthropy, ). Indeed, the increased the need of food assistance. Addition- Government was able to re-establish the daily ally, given that crops like banana and coconut take school routine within only one month. However, years to grow, the gravity of the agricultural dam- many challenges remained, such as the lack of ages is tremendous (Shelter Cluster, ; OCHA, electricity, school records and teaching mate- : ). rial (OCHA, ). Many health care facilities were also destroyed, decreasing dramatically the .. Housing and infrastructure number of available health care providers in the light of the high number of physically injured and The National Disaster Risk Reduction and Manage- psychologically distressed people (OCHA, ). ment Council (NDRRMC) reported that more than Bopha had also detrimental impacts in regards , shelters were affected by the typhoon to safety issues. An increased number of gender- (IOM Philippines, ). Given that many houses based violence cases was reported (OCHA, b). were built with wood and other vulnerable mate- Additionally, many people, particularly indig- rials, numerous houses lost their roofs or were enous people, living in the hinterlands were not completely flattened. Yet, it was reported that able to access basic government services, such as

IDDRI STUDY 09/2013 37 The State of Environmental Migration 2013

social welfare programs, because they had never .. Number and location possessed birth certificates (Protection Cluster, of people displaced ). from their homes

The reported number of people displaced from 4. THE VARIOUS FORMS their homes significantly increased after Bopha AND IMPACTS OF PEOPLES’ hit the Philippines on  December. Initially, news articles, international organizations and national DISPLACEMENTS authorities reported that tens of thousands people had been evacuated or fled by themselves to .. Preventive evacuations various evacuations centers (The Telegraph, ). The peak was achieved shortly after Bopha’s land- Before Bopha hit the country, many people had fall (OCHA d). The reported number of people already left their homes to flee the threats of displaced from their homes further increased to typhoon. Between  November and  December, about one million by taking into account people national and international disaster authorities who were inside evacuation centers and outside encouraged communities to take responsibility for evacuation centers as Bopha had, for instance, their safety and ordered pre-emptive evacuations damaged their houses (OCHA ). The latest in areas that were expected to be hit the hardest figures report that Bopha induced the displace- (IFRC, ; Escalante, , OCHA, :  ). As ment of more than . million people in total a response, residents were evacuated or fled by (Yonetani : ). themselves to safer grounds, which included evacuation centers set up in schools, bus terminals and Figure 2. Displaced People inside and outside Evacuation gyms. People also fled to relatives who lived on Centers higher ground to prevent being affected by flash Number of People displaced from their homes floods and landslides that were expected to result 1,000,000 from the typhoon (Escalante, ; NDRRMC, a). 800,000 Officials have struggled to measure the scale of the pre-emptive evacuations and escapes. On the 600,000 one hand, authorities stated that more people had Outside Evacuation Centers been encouraged to leave typhoon-hazardous ar- 400,000 eas compared to the previous year when Typhoon 200,000 Washi hit the country (Mullen, ). On the other Inside Evacuation Centers hand, they emphasized that there were important differences between areas. In the more northern 4 Dec 11 Dec 18 Dec 25 Dec 1 Jan 8 Jan areas, where Washi hit Mindanao, the number of Source: OCHA, 2013: 3. Map provided with the courtesy of the UN Office for the Coordination of Humanitarian Affairs. The boundaries and names shown and the people who left their homes was much larger than designations used on this map do not imply official endorsement or acceptance by in the southern areas, where Washi had not made the United Nations. landfall in . Ortigas () reported that the population in the north had taken the warning More specifically, in January , the Philip- signals more seriously because Bopha had been pines’ Department of Social Welfare and Devel- expected to hit this area strongly and people were opment (DSWD) estimated that, at the height of still traumatized through Typhoon Washi. In con- the disaster, about , families, or , trast, the population in the south paid less atten- people, stayed in about , evacuation centers tion to the warning signs because Bopha had been (OCHA, : ). Indeed, these numbers exceed expected to hit this region less and people “didn’t those that the Office of the Coordination of Hu- even know what the warning signals meant (…)” manitarian Affairs (OCHA) reported and are illus- because “typhoons always happen somewhere trated in figure . OCHA (d) estimated a peak else, not here (…)” (Ortigas ). Eventually, Bo- of , evacuees staying in  evacuation pha hit the south stronger than anticipated, caus- centers on  December . Many reasons could ing a higher number of deaths than in the north explain this significant difference in data, such as (Ortigas ). that the DSWD may have used additional sources that were not known to OCHA in December .

. Personal communication with OCHA Head of Office, . Ibid. conducted on  June  via email.

38 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Evacuation centers were increasingly established people from their familiar environment, which in different locations, including local government would make it difficult for them to get back into buildings, gyms and stadiums (OCHA, : ). their normal life once they have left these tempo- Soon after Bopha had hit the country, the number rary solutions. of people who stayed in these evacuation centers While many people were displaced in evacu- started decreasing (IOM, a). By February ation centers and transitional sites, the over- , less than , people remained in about whelming majority, numbering about , one hundred evacuation centers (OCHA, e). people as of February , remained with host Many factors explain why these numbers started families and in so-called “spontaneous settle- declining quickly. First, numerous families wanted to ments”, where people have set up their own return to their homes to start rebuilding their houses makeshift shelters (OCHA, e; IOM Philip- and save whatever they could (IOM, a). Second, pines a). This number had stayed relatively many people left because of the bad conditions that consistent from mid-December , as shown in often prevailed in evacuation centers. Many centers figure . However, one may observe a significant were overcrowded and could not provide adequate drop in the number of people staying outside protection that people required, such as material for evacuation centers around  December. OCHA emergency shelters, adequate sanitation and pota- had used only data that was available at the ble water (IOM, a). Several evacuation centers time. However, data from some municipalities were also roofless after having been hit by the ty- had not arrived, which explains the sudden drop phoon and some evacuation centers in Compostela in numbers. The non-supply of the data could Valley were badly flooded (IOM, ). These condi- have been due to numerous reasons, such as lo- tions also led some people to move to other places, gistical challenges to obtain data. Spontaneous such as to transitional sites (IOM, ). settlements were mainly found along roadsides, Transitional sites included tent cities and bunk- near the original houses of displaced people houses that were set up immediately after the ty- and in affected communities after the flood- phoon in order to provide a sanctuary for displaced water receded (IOM Philippines, a). While people. By March , IOM Philippines () re- people received similar assistance as those who ported that  bunkhouses were built in Davao remained in registered sites, providing the nec- Oriental and Compostela Valley. The construction essary aid to them was more challenging (IOM of bunkhouses mainly happened under the con- Philippines, a). trol of national authorities, with the financial and technical support of international and national . Migration and trafficking humanitarian organizations and the assistance of local people who were hired to build these houses After  December, an increasing number of people (IOM, ; OCHA, : ; OCHA, d). left their homes to better deal with the impacts However, transitional sites were a highly contro- of Typhoon Bopha. In order to cope with the versial topic. Similar to many evacuation centers, financial difficulties, for example, several people, transitional sites were often in bad conditions and particularly men, migrated to other areas to look lacked, for instance, adequate water, sanitation for work. Although this implied that many were and hygiene facilities (IOM, ). In addition, separated from their families and usual environ- newspapers criticized the construction of bunk- ments, they saw this as their duty. As noted earlier, houses and accused the government of managing women who stayed home consequently emerged this process in a corrupt manner (Escalante, ). as the main provider for their families (Protection For instance, local newspapers reported that the Cluster, ; UNICEF, ). Similarly, adolescent payments of those who were hired to build bunk- boys and girls started going to cities to look for houses were dubious because some workers had jobs. Given the high incidence of child labor in the received more money than they were entitled Philippines, this was not considered an unusual (Romero, ). Furthermore, the construction of phenomenon. However, travelling as child laborers transitional sites had often not been regarded as a to other regions made children more vulnerable to best practice. Specialists on post crises and emer- violence, abuse and trafficking (Protection Cluster, gency issues noted that efforts should have been ; UNICEF, ). directed toward making people stay in their trusted environment so that they can recover as quickly  . Interview with IOM Regional Emergency and Post Crisis as possible. Transitional sites would disconnect Specialist at IOM Thailand, conducted on  March  via Skype. . Interview with IOM Regional Emergency and Post Crisis Special- . Personal communication with OCHA Humanitarian ist at IOM Thailand, conducted on  March  via Skype. Affairs Officer, conducted  June  via email.

IDDRI STUDY 09/2013 39 The State of Environmental Migration 2013

Hence, UNICEF () raised the concern that complicate whether people will be able to rebuild the incidence of trafficking could increase as a their houses. Assessments show that many of the result of Typhoon Bopha. Already before Bopha, worst affected people were living in informal ar- trafficking had been a common issue in the Mind- rangements without any legal documentation on anao region (UNICEF, ). Due to the clandes- land ownership. However, such documentation is tine nature of trafficking, however, reliable and important in order to receive government assis- up-to-date data have not yet been available that tance to rebuild houses; but it may take “some time” would allow better assessment of whether or not to obtain them (Kelly : ). Additionally, going the issue of trafficking has increased due to Bopha. back to normal will be difficult for everyone who Nevertheless, the concern is reasonable because suffered a loss, especially for dependents. Whole children that resulted from the typhoon are likely communities will also need to be rebuilt to enable a to be more vulnerable to child trafficking. For ex- similar life before Bopha. The question of whether ample, the rising number of migrating child labor- the Government will actively support people return ers due to Typhoon Bopha could add to the issue of to their homelands and, instead, relocate them to child trafficking (UNICEF, ). Furthermore, the less disaster prone areas is of high importance. growing number of women and girls who were left behind or looking for work could exacerbate the problem of sex trafficking. The fact that traffick- 5. POLICY RESPONSES ing increased by about  per cent after Typhoon AND CHALLENGES Washi had devastated the Philippines in  indicates that trafficking could intensify (UNICEF, ). Consequently, several humanitarian actors .. The questioned efficiency have mainstreamed awareness-building programs of preparedness measures on trafficking issues amongst governmental officials and regional anti-trafficking networks since Many actions that aimed to prepare for any eventu- Bopha (IOM Philippines, a). alities related to Bopha started on  November  when the Philippine Atmospheric, Geophysical and . Return of the Astronomical Services Administration (PAGASA) displaced people and the NDRRMC began to issue regular updates on the typhoon. Bopha was expected to reach the The question of when and whether displaced country by  December with wind strength between people can go back is difficult. Indeed, several  km/h and  km/h. Local authorities also people returned to their trusted environment expected that Bopha would have more devastating to continue their life, or start a new life, as early impacts than Washi and therefore, aimed to inten- as possible. However, even if people were able sify preparation and evacuations efforts. (IFRC, to return and live in their houses, the conditions ; PAGASA, n.d.; OCHA, ) in which they live are often very poor. A study Both governmental and non-governmental ac- conducted by REACH in March  had shown tors implemented emergency preparedness ac- that  per cent of families, or nearly a half a tions. This included that PAGASA and NDRRMC million people, lived in houses that were classified classified different regions under different public ”uninhabitable” (Bamforth, ; Reach, ). warning signals. In the view of these signals, au- Updated data suggests that, as of  April , thorities directed pre-emptive evacuations of fami- , people displaced by Bopha still remain in- lies that resided in areas that were anticipated to side evacuation centers and , outside of be hit the strongest. Furthermore, they advised the evacuation centers (OCHA, a: ). While population to take precautionary measures, which many live in tents and bunkhouses, the majority included cancelling travel and outdoor activities, remains in spontaneous settlements (IOM Philip- seeking shelter in stronger buildings, evacuat- pines, ). Those who live in spontaneous settle- ing to safer areas, calling off school classes and ments are often considered more likely to return staying in shelters until the typhoon completely to normality than people residing in other sites. passed (IFRC, , PAGASA n.d.). Furthermore, Many spontaneous settlements are close to the humanitarian stakeholders established evacuation original homes of displaced people where families centers, pre-positioned relief items and mobilized can live together and the community life can start the emergency responders (IFRC, ). People again (Bamforth, ). Yet, it may take years until all the people will be . Interview with IOM Communications Specialists, Inter- able to return and live in their homes as it was be- view with IOM Regional Emergency and Post Crisis fore Bopha. For example, land tenure issues might Specialist

40 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

generally took the warnings seriously and fol- : -). lowed the recommended precautionary measures. As planned by the responsibility sharing of the Given the high number of casualties that Bopha cluster approach, specific actors were in charge of caused, it appears that the implemented prepar- specific sectors, according to their mandate, com- edness measures were insufficient and ineffec- petences and field presence. The DSWD and the tive. Therefore, one may interogates the extent to United Nations Development Program (UNDP) which the Philippines could still be seen as a mod- took the lead for the early recovery cluster (OCHA, el in addressing issues related to DRR (Ginnetti et : ). Cluster outputs included the distribution al., ). However, Bopha happened in a region of food packs and the realization of cash-for-work where a typhoon of such great strength had no such as hiring workers to clear debris, rebuilding precedent. This raises the question as to whether homes and infrastructure (OCHA : ). Bunk- people can be expected to effectively and success- houses were constructed and emergency shelter fully prepare for something that never happened kits were distributed to address the needs for shel- before and that no one genuinely expected to ever ter (OCHA, ). happen. However, significant gaps remained in the provision of shelter (OCHA : , OCHA d). A .. Declaration of the “state comprehensive shelter policy for Bopha had still of national calamity” to be drafted in March  (Scheidler & Hilmi, ). This lack of policy hampered humanitarian On  December, after Bopha’s landfall, the Presi- assistance because local authorities, communities dent of the Philippines signed Proclamation No. and other humanitarian stakeholders were await-  to declare a state of national calamity. In doing ing clarity in how to address certain shelter issues. so, he aimed to speed up the rescue, relief and However, a comprehensive shelter policy could rehabilitation plans implemented by the national have significantly accelerated the provision of as- and international humanitarian assistance sistance to those living in evacuations centers, tran- (OCHA, ; Kelly ). While President Aquino sitional sites, spontaneous settlements and in other stressed that government actors would take the vulnerable conditions (Scheidler & Hilmi, ). lead in addressing the needs that had resulted This had been seen in the previous year when Ty- from Bopha, he also emphasized the importance of phoon Washi hit the country. In its aftermath, a Lo- a coordinated approach between different human- cal Inter-Agency Committee (LIAC) had been cre- itarian actors (OCHA, : ). ated that developed a shelter strategy to coordinate and leverage national and international assistance .. Humanitarian response more effectively (Scheidler & Hilmi, ), and the clusters approach The environmental disasters that happened in the months after Bopha added to the challenge By supporting the UN clusters approach to address of implementing the clusters approach and ad- the particular needs of the displaced population, dressing the needs of the displaced population. In the Philippines aimed to strengthen the effective- particular, rains did not stop for months and dan- ness of the humanitarian response by making sure gerous floods occurred in Bopha-affected regions that high standards of predictability, account- in January  (IRIN, ). The monsoon and ability and partnership in all activity sectors were typhoon season that started in May/June were ad- achieved (DRC, n.d.). Needs assessments were ditional challenges (Scheidler & Hilmi, ). conducted by national authorities in coopera- The inaccessibility of areas, politicization of aid tion with other humanitarian organizations and and security have also been major challenges to ad- completed on  December (OCHA, : , ) and equate and nondiscriminatory aid provision. Many identified key clusters to be implemented - camp areas were hardly accessible and sometimes only coordination and camp management (CCCM), accessible with helicopters due to the destruction early recovery, food security and agriculture of numerous roads and bridges. Additionally, local (OCHA, ). The clusters approach was imple- groups and typhoon survivors argued that the pro- mented at the regional, provincial and municipal vision of aid would have been politicized due to the levels on a needs basis. This was significant given elections in May . This would have meant that that the majority of the local governments in the those who were considered potential voters would affected areas had little experience in addressing have been provided with humanitarian assistance both natural and man-made disasters (OCHA, before others that were not considered potential

. Interview with IOM Regional Emergency and Post Crisis . Ibid. Specialist.

IDDRI STUDY 09/2013 41 The State of Environmental Migration 2013

voters (IRIN, a). Furthermore, humanitarian programs, local governments have to obtain an assistance had to take special security measures due authorization from the MGB to ensure that these to the conflict-related situation in the region. areas are safe (Kelly, : ). If areas have pre- Funding has been a last major challenge to hu- viously been classified as a NBZ, authorization manitarian response efficiency. Even though con- forms are not issued. stant appeals were made to the international com- A major reason why the demarcation of NBZs is munity, the humanitarian response to Bopha has still an unofficial policy is that it is a highly contro- generally been underfunded, especially for emer- versial topic. This has been discussed at numerous gency shelter needs (OCHA, d). Out of the levels within the Philippines for many years and nearly one hundred thousand people who were Bopha has, all over again, clarified the contentious helped by agencies in the shelter cluster, for ex- nature of this topic. On the one hand, the realiza- ample, only half had obtained a tarpaulin because tion of NBZs could significantly lessen Mindanao’s of the agencies’ inability to obtain the sufficient risk of people displacements because of natural financial resources (Bamforth, ). disasters. On the other hand, important challenges to implement such policy exist. .. No-Build-Zones (NBZs) For example, challenges can be associated with indigenous people. Many indigenous communi- Besides the clusters approach that has particularly ties live in areas that are likely to be classified as focused on addressing the needs of the displaced NBZs. However, they may be reluctant to move population in the short-term, the Government has from their land due to their ancestral relation- implemented actions that aim to reduce the risk ship with the land (Kelly, : -). In addi- of people’s displacement due to natural disasters tion, their land and property rights are legally in the long-term. A particularly interesting and recognized by the Indigenous People’s Rights Act controversial strategy has been the identification (IPRA) of , complicating issues around relo- of so-called ‘No-Build-Zones’ (NBZs). To do so, cating them (Kelly, : -). Indeed, Section the Government’s Mines and Geosciences Bureau  of the IPRA that shows that indigenous people (MGB) has conducted geo-hazard risk assessments are only entitled to return to their land if the State in Bopha-affected provinces. In these assessments, considers it to be “safe”. However, many indige- areas are classified either as low-risk areas where nous people are likely to not accept this clause if it is safe to build, medium-risk areas where miti- it is to be enforced. Numerous field consultations gation measures may be appropriate or high-risk conducted after Bopha revealed that indigenous areas where it is unsafe to build. High-risk areas people have an extremely strong desire to return are interpreted as NBZs. Depending on the risk to their land. Although indigenous people went category, the MGB recommends actions that can through trauma, experienced severe losses and be either moderate or radical. For example, in are aware that such disasters can happen again, some cases, it is recommended to move houses they still want to go back to their old homes and for a few meters away. In other cases, it is recom- regions that their ancestors have inhabited for mended to relocate whole communities. (IRIN, centuries. ; Kelly, : ) People’s reliance on certain livelihoods and Until today, the Government has not yet released their willingness to change particular life patterns a NBZ policy. The MGB recommendations are not can also complicate the creation of NBZs. Given compulsory and cannot be legally enforced on lo- that many people rely on subsistence and com- cal governments (Kelly, : -). Neverthe- mercial farming, their relocation to land unsuit- less, the identification of NBZs has had important able for such activities can be devastating (Schei- consequences. For example, President Aquino dler & Hilmi, ). Furthermore, several people has urgently advised local governments to adopt have expressed their unwillingness to transform the recommendations issued by the MGB (Kelly, their life styles because as migrating would re- : ). In addition, several national authori- quire them to do so (Kelly, : ). For example, ties have avoided providing rebuilding assistance people who have a long family tradition of farm- to areas that were classified as NBZs (Kelly, : ing may now have to be a fisher due to the new ). Furthermore, in order to realize resettlement living circumstances.

. Interview with IOM Communications Specialist. The . Interview with IOM Communications Specialists, Inter- rural-based guerrilla force, New People’s Army (NPA), view with IOM Regional Emergency and Post Crisis which is the military wing of the Communist Party of Specialist. the Philippines (CPP), has been fighting in eastern Min- danao against the Philippine government for decades . Interview with IOM Communications Specialist. (OCHA, : ; BBC, ). . Ibid.

42 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Furthermore, there are important logistical and of people due to natural disasters. Indeed, many cost challenges. In some instances, the MGB has have considered the country as a global leader in identified many and wide-ranging areas as NBZs. enacting policies and legislation to reduce the risk For example,  per cent of New Bataan has been of natural disasters. However, given the detrimen- classified as such. Clearly, this raises questions tal impacts that Bopha had, more efforts need to around whether the relocation of a great number be taken to address the issue of natural disasters of people is feasible to implement and if there is and its consequences on the displacement of peo- a viable alternative to which they could be relo- ple more effectively. cated. Another challenge is that the relocation Policy recommendations are suggested to ad- of people is often a slow process. In particular, the dress how to more efficiently prepare and manage identification of relocation sites has seen signifi- forthcoming disasters, and address the particular cant delays. While government land has been ac- needs of those displaced. quired relatively fast, negotiating relocation sites that are privately owned has taken far longer (Kel- Scale up current policy responses ly, : -). In addition, past experiences have The humanitarian assistance should scale up the shown that the relocation of people has been a current policy responses. Still today, there are very costly policy (Scheidler & Hilmi, ). Given many gaps that need to be filled in order to address that the Philippines are ranked as lower-middle in- the needs of the displaced people in a more effec- come country, associated costs can be a difficulty tive and efficient manner. The implementation of to finance (World Bank, ). a comprehensive shelter policy should be consid- Finally, political issues can represent obstacles ered as a priority. Furthermore, it should include to put a NBZ policy into practice. Although the increasing the financial support for the different national government is generally in favor of im- clusters, especially for the shelter cluster that has plementing NBZs, local governments do not often experienced insufficient funding (OCHA, c). agree with the realization of NBZs, mainly because this would complicate the usual voting proce- Implement standardized preparedness dure. In particular, population shifts may mean measures that certain voting districts will cease to exist. As Standardized preparedness measures should a consequence, the existence of local governments be implemented countrywide. In an interview, in these areas may be affected. specialists on post crises and emergency issues noted that the Philippines were prepared for a typhoon of such great strength; however, it was 6. CONCLUSION AND POLICY not prepared for a typhoon that hit the country RECOMMENDATIONS so far in the south in such enormous power. To implement such measures, offices that exclusively This study has shown the extent to which the handle issues related to natural disasters could Bopha typhoon has caused large-scale displace- be established in each province; governmental ments of people. These displacements have been actors from the regional, municipal and village complex because they occurred before and after level should participate in regular and compulsory the typhoon in many different forms. Further- trainings; and national campaigns that inform more, the typhoon has decreased standards of the population on appropriate preparation meas- living of the displaced people both in the short and ures and emphasize that early warning systems long term. The economic status of the Philippines must be taken seriously should be conducted in all may drag out the negative impacts of the storm provinces. The case of Indonesia could serve as an because of the country’s financial limitations. This example of how a best practice approach could be already difficult situation is additionally compli- implemented, given that this country has realized cated because climate patterns are changing and a standardized way of response.  similar disasters may increasingly happen in the future in the country. Develop clear strategies that address the The above-explained situation has confronted needs of those who live in so-called ‘NBZs’ the Philippine government with new challenges It is recommended to develop clear strategies that that are related to the issue of the displacement address the needs of those who live in areas that are to be classified as NBZs, which still appears . Ibid. to be a grey area in existing policy responses. . Ibid. . Interview with IOM Regional Emergency and Post Crisis Specialist. . Ibid.

IDDRI STUDY 09/2013 43 The State of Environmental Migration 2013

Strategies should include developing clearer legal changing climate patterns affecting the Philip- instruments that describe the process of declaring pines. Indeed, in February , the Philippines’s and enforcing NBZs (Kelly, : ). Furthermore, Congress took already the right step in the right people living in NBZs should be clearly informed direction by launching a landmark bill that spells about the risks they face and receive appropriate out the particular rights of IDPs before and after assistance that support their decision – whether it displacement and allows IDPs to claim monetary is to remain in the NBZ area or move away (Kelly, assistance and compensation (UNHCR, ). The :). challenge is now one of its implementation. 

Ensure that mechanisms to protect the rights of people displaced by environmental events are put into action . National and international actors praised this bill, which represented the first of its kind in Asia, as a milestone for It is imperative to ensure that appropriate mecha- the protection of IDPs in the Philippines (UNHCR ). nisms to protect the rights of people displaced However, in June , the President of the Philippines by environmental events are put into action. vetoed the Rights of the IDP Act (NRC&IDMC ). People displaced by natural events do not only Clearly, this emphasizes both the challenge to implement mechanisms to protect the rights of people displaced by have particular needs, but their numbers may environmental events, as well as the need to continue also be increasing in the coming years due to the campaigning for it.

44 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Reach. 2012. Shelter Cluster Assessment in Mindanao, Philip- BIBLIOGRAPHY pines. Shelter Cluster Report, Final Assessment Report. Chatelaine, Switzerland: Shelter Cluster. R   Scheidler, Kim and Mohamed Hilmi. 2013. Calling for a Com- Alliance Development Works. 2012. World Risk Report 2012. prehensive Shelter Strategy in Mindanao. Berlin: Alliance Development Works. Field Report. Washington DC: Inter Action. DRC Danish Refugee Council. n.d. Humanitarian Reform and the Global Cluster Approach. DRC Danish Refugee Coun- H B cil. GFDRR Global Facility for Disaster Reduction and Recovery. IFRC. 2012. “Philippines: Typhoon Bopha”. Information Bul- 2011. Disaster Risk Management letin No.1. 03 December. Programs for Priority Countries. The World Bank, GFDDR, OCHA. 2012. “Philippines”. Humanitarian Bulletin. 11(1): ISDR International Strategy for Disaster Risk. 1-4. Ginnetti, Justin; Dagondon, Butch; Villanueva, Clarisse; En- OCHA. 2013b. “Philippines”. Humanitarian Bulletin. 12: riquez, Juanito; Temprosa, Francis Tom; 1-4. Bacal, Christie and Fr. Norberto L. Carcellar. 2013. Disaster- OCHA. 2013c. “Philippines.” Humanitarian Bulletin. 14: 1-4. Induced Internal Displacement in the Philippines – The Case OCHA. 2013d. “Philippines”. Humanitarian Bulletin. 16: of Tropical Storm Washi/Sendong. Geneva: IDMC Internal 1-5. Displacement Monitoring Centre & NRC Norwegian Refu- gee Council. National Documents IFAD International Fund for Agricultural Development. Proclamation No. 522 Declaring a State of National Calamity. 2009. Enabling Poor Rural People to Overcome 2012. Malacañan Palace, Manila, Philippines. Poverty in the Philippines. Rome, Italy: IFAD. Newspaper Articles and Public Statements Kelly, Tessa. 2013. Legal and Regulatory Issues: Typhoon Bo- Bamforth, Tom. 2013. “Typhoon Bopha survivors languish in pha. Shelter Cluster. ‘uninhabitable’ homes as funding Yonetani, Michelle. 2011. Displacement Due to Natural Haz- shortfall stalls recovery efforts.” IFRC. 25 March. Available: ard-Induced Disasters – Global Estimates http://www.ifrc.org/en/news-and-media/opinions-and- positions/opinion-pieces/2013/typhoon-bopha-survivors- for 2009 and 2010. Geneva: IDMC & NRC. languish-in-uninhabitable-homes-as-funding-shortfall- Yonetani, Michelle. 2013. Global Estimates 2012 - People Dis- stalls-recovery-efforts-61108/, Accessed: 05.05.2013. placed by Disasters. Geneva: IDMC and NRC. BBC. 2012. “Guide to the Philippines conflict.” BBC News Asia. 08 October. Available: http://www.bbc.co.uk/news/ world-asia-17038024 Accessed: 05.04.2013 H A R  A P Escalante, Shirley. 2012. “Palau, Philippines brace for Ty- phoon Bopha”. Australia Network News. 03 DSWD Department of Social Welfare and Development, Re- December 2012. Available: http://www.abc.net.au/ public of the Philippines; Department of news/2012-12-02/an-phils-on-typhoon-alert/4403512, Health, Republic of the Philippines; UNDP; ILO; UNICEF; Accessed: 05.05.2013. FAO; World Vision International; AFC Escalante, Shirley. 2013. “Bopha survivors protest Philip- International and WFP. 2013. Emergency Food Security, Nu- pines aid program.” Australia Network News. trition and Livelihoods Assessment for Bopha (Pablo) Philip- 14 February. Available: http://www.abc.net.au/ pines. Key Findings. January. news/2013-02-14/an-phils-suvivors-protest-bopha-aid-pro- IASC Inter-Agency Standing Committee. n.d. Developing gam/4519480, Accessed: 21.05.2013. a Humanitarian Dashboard. Guidelines for COs Using the Gutro, Rob. 2012. “Hurricanes/Tropical Cyclones: Typhoon Dashboard to Monitor the CAP Cycle. Available: http://as- Bopha – Western North Pacific Ocean.” sessments.humanitarianresponse.info/files/2012_guid- ance_dashboard_MYR.pdf, Accessed: 13 August 2013. NASA News. 10 December. Available: http://www.nasa. gov/mission_pages/hurricanes/archives/2012/h2012_Bo- NDRRMP National Disaster Risk Reduction and Manage- pha.html, Accessed: 05.05.2013. ment Plan, Republic of the Philippines. 2011. IRIN. 2013. “ ’Building back worse’ after Philippines ty- The National Disaster Risk Reduction and Management phoon”. IRIN News. 08 February. Available: http://www. Plan 2011-2028. irinnews.org/report/97435/Building-back-worse-after- OCHA. 2013. Philippines (Mindanao) – Humanitarian Ac- Philippines-typhoon, Accessed: 05.05.2013. tion Plan 2013. Typhoon Bopha/Pablo IRIN. 2013a. “Allegations of aid politicization in Philippines Response – an Action Plan for Recovery. Revision. Geneva, ahead of elections.’ IRIN News. 22 New York: OCHA. February. Available: http://www.irinnews.org/re- OCHA. 2013a. Philippines (Mindanao) Humanitarian Action port/97529/Allegations-of-aid-politicization-in-Philip- Plan - Mid-Year Review. Geneva, New York: OCHA. pines-ahead-of-elections , Accessed: 05.05.2013.

IDDRI STUDY 09/2013 45 The State of Environmental Migration 2013

Met Office. 2012. “Typhoon Bopha on course for south- OCHA. 2012a. Palau Typhoon Bopha. Situation Report No.1. ern Philippines.” Met Office News. 03 December. Avail- 03 December. able: http://metofficenews.wordpress.com/2012/12/03/ OCHA. 2012b. Philippines: Typhoon Bopha. Situation Report typhoon-bopha-on-course-for-southern-philippines/, Ac- No. 1. 04 December. cessed: 05.05.2013. OCHA. 2012c. Philippines: Typhoon Bopha. Situation Report Mullen, Jethro (2012) “Typhoon Bopha carves across Phil- No. 2. 05 December. ippines, killing scores of people”. CNN News. 05 December. Available: http://edition.cnn.com/2012/12/04/world/ OCHA. 2012d. Philippines: Typhoon Bopha. Situation Report asia/philippines-typhoon, Accessed: 05.05.2013. No. 1. 06 December. NRC & IDMC. 2013. A bend in the road is not the end of the OCHA. 2012e. Philippines: Typhoon Bopha. Situation Report road – the veto of the IDP Act in the No. 1. 18 December. Philippines. Public Statement. 14 June. Available: http:// OCHA. 2012f. Philippines: Typhoon Bopha. Situation Report www.internal-displacement.org/8025708F004CE90B/ No. 11. 27 December. (httpDocuments)/572DFD30C7E43D2AC1257B89004 OCHA. 2013e. Philippines: Typhoon Bopha. Situation Report 0278B/$file/IDMC-NRC-public-statement-Philippines- No. 19. 12 February. 13-june-2013.pdf, Accessed: 21.06.2013. Protection Cluster. 2013. Protection Cluster Fact Sheet. Ty- Ortigas, Marga. 2012. “Typhoon Bopha: A tale of two sto- phoon Bopha Response. 14 January. ries.” Aljazeera. 14 December. Available: http://blogs.alja- Mindanao, Philippines: Protection Cluster. zeera.com/blog/asia/typhoon-bopha-tale-two-stories. Ac- cessed: 13 August 2013. Shelter Cluster. 2012. Emergency Shelter Assessment Ty- phoon Bopha. Fact Sheet # 1. 16 December. Romero, Alexis. 2013. “DSWD denies irregularities in building of bunkhouses”. Nation. 22 February. Available: UNICEF. 2012. Philippines (Typhoon Bopha). Humanitarian http://www.philstar.com/nation/2013/02/22/911560/ Situation Report 4. 18 December. dswd-denies-irregularities-building-bunkhouses, Accessed: UNICEF. 2013. Summary Report of Child Protection Priority 05.05.2013. Needs and Responses. The Telegraph. 2012. “Philippines: Typhoon Bopha forces 40,000 to evacuate.” The Telegraph. 04 W December. Available: http://www.telegraph.co.uk/news/ worldnews/asia/philippines/9720519/Philippines-Ty- Center for Disaster Philanthropy. 2013. Typhoon Bopha, phoon-Bopha-forces-40000-to-evacuate.html, Accessed: Philippines. Available: http://disasterphilanthropy.org/dis- 05.05.2013. asters/typhoon-bopha-philippines/ Accessed: 06.04.2013. UNHCR. 2013. “Philippines passes historic bill to protect in- Morrison, John. 2013. Mindanao Montane Rains Forests. ternally displaced”. UNHCR News Stories. WWF. Available: http://worldwildlife.org/ecoregions/ 08 February. Available: http://www.unhcr.org/5114dd5c9. im0128, Accessed: 05.05.2013 html, Accessed: 05.05.2013 National Statistics Office. 2012. Population Density Increase by 53 Percent per Square Kilometer. Republic of the Philip- pines. Available: http://www.census.gov.ph/content/ S R  F S population-density-increase-53-persons-square-kilometer. Accessed: 12 August 2013. Haeseler, Susanne. 2012. Typhoon BOPHA Across the Philip- pines at the Beginning of December 2012. Deutscher Wetter PAGASA Philippine Atmospheric, Geophysical and Astro- Dienst. nomical Services Administration. n.d. The IOM. 2012. Bopha Response. Compostela Valley-Mindanao. Philippine Public Storm Warning Signals. Available: http:// 10 December. www.pagasa.dost.gov.ph/genmet/psws.html#psws3, Ac- cessed: 05.05.2013. IOM. 2012a. DTM Displacement Tracking Matrix Report. Ty- phoon Bopha. 14 December. World Bank. 2013. Philippines - Data. Available: http://data. worldbank.org/country/philippines, Accessed: 05.05.2013 IOM Philippines. 2012. Tropical Cyclone ‘Bopha’. Situation Report. 04 December. IOM Philippines. 2012a. Tropical Cyclone ‘Bopha’. Situation Report. 05 December. IOM. 2013. DTM Displacement Tracking Matrix Report- Ty- phoon Bopha Response. 14 January IOM Philippines. 2013. Typhoon Bopha Emergency Response. Situation Report. 05 March. IOM Philippines. 2013a. Typhoon Bopha Emergency Re- sponse. Situation Report. 20 February. NDRRMC, Republic of the Philippines. 2012. Effects of Pablo (Bopha). Situation Report No. 38. 25 December. NDRRMC, Republic of the Philippines. 2012a. Preparedness Measures for Typhoon Pablo (Bopha). Situation Report No. 02. 03 December.

46 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

IN FOCUS Cyclones Giovanna and Irina make Long-term challenges - Reducing Vulnerabilities Madagascar’s current political crisis has complicated the national thousands of people homeless response to natural disasters. Since 2009 and the deposition of the elected President Marc Ravlomanana by President Andry Rajoelina Maëlle Robert deposed, great troubles have shacked the country. Today, emergency assistance is mainly provided by international institutions or NGOs and Each year, an average of three to four strong tropical storms formed Malagasy rulers have become dependent on foreign donors to finance in the Indian Ocean hit Madagascar. In the night from 13 to 14 Febru- any action. After Giovanna and Irina, most of the outside funding has ary 2012, Cyclone Giovanna landed on the eastern coast of the island, been used for rebuilding, rather than disaster risk reduction. Yet, risk moved inland and crossed the country from the east to the south-west, could be minimized. bringing strong winds, high waves, heavy rains and flooding. The The government has to invest to improve the early warning system eastern districts were the most affected. Two weeks later, on the 26 of by using new telecommunication means – especially phone messages. February, another tropical storm, Irina, hit the island. The two disasters This system is efficient; however, it is difficult to apply in the rural caused 112 deaths, injured 299 people and left three people presumed areas, where people don’t necessarily have mobile phones and a func- lost1. According to the National Office of Risk and Disaster Manage- tioning network. ment (BNGRC), an estimated 60,000 houses, in addition to schools, As traditional huts are mainly composed of wood, reeds, bamboo health centres and vital infrastructure, were destroyed, damaged or and palm leaves that hardly withstand storms, community leaders flooded2. The BNGRC, supported by non-governmental organizations should encourage people to build durable structures with baked stones (NGOs), quickly organized humanitarian assistance to provide people or concrete. Currently, they are also trying out new ways of construct- drinking water, food stoves and/or sanitation products. Efforts were ing, particularly using crossing houses frame, an anti-storm architec- also made to rebuild the damaged infrastructure (roads, bridges, tel- tural style inspired by the anti-earthquake Japanese constructions, in ecommunication installations, schools, houses). order to make the buildings more flexible and resistant to shocks and Temporary migration: how far, how long? strong winds6. Overall, the two storms displaced about 55,000 people,3 mostly tempo- Local leaders need to develop collective food reserves, raised or shel- rarily and close to their homes. Most of them chose to move just before tered in safe places, in order to prevent them from being flooded and to the cyclone: they took shelter in neighbors’ solid houses or evacuated ensure food security notwithstanding environmental disruptions. to more resistant structures, such as schools or churches4.Their return As a global issue, the national government has to keep on fighting usually occurred after a couple of days, and they often rebuilt their deforestation to make trees serve as buffers to slow winds and prevent huts at the same place as before. Yet, as Giovanna and Irina mainly hit floods by increasing soil’s absorption capacity7. isolated and landlocked villages, where many people did not receive In any case, it is essential to include the people, so that they behave the alert messages broadcasted on the radio stations or the warning in a way that best reduces their vulnerability8. campaign, a majority of the people did not evacuate. In the following months, because Giovanna and Irina flooded private food reserves and destroyed more than half of the crops, food prices rose in several places. It is doubtful however that this growing food insecurity induced further migration across the region5.

1. Data from the National Office of Risk and Disaster Management (BNGRC) actions in CARE France and Responsible for an assessment on the CARE – From an email interview with Katia Rakotobe from CARE Madagascar, in projects in Madagascar after the passage on Giovanna on site, 16 April, mission in Antananarivo, 16 - 17 April, 2013. 2013. 2. National Office of Risk and Disaster Management (BNGRC), Executive Sec- 6. Interview with Emmanuel Lan Chun Yang, Responsible for emergency retariat and Malagasy Home Office : “Bilan provisoire des dégâts suite au actions in CARE France and Responsible for an assessment on the CARE passage du cyclone tropical intense “Giovanna””, 14 March 2012, http:// projects in Madagascar after the passage on Giovanna on site, 16 April, www.bngrc.mg/documents/giovanna/SITU%20Giovanna%2014_03-15h. 2013. pdf 7. Interview with Emmanuel Lan Chun Yang, Responsible for emergency 3. Data from the National Office of Risk and Disaster Management (BNGRC) actions in CARE France and Responsible for an assessment on the CARE – From an email interview with Katia Rakotobe from CARE Madagascar, in projects in Madagascar after the passage on Giovanna on site, 16 April, mission in Antananarivo, 16 - 17 April, 2013. 2013. 4. Email interview with Thierry Razanakoto, National Consultant and Research 8. “Cyclone Giovanna struck with little warning”, IRIN, 20 February, Assistant, IOM (Madagascar), 17 - 18 April, 2013. 2012, Brickaville http://www.irinnews.org/report/94911/madagascar 5. Interview with Emmanuel Lan Chun Yang, Responsible for emergency -cyclone-giovanna-struck-with-little-warning

IDDRI STUDY 09/2013 47

The State of Environmental Migration 2013

CYCLONE EVAN IN SAMOA ANNE MELDAU

INTRODUCTION: NATURAL of risk were based on the likelihood of a disaster HAZARD AND DISASTER EVAN caused by a hazard and the consequences of the hazard. In December , a tropical depression formed in the South Pacific, which was subsequently named Table 1. Highest risk hazards for Samoa Tropical Cyclone Evan and eventually developed Risk Level Type of Hazard into a Category  tropical cyclone. In its path, Evan Extreme Cyclone (inluding stormsurge) brought heavy winds and rain to the island nations Extreme Volcanic Eruption of Tonga, Fiji, Samoa and the French territory of Extreme Tsunami Wallis and Futuna. This case study will present an Extreme Apia: Urban Fire analysis of the cyclone’s impacts on Samoa with Extreme Public Health Crisis a particular focus on migratory effects including Extreme Environmental Crisis: Invasive Species evacuations, shelter, reconstruction and possible High (Inland) Flooding (due to heavy rain) relocation in the Small Island Developing States High Earthquake (SIDS). High Landslide Pacific SIDS, such as Samoa, are often portrayed High Forest Fires as the first victims of climate change in interna- High Airport: Aircraft Emergency tional negotiations, especially in the context of sea-level rise. Apart from this, they rarely make High Hazchem Incident: Marine news headlines in Europe or the United States Source: based on “Table 1: Highest risk hazards for Samoa” (National Disaster of America in case of a disaster. This is arguably Council 2011:18) due to their isolation and small size, particularly Tropical Cyclone Evan is a typical example of a with respect to their populations. Yet, the size of natural hazard that has turned into a disaster. It Samoa’s population – , (OCHA, b) – will be analysed as such in this case study. First, should not impede their inclusion in analyses of the essential facts on Cyclone Evan and its impact natural disasters and related migration. The South on Samoa will be presented. The section will in- Pacific is one of the most natural-disaster-prone clude a description of the storm, information on its regions of the planet and therefore a “hotspot” of economic effects and initial disaster response and climate change not only with respect to rising sea management as well as insights into the context of levels. For Samoa, sea-level rise, coastal erosion natural disasters in Samoa and the country’s vul- and climate change are mentioned, among others, nerability to them. Second, the migratory effects as key issues related to the country’s natural en- of the cyclone will be addressed including evacu- vironment in the National Disaster Management ation and return of the affected population, re- Plan (National Disaster Council, :). Addi- building and relocation. Finally, a conclusion will, tionally, Samoa’s exposure to natural hazards is finally, summarize the findings and put them into highlighted, including ever-present threats such as their broader context, the state of environmental volcanic eruptions, earthquakes and tsunamis as migration in . well as seasonal threats such as droughts, floods and tropical cyclones (ibid.:). The following table shows the “extreme”- and “high”-risk hazards identified in a risk assessment for Samoa. Levels

IDDRI STUDY 09/2013 49 The State of Environmental Migration 2013

Map 1. Map of Islands Affected by Evan

Source: OCHA 2012b

1. A STORMY AFFAIR: EVAN’S issued a special weather bulletin including a MARK ON SAMOA storm warning, hurricane watch and flood advisory (Samoa Meteorological Services, ; Radio New Zealand, ). Evan was stationary over .. Natural hazard Evan: the country for about  hours, where it intensi- wind, rain and floods fied into a Category  tropical cyclone; the RSMC upgraded it at  p.m. on  December (OCHA, On  December, a weak tropical depression a; Budvietas, ). The main damage was was noticed about  km northeast of Suva, done during the late afternoon of  December Fiji (Government of Samoa, :). At  a.m. in a two-hour period from  p.m. to  p.m. (ABC (Samoan time) on  December , the US Joint Radio Australia, a). Of the two main islands Typhoon Warning Centre issued a tropical cyclone of the country, Upolu and Savai’i, the former was advisory. Around  p.m. the Tropical Depression particularly affected with significant damages F was upgraded to a Category  tropical cyclone particularly on the south coast and in the capital and named Evan by the Regional Specialized Mete- Apia (OCHA, a, b). After leaving Samoa, orological Centre, RSMC (Budvietas, ; Fiji Evan continued towards Fiji, transforming into Meteorological Service, a). Evan intensified a Category  tropical cyclone (OCHA, d). At and was upgraded by the RSMC to a Category  its peak, Evan reached wind speeds of  km/h storm at  a.m. on  December (Budvietas, ). according to the US Joint Typhoon Warning That night, the Samoa Meteorological Services Centre (Budvietas, ).

50 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 2. Tropical Cyclone Evan Forecast Track Map

Samoa

Source: Fiji Meteorological Service, 2012b

In Samoa Evan brought winds of up to  km/h missing persons was called off in mid-January and a total rainfall between  a.m. on  December (Press Secretariat Government of Samoa, a). and  a.m. on  December that amounted to  These included fishermen lost at sea (OCHA, mm in Afiamalu and  mm in Nafanua, both on c; Budvietas, ). Upolu Island (OCHA, a). Apart from the capi- In the immediate aftermath of the cyclone, the tal, the rural areas of western Upolu and the south- whole of Samoa was without electricity and run- coast villages Siumu, Safata and Lefaga were ning water (OCHA, a). There was substantial most severely affected (Government of Samoa, damage to three hydropower plants, minor dam- :). Flooding of the banks of the Vaisigano age to two others (Government of Samoa, :) River caused by the cyclone also led to major dam- and a significant amount of debris-restricted ac- ages (ONE News, a). For example the floods cess to affected areas (OCHA, g). of the river, which runs into Apia, completely de- There was severe damage to infrastructure, in- stroyed the village of Magiagi with water rising up cluding roads and buildings, and many trees were to five metres high (ONE News, ) and led to downed (Budvietas, ; Government of Samoa, severe flash floods in the capital (OCHA, a). :). To illustrate this, the Post Disaster Needs Assessment (PDNA) states that “the estimated val- .. Natural disaster Evan: ue of destroyed physical assets represents  per destruction, damages and cent of the normal value of construction activities loss of human lives in Samoa” (Government of Samoa, :xiii). The tourism industry stated that damage was A total of five people were killed by Cyclone worse than after the  tsunami (OCHA, e) Evan in Samoa (Government of Samoa, :) and amounted to USD . million. including two children (Budvietas, ) and it has Additionally, revenue losses are projected to been pointed out that this number stayed compa- sum up to USD . million for the time period from rably low because the storm hit during daytime  to  (Government of Samoa, :f). (ABC Radio Australia, a). The search for  For the agriculture and fisheries sector damage

IDDRI STUDY 09/2013 51 The State of Environmental Migration 2013

and loss are estimated at about USD . million to coordinate the emergency response (OCHA, (ibid.:). The worst hit areas suffered losses of a). It included four sub-committees: First up to  per cent of banana trees,  per cent of Response, Community Welfare and Internal Dis- breadfruit,  per cent of other economically im- placed Persons, Housing Reconstruction and portant crops such as cocoa, citrus and coffee as Settlement, Early Recovery and Recovery Needs well as  per cent of root crops (Press Secretariat Assessment (Press Secretariat Government of Sa- Government of Samoa, ). These losses were moa, ). concentrated on Upolu Island, where the worst-hit Restoration of the water and power supply as areas were the south-western, central and south- well as providing support for the shelters were ern parts and the central highlands (OCHA, c; identified as the most critical humanitarian priori- Government of Samoa, :). On the contrary, ties (OCHA, g). Debris was cleared by  Janu- Savai’i Island was hardly touched and could pro- ary and water supply restored up to  per cent by vide food supplies to the rest of the country (ABC  January (Press Secretariat Government of Sa- Radio Australia, a). However, many families moa, a). The Electric Power Corporation had lost their private gardens, which are used for per- restored % of electricity on Upolu Island by  sonal consumption, as well as parts of their root January (Press Secretariat Government of Samoa crop productions, used for subsistence and sell- b). With respect to the emergency centres, the ing (European Commission, ). About , situation in Apia was reported as good with less in- households, equivalent approximately to , formation on those shelters located in rural areas people, were affected and many small farmers are (OCHA, f). Issues raised regarding the shel- expected to lose around half of their  annual ters were mainly the supply of drinking water and income (Samoa Observer, a). Consequently, sanitation (OCHA, b), food (OCHA, d) people with low incomes and subsistence farm- and later on also safety due to a lack of police pres- ers were most adversely impacted by the cyclone ence (OCHA, e). The government provided (Government of Samoa, :xiv ff.). water, food and non-food items (Press Secretariat The total economic effects of the cyclone were Government of Samoa, ) and installed pit la- also substantial. Initial estimates of total damages trines (Press Secretariat Government of Samoa, were made at more than USD  million (ABC Ra- a). One issue reported in the PDNA was that dio Australia, b; Samoa Observer, a) and shelters were prioritized for distribution of relief the PDNA set the total effects to USD . million items, which was particularly problematic as peo- (Government of Samoa, :xiii). For a country ple were not aware of this prioritization (Govern- with a GDP of approximately USD . billion (CIA, ment of Samoa, :). ), this represents a substantial sum amounting In addition to provisions from the government to .% of GDP. The public sector incurred  per and local religious and humanitarian organiza- cent of the disaster effects and a total loss of about tions, international assistance was important. , equivalent jobs was estimated (Government Foreign governmental help was received from of Samoa, :xivff). As a consequence of pro- New Zealand, Japan, Australia, the United States, duction losses and increased expenditures, GDP American Samoa and the EU. International organ- forecasts for the next years were lowered. Growth izations including UNDP, WHO, OCHA, UNESCO, might be zero or negative for  and might not the Asian Development Bank and the World Bank fully recover in  (ibid.:). also provided assistance (OCHA, b, d, g; European Commission, ; ADB, ; .. Initial disaster Samoa Observer, a). International help most- response and subsequent ly encompassed shelter, sanitation, tools as well as disaster management technical expertise (Press Secretariat Government of Samoa, ). A declaration of disaster was issued on the national Damage assessments commenced directly after radio around  p.m. on  December by the deputy the storm. Initial ground assessments were com- prime minister who was also the acting chairman pleted for Upolu Island on  December and in- of the National Disaster Council (Tusani, ). dicated considerable damage while reports from When it expired after  hours it was renewed Savai’i Island suggested that damage was minimal until midday on  December (OCHA, b). there (Budvietas, ). The Post Disaster Needs That day, a state of emergency was declared for  Assessment and Damage and Loss Assessment mis- days (OCHA, c). Parliament reconvened on  sion took place from  to  January with the infra- December (Samoa Observer, ). structure and shelter parts starting on  January In the morning of  December the National (OCHA, a). For the first time ever UNESCO Emergency Operations Centre was activated additionally undertook an assessment of the

52 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

impacts on the cultural sector, especially concern- In regard to other disasters, the  tsunami ing cultural heritage, which were moderate with destroyed  villages, killed  people, displaced the exception of some landmarks suffering severe , and caused damage of approximately damage (UNESCO Apia, ). USD  million (Fickling, ). Therefore, though less deadly, the economic damage of Evan is probably at least as severe. A staff member of .. Putting the storm into the International Monetary Fund put the two into context: Evan and other relation: “The disaster caused a significant disrup- natural hazards in Samoa tion to economic activity, which had just recov- ered following the devastating effects of the  Samoa is located in Polynesia in the South Pacific. earthquake and tsunami” (IMF, ). It is in the This region is one of the most vulnerable places context of recurring disasters that policy responses in the world to natural disasters with up to today to Evan need to be evaluated. millions of people affected by them and billions of dollars of economic losses (Bettencourt et al., :viii; The International Bank for Recon- 2. EVAN AND ENVIRONMENTAL struction and Development and The World Bank, MIGRATION :). Pacific nations figure prominently among the world’s countries with the highest annual losses from disasters as a proportion of .. The difficulty of keeping GDP (The World Bank, ). For Samoa, the track: evacuation and average annual impact on GDP in disaster years damage to houses (for the period from  to ) was . per cent with . per cent average impact in all years As a consequence of damages and destruction (Bettencourt et al. : ) and the country is caused by high winds and the flooding of the expected to suffer USD  million/year in losses Vaisigano River banks, about , people were caused by earthquakes and cyclones (SPC/SOPAC evacuated (Government of Samoa, :). Giving et al., :). In the region, cyclones are the most concrete estimates of the people subsequently common type of disaster and climate change is staying in evacuation centres is difficult due to a intensifying the situation. Over the past  years, number of reasons. the overall number of hurricane-strength cyclones First, people went to official evacuation centres has increased in the region. Today, an average of in rural areas and in Apia but also turned to infor- four of these cyclones occurs in the south-west mal centres, which were often managed by denom- Pacific every year. Additionally, the intensity of inational groups (OCHA, b). This is explained cyclones increases. Wave heights, for example, as Samoans have strong religious ties and Church have exceeded the predictions made in climate plays a crucial role in everyday life (National Disas- change models (Bettencourt et al., :). ter Council, :). Second, some families went These observations also apply to Samoa, which to stay with relatives or neighbours instead of turn- lies in the South Pacific tropical cyclone belt ing to an evacuation centre (Press Secretariat Gov- (Sutherland et al., :) and whose tropical ernment of Samoa, ). This was encouraged by cyclone risk is classified as “extreme” (National the National Emergency Operations Centre, which Disaster Council, :). The / official requested people in rural communities to accom- cyclone season was  November  to  April modate evacuees (OCHA, a). Third, while  with mid-February to mid-March historically numbers seemed to be decreasing in the days im- being the period with most tropical cyclones in the mediately after the storm as people returned home South Pacific; between  and  named cyclones (OCHA, b), they started to increase again at were forecast (OCHA, d). Therefore, expe- the end of December. Apparently, some “families riencing a damaging storm as early as December that initially evacuated to family/friends […] de- means that the affected areas face the risk of being cided that their needs are better met at the centres hit again and enduring severe damage in the fol- (timely food and supplies which the distribution lowing months. With respect to Evan, an ensuing evacuation centres have prioritised)” (Press Sec- tropical depression brought more rain, wind and retariat Government of Samoa, ). These three flooding to Apia in mid-January (Tupufia, a). reasons make that different numbers were given If Samoa have experienced various natural disas- by the government and international organiza- ters including storms, Evan has caused devastation tions such as the UN Office for the Coordination of that had not been seen since Cyclone Val in  Humanitarian Affairs (OCHA). The following dia- (Budvietas, ; Government of Samoa, :). gram displays the statistics provided by the latter.

IDDRI STUDY 09/2013 53 The State of Environmental Migration 2013

According to OCHA, the number of people in is essential for targeted disaster management and evacuation centres/shelters (see Side Note: Ter- the coordination of relief efforts. minology for a discussion of terms) therefore In addition to the destroyed houses,  were equalled . per cent of the total population of Sa- partially damaged and therefore unsafe so that moa (,) at its peak (OCHA, b). a total of  houses were left in need for recon- The actual number can be considered to be high- struction. The remaining , affected houses suf- er as neither all informally-run centres nor those fered minor damages. The overall damage and loss people seeking shelter with relatives or friends are including household goods is estimated at USD  included in these official counts. Furthermore, the million. Most of the damages occurred on more figures should, by no means, be seen as definite populous Upolu Island in Vaimauga West (north); and complete; additional information and compet- followed by Safata, Siumu (both south); Falealili, ing numbers for the same dates are also available. Lefaga ma Faleseela (both south); Anoamaa West For example, Caritas Australia () indicates and Vaa o Fonoti (both north-east) (Government ,+ evacuees staying at the basement hall of of Samoa, :ff.). the Caritas Samoa office on  December. For  Thus, following the cyclone an estimated , December, the ADB () states ,+ people people (based on average household size of sev- evacuated. Therefore, all figures should be seen en members) were in immediate need of finan- as incomplete estimates. They can nevertheless cial and technical assistance with reconstruction give an indication of the number of people that (ibid.:xvii). Applying this average household evacuated to various formal or informal places. size to the total number of house damages, an Additionally, the differing numbers illustrate the estimated total of , people were affected, challenge of keeping track with a sometimes rap- which represents about . per cent of the total idly evolving natural disaster situation, especially population. if people resort to informal solutions, such as help from family and if needs differ largely. This was for .. An official return of example the case with respect to the period people the affected population spent at the shelters. While some stayed only one night or a couple of days, others remained at the Returning home to rebuild was encouraged by shelters for the entire period they were open (Gov- the government already few days after the storm ernment of Samoa, :). (OCHA, d,  g). The formal evacuation The main evacuation centres were managed centres were closed at noon on  January, based by the Samoa Red Cross Society, the Disaster on a decision by the National Disaster Council on Advisory Committee, local government, Cari-  December, partly to prepare school buildings tas, Church Jesus Christ of Latter-day Saints and used as shelters for the beginning of the school Seventh Day Adventist churches, to give some year at the end of the month (Press Secretariat examples (Press Secretariat Government of Sa- Government of Samoa, b). However, as some moa ). This means that three main groups of people had nowhere to go because their houses actors were involved, the Samoan Government had been destroyed, it took in practice a couple (local and national level), humanitarian organi- of weeks until everybody had left. By March zations and churches. After Cyclone Evan was most of the temporarily displaced had returned the first time there were officially designated and lived in temporary homes constructed of emergency centres. These official shelters were tarpaulins, plastic sheeting, and timber. Some located in urban areas while people in rural com- people still lived with relatives or friends in over- munities sheltered in make-shift refuges such as crowded housing units (Government of Samoa, church halls or with other families (Government :). of Samoa, :). Thus, the closing of the official evacuation The strong winds and floods brought by Evan centres should not be interpreted as a change affected a total of , houses including  back to normal. The tropical cyclone season was that were completely destroyed (Government on-going making living conditions in temporary of Samoa :ff.). This number was reduced housing difficult and the returned population from originally almost  houses reported as de- vulnerable to new rains and flooding. stroyed (Press Secretariat Government of Samoa ), apparently because some people had put in multiple claims under different names or for outdoor toilets (ONE News, b). As was the case for evacuee numbers, this indicates the dif- . Interview with Dr Sue Vize, Officer-in-Charge, UNESCO ficulty of keeping track of disaster effects, which Apia, conducted on  March  via email.

54 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 3. Villages Affected by Cyclone Evan on Upolu Island

Source: Government of Samoa, 2013:12

.. Recovery in Samoa: Suggestions also included homes elevations to reconstruction and relocation protect from flooding (ibid.:). Medium-term requirements for the building sector proposed in Disaster response includes recovery, but can also the PDNA also included an update of the building be used to prepare for future developments such code (and ensuring it is correctly implemented), as the recurrence of natural disasters through risk retrofitting and an increase in home-insurance reduction measures. For cyclone-prone Samoa, coverage. During Evan it also became apparent not only reconstruction is an issue, but relocation that traditional Samoan houses were less damaged may become an adaptation strategy to environ- than European-style ones, something that was mental stressors as well. encouraged to be considered when rebuilding (ibid.:f.). Reconstruction constrained: financial Following the disaster, the Prime Minister stated challenges for an effective recovery of the that government has “been asking people to help most vulnerable households themselves rebuild their homes” (ABC Radio Aus- The PDNA estimates the total amount of money tralia, a). Similarly, the PDNA accords only a needed for recovery and reconstruction in the catalyzer and facilitator role to the government and housing sector alone at about USD . million focuses on technical advice ensuring that home and reconstruction to take two years. With respect owners “build back better using disaster-resilient to how reconstruction should go about, the PDNA standards” (Government of Samoa, :). recommends monitoring to ensure pre-disaster However, many residents had difficulties financ- vulnerabilities are not recreated (Government ing the rebuilding of their homes and sometimes of Samoa, :). For infrastructure some could not afford it at all (ONE News, a). Over- proposals include digging electricity poles deeper all access to financial means to purchase materials into the ground (ABC Radio Australia, a), for reconstruction was identified as an important changing the alignments of water pipes that constraint (Government of Samoa, :). typically follow river flows, use below-ground Assistance provided by the government included pipework and create upstream flood protection relief supply for reconstruction such as tents, tar- (Government of Samoa, :). With respect to paulins, tool kits and household items. While the trees both cutting those close to homes, electricity Disaster Advisory Committee’s Sub-Committee lines (ABC Radio Australia, a) and fences for Housing Reconstruction and Settlement had holding livestock (Government of Samoa, :) originally recommended considering only fami- as well as replanting them selected locations were lies living in shelters or camps for this distribu- proposed. The latter is due to the fact that defor- tion (OCHA, g) hand-outs were eventually estation near rivers and streams increased the given to families leaving evacuation centres as impact of flash floods (Samoa Observer, ). well as to those whose houses were destroyed or

IDDRI STUDY 09/2013 55 The State of Environmental Migration 2013

severely damaged. The total cost of the distribution c). This contrasts the  tsunami response amounted to USD , (Government of Samoa, because there were a lot of overseas donations :; OCHA, b). Furthermore, the Min- making the hand-out of rebuilding grants possible istry of Natural Resources and Environment and after the tsunami, but the government could not the Fire and Emergency Services helped with the afford such an approach following Cyclone Evan tree cutting and the latter continued to assist with as the international aid levels were different (ONE cleaning houses until mid-January (Press Secretar- News, b). In such a situation, non-govern- iat Government of Samoa, a). Despite these as- mental assistance becomes more important. With sistance efforts some media reports suggested that respect to rebuilding after Evan, examples of this little seemed to have been done in some areas such include the OIL group who handed out cheques to as the Magiagi village (ONE News, b). assist employees with reparation (Samoa Observ- Apart from this direct assistance, the government er, c) and the Fellowship of Ministers Samoa also provided USD  million in housing assistance New Zealand who provided vouchers for the pur- in the form of a loan scheme at the end of Febru- chase of building materials (Ta’ateo, ). Finally, ary. The loan was provided interest-free during the remittances from family members abroad, which first year with a subsequent interest rate of  per are important for many households’ incomes in cent. Mortgage loans could amount to USD , Samoa, have been identified as important for re- and personal loans to USD ,. The money is building as well (Government of Samoa, ; Es- part of a USD  million package whose remain- era, b; Samoa Observer, d). der was given to the Development Bank of Samoa Rebuilding faces many challenges and notable to help the business community. It is lent to the financial constraints, making recovery particularly Samoa Housing Corporation by the Central Bank. difficult for low-income households. This is prob- The CEO of the corporation stated that they were lematic as they are among the most vulnerable of “barely making enough to cover our administra- the population. Therefore it is also questionable tion costs” (Samoa Observer, b). While the low whether high expectations to build back better and interest rate is attractive, it has to be pointed out in a resilient way will be widely met. that on the one hand some families require more money which would have to be loaned at the nor- The complexity of relocation mal interest rate of  per cent (Tupufia,  b). One motivation to leave one’s home could be On the other hand, few households took loans to the loss of livelihoods. It has been observed that cope with their recovery needs (Government of seasonal employment schemes in New Zealand and Samoa, :). Additionally, rumours emerged Australia have attracted farmers to leave Samoa that only government employees were eligible for job opportunities with higher incomes (Esera, for the scheme (Esera, a). Even though these c). Though no rural household interviewed rumours were promptly dismissed by officials, for the PDNA indicated the intention to change the controversy nevertheless illustrates problems their livelihoods (Government of Samoa, :) with the loan scheme approach (Samoa Observer, these potential migration trends might be initiated b). Its main problem, however, is that while by the effects of Cyclone Evan. Another motiva- the scheme might be well-intended it does not action could be the changed course of the Vaisigano tually help those most severely impacted by Evan, River, that causes continued flooding of some namely low-income families and poor subsistence houses especially in the Magiagi Valley whenever farmers in rural areas. It also highlights a govern- there is heavy, constant rain (ONE News, a; ment dilemma that also applies to other develop- Samoa Observer, e). ing countries faced with a disaster, namely limited Finally, moving can be motivated in anticipation financial resources. This problem was highlighted of future impacts. Related to flooding and going by the Prime Minister (ONE News, a), who also beyond it the Prime Minister stated that “the valua- pointed out that a lot of aid received from donors ble lesson following this cyclone is that we need to is earmarked and cannot simply be used for other warn people that are living close to rivers that they purposes by the government (Hazelman-Siona, ought to shift, as well as those who live very close ). The Prime Minister also defended the gov- to the coastal areas, that they need to shift” (ABC ernment’s prioritization: “In regards to housing not Radio Australia, a). This call was reiterated by the entire country was affected or had their houses destroyed – only a small number due to strong . This illustrates the blurry distinctions between differ- winds and flooding but the entire island of Upolu ent types of migration. While the seasonal employment was affected when the power went out” (ibid.) schemes could be seen as economic/work migration some of it might in fact be motivated by the environ- This line was subsequently promoted by the mental impacts of the storm on the livelihoods of the Finance Minister as well (ABC Radio Australia, migrants and their relatives who stay behind.

56 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

the Acting Chief Executive Officer of the Meteorol- rural environments, namely the ties not only be- ogy Office in light of mid-January floods caused by tween a family and its ancestral land (and links to a subsequent tropical depression (Tupufia, a). current livelihoods) but also within communities. New Zealand engineers carried out a survey of the These community ties would make it extremely flooded area of the Vaisigano River. They recom- difficult to live in a village one does not belong to mended relocation aside reinforcing riverbanks (ibid.:). and forcing buffer zones (ONE News, a). Some Therefore different attitudes towards relocation people have already moved to higher ground (ONE can be identified among the population of Samoa. News, a). With approximately  per cent of Simply speaking some people want to stay while the population living within one kilometre of the others want to move (Esera, b). But a more coast and critical infrastructure being primarily nuanced evaluation can be made: there are indeed located in coastal areas (Government of Samoa, some, notably from rural areas, that do not want to :), Samoa is highly vulnerable not only to relocate due to their attachment to their land and flooding but also to tsunamis and eventually sea- community and prefer rebuilding in a way that level rise. Therefore it is understandable that re- would be more resilient (Government of Samoa, location was highlighted by communities and the :). A second group has already moved to government as a key issue in the PDNA (ibid.:). higher ground (ONE News, a). A third group With respect to infrastructure, the assessment indicates an interest in relocating but is unable to recommended the inland relocation of some do so mostly due to financial constraints (Govern- health facilities (ibid.:), the reconstruction of ment of Samoa, :). Thus the remark in the schools while ensuring that they are not – as was PDNA that most of the households seem to favour previously the case in Vaimauga West – located on local reconstruction as the easiest and most afford- an ancient riverbed or in a floodplain (ibid.:) able option (ibid.:) obscures the fact that “easy” and the relocation of damaged water and sanita- and “affordable” do not always equal “preferred”. tion stations from vulnerable to safer locations Finally, a fourth group of people interviewed for (ibid.:). Additionally it was proposed that live- the PDNA was struggling with their daily survival stock be moved to higher ground and away from or still reeling psychologically after the disaster rivers and watershed areas (ibid.:). so that they had not thought about relocation With respect to those recommendations, the (ibid.:). This means that proportions of those feasibility of relocation appears as one of the main who do not want to leave, those who actually leave challenge. First, the availability of land – espe- and those who would like to leave but cannot, cially inland (ibid.:) – is limited in Samoa and might shift over time. where available, its purchase requires substantial While the PDNA highlights the important point financial and time resources (ibid.: & ). The that “relocation cannot be imposed” (ibid.:) the land on higher grounds is particularly expensive situation in Samoa post-Evan therefore illustrates and has oftentimes already been bought up by the relationship between migration (relocation) well-off people (Lesa, ). This makes it diffi- and adaptation. Each individual should have the cult, if not impossible, for low-income families, the possibility to choose its adaptation strategy, that most adversely affected by the cyclone, to relocate means between staying (and rebuilding in more (Government of Samoa, :). Second, inherit- disaster-resilient ways) or leaving. Both policies ance is a problem. On the one hand people do not should be supported by government policies. This necessarily want to leave what they inherited from case study indicates that government actions fa- their ancestors (Lesa, ). On the other hand, in vour rebuilding locally over relocation. The recom- Samoa customary land cannot be sold. This poses mendation of the PDNA to “conduct a broad-based problems for affected families in vulnerable rural community consultation process for relocation settings who have houses on their own land and that includes community wishes and needs as would not necessarily find alternatives (Govern- well as government considerations for change” ment of Samoa, :). Furthermore disaster (ibid.:) does not seem to reflect the current threats often exist for entire villages so that mov- situation but is a step in the right direction. ing within the village would not reduce the level of risk. This hints at a third constraint, especially for . This does not only refer to the reconstructed houses but also the livelihoods. One of the main lessons from cyclone Evan according to the Prime Minister was for . It has however been highlighted that major ongoing example the need to place more emphasis on root crops infrastructural projects such as the Apia Waterfront that resist better to storms and could therefore limit food Redevelopment or a new government building were con- shortage as well as negative impacts on the livelihoods of tinuing (Press Secretariat Government of Samoa b) (subsistence) farmers (ABC Radio Australia a; Gov- while they might better be shifted inland (Lesa ). ernment of Samoa :).

IDDRI STUDY 09/2013 57 The State of Environmental Migration 2013

CONCLUSION: THE STATE OF individuals with respect to natural disasters, such ENVIRONMENTAL MIGRATION as Cyclone Evan, varies largely, which can result from personal choice. Following Evan and with IN SAMOA AFTER EVAN respect to relocation these were returning/stay- Evan was the most damaging natural disaster ing and leaving. Ideally policies take this into ac- to hit Samoa in recent years. It left  houses count. Yet, this case study shows that government in need of reconstruction, caused an estimated strategies are not always capable of doing so. With value of USD  million of post-disaster finan- respect to Cyclone Evan, the problem was particu- cial requirements and displaced , people to larly financial constraints. International assistance evacuation centres at peak displacement periods was often earmarked to particular humanitarian (Government of Samoa, :xiv) This highlights or recovery objectives and the government did not a couple of aspects of the state of environmental have the financial resources to fund large rebuild- migration in . ing or even relocation projects. This was highlight- On the one hand this case study underscores ed by the IMF, who suggested government caution the challenges faced by private households and against external loans and prioritization of grants the diversity of individuals’ situations and needs. as the public debt is high and the fiscal deficit Some are stuck in temporary structures next to would be increased by loans (Samoa Observer, their destroyed houses because they cannot afford f). In turn, this implies that if no grant is pro- to rebuild. But even if reconstruction is affordable, vided the government faces the choice between in- this might not be what individuals actually want. creasing public debt or limiting disaster response The analysis of relocation attitudes suggests that measures. Therefore, much hope is placed in an some Samoans would prefer to relocate to less dis- insurance pilot launched by the World Bank in aster-prone places but are unable to do so because January . Samoa, the Marshall Islands, the of high relocation costs and no viable alternative Solomon Islands, Tonga and Vanuatu participate destinations. They could, therefore, be considered in the Pacific Catastrophe Risk Insurance Pilot that as being forced to stay and policies should ide- will test whether a risk transfer arrangement mod- ally facilitate out-migration. But at the same time elled on an insurance plan can help small-island many Samoans affected by the cyclone want to nations cope with the financial effects of natural stay. For them, adaptation to disaster risks means disasters (The World Bank, ). For Samoa, the building back better, adopting more resilient live- pilot programme would be particularly valuable lihoods and increasing disaster preparedness. In since the government has indicated that it does this context, the fact that none of the communities not plan to request a postponement of Samoa’s visited for the PDNA had a general disaster prepar- graduation from LDC (Least Developed Country) edness plan is worrying (Government of Samoa, status scheduled for  (Press Secretariat Gov- :). Similar to post-tsunami disaster-prepar- ernment of Samoa, b). As many developed edness education initiatives (Nemerever, ) cy- countries set aside financial support particularly clones need to be addressed by awareness-raising for LDCs, graduation could intensify public budget projects especially as they present the biggest haz- constraints of the Samoan government and there- ard for the country. Government policies can sup- fore make it more difficult for the country to deal port all of these elements through well-enforced with future disasters. and strict building codes or disaster preparedness For the future, Samoa must find resources to campaigns. With respect to changing livelihoods, adapt to environmental challenges. With total eco- Samoa already has in place a stimulus package to nomic loss and damage amounting to USD . incentivise farmers to plant traditional crops (Es- million (Government of Samoa, :xiii) and thus era, c). This could be modified to particularly .% of the country’s GDP (CIA, ), cyclone encourage root crop farming. Evan demonstrated that financial constraints are Beyond this, this case study shows the need a main explanatory factor for response or rather for nuanced policies. The coping strategy of non-response measures in Samoa.

58 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

OCHA. 2012d. Pacific: Tropical Cyclone Evan: Situation Re- BIBLIOGRAPHY port No. 4. 19 December, http://www.phtpacific.org/disaster-emergency/tc-evan-12-td-04f (consulted on 26 April A  R 2013). OCHA. 2012e. Pacific: Tropical Cyclone Evan: Situation Re- ADB. 2012. “ADB provides $ 500,000 to Samoa for cyclone port No. 5. 20 December, http://www.phtpacific.org/dis- relief efforts.” In News from country offices, 28 December, aster-emergency/tc-evan-12-td-04f (consulted on 26 April http://www.adb.org/news (consulted on 26 April 2013). 2013). Bettencourt, Sofia et al. 2006. Not if but when: adapting to OCHA. 2012f. Pacific: Tropical Cyclone Evan: Situation Report natural hazards in the Pacific Islands Region. A policy note. No. 6. 23 December, http://www.phtpacific.org/disaster- The World Bank. emergency/tc-evan-12-td-04f (consulted on 26 April 2013). Caritas Australia. 2012. “Samoa: Caritas provides emer- OCHA. 2012g. Pacific: Tropical Cyclone Evan: Situation Re- gency shelter for thousands in cyclone battered Samoa.” In port No. 7. 28 December, http://www.phtpacific.org/dis- Media release, 16 December, http://www.caritas.org.au/ aster-emergency/tc-evan-12-td-04f (consulted on 26 April (consulted on 26 April 2013). 2013). CIA. 2013. Samoa. The World Factbook, https://www. OCHA. 2013a. Pacific: Tropical Cyclone Evan: Situation Re- cia.gov/library/publications/the-world-factbook/geos/ port No. 8. 3 January, http://www.phtpacific.org/disaster- ws.html (consulted on 26 April 2013). emergency/tc-evan-12-td-04f (consulted on 26 April 2013). CRED. 2012. “Samoa country profile: natural disasters.” OCHA. 2013b. Pacific: Tropical Cyclone Evan: Situation Re- EM-DAT: the OFDA/CRED International Database, http:// port No. 9. 11 January, http://www.phtpacific.org/disaster- www.emdat.be (consulted on 26 April 2013). emergency/tc-evan-12-td-04f (consulted on 26 April 2013). European Commission. 2013. Commission Implementing OCHA. 2013c. Pacific: Tropical Cyclone Evan: Situation Re- Decision of 13.2.2013 financing emergency humanitarian port No. 10. 25 January, http://www.phtpacific.org/dis- actions in Fiji and Samoa from the general budget of the Eu- aster-emergency/tc-evan-12-td-04f (consulted on 26 April ropean Union. C(2013) 950 final. 2013). Fiji Meteorological Service. 2012a. Media Release: Tropi- OCHA. 2013d. Climate Update Pacific Issue 02 January 2013, cal Cyclone ‘Evan’. 12 December 2012, from www.met.gov.fj http://reliefweb.int/report/samoa/climate-update-pacific- (consulted on 26 April 2013). issue-02-january-2013 (consulted on 26 April 2013). Fiji Meteorological Service. 2012b. Severe Tropical Cyclone EVAN Forecast Track Map issued 1703 UTC. 14 December Press Secretariat Government of Samoa. 2012. Press Release 2012, https://www.facebook.com/pages/Fiji-Meteoro- No. 10: Cyclone Evan relief efforts. 28 December, https:// logical-Service/112621892096909 (consulted on 26 April www.facebook.com/PacificTV.info (consulted on 26 April 2013). 2013). Government of Samoa. 2013. Samoa post-disaster needs as- Press Secretariat Government of Samoa. 2013a. Press Release sessment cyclone Evan 2012. March 2013. No. 13: Cyclone Evan relief efforts. 15 January, https://www. facebook.com/PacificTV.info (consulted on 26 April 2013). IMF 2013. “IMF reaches staff-level agreement with Samoa on economic recovery program to be supported under rapid Press Secretariat Government of Samoa. 2013b. Press Re- credit facility.” Press release 13/95, 26 March, http://www. lease No. 11: Cyclone Evan relief efforts. 6 January, http:// imf.org (consulted on 26 April 2013). www.savalinews.com/ (consulted on 26 April 2013). National Disaster Council. 2006. Samoa National Tropical Samoa Meteorological Services. 2012. Special weather bulle- Cyclone Plan. tin number three (3) for Samoa. 13 December, http://www. webcitation.org/6CrPHSKmf (consulted on 26 April 2013). National Disaster Council. 2011. Samoa’s National Disaster Management Plan 2011-2014. SPC/SOPAC et al. 2011. Country risk profile: Samoa. Pacific Catastrophe Risk Assessment and Financing Initiative, http:// Nemerever, Zoe. 2012. Preparing for the worst: disaster www.worldbank.org/ (consulted on 26 April 2013). preparedness education in Samoa. Independent Study Pro- ject (ISP) Collection 1281, http://digitalcollections.sit.edu/ Sutherland, Karen et al. 2005. Vulnerability in Samoa. Tiem- isp_collection/1281 (consulted on 26 April 2013). po 54, www.tiempocyberclimate.org (consulted on 26 April 2013). OCHA. 2012a. Pacific: Tropical Cyclone Evan: Situation Re- port No. 1. 14 December, http://www.phtpacific.org/dis- Ta’ateo, Jonathan. 2013. “Ministers Fellowship lends help- aster-emergency/tc-evan-12-td-04f (consulted on 26 April ing hand.” In Samoa Observer, 22 March, http://samoaob- 2013). server.ws/ (consulted on 26 April 2013). OCHA. 2012b. Pacific: Tropical Cyclone Evan: Situation Re- The International Bank for Reconstruction and Develop- port No. 2. 15 December, http://www.phtpacific.org/dis- ment and The World Bank. 2010. Natural hazards, unnatu- aster-emergency/tc-evan-12-td-04f (consulted on 26 April ral disasters: the economics of effective prevention. The World 2013). Bank, Washington DC. OCHA. 2012c. Pacific: Tropical Cyclone Evan: Situation Re- The World Bank. 2013. “5 Pacific island nations to be in- port No. 3. 17 December, http://www.phtpacific.org/dis- sured against natural disasters.” Press release, 18 January, aster-emergency/tc-evan-12-td-04f (consulted on 26 April http://www.worldbank.org/en/news (consulted on 26 2013). April 2013).

IDDRI STUDY 09/2013 59 The State of Environmental Migration 2013

Tupufia, Lanuola. 2013a. “Tropical depression brings heavy ONE News. 2012. “Recovery effort builds in Samoa.” In rain.” In Samoa Observer, 13 January, http://samoaobserv- TVNZ ONE News, 15 December, http://tvnz.co.nz/news er.ws/ (consulted on 26 April 2013). (consulted on 26 April 2013). Tupufia, Lanuola. 2013b. “Housing loan limit, $30,000.” In ONE News. 2013a. “Samoa struggling to recover from Cy- Samoa Observer, 5 March, http://samoaobserver.ws/ (con- clone Evan.” In TVNZ ONE News, 28 February, http://tvnz. sulted on 26 April 2013). co.nz/news (consulted on 26 April 2013). UNESCO Apia. 2013. Cyclone Evan in Samoa. UNESCO Of- ONE News. 2013b. “Samoans told to prepare for more natu- fice in Apia, 5 February, http://www.unesco.org/new/en/ ral disasters.” In TVNZ ONE News, 27 February, http://tvnz. apia/ (consulted on 26 April 2013). co.nz/news (consulted on 26 April 2013). Radio New Zealand. 2012. “Samoa and American Samoa is- P A: sue alerts for tropical cyclone Evan.” In RNZI, 12 December, http://www.rnzi.com (consulted on 26 April 2013). ABC Radio Australia. 2013a. “Samoans urged to move away Samoa Observer. 2012. “Prime Minister addresses nation.” from coast and riverbanks.” In ABC Radio Australia, 11 In Samoa Observer, 18 December, http://samoaobserver. January, http://www.radioaustralia.net.au/international/ ws/ (consulted on 26 April 2013). (consulted on 26 April 2013). Samoa Observer. 2013a. “World Bank grants $ 44m.” In Sa- ABC Radio Australia. 2013b. “Australia announces further moa Observer, 15 March, http://samoaobserver.ws/ (con- aid for Samoa rebuild.” In ABC Radio Australia, 12 February, sulted on 26 April 2013). http://www.radioaustralia.net.au/international/ (consult- Samoa Observer. 2013b. “Housing boss sets record straight.” ed on 26 April 2013). In Samoa Observer, 13 March, http://samoaobserver.ws/ ABC Radio Australia. 2013c. “Samoan rebuilding loans tar- (consulted on 26 April 2013). geted at low income earners.” In ABC Radio Australia, 11 Samoa Observer. 2013c. “Oil Group helps Evan victims.” March, http://www.radioaustralia.net.au/international/ In Samoa Observer, 17 March, http://samoaobserver.ws/ (consulted on 26 April 2013). (consulted on 26 April 2013). Budvietas, Sophie. 2012. “Evan’s destructive path.” In Sa- Samoa Observer. 2013d. “Evan helps tourism, remittances.” moa Observer, 18 December, http://samoaobserver.ws/ In Samoa Observer, 13 March, http://samoaobserver.ws/ (consulted on 26 April 2013). (consulted on 26 April 2013). Esera, Unumoe. 2013a. “Hardship prevails.” In Samoa Ob- Samoa Observer. 2013e. “Samoa struggling to recover, says server, 12 March, http://samoaobserver.ws/ (consulted on TVNZ.” In Samoa Observer, 5 March, http://samoaobserv- 26 April 2013). er.ws/ (consulted on 26 April 2013). Esera, Unumoe. 2013b. “Victims embrace government Samoa Observer. 2013f. “IMF cautions government against help.” In Samoa Observer, 3 March, http://samoaobserver. loans.” In Samoa Observer, 4 April, http://samoaobserver. ws/ (consulted on 26 April 2013). ws/ (consulted on 26 April 2013). Esera, Unumoe. 2013c. “Local farmers rewarded.” In Samoa Tusani, Lanuola. 2012. “Cyclone Evan stops country.” In Observer, 4 March, http://samoaobserver.ws/ (consulted Samoa Observer, 14 December, http://samoaobserver.ws/ on 26 April 2013). (consulted on 26 April 2013). Fickling, David. 2012. “Cyclone Evan impact on Samoa’s economy as bad as 2009 tsunami.” In Bloomberg, 22 Decem- ber, http://www.bloomberg.com (consulted on 26 April I: 2013). Dr Denis Chang Seng, Natural Sciences Program Adviser, Hazelman-Siona, Niccola. 2013. “Aid tagged for specifics, UNESCO Apia, conducted on 11 March 2013 via email. says P.M.” In Samoa Observer, 5 March, http://samoaobserver.ws/ (consulted on 26 April 2013). Dr Sue Vize, Officer-in-Charge, UNESCO Apia, conducted on 11 March 2013 via email. Lesa, Mata’afa Keni. 2013. “It’s not that simple.” In Samoa Observer, 16 January, http://samoaobserver.ws/ (consulted on 26 April 2013).

60 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

IN FOCUS Responding to Hurricane Isaac: strongly improved since Katrina, the practical access to evacuation was not provided to all. In some parishes, such as Terrebonne and assessing evacuations and federal levee Lafourche, inaccurate predictions caused hasty last minute evacuations. Thousands of residents who had not evacuated needed to be seven years after Katrina searched for and rescued. Tania Boulot Reality check for the federal levee Hurricane Isaac is an especially noteworthy case, due to the fact that it Hurricane Isaac originated near the Lesser Antilles on the 21 August was the first real test for the newly reconstructed federal levee system 2012 and was quickly upgraded to a tropical storm. On its way across protecting New Orleans. As a consequence of Hurricane Katrina, U.S. the Atlantic, Hurricane Isaac caused significant damage and flood- Congress approved plans for upgrades to the federal levee system in ing in Haiti and the Dominican Republic. The Gulf Coasts of Florida, the value of approximately 14.5 billion dollars, now called the 2012 Alabama and Mississippi had been affected as well. Isaac reached greater New Orleans area 100-year Hurricane and Storm Damage Risk Louisiana on the morning of the 28 August, hitting the coast seven Reduction System (HSDRRS). The major complex includes one of the years to the day after Hurricane Katrina. Even though Hurricane Isaac most extensive storm surge barriers in the world and is today 133 miles had been predicted as a Category 2 hurricane, it was downgraded to long. Many had been anxious about the levees being able to hold back a Category 1 hurricane shortly before reaching the Gulf Coast, due to Isaac’s storm surge, but it successfully withstood the test and shielded decreasing winds. However, this downgrading proved to be treacher- the inner parts of the New Orleans. ous, because it led to an underestimation of the effects of Isaac’s weak However, Isaac revealed that the protection of those inside the lev- forward momentum, namely longer exposure, higher rainfalls and ees inflates the risks for those living outside the levees. As the U.S. flooding totals. Additionally, this downgrading considerably reduced Army Corps of Engineers confirmed, “in many areas outside the 100- the willingness of the population in high-risk parishes to evacuate. year HSDRRS, water levels exceeded those from Hurricane Katrina and 3 Measures of Evacuation Gustav” . Isaac inundated some parts of greater New Orleans, such The mayor of New Orleans, Mitch Landrieu, had not ordered manda- as Laplace in St. John the Baptist Parish, which had previously expe- tory evacuations but had strongly encouraged those outside the levees rienced little to no flooding. For many in those places, the extent of to evacuate, making therefore the choice not to “err on the side of Isaac’s effects came as an unprecedented surprise, and demonstrated caution”1, thus saving a definite amount of chaos. On the contrary, the the lack of preparation in the region. The increased exposure of the Louisiana Government ordered mandatory evacuations for a number areas outside of the federal levee disproportionately affects those citi- of communities in parishes likely to be the most affected, which to a zens who are already disadvantaged, socially and economically. large extent included those living in low-lying and sparsely populated Lessons learned areas. Overall, the emergency status was declared in 41 parishes, 7 Isaac showed that the future of people living in these low-lying areas of which were at least partially evacuated. Thousands of citizens were outside of the federal levee system is hardly predictable and sus- put in shelters across the State of Louisiana. The State government, tainable if no decision is taken soon to extend the federal levee or to with the help of the Department of Education, provided approximately implement efficient and extensive mitigation and adaption measures. 350 buses, evacuating those in need of shelter in large parts to the Louisiana’s low elevational coastal zone is becoming increasingly vul- Jewella Shelter in Shreveport, or to a building close to Louisiana State nerable in the face of constant sea level rise, leading to an erosion of University at Alexandria2. the soil and a thus increased frequency of floods and storms. Con- Nevertheless, a considerable amount of people stayed in their sidering the risks and costs of living these low-lying delta regions, it homes, often due to a lack of possibilities for fleeing and very lim- is essential that changing residence or migrating be considered and ited financial resources. This demonstrates the insufficiencies of the proposed as an adaptation strategy. evacuation and sheltering plans. Even though the mandatory evacuations were proclaimed in advance, and the evacuation measures had

1. Campanella, Richard: “What the Nation’s best-educated Amateur Plan- ners Learned from Hurricane Isaac. And Gustav. And Rita and Kat- rina. And Cindy, Ivan, Lili, Isidore, and Georges.” The Design Obser- ver Group, Nov. 2012. http://places.designobserver.com/feature/ what-we-learned-from-hurricane-katrina-new- orleans/36148/ 2. Governor’s Office of Homeland Security and Emergency Preparedness: 3. U.S. Army Corps of Engineers: “Modeling Hurricane Isaac with and without “Hurricane Isaac after Action Report and Improvement Plan”. State of Lou- 2012 100-year HSDRRS”. 9 November 2012. http://www.slfpae.com/pre- isiana, December 2012. https://www.llis.dhs.gov/sites/default/files/Hurri- sentations/2012%2011%2015%20-%20USACE%20Post- Hurricane%20 cane_Isaac_After_Action_Report_1_0_v8_1.pdf Isaac%20Assessment.pdf

IDDRI STUDY 09/2013 61

The State of Environmental Migration 2011

ASSAM AND THE BRAHMAPUTRA: RECURRENT FLOODING AND INTERNAL DISPLACEMENT SABIRA COELHO

Assam epitomizes images of natural beauty: wild- flood-prone regions of the country and even for life sanctuaries, tea estates and lush rainforests. It other countries in South Asia, like Bangladesh, who is the largest of the “seven sisters”, the states that confront the same issue. Also, solving the conun- make up the north-eastern wing of India. Right drum faced by Majuli, a sinking river island, would through the heart of this state runs the Brahma- contribute to possible solutions for island nations putra, the second largest braided river in the world, that could face the same predicament in the future. known for its meandering and frequent changes The first section of the paper will give an outline of course (Brahmaputra Board, ). The name of the demographics of the region, including the Brahmaputra means “the son of Brahma”, who existing migration patterns. The second section in Hindu mythology is the creator of all humans will discuss the floods (focusing on ) and con- and along with Vishnu, the preserver, and Shiva, sequent environmental degradation, its influence the destroyer, forms the “Great Trinity”. The Brah- on demographic trends and the prospects in light maputra created Majuli, a river island situated of climate change. The final section will assess the mid-stream, by gradually depositing sediment to policies relating to flood control and dealing with form an alluvial plain (ASI, ). However, the internal migration. Brahmaputra over time is transforming into Shiva, evidenced by the destruction it has caused in the Assamese part of the Brahmaputra river valley. The 1. KEY DEMOGRAPHIC most vulnerable to this destruction are the , CHARACTERISTICS OF ASSAM residents of Majuli, who have been the victims of constant flooding of the river for decades (Census STATE AND MAJULI RIVER of India, ). The communities struggle to save themselves from the wrath of the river; some .. Background migrate out of the flood-prone region while others remain in temporary settlements along the river- The State of Assam has a population of  million banks. Despite financial losses to the tune of INR people, of which the majority is concentrated in the  billion per year, government attempts to curb the two river valleys of the region: in the Brahmaputra damage are both negligible and ineffective (Kalita, valley, that covers  districts, and in the Barak ). River, that covers  districts. Only  per cent of the In the context of climate change, the frequency total population lives in urban areas, far lower than and intensity of flooding is likely to increase, imply- the Indian average of  percent (Census of India, ing that best practices for flood control need to be identified immediately; in cases where flood con- . Since compiled data on the floods and their impact on trol is not possible, plans to minimize the impact Assam (such as numbers affected, people displaced, etc.) on local populations need to be developed. Given are scarce, the paper will rely upon news reports and grey literature. It is interesting to note that these flood- that  per cent of the  million hectares of land related news reports are prominent only in the regional prone to flooding in the country is located in As- newspapers rather than the national the problem is sam, tackling the problem posed by the Brahmapu- perceived in India; it is newspapers, implying that access tra is critical for India (Kalita, ). Further, any to information on the area is limited even in the Indian capital. This reflects how reduced to a ‘regional’ issue effective policies in disaster management imple- instead of a ‘national’ problem, much to the frustration mented in Assam could act as a prototype for other of locals in the region.

IDDRI STUDY 06/2012 63 The State of Environmental Migration 2013

). The population density in Assam of  settled down in the Brahmaputra valley to cultivate people per square kilometer exceeds the nation- the char areas (riverine land), which continues wide rate of  persons per square kilometer to influence the migration patterns of today. Two (Census of India, ). subsequent waves of Bengali immigrants followed The main occupation of the Assamese people is Indian independence and partition in , and farming (mainly of paddy) followed by work in the Bangladeshi independence from Pakistan in . service sector or in crafts such as pottery. This high Like before, these newer Bengali immigrants dependence on agriculture, where over  per cent settled in the floodplains of the Brahmaputra River of the people rely directly or indirectly on the land, (Singh, ). is risky considering that  per cent of the state’s Over the past four decades, the influx of immi- cultivable land is vulnerable to floods or drought. grants has persisted. Assam’s population has been As it is, Assam witnesses a higher incidence of pov- growing at a rate faster than the whole of India, erty of . per cent, (in comparison to the rest of despite a falling birth rate, which is attributed to the country where the rate is . per cent) par- heavy immigration flows from neighboring regions ticularly in rural areas where . per cent live in (Directorate of Economics and Statistics, Assam, poverty (Directorate of Economics and Statistics, ). Amongst the immigrants, there are second/ Assam, ). third generation Bengali immigrants whose fami- Within the state of Assam is the small river island lies have been in Assam for decades and have in- of Majuli, bound by the Brahmaputra to the south tegrated with the Assamese culture, and recent and one of its larger tributaries, Subansiri, to the immigrants whose culture is distinct from that of north. The island houses around seven different the older immigrants (Dasgupta, ). However, tribes who live in  small villages spread across precise data concerning both recent and historical the island (ASI ). Majuli is also the nucleus of immigration flows remains scarce or completely Vaishnavite faith (a main branch of Hinduism), as inexistent in Assam. Though there are assump- it hosts  Vaishnavite monasteries. Each of these tions concerning high levels of illegal immigration monasteries, known as sattras, imparts teachings from Bangladesh to Assam, the numbers are mainly to disciples, while also fulfilling civic functions like speculated. the settlement of disputes. As a consequence of immigration, tension has Majuli frequently faces the rage of the river. Fol- been escalating in the region. Assam has become lowing an earthquake in , the Brahmaputra a hotbed of contention between the indigenous changed its course, eroding the island’s area from tribal population and groups of Muslims, who are , sq km to  km². The river’s potential for believed to be illegal Bangladeshi immigrants (ir- destruction and creation of fertile lands through respective of their actual migration status, nation- inundation has made it an integral part of Majuli’s ality and timing of migration). This friction has spiritual culture. It is for these reasons that the is- intensified to violent levels on several occasions, land lies at the heart of the Assamese civilization, most recently in  with the riots between the with its intricate web of interaction between indig- Bodos, the chief tribe in the state, and the Mus- enous groups, the environment, spirituality and lims (Bhattacharyya & Werz, ). This particular culture (Pisharoty, ; Choudhary, ). outbreak led , people to flee their home, considered as internally displaced persons (IDPs) .. Migration trends in (IDMC, ). Assam: Growing in-migration Many speculate that the influx of Bangladeshi fuels ethnic tension immigrants is a consequence of adverse environmental factors, indicating that environmentally The state of Assam is extremely ethnically and reli- induced migration is already underway. According giously diverse. This plurality is reflected in the to Reuveny (), “when asked why they (Bangla-  different mother tongues that were spoken in deshi migrants) moved, they often provide natural  (Assam Online Portal, ). This diversity disasters, land scarcity and degradation and pov- can be explained by immigration into the region, erty as reasons”. These immigrant communities which is facilitated by the state’s geographical are vulnerable to environmental catastrophes as, position between seven Indian states, Bangladesh according to a study conducted by Shrivastava and and Bhutan. Entry points are therefore numerous Heinen (), many of them continue to establish for migrants who wish to benefit from the state’s temporary settlements in low-lying areas highly increasing prosperity. Yet, immigration can be prone to flooding. This is a consequence of the ten- traced back to the colonial period, when migrants sion as immigrants are not welcome in the main- from Nepal, Bihar and Bengal moved to Assam land towns and resort to setting up homes only in to work on the tea estates. Many of the Bengalis regions not densely inhabited by local populations,

64 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

which are, incidentally the low-lying char areas. blame for the river’s deluge. Landslides and flash On one hand, by settling there, these immigrants floods occurring due to a combination of all these free themselves from the risk of ethnic conflict as factors are, as a natural consequence, extremely these regions are mainly traditional “ghettos” of common. Unfortunately, the disaster does not end the Bangladeshi communities, but on the other there; a vicious cycle follows the onset of land- hand, they expose themselves to natural disasters slides, as the falling debris block the course of the like flooding (Dasgupta, ). river, forcing it to flood elsewhere (TERI, ). Emigrants from the state of Assam represent a Due to these factors, the flood-prone area in As- negligible amount of all internal migrants in India, sam totals to . million hectares, which is  per only . percent. Over half of the , Assa- cent of the state’s area, of which,  villages dot- mese emigrants were women, migrating for more ting the banks of the Brahmaputra are particularly than half of them for marriage purposes (Census vulnerable (TERI, ; TNN, ) Every mon- of India, ). However, there is considerable ru- soon, the state experiences flooding, with major ral-urban migration within Assam. This increased floods occurring at least once every four years urbanization, coupled with industrialization has (Directorate of Economics and Statistics, Assam, caused significant environmental degradation that ). Despite the suddenness of such events, the correlates with the increasing destruction caused government has ample warnings of their potential by floods (TERI ). magnitude and should be therefore able to implement flood-control measures to minimize the destruction. However, they still cause immense 2. FLOODING OF THE destruction: , hectares are affected each BRAHMAPUTRA RIVER year, eroding at least , hectares annually. Ac- cording to the Assam State Government (), at least . per cent of the total landmass of the state .. The Brahmaputra’s has been eroded since . Financially, the cost of geomorphology and the major floods is extremely high - INR . billion in impact of the floods damages to utilities, crops and houses when Assam was badly hit in  (Kalita, ). Majulihas also Originating from the Kanglungkang Glacier in been heavily eroded due to the annual inundation Tibet, China, the Brahmaputra River traverses the of the Brahmaputra and the migration of the riv- Tibetan Plateau before bending into Arunachal ers bounding the island. Between  and , Pradesh in India, after which it runs through the island lost an average annual amount of , Assam, merges with its tributaries in Bangladesh hectares, while , hectares of land remain con- and drains out into the Bay of Bengal. Along the stantly under water (Dutta, Barman, & Aggarwal, rumbling journey from a high altitude in Tibet till ). Subtracting this land from the island’s total, the Assam valley, the river picks up a large load of along with the land unsuitable for agriculture and sediment, making the river highly unstable in the the land reserved by the government, only , upper reaches of the Assam Valley. With deforesta- hectares, or  per cent of the island’s total area re- tion caused by shifting cultivation - a particular mains available for cultivation (ASI ). method of cultivation, often employed in South Asia, soil erosion has increased, which in turn has .. The floods in : How increased the amount of sediment carried by the relief efforts were obfuscated river (Sharma, ). In addition to a large amount by outbreaks of violence of sediment, the river also carries a large volume of water. The Brahmaputra has over a  tribu- In , a first wave of floods occurred from April taries, of which  large ones originate in the north to June as it crossed the danger mark by . cm, due to the melting of snow on the Eastern Hima- with subsequent waves of flooding in the following layas (Ghosh & Dutta, ). Once the river enters months. As of the  October , the Assam State the Assam Valley, the silt is deposited, leading to Disaster Management Authority (ASDMA) stated a change in the slope of the river, erosion of river that . million people were affected, from , banks, frequent migration of the river’s course and villages (ASDMA, ). Dhemaji, Morigaon, heavy floods as the river cannot contain the volume Dhubri, and Lakhimpur, the most densely popu- of water (Sharma, ). To exacerbate the situa- lated districts in Assam were worst affected by tion, the region receives high amounts of precipita- these floods. The government set up  relief tion (in the range of -mm) and is a seismi- cally active zone (NDMA ). This high amount of rainfall during the monsoon season is equally to . No updated date could be found.

IDDRI STUDY 09/2013 65 The State of Environmental Migration 2013

camps at the peak of the flooding, mainly in school funds became scarce since donors believed that buildings, to provide assistance to , people. flood-affected populations returned to their homes Furthermore,  raised platforms and  shelters as the displacement was supposed to be only short were also constructed. On the  August ,  term (Joseph, ). In reality most families have of the  relief camps sheltered a majority Muslim nothing to return to as the lands are still submerged. displaced persons. Despite governmental efforts, For those able to go back to their re-merged areas of many communities remained outside the relief origins, they often face challenges in getting back camps because the government didn’t have the their lands due to unfair relocation processes (Das- capacity to provide shelter to all of them and were gupta, ). Many of them may remain perma- struggling to survive by setting up temporary shel- nently displaced persons within Assam, yet close to ters. These populations were even more vulner- their original villages (Dasgupta, ). able to subsequent waves of flooding than those in A great polarization on the basis of different eth- the camps (ASDMA, ). nicities is found amongst IDPs, due to the ethnic Following the floods in , , people were tension that has been bubbling in Assam for years actually living on the dykes constructed to protect (Dasgupta ). Based on this diversity division Majuli and that this number had risen by , among the communities themselves, camp man- from the previous year (TNN, ). As floods had agers have divided the displaced population be- deluged the whole island, the government needed cause of fears that violence would erupt between to evacuate large numbers of people living on these the polarized groups of IDPs. The composition of embankments. However, relief efforts were obfus- the camps reveals that majority of the displaced, cated due to the geography of the island. As a sign are Muslims followed by members of marginalized of the government apathy, the President of Con- tribes, acknowledgement that interrogates on the gress Sonia Gandhi and the state minister aerially linkages between daily persecutions and marginali- surveyed the flood affected areas (ACTED, ). zation and vulnerability to natural disasters Deal- Damage to infrastructure like highways further dis- ing with the tension between the originating tribes rupted any quick attempts at providing relief to the and the Muslims is therefore very sensitive during population. To aggravate the flood situation, riots the relief phase, as resettlement and return always also broke out in July, creating an additional chal- raises issues regarding land ownership and legiti- lenge during the relief efforts and leading the state macy. Several tribes argue that the land is their into a complex crisis. rightful property and that the Muslims should va- According to situation reports by the Interna- cate it, while the latter refuse to do so arguing that tional Federation of Red Cross (IFRC, ), the they have been farming there for years. If Anandita IDPs on the riverbanks and trapped populations Dasgupta () professor at Guwhati University in Majuli were facing dreadful living conditions as argues that “the use of violence, dispossession, food, water, shelter, health and sanitation facilities murder, and confiscation of crops and animals have remained lacking. The government provided them almost become established patterns of char life”, with staples of rice and dal (lentils), nutrient sup- they seem particularly sensitive during post-disas- plements and baby food for children. The floods ter relocation processes. Considering therefore en- had destroyed the hand pumps, the main source vironmental factors but also ethnic tensions as in- of clean water, and the provision of purified water centive for displacement, it is difficult to ascertain was therefore extremely critical. The government the exact degree of environmental influence in the also provided tarpaulin sheets that were however displacement processes – an important considera- insufficient for the large numbers of IDPs and cer- tion to keep in mind in the labeling process of such tain relief camps were extremely make-shift. Fortu- displaced groups. nately, there was no epidemic outbreak following the floods, though IDPs suffered from malnutrition .. Is growing out-migration a (IFRC, ). In the context of humanitarian cri- consequence of recurrent floods? sis, provision of education was extremely difficult (TNN, ). Relief efforts following the retreat With the flooding of the Brahmaputra over the years of the monsoon seemed to slowly dwindle down; and the subsequent slow-onset event of erosion of riverbanks, there has been large displacement . Usually one needs to take two ferry rides across the river of people every year. Whole villages have been to get to Majuli, but during the floods when the water erased or washed away due to the force of the river. was at dangerously high levels, traversing the river Unfortunately, official statistical databases for the was near impossible. Army helicopters were thus used number of villages displaced and details regarding for dropping essential supplies like packaged food and drinking water to the marooned population, but could the relocation of the population - where they move not land on the island because it was submerged. to and how they survive - remain vague.

66 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Table 1. Number of erased villages in ASSAM region (Das et al. ). These youth migrate on District Number of “erased villages” a temporary basis to earn a modest income, which Dhubri 71 they remit back to their families for making repairs, Jorhat 2 buying food and maintaining livestock. Though Kamrup 14 these remittances are useful, there is a decline in Goalpara 75 the workforce for agriculture, which increases the Chars (Riverine areas) 181 burden on the older generation who stay behind (Das et al. ). However, the youth work ini- SOURCE: TNN tially as seasonal farmers on the less flood-prone, fertile lands, before migrating for the rest of the Revenue Minister, Prithivi Majhi, though, states season. (Das et al. ) Others migrate over short that data collection is underway (TNN, ). This distances to northern districts to engage in menial collection of information is likely to be an arduous labour like pulling rickshaws (Dasgupta, ; Das task as with recurrent inundations, villages located et al. ). In Majuli, the youth, often more edu- in low-lying char areas, separate over time to com- cated than the older generation, have migrated to bine with different villages in the high lands, as the southern city of Hyderabad in search of work. seen in a study conducted by the NGO Arayanyak At the same time, many in Majuli do not have the (Das, Chutiya, & Hazarika, ). Similarly, in Ma- resources to migrate and remain trapped on the is- juli, according to the state government, since  land (Pisharoty, ). approximately , families have been left home- Following the Bodoland riots in , the visibil- less due to land erosion. The government has relo- ity of the Assamese “ghettos” in other parts of India cated only  families, leaving the approximately became more visible, as their population increased , remaining displaced peoples to fend for due to a mass exodus of Assamese people (Devu- themselves (Choudhary, ). Since government lapalli, ). This large number of Assamese work- resources provide no information on this reloca- ers in other Indian states suggests that first, people tion, Mr Choudhary was interviewed for an insight are emigrating from the state not only in search of into the state of the displaced persons. These popu- jobs, but perhaps also as an adaptation strategy to lations, according to the journalist, have been re- deal with the floods and infertile lands. Second, it located to “some other villages. They now survive suggests that though census data reflects low out- on farming, a few on fishing; others have taken up migration from Assam, the figure is possibly higher petty jobs. They have coped since survival is inher- than estimated. Migration flows, therefore, are ent to them.” changing in Assam due to the impact of the Brah- The damages of floods on livelihoods are tre- maputra on the lives of the local population. mendous. With increased siltation, fertile land is Based on the newspaper reports and other no longer productive, rendering Assamese farmers sources of data, the state’s migration flows can be landless with very limited opportunities to find re- summed up as follows: first, temporary intrastate munerative employment (TERI, ). Thus, these displacement during the monsoon seasons with re- populations, dependent on agriculture, livestock turn to the homeland when the lands re-emerge; production and fishing, either abandon their lands second, permanent intrastate displacement to em- and homes in search of work by migrating to other bankments faced by the poorer and marginalized cities where they become rickshaw-pullers or cart- IDPs; third, permanent migration of whole villages, pullers earning dismal daily wages, or remain dis- which are proactive in order to avoid the floods; placed (Pisharoty, ; Dasgupta, ). Many of and fourth, long-distance, interstate, temporary these displaced persons especially vulnerable to la- and seasonal migration for additional economic bour exploitation and trafficking (Dasgupta, ). reasons. It is interesting to note that, apart from the Information available on relocation of displaced last group of migrants, most others remain fairly villages and families and on out-migration of As- close to their original locations. samese living on the banks of the Brahmaputra is The likely reasons behind migration, and how based on newspapers. As reporters note, many of one is affected by the floods in Assam, include the youth leave the difficult flood-ravaged regions level of education, income and occupation. Pub- with few employment opportunities to work in big lic sector employees are less likely to be displaced cities in other states like Kerala, in the south or because they have a secure livelihood despite fre- Nagaland, which is also within the north-eastern quent inundations. Furthermore, those who have higher levels of education are more adaptable to migration, as they have better skills, making the . Interview with Ratnadip Chaudhary, Principle Correspondent, Tehelka, conducted on  March  probability of finding employment higher. Income via email. is also important in migration as the people in the

IDDRI STUDY 09/2013 67 The State of Environmental Migration 2013

high- and middle-income groups sometimes al- Second, increased out-migration from the flood- ready have property in safer areas to where they plains is also likely to create a labour shortage for relocate (Choudhary, ). Thus, those who can- the farms that are not submerged. Furthermore, not move away from the floods are the peasants it is likely that this out-migration will be a genera- and the indigenous tribal groups who are bound tional issue as the youth is more willing to migrate to the land. Culture is a second important deter- in search of work, leaving their parents behind to minant in migration. For the tribal populations in- work on the land. Third, given the ethnic divisions habiting the riverbanks, coping with the flooding in Assam’s society and the hostility faced by IDPs is natural to them. The Mishing tribe for instance, and immigrants, increasing migratory movements builds houses on stilts, known as ‘Char ghars’ that due to environmental factors is likely to be a huge have been emulated by governmental relief pro- challenge. As it is, when the Bengali Muslims liv- grammes. This riparian culture of the tribes often ing in riparian areas migrate upwards during the impedes them from migrating too far from the floods, locals view this as an influx of new illegal river, owing to sentimental attachment, implying immigrants from Bangladesh (Dasgupta, ). In that these communities are highly unlikely to mi- the face of climate change, dealing with these chal- grate long-distances and for the long-term (Das et lenges will be essential. al. ).

.. The climate change challenge 3. POLICY ANALYSIS

Climate change is likely to have an extremely Given the destruction caused by the floods each adverse effect on flooding and erosion in the region. year, the state and central government have imple- First, since rainfall is currently one of the root causes mented a series of disaster management policies in behind overflowing of the river, any increase in order to curb the harm caused by the floods. Large- precipitation as predicted in a few studies outlined scale floods that occurred across the country in  in the State Action Plan on Climate Change is likely prompted the Indian Government to look into flood to make the floods more extreme (TERI, ). control measures. Since then, the government has Second, since the river also receives glacial run-off taken steps to implement this objective by creating through its tributaries, the rapid melting of snow the High Level Committee on Floods (), the due to global warming will also increase the volume Ministers Committee on Flood Control (), the of water that the river holds (Sharma, ). Hence, Rashtriya Barh Ayog (National Flood Commis- floods are likely to become more frequent and sion, ) and Task Force on Flood Management/ more destructive in the near future, which may also Erosion Control () (NDMA ). In December accelerate out-migration from the region. Though , the government enacted the Disaster Manage- this latter process allows people to cope with the ment Act, which established the National Disaster disaster, there are many questions concerning relo- Management Authority (NDMA), which presides cation that need to be addressed. over the State Disaster Management Authori- First, in the case of Majuli, the protection of cul- ties (SDMA), Disaster Response Forces and the tural integrity is an imperative. Due to the changing State Funds for Disaster Relief. This approach was course of the river, the people living in Assam have expected to ensure an integrated disaster manage- imbibed migration as a trick for survival, migrating ment approach in the whole country (NDMA ), within ten kilometers or to entirely new districts in while the direct responsibility of relief, rescue and the upper reaches of the river to avoid flooding. This rehabilitation measures lies in the hands of the adaptation mechanism has a detrimental impact on state governments concerned. culture, as villages do not move in totality, because In addition to the SDMA, each state must adopt of population density constraints (Das et al. ). a State Disaster Management Act to identify re- In Majuli, the number of sattras has declined from sponsibilities and coordination mechanisms. For  to , and in the event that the island becomes flood management, State Governments and lo- engulfed by the river and its inhabitants have to cal authorities need to refer to the guidelines put relocate, it is likely that the translocation of these down by the NDMA when developing action plans. monasteries will not be successful (Choudhary, Below the state level, district-level and village-level ). In this context, believers of the Vaishnavite bodies have been created with the responsibility of faith fear the erosion of an important part of their disaster response and relief, mitigation and preven- culture along with the river island. Consequently, tion at the district level. In case of a disaster, it is it is important to think not only of how to preserve these bodies that coordinate the functions of local populations, but also their unique cultural identity authorities, health facilities and primary schools when formulating rehabilitation plans. (Ministry of Home Affairs, ).

68 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

In Assam, the ASDMA (Assam State Disaster migration call for the detection and deportation of Management Authorities) works within this insti- illegal immigrants, implying that cross-border mi- tutional framework. The ASDMA has taken both gration occupies the public discourse far more than structural and non-structural measures for flood internal migration (Singh, ). preparedness. Structural measures to mitigate flooding include the construction of embankments, water reservoirs, retention basins and buildings on 4. KEY CHALLENGES IN elevated areas. Non-structural measures include ADDRESSING THE FLOODING AND flood-forecasting mechanisms and floodplain zoning that check uncontrolled settlement in vulner- DISPLACEMENT IN ASSAM able areas (ASDMA, n.d.). A flood-hazard atlas has Firstly, the institutional capacity and resources already been developed for each of the districts in among the state and district authorities may be coordination with the Indian Space and Research too limited to implement the outlined priori- Organization. An Assam Relief Manual also exists, ties. The ASDMA’s Annual Report  outlined though it was created in  and does not include that State Disaster Response Force consists of preparedness strategies (TERI, ). Additional only  personnel, with roughly  from each measures include building community awareness, district. In , ASDMA conducted , training stockpiling emergency essentials, creating early programmes and workshops to build capacity warning systems, identifying areas for settlement, among different demographic groups like health conducting drills, and developing coordination professionals, students and teachers, volunteers plans with NGOs and other local organizations and NGOs. Though such programmes indicate a (ASDMA, n.d.). positive step, they were able to reach only , In addition to the ADSMA, developing a master participants, which is very low given that . plan for flood control in the Brahmaputra valley is million people were displaced in  (ASDMA, under the jurisdiction of the Brahmaputra board, ). This dearth in capacity leads to a depend- created by the central government through the ence on external assistance from non-local bodies, Brahmaputra River Act in . This master plan which do not coordinate well with the existing for the control of floods and erosion entails the NGOs and other informal institutions set-up in the construction of “multipurpose dams”, which would state (ASDMA, ) provide extra benefits like hydroelectricity and ir- Secondly, existing policies do not take into ac- rigation. In , protection of Majuli was placed count ground realities. The training programmes, under the responsibility of the Brahmaputra board for example, seem to target very specific communi- (Brahmaputra Board ). ties, i.e. those people with a basic level of education, When developing an effective strategy for flood with the risk of marginalizing communities living in control in Assam, it is pertinent to look at policies rural or remote areas. In a state like Assam where influencing migration trends in the region. The only  per cent of the population is literate, devel- Mahatma Gandhi National Rural Employment oping effective training programmes that empower Guarantee Act (MGNREGA) has been reported to those who are illiterate is crucial (Directorate of alter the rural-urban migration flow in other states Economics and Statistics, Assam, ). Addition- of India as it provides  days of guaranteed em- ally, floodplain zoning, which prevents people from ployment in rural India. However, according to re- setting up houses on the riverbanks, is not imple- search conducted by the Agro-Economic Research mented in Assam due to the state government’s Centre, the scheme has had little impact on slow- argument that the thick population density in the ing down this trend in Assam, largely due to the state impedes the implementation of such a policy fact that the days of work promised by the act was (Planning Commission, ). Though the govern- not yet a reality in the state and, as a consequence, ment’s claim has merit, a check on human activity wages in urban centers still remained higher than in these floodplains must be ensured in order to those provided by the MGNREGA (Bordoloi, ). prevent such a damaging effect year after year. Thus, the rural populations in Assam (and perhaps Thirdly, the construction of multipurpose dams many living on flood ravaged river banks) consider for flood control is also questionable. In the midst migration to cities as a better option, adding to the of frantic construction of three dams upstream on rapid urbanization of Assam. In the absence of ur- the Brahmaputra by China, India seems to have for- ban policies or policies protecting internal migrants gotten the central objective behind building dams in India, these populations are often very vulner- of flood control, and focuses instead on ensuring able, given a low priority by the government and the hydroelectric opportunities of the river (Or- have no social or legal protection (UNESCO, ). land, ). Until now, embankments have not only In fact, the only policies in Assam dealing with been ineffective in controlling the damage, but

IDDRI STUDY 09/2013 69 The State of Environmental Migration 2013

actually further aggravate the situation by creating Assam in the first meeting for the working group drainage issues in areas outside the dam (Das et al., on flood management for the twelfth five-year ). In addition to creating instances of dam-in- plan was also symptomatic of the apathy on behalf duced flooding, these dams have also been respon- of the state government (Planning Commission, sible for displacing populations (mostly indigenous ). Another indication of the ASDMA’s lackadai- people, sometimes comprising of whole tribes) but sical attitude is evidenced by fact that the  draft this is often brushed aside as the displacements in of the “Assam Relief Manual” was revised for the the north-east are “relatively small” in comparison first time in  (Henriques, ). to other parts of India (Vagholikar, ). Further- Coordination between ASDMA and NDMA is also more, apart from the construction of new dams, old lacking. As a recent audit by the Comptroller and embankments need to be repaired. In ,  em- Auditor General on the NDMA revealed, it is “inef- bankments were breached during the flooding ac- fective in its functioning in most of the core areas.” cording to Water Resources Minister, Rajib Lochan (Supreme Audit Insitution of India, ). Key pro- Pegu. The minister also added that, “,. km of jects like hazard mapping have not been complet- a total ,. km of embankments have surpassed ed, while poor inter-governmental agency commu- their effective life-span”, implying that  per cent nication is a hindrance to the disaster management of the embankments in Assam were technically be- process. Furthermore, rescue and relief teams were yond their expiry date (Kashyap, ). Lastly, such inadequately trained, while the disbursement of dams do little to control siltation, which causes as funds was delayed (Supreme Audit Insitution of In- much damage as inundation (Sharma, ). dia, ). Though the report included eight state- In Majuli, despite the construction of a few em- specific observations, an audit of ASDMA was not bankments and the recovery of land in the form of undertaken. However, given the lacking capacity sandbars, no concrete moves have been taken to of the national agency, ASDMA has little support avoid further erosion. In order to ensure the cul- and guidance from the central government to deal tural identity of Majuli, the State Government ap- with disasters effectively. The Brahmaputra Board plied to UNESCO for the status of “World Heritage also lacks coordination with the state government: Site”, on the premise that recognition of the island’s set up by the central government, the control of its cultural value would increase both national and in- activities remains in New Delhi (Government of In- ternational attention for its protection (ASI, ). dia, ). States should be given increased control Unfortunately, the island remains on the list of and autonomy over the determination and enforce- nominated sites. Assam’s flooding problem also gets ment of their flood control master plans. lost in the politics between the central and the state International coordination to manage the Brah- government. The Assamese believe that the cen- maputra’s water resources is fairly tense, despite tral government remains apathetic to the situation. cooperation on other riparian systems in India, as Protests led by the Assamese Student Union against suggested in the rejection of a recent Indian pro- negligible budget allocation to the strengthening of posal for a new mechanism on water sharing by embankments and for the declaration of annual As- China (PTI, ). Despite this attempt by India sam floods as a national problem, rather than a state (advocated by the central government) to improve problem are frequent (TNN, ). The central gov- international coordination, political will to do the ernment reiterates however, that this is a state issue, same amongst north-eastern ministers is weak. despite Chief Minister Tarun Gogoi’s assertion that For example, Mr Gogoi, Assam’s state minister ar- the debt burden of the floods is too immense for the gued that when flood management coordination Assamese government to handle on its own (North between north-eastern states remains deficient, in- East News Agency, ). However, the state govern- ternational cooperation on the matter remains out ment has received INR . billion from the central of the question (North East News Agency, ). government to tackle the flood challenge, which is A last challenge with the floods occurs with the far more that the INR . billion recommended by retreat of the monsoons. As the rains disappear, so the Task Force on Floods (Kashyap, ). Further- does attention to the issue. However, many IDPs more, the state government has demanded funds to continue to live in relief camps and issues with nu- respond to annual floods but despite receiving the trition, health and sanitation remain. Some IDPs requested money, projects are still delayed. Skeptics in camps are even forced to leave, since the camps question whether the funds are actually being em- located in school buildings need to reopen (Joseph, ployed, or simply siphoned off (Chaudhary, ). ). The crux of the issue lies in the fact that the This slow implementation of schemes, according IDPs are not legally recognized. The Guiding Prin- to a member of the Central Water Commission, is ciples on IDPs as formulated by the UN are not rec- supposedly the reason behind the slow release of ognized by India and not implemented by the As- funds. The absence of state representatives from samese Government (Dasgupta, ).

70 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

CONCLUSION aspects of life from education to disputes with its giant halls called ‘namghars’ where the villagers Though the basic structure for flood management collect. Thus, imparting knowledge of flood pre- has been established, policies are often misguided paredness from these institutions would ensure lo- and worsened by the lack of attention given to cal mobilization. The community-run early warn- north-eastern states due to the region’s limited ing systems for floods in the disaster prone area political clout. Instead, both the central and state from Assam to Bangladesh is an example of the role governments need to coordinate and strengthen informal institutions can play in disaster manage- their institutional capacities in order to success- ment (NDMA, ). Effective policies need to in- fully employ a multipronged approach looking at clude civic participation and since “NGOs are ready different dimensions - social, economic and envi- to take up the issue but government is in denial” as ronmental - that will be most effective at allevi- Mr Choudhary argues, it is useful to increase their ating the situation. At the outset, it is essential for engagement prior to and during the floods. the government to safeguard the rights of IDPs, As for structural measures, steps to deal with the which includes recognizing them, particularly in excessive silt carried by the river also need to be the context of ethnic violence. The ethnic chal- taken when constructing dams. As Nayan Sharma lenge acts as an impediment to relief efforts and so (), a professor at IIT Roorkee asserts, “a mas- policies to encourage economic and social devel- sive soil conservation campaign is needed to effec- opment among tribal and marginalized Muslim tively reduce the sediment volume of the Brahma- populations should be a priority. Furthermore, putra, along with the creation of flood detention livelihood support for those whose lands have been reservoirs, river training for erosion control, land destroyed needs to be provided by the govern- reclamation and channelling.” Sharma also suggest ment, rather than just provision of food supplies that India takes a leaf out of China’s flood manage- following the floods. The existing policy misses the ment book by using similar technology for sedi- point that crops were sold for monetary purposes ment control as used on the Yangtze River. and not merely for sustenance (Talukdar, ). The lack of accurate data is a huge challenge, To prevent the problem of food and job insecu- which mires the whole relief operation as the num- rity following the floods, the government should bers affected are undercounted. The state govern- consider providing training for agriculture in flood- ment needs to update its databases and keep a affected areas, and targeting particularly and youth record of the number of inhabitants of all villages to increase the labour force in case of young males (TERI, ). A good practice of ASDMA was the migrating (Das et al., ). daily reports published on the floods in , which With regard to migration, policy makers are kept an account of numbers in relief camps and oblivious of the temporary migration among the those affected including livestock. By looking at tribes that allow them to survive (Das et al., ). these reports, it is easier to determine trends and The government also fails to recognize migrant best practices that can help formulate effective pol- flows- both into and out of Assam - that are crucial icies for the years to come. in order to draft policies dealing with unplanned ur- Finally, flood preparedness and disaster risk mit- banization and to effectively manage flood control. igation needs to be streamlined into development To begin, the Government could consider facilitat- policies. There is hope that the government is moving migration of these tribes. It should also create ing in this direction, as it recently approved a IDR a temporary and circular migration scheme within . billion project to ensure economic develop- the country that may also help redistribute income ment, social improvement and preservation of the to the north-east through remittances. They could remaining  sattras in Majuli (PTI, ). Thus, be used to ensure that these tribes have sufficient for the people of Assam, while the Brahmaputra resources in the face of flooded agricultural land. continues to destroy, preservation and creation lies Local knowledge, particularly among the tribes within the hands of the government. with a strong riparian culture, should also be tapped into by ASDMA (Das et al., ). The AS- DMA could utilize the existing networks to more efficiently provide aid to those in needs. In Majuli, . Interview with Ratnadip Chaudhary, Principle in particular, the sattra is a key informal institu- Correspondent, Tehelka, conducted on  March  tion that is non-discriminatory and overarches all via email.

IDDRI STUDY 09/2013 71 The State of Environmental Migration 2013

The Energy Resources Institute (TERI). 2008. Draft Report: BIBLIOGRAPHY Assam State Action Plan on Climate Change, 2012 - 2017. Dis- pur: Department of Environment & Forest, Government of B   Assam. UNESCO and UNICEF. 2012. For a Better Inclusion of Inter- ACTED. 2012. Fighting the floods on the Island of Ma- nal Migrants in India. New Delhi: UNESCO and UNICEF juli, http://www.acted.org/en/fighting-floods-island- majuli (consulted on April 30, 2013). Archaeological Survey of India. (ASI) 2004. River Is- A land of Majuli in midstream of Brahmaputra River in As- Das, P, D Chutiya, and N Hazarika. 2009. Adjusting to floods sam. Paris: UNESCO on the Brahmaputra plains in Assam, India. Kathmandu: Assam State Disaster Management Authority. 2013. An- ICIMOD. nual Report 2012. Dispur: Assam State Disaster Man- agement Authority. Dasgupta, Anandita. 2001.”Char’red for a lifetime : Internal Displacement in Assam Plains in India. South Asia Refugee Assam State Disaster Management Authority. July 31, Watch Vol. 3 & 4: 1-13. 2012. “ASSAM STATE DISASTER REPORT -2012.” As- sam State Disaster Management Authority. http:// Dutta, M.K, Swapnali Barman, and S.P. Aggarwal. 2010. “A sdmassam.nic.in/download/Assam_Flood_Re- study of erosion-deposition processes around Majuli Island, port_31.07.2012.pdf (consulted on April 30, 2013). Assam.” Journal Earth Science India: 206-216. Assam State Disaster Management Authority. Decem- Ghosh, Shymal, and Subashisha Dutta. 2012. Impact of cli- ber 31, 2012. “CAMP & CAMP INMATES IN KOKRA- mate change on flood characteristics in Brahmaputra basin JHAR &NEIGHBOURING DISTRICTS.” Assam State using a macro-scale distributed hydrological model. Journal Disaster Management Authority. http://sdmassam.nic. of Earth System Science: 636-657. in/download/31-12-2012.pdf (consulted on April 30, Reuveny, Rafael. 2005. Environmental Change, Migration 2013). and Conflict: Theoretical Analysis and Empirical Explora- Assam State Disaster Management Authority. “Camp tions . Human Security and Climate Change . Oslo: 1-25. Report 2012.” Assam State Disaster Management Au- Shrivastava, Rahul, and Joel Heinen. 2005. “Migration and thority. 27 August, 2012. http://sdmassam.nic.in/ Home Gardens in the Brahmaputra Valley,Assam, India.” download/CAMP_REPORT.pdf (consulted on April 30, Journal of Ecological Anthropology: 20-34. 2013). Singh, Gurinder. 2008. “Illegal Migration, Insurgency, and Assam State Disaster Management Authority. Flood Pre- the Political Economy of Assam.” Strategic Analysis : 305- paredness Planning. Dispur: Assam State Disaster Man- 315. agement Authority. Vagholikar, Neeraj. 2011. “Dams and Environmental Gov- Bhattacharyya, Arpita, and Michael Werz. 2012. Cli- ernance in North-East India.”Kalpvriksh: 360-369. mate Change, Migration, and Conflict in South Asia: Ris- ing Tensions and Policy Options across the Subcontinent. Washington D.C.: Center for American Progress. P  Bordoloi, Jotin. 2011. Impact of NREGA on Wage Rates, Assam Times. “Relief to flood affected people in Majuli.” In Food Security and Rural-Urban Migration – A Study In Assam Times, November 27, 2012. Assam. Jorhat: Agro-Economic Research Centre for North East India, Assam Agricultural University Chaudhary, Ratnadip. 2012. “A Vicious Floods And Funds Cycle.” In Tehelka, October 1, 2012. Directorate of Economics and Statistics, Assam. 2011. Economic Survey of Assam, 2010-2011. Dispur: Govern- Choudhary, Ratnadip. 2012. “Majuli has shrunk by half. ment of Assam. Leaving farmers landless, hopeless.” In Tehelka, October 29, 2011. Government of India. 1980. The Brahmaputra Board Act, 1980. New Delhi: Government of India Chaudhary, Ratnadip. 2012. “To Majuli with tears.” In Tehe- lka, September 24, 2012. IFRC. 2012. Disaster relief emergency fund (DREF) In- dia: Assam Floods. IFRC. Devulapalli, Rahul. 2012. “Rumours trigger exodus of As- sam people from Hyderabad.”In The Times of India, August Ministry of Home Affairs. 2005. Disaster Management 15, 2012. in India. New Delhi: Government of India, Henriques, Jude. 2011. “Rising above it all.” In The States- National Disaster Management Authority. 2008. Na- man, November 6, 2011. tional Disaster Management Guidelines: Guidelines for Floods. New Delhi: Government of India. Internal Displacement Monitoring Center. 2012. “News Alert Northeast India: 400000 people internally displaced Planning Commission. 2011. Report of Working Group on Flood Management and Region Specific Issues for XII following violent riots.” In IDMC website, August 8, 2012. Plan. New Delhi: Government of India. Joseph, Ammu. 2012. “Where camp is a four-letter word.” In Supreme Audit Insitution of India. 2013. Performance The Hindu, October 9, 2012. Audit of Disaster Preparedness in India. New Delhi: Kalita, Pratiba. 2013. “Floods rob Assam of 200 cr every Comptroller and Auditor General. year: Economic survey.” In The Times of India, March 22.

72 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Kashyap, Samudra. 2013. “82% embankments in Assam I have outlived their utility.” In The Indian Express, March 29. Interview with Ratnadip Chaudhary, Principle Correspond- Kashyap, Samudra. 2013. “Assam’s cycle of floods and promises.” The Indian Express, September 25 ent, Tehelka, conducted on 30 March 2013 via email. North East News Agency. 2003. “Sorrow of Assam.” In North East Enquirer W Orland, Brian. 2013. “Brahmaputra Dams Promise Prosper- Assam Online Portal. 2013. People: Languag- ity But at a Big Cost.” The New York Times, March 5, 2013. es. http://online.assam.gov.in/web/guest%20/ Pisharoty, Sangeeta. 2011. “The Sorrow of Majuli.” The Hin- people?webContentId=174762 (consulted on April 30, du, December 24, 2011. 2013). Press Trust of India (PTI). 2012. “Plan panel approves Rs Brahmaputra Board, Ministry of water Resources. Activi- 116-cr flood control project in Majuli.” The Hindu Business ties, Brahmaputra Board. http://brahmaputraboard.gov. Line, February 1, 2012. in/NER/Activities/activities.html (consulted on April 2013, Press Trust of India (PTI). 2013. “China spikes India’s pro- 30). posal for joint mechanism on Brahmaputra.” The Hindu, Census of India. 2011. “Migration Tables.” Census of India, April 17, 2013. 2011. http://censusindia.gov.in/Tables_Published/D- Sharma, Nayan. 2012. “When the Brahmaputra flooded.” Series/Tables_on_Migration_Census_of_India_2001.aspx The Indian Express, July 17, 2012. (consulted on April 30, 2013). Talukdar, Ratna Bharali. 2013. “Opting out of the chaos.” In Census of India. 2011. “Provisional Population Totals Paper India Together. 2 of 2011 : Assam.” Census of India. http://www.censusin- TNN. 2012. “21 killed, 4 lakh displaced by floods in north- dia.gov.in/2011-prov-results/paper2/prov_results_pa- east.” The Times of India, September 26, 2012. per2_assam.html (consulted on April 30, 2013). TNN. 2013. “Aasu stir to get national problem tag.” The National Disaster Management Authority. 2013. National Times of India, March 23, 2013. Disaster Management Authority: Evolution. http://ndma. TNN. 2011.”Hope floats for displaced people of Majuli.” The gov.in/ndma/evolution.html (consulted on April 30, 2013). Times of India, December 14, 2011. TNN. 2013. “No information on displaced from 181 flood- V hit villages: Govt.” The Times of India, March 5, 2013. Map of Majuli. TNN. 2012. “Rise in homeless living on Majuli dykes.” The Google maps. 2013. Times of India, July 12, 2012 Maps of India. 2013. Map of Assam.

IDDRI STUDY 09/2013 73 The State of Environmental Migration 2013

IN FOCUS Australia and its swift disaster responding quickly to the disaster, is one of the factors that was handled particularly well during both emergencies (Clarke, 2013). response management The good working relationships among the police, emergency service personnel and local councils were considered to be very effective. Nicole Schmidt Certainly, the special establishment of the Queensland Recons- truction Authority overlooking the emergency management for the Australia is no stranger to natural disasters. However, the country’s disaster in 2011, and later in 2012 too, has been a key factor of a environmental vulnerability to climatological and hydro-metrologi- well-organized coordination of relief and recovery disaster manage- cal disasters has been dramatically increasing since 2010 with the ment. Funds were made available immediately for those who had recurrence of one major flood every year. The Queensland floods in not sufficient insurance coverage and there was an immense natio- 2010/2011 inundated 80 per cent of the state and heavily impacted nal outpour of help from fellow Australians in general. Most people the national economy as well as local communities. The Internatio- whose houses had been flooded stayed with family or friends but the nal Disaster Database EM-DAT estimated the damage to be USD 7.3 government also set up additional emergency crises camps. billion. 35 people died during the incident; 175,000 were affected Make use of social media and in immediate need for assistance such as food, water or shelter The increasing use of social media has been another driving factor supplies (The OFDA/CRED International Disaster Database, 2013). for making the evacuation process both in 2011 and 2012 success- Just one year later, in February and March 2012, Victoria, Queens- ful. As the affected areas were surveyed, it became clear that many land and New South Wales were hit by yet another flood commonly households were unable to watch TV or make use of their landline referred to as the Eastern Australian floods. Fortunately, the number phones due to the flooding. Therefore, Australia’s emergency mana- of people affected was significantly smaller (13,300) and no casual- gement began spreading information and updates on the current ties were reported. disaster management situation through different channels such as “Global best practice” Facebook and Twitter because people still had their mobile phones. Operating under the internationally recognized principles of “Pre- The Queensland police for example began spreading information as vent, Prepare, Respond and Recover”, the management of Austra- soon as people had to be evacuated and at the height of the Queens- lia’s evacuation process, especially during the Queensland disas- land flood disaster, more than 35,000 tweets containing the hash- ter, has been quite frequently referred to as “global best practice” tag ‘qldfloods’ were sent via Twitter (Mellick, 2013; The Brookings (Arklay, 2012) in recognition of their particularly well handled crises Institute, 2012). Since then, social media has become a key factor management. in providing useful information to the general public before, during The advantage of being developed and after a disaster. Australia has access to several types of ressource that make it easier to act quickly and efficiently. However, the magnitude of the Queensland floods in 2011 exceeded the country’s capacity during Bibliography the evacuation process and called for additional help. Therefore, the Arklay, Tracey. 2012. Queensland’s State Disaster Mangaement Group: An all Australian Government’s Department of Defense provided an Austra- agency response to an unprecedented natural disaster. lian Division Force (ADF) team to help assist in the recovery efforts. Lahey, Kate. 2011. Defense force helps Queensland on road to recovery. The Aus- The defense personnel carried out a number of different tasks: from tralian Journal of Emergency Management 26 (Nr. 3) : 14-19. searching for bodies and performing rescues, to airlifting people to The Brookings Institution. 2012. The Year that Shook the Rich: A Review of Natural safety or delivering medical supplies, to the distribution of food and Disasters in 2011. London School of Economics Project on Internal Displacement. more than two million liters of purified water, especially in isolated The OFDA/CRED International Disaster Database – www.emdat.be – Université communities to prevent a future health crisis (Lahey, 2011: 15-16). Catholique de Louvain – Brussels – Belgium. Accessed 25 April 2013 Thus, the ADF was effectively coordinating the evacuation, relief Interview with Steve Mellick, Detective Sergeant & Civil-Military Project Manager, efforts and recovery support. conducted on 2 May 2013 in Paris (via Skype). Swift coordination Interview with Clarke Matthews, Professor at Deakin University, Head of School of Humanities and Social Sciences, conduced on 29 April 2013 in Paris (via Email). The strong coordination, not only within the ADF, but also among all branches of the government – local, state and federal – in

74 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

IN FOCUS Mozambican floods and resettlement Patt and Schroter (2008)2 related this phenomenon in Mozambique to differing perceptions of climate risk among resettlers, policymak- processes ers, and program managers. This study suggests that many farmers were aware of the risks that flooding posed but returned to the Geraldine Zambrana floodplains mainly because they were unable to establish a viable livelihood in their new locations. This finding underlines the impor- In early January 2013, Mozambique suffered from the most cata- tance of ensuring that resettlement communities in Mozambique are strophic floods the country had seen since 2000. The floods affected socially and economically viable as well as physically robust in an 300,000, killing 117 and displacing approximately 200,000 people infrastructural sense, echoing the limitations of state-led relocation (INGC, 12/03/13). The disaster affected the provinces of Zambezia programs in earlier periods. Displaced people have a connection with and Gaza. If the government does not effectively respond, the migra- the land and are reluctant to move because they also do not want to tion will cause overcrowded cities, like Maputo. become dependent on public projects. Case Study: The resettlement site in Chiaquelane The government’s goal is to identify and give away 4,830 plots, Chiaquelane, the largest long-term resettlement site, consists of the majority of them on the higher ground of Chiaquelane (IRIN 1,123 families divided into 5 “Bairros”. 1 The site lacked adequate 13/02/13). Kaisa Nugin an expert from OCHA stated that “Chi- security and electrical, health and WASH provisions. In contrast, aquelane was used for resettlement during earlier flood events most of “Bairros” from Lionde – the affected town from which most (2000/2001). By combining the old village with the prior reset- of people come to Chiaquelane - have adequate facilities, though tlement, Chiaquelane will ultimately develop into a larger urban not all of them are close to the resettlement site. This could repre- area”3. We might ask ourselves if there could be a possibility that sent a problem in convincing people to resettle. In some cases, the Chiaquelane become a city of environmental migrants. As people are site of origin is more than 20 km away, and the population does not returning, there are also people considering staying at the resettle- have access to farming land close by. They must therefore go back ment site. to their site of origin in order to cultivate. Populations indicated that At any rate, if the Government intends to make Chiaquelane a they return on a weekly basis to their site of origin. There is a high resettlement site, it has to make the space “a good place to live”, rate of return. In fact, Chokwè is historically marked by migration with basic infrastructure and public services, as well as proper live- flows. Chiaquelane’s population increased rapidly during the 2000 lihood support. NGOs and development actors should accompany floods that struck Chokwè which is historically marked by migration this action by raising awareness about environmental disaster risks, flows. But once the floods ended, the temporary settlements were living in community and political and historical influences in devel- disassembled and families returned to their old residencies in the oping attachments to a particular territory. lowlands. A general reluctance among affected populations to relocate after a natural disaster has been recognized in the literature. More recently, 2 Patt, A. and D. Schroter (2008) ‘Perceptions of climate risk in Mozam- bique: implications for the success of adaptation strategies’. Global Environmental Change. 18(3). pp. 458– 467. 3 UNEP/OCHA JEU / Kaisa Nugin, MSB, ENVIRONMENTAL ASSESSMENT 1 « Report matrix origin- Ressettlement site » by IOM Mozambique, April Flooding in the Gaza Province, Limpopo River Basin, Mozambique, April 2013. 22, 2013.

IDDRI STUDY 09/2013 75 The State of Environmental Migration 2013

IN FOCUS

Disaster response in Nigeria: Managing agreed on training 21 national migration officials with the goal of creating the National Intelligence Unit (NIU). In 2011, IOM launched the largest displacement of 2012 two-week training, Promoting Better Management of Migration in Nigeria, providing interview techniques, data collection and man- Sora Kim agement. Also, International Federation of Red Cross and Red Crescent Societies (IFRC) is helping the NRCS distribute aid more In 2012, Nigeria experienced massive rainfall from July to Novem- effectively by providing health care, water, sanitation and hygiene of ber. The unexpected floods killed hundreds, injured and displaced makeshift camps and distributing non-food. (Sep, 2012). millions. Even if this accident needed prompt actions to deal with The government needs to address several challenges such as the the emergency, the government failed to respond effectively because lack of data and the shortage of camp management capacity. Set- they were unprepared -lacked the data and plans- for the large num- ting up an accurate terminology for the displaced people and provid- ber of displaced. ing legal support are essential. In addition, the government should Inundation starting from July brought about release water from train officials to carry out timely interviews of the displaced people, four major dams in the Niger and Benue Rivers. Floods resumed gather information, and manage the aftermath of a disaster. Inter- in September and further hampered relief work and the rehabilita- national cooperation between NGOs, UN Agencies and the private tion process. At the end of September, most states reported reced- sector is key to react properly and effectively to disasters. ing water levels whereas delta and Bayelsa states had no signs of decreasing water. It took much more time for water to recede given that water corridors were larger than any other year and drainage channels were not working properly. Heavier seasonal rains coupled References with a poor drainage have threatened 19 states out of 38 mostly in IFRC International Federation of Red Cross and Red Crescent Societies the southern region of the country. Unexpected water run-off was not 2013 Emergency appeal operation update Nigeria (January): Floods. International timely managed with contingency measures and severe rain storm Federation of Red Cross and Red Crescent Societies. caused the overflow of water reservoirs. IOM International Organization for Migration The National Emergency Management Agency (NEMA) has prior- 2012 Press Briefing Notes. Récupéré sur Nigeria Immigration Service, IOM Esta- itized data collection and provision of relief supplies in cooperation blish HQ National Intelligence Unit: http://www.iom.int/cms/en/sites/iom/home/ with the Nigerian Red Cross Society (NRCS). NCRS is working jointly news-and-views/press-briefing-notes/pbn-2012/pbn-listing/nigeria-immigra- with the Nigerian Government urging better trained responders to tion-service-iom-establi.html manage emergency response. Mainly, the Nigerian government NEMA National Emergency Management Agency provided the contingency plan in response to sudden onset natural 2012 Emergency Zone. Récupéré sur Flood: more than 25,000 displaced in Benue disasters including camp management, emergency shelter and non- as many communities remain adamant: http://www.nema.gov.ng/emergency- zones/floods.aspx food items. Moreover, the Nigeria Immigration Service (NIS) and IOM

76 STUDY 09/2013 IDDRI PART 2 MOBILITY, RESETTLEMENT AND RETURN

The State of Environmental Migration 2013

DESERTIFICATION AND DROUGHT RELATED MIGRATIONS IN THE SAHEL – THE CASES OF MALI AND BURKINA FASO NAKIA PEASRON AND CAMILLE NIAUFRE

INTRODUCTION This paper considers how mobility patterns, already part of a traditional survival strategy in this Migration on the African continent occurs mainly climatically adverse region, are affected by recur- internally within countries as well as subregion- ring droughts and ongoing desertification. The ally across borders, with West Africa representing first section describes the climatic, environmental  per cent of all regional movements (Black et.al, and population conditions in the Sahel region, and ). Farmers have historically moved mainly briefly shows how closely their impact on migra- between neighboring countries due to “artificial tion is related to food security issues. Two succes- boundaries demarcating socially homogeneous sive case studies of migration patterns in Mali and units into separate states” (Adepoju :; Burkina Faso are here presented, including data qtd. Ammassari, Black, ). Both Malian and collected from Pearson’s field study in Burkina Burkinabé farmers that straddle agricultural and Faso in . The final section summarizes existing political borders have long employed a substan- policies related to migration and environmental is- tial part of their household labor force to work sues and provides some recommendations. on secondary cocoa and coffee fields in the Ivory Coast during the Sahelian dry season (De Haas et.al, ; Konseiga, ). 1. ENVIRONMENTAL While having long engaged in migration as a THREATS AND MIGRATION livelihood diversification strategy against the fragile Soudano-Sahelian climate, low-income and landlocked Mali and Burkina Faso have found it .. Variability of Sahel difficult to liberate their populations from the de- rain conditions bilitating episodes of recurrent droughts and erratic rainfalls that have plagued the Sahel for the Normal Climate Variability last half-century. Since the - droughts, both Situated in the Western Sahel, between the Sahara countries have experienced increased north-south Desert to the north and coastal rainforests to the internal migration, as farmers and herders escape south, Burkina Faso and Mali are intertropical the descending “sahelization” of the countries’ countries marked by a Sudano-Sahelian climate, northern regions (Albergel, Valentin, ). and are prone to strong geographic variation in That the two neighbors rely on rain-fed agricul- annual rainfall as presented in the table below: ture for sorghum and millet staples as well as for cotton exports, makes their economies and food Table 1. Annual rainfall in Mali and Burkina Faso security highly vulnerable to changes in tempera- Regions Mali Burkina Faso ture and rainfall. In light of the  Sahelian food Annual rainfall in North 50 mm 350 mm crisis following a major drought that has thus far Annual rainfall in the More than 1000 mm More than 1000 affected  million people in the region, and a pro- South mm jected  million people worldwide at risk of be- Source: Authors; Data Source: Hummel et al., 2000; World Bank Dashboard Overview ing displaced from desertification, land degrada- Burkina Faso, 2013 tion, drought (DLDD) (Almeria ), the need to Rain is mostly concentrated in - months of bring to the forefront the complex, yet widespread summer rainfall (May-September), when tem- impacts of slow-onset climate events, is crucial. peratures are high, thus accelerating evaporation rates. Intra-seasonal droughts are common, . They would return to cultivate cotton and grain at home since rainy seasons consist of many high intensity during the rainy season July-September. storms (OSS, ).

IDDRI STUDY 09/2013 79 The State of Environmental Migration 2013

Figure 1. Annual rainfall variability in Sahel

Time-series of Sahel (10°N–20°N, 18°W–20°E) regional rainfall (April–October) from 1920 to 2003 derived from gridding normalised station anomalies and then averaging using area weighting (adapted from Dai et al., 2004a). Positive values (shaded bars) indicate conditions wetter than the long-term mean and negative values (unfilled bars) indicate conditions drier than the long-term mean. The smooth black curve shows decadal variations. [WGI Figure 3.37] Extracted from IPCC, 2008:80, fig. 5.2 Figure 2. Annual mean of daily mean temperatures in Western Africa

Extracted from Hummel et al., 2000: 22

Increasing climate variability On the temperature side, the Intergovernmental UNEP () classifies the major drought events Panel on Climate Change (IPCC) predicts a warm- of - and - as part of an ongoing ing of Africa by .°C to .°C per decade, with drought period starting in the s, and inter- even higher temperatures expected in Sahelian, rupted occasionally by one-off seasons of Central and Southern Africa (IPCCC, ). In the adequate rainfall. But while precipitation rates Sahel, the rise of temperature could reach +. to have improved since , there have been longer +.°C between  and . These predictions successions of dry years, and single humid years, are still uncertain, and their distribution unequal particularly , , and  (Ali ). between Sahelian countries, but the potential for Furthermore, this increased humidity has not been amplified evaporation of surface water and penu- uniform across the Sahel. Rain indices show that ry of water resources would strongly impact food isohyets have increased in the eastern Sahel (from security, whether in agricultural, pastoral or fish- Chad to eastern Niger) in the period -, ing sectors. In Mali, the production of rice by con- while they remain low in the western Sahel (from Senegal to western Mali) (Ibid). Additionally, the .  per cent of the rainwater in the Sahel is not absorbed timing of annual rain cycles is increasingly vari- by the land and runs off. able as precipitation may decrease only at the end . For instance, fish captures in the Malian Niger Delta of dry seasons rather than in the beginning (Ibid). went from , tons in an average year (as in ) to

80 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

trolled submersion, organized mainly by the Of- Figure 3. Evolution of the Malian population by region fice du Niger in the Ségou, Mopti and San Region, is bound to disappear by  if the IPCC predic- 12 tions of rain reduction are realized (UNFCCC Mali, Nothern regions ). 10 Southern regions

.. Desertification and 8 land degradation 6 millions Arid West African soils inherently suffer from poor fertility, exhibiting limited water retention 4 capacity, limited nutrient value, and reduced depth for root extension (Lahmar, ). When 2 erosive crusts, or zipelle, develop, they block infil- tration, initiating a vicious cycle of desertification 0 Population 1976 Population 2009 (Valentin, Casenave, ). The chemical depletion of soils sparked by the human elements of deforestation, bush fires, and over-cultivation, all Similarly, Burkina Faso has one of the fastest help to reduce organic material through oxidation growing populations in Africa, with an annual in- and the leaching of unused nutrients (PAN-LCD, crease of more than  per cent in - (UNEP, ). Such soils, having lost their physical and ). As in Mali, migrants move southward from chemical integrity, are more easily swept away in the populous Mossi plateau in northern and cen- the dust storms or violent rains that occur during tral Burkina, where naturally erosive soils are fur- the later part of the rainy season (World Bank ther degraded by longer dry seasons, driving farm- Dashboard Overview Burkina Faso, ). ers to the fertile East and West (PAN-LCD, ; According the UNCCD (),  per cent of Brown, Crawford, ). But while rural regions Burkina Faso’s land is degraded, while  per cent are emptying out in Mali, thereby posing a threat is at high risk of degradation, most of which in the to its primary sector that constitutes  per cent of densely populated Central Plateau. Likewise, Mali the country’s GDP, studies have shown that rural has witnessed a strong degradation of the vegetal out-migration is beginning to stagnate in Burkina cover since the s, leaving only  per cent of its Faso, while internal urban out-migration is on the total surface arable (UNFCCC Mali, ). increase (Beauchemin, ). These trends put a serious additional pres- .. Population Pressure sure on land resources and production. Sahelian farmers have had to expand their farmland de- Following the trajectory of Sub-Saharan Africa, spite declining soil fertility and increasing weath- the total Sahelian population has increased four- er vagaries. Researchers have linked the rate of fold since the s, growing at a . per cent rate rural population growth (.% in ) to the between  and , and will likely double rate of increase of cultivated surfaces (.%) between now and  (Ozer and al., ). (Cambrézy and Sangli, ). Ouedgraogo et. al’s In Mali, the total population grew from . mil- () study in Sissili province in southern Burki- lion in  to more than . million today (UN na Faso found that the annual rate of conversion World Population Prospect, ). If this extreme- of forest land to cropland at . per cent has ly rapid rate is maintained, the Malian population increased simultaneously alongside population could reach  million by  and more than  density from  inhabitants/km in  to  million in . This population is increasingly hbts/km in , due in part to migration since concentrated in the cities, with the capital city of the s droughts. Bamako having already reached  million inhabit- This phenomenon is also present in Mali where ants. Southern regions are the ones experiencing the FAO estimates a loss of forests of , ha/ the fastest growth, as shown in Figure . year. Here, it is mainly due to biomass energy

. The primary sector employs over % of their active populations and represents % GDP in Burkina Faso and % GDP in Mali. This share is bound to be reduced as the rural regions are emptying out (see Fig. ). Food insecurity could be experienced in the coming years, , tons in a dry year (as in ), which represents a further triggering migrations, whether from northern to diminution by % (UNFCCC Mali, ) southern regions or towards foreign countries.

IDDRI STUDY 09/2013 81 The State of Environmental Migration 2013

consumption: wood and wood coal covers almost Migration flows in Burkina Faso and  per cent of energy production in Mali, and con- Mali - Key figures sumption is rapidly growing alongside population Mali: rates (respectively +% and +% between  Around 200,000 permanent migration flows per year and ). Additionally, landclearing thousands Almost 300,000 displaced by the 2012 conflict and of hectares of land by brush fires contributes to 177,000 refugees (including 48,000 in Burkina Faso). rapid deterioration of soils (UNFCCC Mali, , 46 000 total international migrant stock 2005 ). Burkina Faso: Furthermore, in Burkina Faso, despite the per- 2,2 million permanent internal migrants total sistent droughts in the region, cattle populations counted in 2006 have doubled from  to , to ,,, 230,000 recent internal migrants 2006 leading to disequilibrium between the number of 773,000 total international migrants 2005 livestock and available resources. In the Central 54,000 recent international immigrants 2006 Plateau where Mossi farmers and Fulani herders have long cohabitated, mounting pressures on natural resources and converging production sys- Sources for Mali: 1992-1993 Malian Survey on Migrations and Urbani- zation (EMMU), quoted by Hummel et al., 2012, UNHCR, OCHA, 2013. tems (mixed cropping and pastoral farming) are Sources for Burkina Faso: RGPH, 2006 increasingly ending in conflict (Breusers, Neder- lof, Van Rheenen, (). The link between slow-onset events and migration .. Environmentally-induced Severe, irreversible forms of slow-onset events migrations are framed by that have a lasting impact on natural resources food security concerns tend to lead to permanent moves while sudden natural disasters spur more temporary migra- Desertification and droughts represents a tion (IOM, ; Tacoli, ). However, the major risk for food security initial coping strategies of seasonal and circular The ‘distress migration’ that followed the s migration during extended periods of drought Sahelien drought was characterized by rural- and loss of agricultural production can very well rural flows over shorter distances, serving as a key evolve into permanent migration as a survival survival strategy for keeping poor households from strategy when food security becomes problematic starvation (Black et al, ). The most recent (IOM, ). Jónsson (:) puts it bluntly: drought of , which has affected . million “permanent abandonment of an environmentally people (FAO, ), on top of the  million people degraded area is rather a solution, and immobility affected by the  drought (Aljazeera, ), would indeed be a major constraint, in some cases may be having the same effect. certainly resulting in continuing degradation and Burkina Faso has been considered one of the death from starvation”. most affected countries of the  Sahel food crisis, which inflicted a  per cent loss in cereal pro- Reactive food security strategies exacerbate duction from the previous year due to droughts and soil deterioration environmental degradation (FAO, ). Similarly, In response to the normal climate variability Mali could lose up to  to  per cent of its agricul- and related food vulnerabilities of the Sahelian tural capacity due to climatic changes (Hummel et region, rural populations tend to resort to adap- al., ). Furthermore, declining national harvests tation strategies to survive in the short term, could raise world market prices, leading to long- hoping for a milder rainy season the following term food insecurity. Already, millet prices had in- year, regardless of the long-term impacts. Smit creased by  per cent in Bamako, and  per cent and Skinner’s () climate change behavior in Ouagadougou between  and , the high- typology differentiates spontaneous responses est increases in the sub-region (Oxfam, ). such as selling off livestock as short-range and operational, while structural changes in management such as livelihood activities is strategic and long term (cited, Stringer et al., ). However, short term adaptive strategies leading up to migration, or failure to migrate, can worsen long- term soil fertility loss. . Given that each square kilometer of national territory Farmers often face deteriorating environmental could nourish  bovins,  ovins, and  caprins. Based on statistics presented by Dr. Augustine Kaboré from conditions by expanding their fields onto marginal INERA, Ouagadougou lands, which has put forests as well as wetlands in

82 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Figure 4. Desertification and reactive responses to food insecurity: a vicious circle leading to migration

Source: Authors, 2013. danger (IUCN, ). Mali counted . million The graph above attempts to conceptualize the ha cultivated in  against . million ha in linkages between desertification, migration and , with the same productivity levels (UNFCCC food security: Mali, ). Likewise, new land cultivation has led Increasing climate irregularities and the result- to a reduction of Burkina Faso’s savannahs from ing soil degradation linked to recurrent droughts , km to , km between  and  and demographic pressures become particu- (Cambrézy and Sangli, ). Earlier planting sea- larly problematic in poor countries like Mali and sons, often during the dry season, precipitate soil Burkina where livelihoods and daily sustenance regeneration. Traditional fallow periods, some- are heavily dependent on natural resources. Mi- times as long as several decades, allowing lands to gration remains a common survival strategy. Yet, “rest” and recover fertility, have been shortened to new challenges such as increased risk of conflict as little as five years, risking complete soil exhaus- resulting from resource competition and tenure tion (IUCN, ). insecurity in destination areas, as well as new vul- Overgrazing is another important cause of envi- nerabilities presented by the feminization of urban ronmental degradation. This relationship remains migrants, are being met with alternative migration unclear: while a stronger concentration of animals patterns and adaptation strategies. on small pastoral areas further deteriorates the vegetation cover, their manure and urine are ex- cellent fertilizers. However, herd movements can 2. IMPACTS ON MIGRATION FLOWS lead to the export of animal manure, thereby de- priving grazing areas of much needed fertilizers Mali and Burkina Faso present highly similar that could compensate for vegetation loss. profiles, on climatic, social and economic levels, and on the effects climate change and environ- . See examples in UNEP ; Mortimore  ment deterioration are impacting their already fragile populations. On the migration aspect, . Phone Interview With Mr. Manda Sadio Keita, Field Pro- gram Officer, And Mr. Modibo Touré, National Expert both countries experience more internal than Supporting The Program, FAO Mali, March th,  international migration, and more short-distance

IDDRI STUDY 09/2013 83 The State of Environmental Migration 2013

Table 1. Destinations Destinations of migrants according to the region of origin in Mali, 2005

Source: Hummel et al., 2012: 45; Database: CFSVA, 2000 than long-distance mobility. They also see their in the countryside, a part of the rural population urban population growing as a result of the cities’ moves to the cities, earns money through short- economic and demographic dynamism. Never- term jobs (Konaté, ), and eventually returns to theless, when looking at migration patterns, their fields to plant during the rainy season. Most divergence is clear. In Mali, most of the mobility often, farmers do not permanently abandon their forms part of a seasonal migration of rural popu- land to live in the cities (Jónsson, ). Today, lations towards the city during the dry season, the majority of the Malian population (%) more of these movements becoming permanent remains in rural areas (Hummel et al., ). and thereby amplifying the phenomenon of rural A large scale study launched in  by the Ma- exodus (Hummel et al., ). In Burkina Faso, lian government (ESBAN, ) showed that  while seasonal migration is increasingly perma- per cent of the households had sent one member nent, rural outmigration is slowing down while away and had received remittances in the previous migrants are starting to leave urban areas for rural six months. Upon return to their villages, seasonal ones (Beauchemin, ; Pearson, ; UNEP, migrants bring home their savings to support the ). family. In this way, seasonal migration serves to al- leviate financial strains on the household budget, .. Internal migration particularly in regards to diversifying livelihood flows in Mali: cyclical and sources and risks like climate vulnerability (Jóns- adaptation mobility son, ).

Seasonal rural-urban migration in Mali: the Origin and destination environment as a normal cause for mobility Permanent internal migration in Mali is inter- Internal migration represents % of total Malian regional (ESBAN, ). Cities are the main desti- migration (EBSAN, ). Since the s and nations for migrants, as illustrated in figure , the rapid growth of the cities (Denis, ), stim- except for the Kidal region, where two-thirds of the ulated both by demographic expansion and the migrants leave to neighboring countries like Niger rural population movements, seasonal migra- or Algeria. Bamako is the principle destination for tion has become part of the agricultural popula- most migrants, who comprised  per cent of its tion’s normal livelihood strategy. During the non- population in  (Hummel et al., ). The harvesting dry season, when less labor is needed second ‘hotspot’ for permanent internal migration is the central region of Ségou, where the activities of the Office du Niger attract more than  . See also Phone Interview With Mr. Manda Sadio Keita, Field per cent of internal migrants. The Office du Niger, Program Officer, And Mr. Modibo Touré, National Expert Supporting The Program, FAO Mali, March th,  . International migration represents thus %, mainly . See also Phone Interview With Mr. Manda Sadio Keita, towards other African countries, and only % for Field Program Officer, And Mr. Modibo Touré, National migration outside of Africa (EBSAN, ). Expert Supporting The Program, FAO Mali, March th, 

84 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

created by the French colonial authorities in , The Tuaregs and the Fulani – is today an , ha irrigated land dedicated Mali and Burkina Faso’s Fluid to ensure Malian food security, mainly through Populations rice-cropping. Its dynamism and job opportunities explain its power of attraction in terms of internal The traditional way of life of the nomadic and semi- migration (Hummel et al., ). nomadic Fulani pastoralists and Tuareg merchants living The Mopti region, situated at the cross-section in both Burkina Faso and Mali, involves transhumance of northern and southern regions, is estimated to migration, i.e., long-distance movement with cattle in bear a moderate migration deficit, compensated search of pasture and water, thought of as a traditional coping strategy in vulnerable climates (De Brujin, Van by immigration from the north. Generally speak- Dijk 2003). In Mali, these movements have occurred ing, a clear north/south movement appears when between the Sahel regions (namely the Kidal, Gao and studying Malian permanent internal migration Timbuktu regions) and areas closer to the Niger River (see map below). This mobility is impacted by during the dry season, whereas in Burkina Faso they three main factors: first, the northern regions are have occurred mainly in the Sahel and Central Plateau  dryer than the southern ones ; second, northern regions. However, the practice of transhumance has been regions, whose population density is lower, are substantially affected by the droughts of 1973 and 1984. poorer than southern ones; third, the majority of Distances have decreased substantially (often to less the economically dynamic cities are in the south- than 10km a year), and movements are more confined to ern half of the country (see figure  below). Sea- the rainy season. Instead of whole families leaving, tran- sonal migration movements follow the same gen- shumance has been limited to a few young men going eral North/South and rural/urban areas pattern alone with cattle. However, with fewer cattle to tend to (Hummel et al., ). in the dry season, seasonal urban migration has been reluctantly pursued by Fulani men as a survival strategy. Consequences of desertification and food In Burkina Faso, this group has engaged the most in insecurity on mobility patterns permanent out-migration from the Sahel due to a failure The consequences of desertification on migra- of pastoralist livelihoods, or feelings of shame, which tion are not immediate, but progressive, which made it difficult for seasonal migrants to return (Hamp- makes it difficult to establish a strong correla- shire, 2006). tion between the two phenomena. Nevertheless, Still, in Mali, a large portion of this population has it has been established that a degradation of become sedentary and diversified its way of life, set- environment conditions do increase migration tling mainly in northern cities where they practice agro- pastoralism (Hummel et al., 2012). Today, the distinc- rates. According to a  study, an estimated tion between farmers and pastoralists is no longer as  per cent of households intensify their seasonal strict as it used to be (Interview FAO Mali, 2013). The mobility in the event of poor harvests, sending herders still following transhumance paths are entering more members on the roads, while  per cent more and more in competition with farmers for the use migrate when there is crop destruction, and  of land, as desertification reduces the available arable per cent leave in the case of strong climatic events and grazing land. or shocks such as extreme droughts or floods (EBSAN, ). Episodes of extreme drought increase the numbers of inter-regional migrants. Moreover, interna- migration movements are occurring within the tional and long distance migration shifted to mi- region (EBSAN, ). The remaining migrants gration over shorter distances in order to sustain are mainly crossing the nearby border with the aggressive cereal and livestock shocks caused Algeria, often for transhumance and nomadic by the - droughts (Findley ). purposes (see table below). The region’s capital city, Kidal, is rather isolated from the rest of the Alternative patterns country, as , km separate it from Bamako and The arid Kidal region, populated by a majority  km from the nearest city, Gao. Tuareg population, is a notable exception to the Specific attention must finally be dedicated to a general migration patterns in Mali:  per cent of relatively new trend in internal migration: women now account for almost half of the migrants . They are also the ones affected by desertification. (Hummel et al., ). Going to work in the cit- Green savannahs surrounding the cities of Kaarta and ies during the dry season is now a general custom Bélédougou in the North-West, the Niger Delta, Dogon for many young women, often between  and country and Liptako-Gourma have become semi-desert. -years-old, to earn financial resources to sup- The Sahelian steppe that used to reach Timbuktu and Menaka are now parts of the advancing desert (UNFCCC port their family, but also to gather the content of Mali, ). their wedding trousseau (Sieveking, ). Local

IDDRI STUDY 09/2013 85 The State of Environmental Migration 2013

Map 1. Migration and Demographic Dynamics in Mali

Source: Authors. Datasource: UN World Urbanization Prospects: the 2009 Revision - Mali Country Profile by the FAO, 2013 - Hummel et al., 2012 - Jónsson, 2010 - UNFCCC Mali, 2000, 2011 - Mortimore, 2000 - OSS, 2007 - UNEP, 2006 - IPCC, 2007 populations appear to be more concerned about Dashboard, ), provoking troubles for the host female migration, as many never return to their regions, including in terms of food security. In native village. March , according to the United Nations Of- fice for the Coordination of Humanitarian Af- The  political and security turmoil has fairs (OCHA), more than , people were weakened the national food system displaced within Mali because of the conflict. The conflict in Northern Mali in  and its occu- Almost  per cent of Internally Displaced Per- pation by Islamic groups have provoked the cross- sons (IDPs) were in the Northern regions of Gao, border movements of thousands of displaced Kidal and Timbuktu, the most difficult ones to populations and refugees. The military confronta- reach by emergency aid and the most unsafe. tions in January , following the French inter- Unlike many other similar situations of internal vention, triggered even more people to move away conflicts, IDPs do not gather in massive humani- from their native region. tarian camps, says Mr Keita from the FAO Mali. In February , the United Nations reported , refugees, mainly in Mauritania (,), . Phone Interview With Mr. Manda Sadio Keita, Field Pro- Niger (,) and Burkina Faso (,) (UN gram Officer, And Mr. Modibo Touré, National Expert

86 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Figure 5. Burkina Faso Migration Balance by Region 2006 also particularly fragile for Malian refugees and Figure 6: Burkina Faso Migration Balance by Region 2006 Burkinabés in Oudalan and Soum provinces of the Sahel region in Burkina Faso, where large animal influxes are putting enormous pressure on areas still recovering from the  Sahelian drought and the resulting Sahelian crisis. The UNHCR Burkina Faso office has issued an appeal for additional international aid, as refugees’ living conditions are in some camps considerably below emergency standards – i.e. receiving less seven liters of water per day (UNHCR, ). The Conflict in Northern Mali in  has provoked the cross-border movements of thousands Source: Authors,Source Data: Authors, Source: DataRGPH –Source: Theme 8, RGPH 2006. – Theme 8, 2006 of displaced populations and refugees. The military battles/conflicts in January , following In the country, solidarity and networks are chan- the French intervention, triggered even more nels that are absorbing the flows of migrants: in people to move away from their native region. southern cities such as Bamako or Ségou, fami- In February , the United Nations reported lies are hosting many of their relatives and for- , refugees, mainly in Mauritania (,), mer neighbors who are fleeing the conflict. In Niger (,) and Burkina Faso (,) (UN Bamako itself, only two small buildings owned Dashboard, ), provoking troubles for the by the Catholic Church and overcrowded, were host regions. In March , according to the converted to host IDPs. This absorption by lo- United Nations Office for the Coordination cal populations makes it extremely difficult to of Humanitarian Affairs (OCHA), more than count and provide the migrants with humanitar- , people were displaced within Mali be- ian support. Local associations of migrants com- cause of the conflict. Almost  per cent of In- ing from the same city (“associations de ressor- ternally Displaced Persons (IDPs) were in the tissants”) are quite active in this process. Many Northern regions of Gao, Kidal and Timbuktu, families left one or two members in the North to the most difficult ones to reach by emergency aid look after the household house and fields. The and the most unsafe. division of the country temporarily disrupted the normal movements of population and distribu- .. Burkina Faso’s Permanent tion of food, seeds and other products between internal flows – Adaptation the two halves of the country. through Migration Mr Keita is convinced that most of the IDPs & Improved Farming are waiting for the end of the conflict and will eventually return home. Following the French in- Increased climate variability as cause for tervention under the “Serval” Operation and the more long term rural-rural migration progressive pacification of the North, this return While Burkina Faso was still an emigration movement has slowly begun. Moreover, thanks country between  and  , internal rural- to the “humanitarian corridor” established by rural migration over short distances played a key the Islamist occupants and according to the na- role in survival strategies to the drought. Rural tional expert of the FAO interviewed, a massive zones represented  per cent of destinations and hunger crisis in Northern Mali was avoided. The  per cent of departures in - (Coulibaly  harvests were considered to be good ones, et. al., ), reflecting the difficulty for house- at sufficient levels (OCHA, ). holds to afford long-distance migration in times Meanwhile, Mali is facing a critical moment of scarcity. Unlike Mali, where most internal in terms of food security, with . million Ma- migrants ends up in cities, internal migration in lian people estimated to be food insecure and Burkina Faso has remained strongly rural-rural, . million children at risk of severe malnutri- with . per cent of migrant settlement in rural tion (UN Dashboard, ). The World Food zones in  and . per cent in  (Coulibaly Program is already assisting , people on et. al ). a daily basis in Mali since the beginning of , While an increasingly variable climate has in- including , in the North.This situation is tensified households’ dependence on extra-farm activity like seasonal migration (Peyraut, ), environmental migration is gradually becoming Supporting The Program, FAO Mali, March th,  a permanent response (UNEP, ). Short-term

IDDRI STUDY 09/2013 87 The State of Environmental Migration 2013

rainfall deficits tend to increase the risk of long- forests into farmland areas, which coincides with term migration to rural areas and decrease the substantial land acquisitions by private inves- likelihood of short-term moves to distant destina- tors encouraged by the government, have also tions (Henry, ). increased pressures on migrants who now have to operate in a harsher tenure environment. Origin and Destination Field studies conducted by Pearson in , in As in Mali, a north-south trajectory describes three villages in the southern, western, and eastern Burkina’s more permanent flows (Lindqvist, destination regions of Burkina Faso for migrants ). . per cent of long-term internal migra- coming from northern and central provinces reveal tion in Burkina Faso occurs between regions, the the migration trajectories and coping strategies of outflows stemming from the dry Central North, migrants at their destinations (Pearson ). North, and the densely populated Central (Mossi) Plateau, (figure ). Migrants are headed mainly Migration Trajectories : fieldwork to seek jobs in the capital Ouagadougou in the conclusions Center (.%). Comparable to Mali’s industrious The studied villages of Bakaribougou, Lan, and Ségou region that hosts the Office du Niger cotton Tagou, are diverse examples of host villages administration, the cotton-producing zone in the located in Burkina Faso’s cotton basin in Haut western region of Haut-Bassins, home to the coun- Bassins, the fertile Central West region that began try’s second biggest city, Bobo-Dioulasso, attracts attracting migrants after the droughts in the early .% of these migrants (RGPH-Theme , ). s (Ouedraogo et.al ), and the East region, whose weak immigration rate allows for greater Figure 6. Burkina Burkina Faso population figures by tenure availability for land-seeking migrants region 1997-2009 (RGPH-Monography of the East Region, ). A host to large migrant flows since the s, Haut Bassins is Burkina Faso’s oldest agricultural colonization zone. However, the rate of migration to the region has since fallen as migrants are heading farther south to less densely populated areas like Comoé, Kénédougou, and Poni (Lignon, Leclère, undated). Migrants now often make up the majority of villages, while some have created new ones. With seven out of its eight hamlets comprised of migrants, Bakaribougou was created by returning migrants forced to work in Mali under the French colonizers (Pearson, ; PCD-AEPA ). Source:Source: Authors,Authors, DataData Source:Source: INSD INSD Online Online Database Database : http://www.insd.bf/fr/: http://www.insd.bf/fr/ In contrast to the populous cotton region, a historically low population density (. hbts/km) Thieba () characterizes the scarcely popu- and abundant forest cover made the Central West lated south-western regions as new recipients of region a preferred destination for herders and migrants, particularly returnees from the Ivory farmers fleeing the droughts in the North and Coast, since the unstable political climate and anti- Central-North (Ouedraogo et. al., ). Rapidly foreigner sentiment in this country have diverted decreasing forest cover has led to many studies many seasonal Burkinabé workers towards the linking population growth to deforestation in the capital (Wouterse, ). Migration has lead to region (Ouedraogo et. al ). In , migrants population growth in the s-s in previ- made up . per cent of Sissili Province, where ously sparsely populated areas like the East and Lan is located (RGPH-Theme , ). Now one Central West. of the most densely populated villages in the rural commune of Léo, Lan grew from , inhabitants Case Studies of the Lan, Bakaribougou, in  to , in  (INSD Onsite Database, Tagou destination villages ), reflecting both natural increase and contin- Migrants in destination areas have been threat- ued migrant flows. ened by tenure insecurity and increasing demo- Also characterized by low population density graphic pressure. The  RAF law, which had (. hbts/km) and relatively abundant land, the intended to allow agricultural land access to all east of Burkina Faso has a migrant population of citizens regardless of their origin (Mathieu et al, only . per cent though this percentage has been ), has been largely ignored by indigenous steadily rising (RGPH-Monography of the East Re- populations. Meanwhile, the rapid conversion of gion, ; INSD Online Database, ). Tagou,

88 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 2. Distribution of the population by region of birth and area of residence in Burkina Faso, 2006

Source: Authors; Data Source: RGPH-Theme 8, 2006 located in Gourma Province, whose population  years for Samos. In Lan, Miriam Barry’s family comprises of  per cent migrants, has attracted of herders had stayed  years in Sapouy, an area much of this flow. Its population – , in  further North in the province, but faced difficulties -- constituted  per cent Fulani migrants, and  when attempting to sedentarise. Souleman Barry’s per cent Mossi migrants, compared to the native family spent  years in Koudougou in Northern Gourmantché population (%) (RGPH-INSD Sissili, before heading to Zoro, where they spent Onsite Database, ).  years before moving even further south to Lan  years earlier, largely because of declining pasto- Situation in the villages ral space for their animals (Pearson ). These observations suggest the need for improved poli- A north-south step pattern of migration due cies to protect migrant pastoralists and to improve to land scarcity in traditional destinations their integration as farmland expansion increases For many study respondents, their current villages in destination areas. were not the first place they had settled. In Lan, migrants had moved in a north-south step pattern, Tenure insecurity forced out of more populous communes like Despite habitual tensions over crop damage Koudougou ( hbts/km) (RGPH-Monography caused by Fulani animals, or field expansion, and of the Central West Region, ) to areas further a fragile hierarchical relationship between the south. This pattern is also reflected in the intra- migrant étranger domicilié and the local tuteur, provincial southward movements within the East region, with migrants leaving heavily populated . Only local villagers belonging to the patrimonial lineage Gnagna province to settle in Gourma, from where of the village (tuteurs) have access to tenure rights, and farmers emigrate south to Kompienga (RGPH- are allowed to delegate them to foreigners (“étrangers Monography of the East Region, ). domiciliés”), who in their turn are expected to demon- In all three villages, the Fulanis were the lat- strate their willingness to integrate in the village by various forms of social commitments. See more in Hochet, est to arrive, with average settlement duration of Peter, Saïdou, Sanou, . « Reconnaissance des droits . years compared to . years for Mossis and fonciers locaux Enjeux opérationnels de létablissement

IDDRI STUDY 09/2013 89 The State of Environmental Migration 2013

fieldwork was part of a larger spirit of coopera- when such actions negatively affect him (Oue- tion, in which migrants were included in village draogo et.al., ; Benoit, ). De Zeeuw’s decision-making. () observations of the high value placed on Located in the densely populated region (. peaceful social relationships in western Burkina hbts/ km) of Haut Bassins, Bakaribougou has Faso helps us to frame the cultural values at work undergone a widespread monetization of land in village structures of authority. rights with most migrants renting their land for , CFA under an annually renewable con- Table 2. Average Total Soil & Water Conservation tract (Pearson, ). However, such land prac- (SWC) Techniques Practiced per Household in the Three tices theoretically allow locals to reclaim land Destination Villages when they need it, by not renewing contracts, or Locals Migrants exchanging good land for marginal land, a process that may intensify inequality between mi-

grants and locals. Native Bolons Samos Mossis Sissala Fulanis Migrant

Moussa Drabo, a migrant farmer in Bakaribou- Gourmantché Gourmantché gou interviewed during Pearson’s field studies, Lan 4.6* 3.8 7.4 explained: Bakaribougou 9 5.3 9.2 18.5 Tagou 7.2 12.3 15 6.7 “Before, during my father’s time, you could Average 4.6 9 7.2 7.1 9.2 13.6 6.7 access land by offering a sacrifice of a chicken. Now it’s money that makes things turn. No Source: Authors; Data Source: Pearson, 2013 money. No field. Before, [land] was free. […] *Numbers based on survey responses from 90 farmers in the three villages. SWC techniques were based on CILSS documents on “Bonnes Pratiques” and informant Each year, you must [now] work in the owner’s interviews with extension agents as well as farmers. Techniques do not include field if you can’t pay the , CFA. You could synthetic fertilizers or pesticides. cultivate one ha of the proprietor’s land yourself, and one ha of the tutor’s land” (Ibid). In general, study results show that migrants tend to be more tenacious farmers. Mossis, who Tenure insecurity may have indirectly lead are particularly perceived by locals in all three to longer land use especially among migrants, villages as problematic due to their aggressive though this has often come with intensification land-clearing and rapidly expanding popula- efforts through the use of manure and SWC tech- tions, engage in the greatest variety of SWC niques, as migrants fear the loss of rights to land techniques introduced by the extension services. re-inserted into the bush for fallow. An agent at the Provincial Agricultural Service in Lan said that Mossis were ‘innovative’: “They Limits to migration: Scarcity leads to come from far to search for their livelihoods…. adaptation The natives are born here so they’re not worried Such efforts to adapt by increasing yields in a about the competition.” The village’s Parcelle drying South-western climate is linked to a sense Vitrine de Producteur, an experimental farming of scarcity in the mind of migrant farmer, Adama plot to demonstrate good practices, is owned by Sawadogo, another study respondent: a Mossi migrant farmer. Indeed, similar results were found by Gray and Kevane () who “Because of the growth of the population, the showed how income levels improve the likeli- cultivatable surface of the land is insufficient. hood of SWC investment, a way to secure land I don’t blame the locals. Now we have to give rights in Haut Bassins. back our land. We’re scared of no longer having In the absence of robust national adaptation any land. But if we lack land, we have to adapt. policies, populations in both Mali and Burkina The problem [of scarce land resources] is eve- Faso are continuously applying autonomous in- rywhere. We can use the manure technique to sights to adapt to an increasingly volatile Sahe- improve the land. We can’t run anymore; there lian climate. Farmers continue to move in search is no more land” (Ibid). of fertile areas abundant in land and better livelihood opportunities, creating new routes where Adama’s account displays a more nuanced un- old ones have become congested, and adapting derstanding of population pressure and the ability to their new environments where fewer resources to contextualize the behavior of local actors even demand better land management and cooperation. National policies have to some extent sup- des accords de prêt prévus par la loi burkinabè portant ported these grassroots responses, though there régime foncier rural N° », GRN Notes, GRET website. remains work to be done.

90 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

3. POLICY RESPONSES AND Assistance of Internally Displaced Persons in THEIR INSUFFICIENCIES Africa. Also known as the Kampala Convention, this mechanism identifies the environment as a Burkina Faso and Mali have taken many steps at cause for forced displacement, but only in the case the international and regional levels to protect of “natural disasters”, without much attention their migrants on the one hand and to deal with dedicated to slow onset events. environmental transformations on the other. But Finally, both Mali and Burkina Faso ratified the despite efforts to take into account the environ-  United Nations Convention relating to the mental cause for migration, they fail to bring an Status of Refugees and the  Organization for integrated response to this multidisciplinary issue. African Unity (OAU) Convention governing the specific aspects of refugee problems in Africa. .. International legal Neither policy instrument explicitly states the en- responses to displacement vironment as a cause for the forced crossing of international borders. Labor migration and development Given the importance of labor migration to income .. Protecting the diversification in the environmentally unstable environment and enhancing Sahel, the ratification approach to international agricultural resilience and legal frameworks is clearly dedicated to the protection of labor migrants and to endorse migra- There is no evidence that national authorities tion as a development strategy. recognize the growing desertification as a direct Ratified by both Burkina Faso and Mali in , “push factor” for migration, whether permanent the  UN General Assembly of the Conven- or seasonal. However, they acknowledge that tion on the Protection of the Rights of All Migrant this phenomenon is a threat for development and Workers and Members of Their Families (entered food security of rural areas and implement poli- into force in July ), while not specifically cies through National Plans to prevent it, in the identifying cross-border migration as an adaptive framework of the  United Nations Convention strategy, does define rights for international sea- to Combat Desertification (UNCCD Mali, ; sonal workers (Article ). UNCCD Burkina Faso, ). Focusing further on facilitating intra-regional Measures to combat drought effects on food se- migration, the  Economic Community of West curity through scientific research have been sub- African States (ECOWAS) Protocol Relating to stantially influenced by the creation of the Club Free Movement of Persons, Residence and Estab- du Sahel/CILSS (Permanent Interstate Committee lishment is a particularly relevant instrument in for Drought Control in the Sahel) partnership in an area where seasonal cross-border is of greater  that, despite its shortcomings, has helped to scope than long-distance international migration. draw attention to agriculture and ecological is- The EU-African migration policy partnerships sues, increase aid flows, and improve dialogue on foster migration as a development strategy to development assistance in the region (OTA, ). reduce intercontinental flows. The Migration While CILSS has found it difficult to infiltrate gov- for Development in Africa (MIDA) programme, ernment policy, the organization’s Director of Nat- launched by the IOM and endorsed by the Afri- ural Resources Management (NRM), Edwige Bot- can Union (AU) in , aims at encouraging Af- oni, has said that both Mali and Burkina Faso have rican nationals and highly qualified profession- actively engaged in efforts to improve agriculture als to directly contribute to the development of to better adapt to climate change (Interview-in- their countries of origin, combating brain drain. Person, ). Along the same lines, the ILO and NEPAD launch Both countries are also rather active in the im- of the  “Africa Labor Migration Policy Initia- plementation of the  United Nations Frame- tive” assists African countries in developing local work Convention on Climate Change (UNFCCC), labor migration frameworks. Additionally, the AU Common Position on Migration and Development and its Strategic Migration Policy Framework, . In article . of the Convention, States commits to prevent internal displacement related to environment adopted in . issues by “devis[ing] early warning systems, in the context of the continental early warning system, in areas of IDPs and Refugees potential displacement, establish[ing] and implement[ing] On the crucial topic of internal displacement, disaster risk reduction strategies, emergency and disaster preparedness and management measures and, both Mali and Burkina Faso signed and ratified where necessary, provid[ing] immediate protection and the  AU Convention for the Protection and assistance to internally displaced persons.”

IDDRI STUDY 09/2013 91 The State of Environmental Migration 2013

with several reports published since the s, Conditions (PC-ACV, ), and the National Envi- detailing the sources of greenhouse gas emissions ronmental Action Plan all tout the development of (GHGs) and how to mitigate them. Given that Mali infrastructure and public services as a way to re- and Burkina Faso contribute very little to global tain migrants (Beauchemin, ). While Burkina GHGs, the reports focus on adaptation strategies Faso’s Poverty Reduction Strategy mentions the rather than mitigation. The link between climate role of migration and/diasporas in stimulating change impacts and environmental migrations is trade, it fails to report any net gain in human capi- not explicitly referred to neither does it appear tal as a result of migration (Black, ), promot- clearly in the targets of the actions implemented ing instead job creation and rural road construc- (UNFCCC Mali, , ; UNFCCC Burkina tion (Beauchemin, ). Faso, ). Furthermore, agricultural intensification is one In , Mali and Burkina Faso presented their of the main lines of action to develop rural areas National Adaptation Programmes of Action (NA- and ensure food security in the long term (FAO PAs) in the framework of the implementation of Mali Interview, ), and prevent north-south the UNFCCC. The primary objective of these NA- migration flows. A  Report by the Malian PAs is to identify a set of priority projects to cope Ministry of Agriculture champions improved ag- with the most urgent dangers of climate change. ricultural mechanization as a means to improve Developing the agro-pastoral sector, protecting the countries agricultural production system for crops and cattle, and promoting food security are which only  per cent of the “Agriculture Pro- among the top five priorities for both countries. duction Units” in the central zones of Sikasso and None of the reports quote migration triggered by Ségou are equipped with elementary material environmental changes as a long-term policy ad- such as animal-towed ploughs. aptation strategy. It is only briefly mentioned in Likewise, access to ploughs in Burkina Faso was the Burkina Faso NAPA as an endogenous practice made available by government programs aimed at to cope with a deteriorating environment (NAPA increasing the production of cotton (Ouedraogo Burkina Faso, ). et. al., ), while the promotion of ‘agribusiness- men stakeholders’ by the government to invest in .. The insufficient management large, mechanized farms (- ha), has been of rural development and touted as a means to improve food security (Oue- seasonal migration draogo, ). Investing generously in agriculture, the government plans to increase sorghum Enhancing the development of rural regions: and maize yields by  per cent and  per cent an indirect tool against migration? respectively by  (AGRA, ). The answer largely proposed to mitigate environmental pressures is the development of rural Adaptation in Burkina Faso: examples of regions and poverty reduction through improved state-led resettlement, improved farming infrastructure and local job opportunities. This and rural development strategies is clearly aimed at reducing rural exodus for Conceptualized in the aftermath of the severe economic reasons. Sahelian droughts of -, the Volta Valley The  Bandiagara PRBP project in Mali Development (AVV) program was primarily an (rehabilitation of dykes, dams and roads) has attempt to resettle families from the densely enabled the opening up of many isolated Dogon populated Mossi Plateau to repopulate the valleys villages and the development of local small busi- of the three Volta rivers that had been ravaged by nesses, with the reduction of permanent migra- the onchocercose cattle epidemic in the s tion as one of its declared objectives. The results, (FAO, ). While the state initially failed to according to the  Impact Report, are ambiva- relocate the desired populations due to high lent: on the one hand, the new resources enabled costs and unresolved tenure tensions between young people to finance their trip to the cities; migrants and locals, these newly developed areas on the other hand, the return rate of seasonal would eventually attract spontaneous migrants migrants is higher since the development of the fleeing the droughts in the s (Thieba, ). region (Sieveking, ). Migrants were responding to what Henry () Across the border, in Burkina Faso, the National analyses as interconnected economic and envi- Programme against Desertification (PAN/LCD, ronmental factors, whereby natural resources ), the Programme for Improvement of Living were economically valorized (cash crops), and development initiatives such as hydroagricultural installations and production systems were organ- .  in the case of Mali,  in Burkina Faso. ized in destination areas. Such areas around the

92 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Kompienga and the Bagré barrages have since However, most of these movements escape the become centers of diversified activities resulting attention of national authorities and rely only on from new farming hamlets, the development individual efforts, networks and solidarity. For of horticulture, tree plantations, etc (Marchal, example, after the last rain of the season, many Quesnel ). young women leave their village as a group, are The importance of natural resource manage- then accommodated by a host family in town and ment in Burkina Faso’s climate adaptation policy come back together after the dry season. This approach, spurred by SWC campaigns led by NGOs system also puts pressure on each member of the and foreign donors since the s (Kondé, ), group to return to the village at the same time as is reflected in its numerous national plans and the others. programs such as the National Biodiversity Strat- Despite these solidarity links and the support egy (), the National Action Plan to Combat of seasonal migrant networks, living in the cit- Desertification (), the National Action Plan ies as housemaids and houseboys can be highly for Integrated Management of Water Resources dangerous, and the youth often experience abuses (), and the Environmental Plan for Sustain- from their employers or new neighbors. There is able Development (PEDD-). SWC campaigns a clear lack of national policy to protect seasonal have lead to at least , ha of rehabilitated migrants: national authorities do not clearly rec- land (CILSS ), while crop yields increased by ognize the social and financial insecurity for often - per cent in fields treated with rock bunds very young migrants. and zaï (Sawadogo ). There is some evidence of rural communities at- Although such scientific studies have been tempting to regulate the migration of their own nationalized under the Institute of the Environ- population, acting on behalf of the national gov- ment and Agricultural Research (INERA), R&D in ernment. According to Sieveking’s () study on Burkina Faso largely depends on financing from the very dry and isolated Bandiagara region, local international donors like the World Bank, and authorities, worried about the development of fe- fluctuations in funding since  have raised sus- male migration and its more and more permanent tainability issues (Stads et. al. ). characteristic, adopted rules, whether financial or While development of rural areas and improved social, sanctioning any woman migrating without farming techniques are perceived as better adapta- prior authorization from their father or husband. tion policies, migration is recognized as one of the In conclusion, the current national and interna- principle risks of natural disasters in the country’s tional policy framework adopted by both Mali and national platform for the Hyogo Framework for Burkina Faso provide some level of protection for Action - the National Council of Emergency Relief migrants abroad, but internal migrants, who con- and Recovery (CONASUR, ; UNISDR, ). stitute most of the environmentally-induced flows, Moreover, the PAN/LCD (), which classifies remain vulnerable. While policy efforts promote migration as a survival and livelihood diversifica- rural development and agricultural resilience to tion strategy, focuses on the need to control migra- climate change, little is done to integrate migration which it links to environmental degradation tion into this formula of adaptation. in destination areas.

Mali: accompanying seasonal migration As previously analyzed, migration is part of a regular livelihood strategy for many rural populations through seasonal mobility between rural and urban areas. This cycle is well known by the Supporting The Program, FAO Mali, March th,  authorities, which try to better accommodate it. . Phone Interview With Mr. Manda Sadio Keita, Field Pro- When the rainy season approaches, national media gram Officer, And Mr. Modibo Touré, National Expert broadcast messages in the cities to encourage the Supporting The Program, FAO Mali, March th,  migrants’ return to the countryside and support . Some rural young men for example travel everyday out- the farmers’ efforts in the fields. This campaign side of Bamako, pick up hay and plants, come back to the city and sell their findings to local livestock farmers for appears relatively efficient: during the harvesting very little money (Konaté, ). period, there is sometimes a lack of houseboys . For instance, the sexual abuse of under-aged girls who and housemaids in the cities, positions that are may end up getting pregnant or ill during their time commonly occupied by temporary rural migrants. in the cities sometimes prevents them from returning to their village at the next rainy season, partly out of shame. Also, recovering their salary is often extremely . Phone Interview With Mr. Manda Sadio Keita, Field Pro- difficult, as employers may take advantage of their youth gram Officer, And Mr. Modibo Touré, National Expert and inexperience (Dembélé, ).

IDDRI STUDY 09/2013 93 The State of Environmental Migration 2013

.. Coping with environmental The emphasis placed on the regulation of in- migrations in the Sahel region: ternal mobility, rather than on the promotion of Policy Recommendations flows, can be understood by the lack of governments’ capacities and funds to support urban pop- The first challenge emerging from the policies’ ulation growth. Additionally, the desertion of poor assessment conducted in this article is the lack of rural areas is not a long-term solution to fighting legal protection for migrants facing the specific chal- climate change or reducing adversities linked to lenges of an environmentally and socioeconomically it. Both local and migrant populations have a ma- fragile Sahel. The insecurity derived from farmland jor role to play in the sustainable management of expansion in destination areas in southern Burkina resources and to stop the advancing desert. Fur- Faso calls for improved policies to protect migrant thermore, the studies we have cited show that an pastoralists. Likewise, in Mali, policy instruments unchecked rural exodus can simply shift environ- must be applied to protect seasonal migrants and to mental degradation from origin to destination reinforce support networks in cities. The few steps areas. taken to protect IDPs are encouraging, but they Yet, examples from Burkina’s AVV program have make no reference to slow-onset events like deser- demonstrated that development of rural areas tification as a cause for displacement; moreover, has indirectly facilitated permanent rural-rural they need to be integrated in national policies in migration, relieving population pressures on en- order to provide concrete protection on the field for vironmentally stressed areas. Field observations Burkinabé and Malian citizens. show how newfound scarcity in rural destina- In both countries, most environmental migrants tions has spurred the cooperation and exchange remain within national borders and are thus not of improved farming techniques between local and covered by international frameworks protecting migrant communities (Pearson, ). These ef- labor migrants and refugees. Furthermore, though forts could be better applied if the potential role instruments like the ECOWAS Protocol provide of environmental migration in climate change was frameworks to foster regional migration, which better studied and more integrated into adapta- may in turn help to relieve population pressures in tion policy. With their use of sustainable strategies countries of poor climate and few natural resourc- (supported by the State) to increase productivity, es like Burkina Faso and Mali, systematic denial environmental migrants appear less as an ecologi- of rights and expulsion of foreign workers in key cal burden and more as a potential solution in re- countries like the Ivory Coast demonstrate how ducing risks and vulnerabilities to climate change. such policy instruments are often at the whims of national politics (Black et.al ). While such regional mechanisms should be improved and ex- CONCLUSION panded to uphold permanent residence rights and to protect environmental migrants, national poli- Evidence from both Mali and Burkina Faso points cies must first address internal migration, as this out that the migration in response to environ- has been the most common trend of environmen- mental problems is almost never entirely forced tally-induced migration in both countries. or voluntary, but rather a gradual response to Simply considering environmental migration a environmental changes that alter the capacity to problem to be fought is counterproductive. Its in- response. tegration, for example, in NAPAs is essential to bet- An even more intense version of this phe- ter understand the phenomenon particularly in the nomenon is occurring today with the recurring context of more voluntary, slow-onset events. Issues Sahelian environmental crises of , , related to permanent internal migration such as ten- , and  continuing to weaken the cop- ure security in destination areas must remain a pri- ing capacity of highly vulnerable populations ority for governments in a way that incorporates de- (FAO, ). Furthermore, the flight of , centralization and participatory processes involving Malian conflict refugees in recent months has both migrant and local decision making and custom- substantially disrupted farming activities in the ary consideration. Policies must be equally accom- drought-sensitive and migrant-senders Prov- panied by effective legal support and the upholding inces of Timbuktu, Gao, and Kidal. Moreover, of property rights, whether this involves property the settlement of these migrants in neighboring titles, monetarised tenure, or temporary contracts. countries has exacerbated local vulnerabilities in Northern Burkina. . See for example Maitre Abdoulaye Harrissou’s Simplified Secure Title, an innovative instrument to provide land “Access to land in Sub-saharan Africa, key to development”, security to African populations (Heard at the conference on June , , at Ifri, Paris)

94 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Less media-grabbing examples of drought- the linkages between the migration and other ef- induced migration in Burkina Faso has shown forts to build resilience against climate change and similar impacts of migration on scarcity in tra- how such strategies may be mutually reinforcing. ditional destination regions. However, the “pop- While the FAO responses to the crisis has at- ulation-degradation” (Kevane, Gray, ) para- tempted to minimize migration, which it per- digm is nuanced by examples of innovation and ceives to be a negative mechanism of adapta- investments in sustainable agricultural among tion (FAO, ), Mr. Modibo Touré of FAO Mali both migrants and local farmers in the face of re- offers that permanent migrations, especially source scarcity, which point further to the need international movements, are also positive as to protect vulnerable groups such as women and they allow for numerous resources that would the poor. otherwise not be generated to pay for seeds and Indeed, it is here where the nexus of mitiga- farming inputs. Moreover, if migration is to be tion and adaptation form a continuum. Whereas, seriously considered in policy responses to envi- the UN Special Reporteur for the Right to Food, ronmental degradation and food security in the Olivier de Schutter, advocates agro-ecological region, care must be taken to coordinate the link- farming methods that are better equipped to ab- ages between migration, survival, mitigation, sorb expected climate shocks (De Schutter, ), and adaptation, but at the same time involving better managed remittances from seasonal and rural communities in the planning process in or- permanent migration may be key funding sources der to help populations build resilience against for scaling up such farming methods and building an ever-evolving climate. up resilience. Furthermore, Pretty et. al’s () study show links between the implementation of SWC practices and decreased rural exodus as young men choose to become day laborers on . Phone Interview With Mr. Manda Sadio Keita, Field Pro- degraded land owned by farmers rather than gram Officer, And Mr. Modibo Touré, National Expert migrate. More efforts should be made to measure Supporting The Program, FAO Mali, March th, 

IDDRI STUDY 09/2013 95 The State of Environmental Migration 2013

Findley, S. E. 1997 “Migration and Family Interactions in Af- BIBLIOGRAPHY rica” in Adepoju, A (ed.) Family, Population and Development in Africa. London: Zed Books A, B Jónsson, G., 2010. “The Environmental Factor In Migration Dynamics – A Review Of African Studies,” Working Paper No. Albergel, J. V. C. «Sahélisation» d’un petit bassin versant sou- 21, International Migration Institute, University Of Oxford danien : Kognere-Boulsa, au Burkina Faso : la dégradation des paysages en Afrique de l’ouest : points de vue et perspectives de Henry, S., 2003. “The Influence Of The Natural Environment recherches, Ministère de la Coopération et du Développement, On Migration In Burkina Faso,” Louvain-La-Neuve: UCL, Presses Universitaires De Dakar 2003.-190 P. Henry, S. et al., 2004. “The Impact of Rainfall on the First Ali, A. 2010. « La variabilité et les changements climatiques au Out-Migration: A Multilevel Event-History Analysis in Burki- Sahel : comprendre la situation actuelle de part l’observation na Faso,” Population And Environment, 25(5):423–60. », Le Sahel Face Aux Changements Climatiques : Enjeux Pour Un Développement Durable, Centre Régional Agrhymet, Bulletin Hummel, D., Doevenspeck, M., Samimi, C., 2012. Climate Mensuel, Permanent Interstate Committee for Drought Con- Change, Environment And Migration In The Sahel, Selected Is- trol in the Sahel (CILSS), 2010 sues With A Focus On Senegal And Mali, Working Paper No. 1, Migration, Climate & Environment, Frankfurt Ammassari, S., Black, R., 2001. “Harnessing The Potential Of Migration And Return To Promote Development, IOM Migra- Kandji, S. T., Verchot, L., Mackensen, J., 2006. “Climate tion Research Series, Geneva, Switzerland Change and Variability in the Sahel Region: Impacts and Ad- aptation Strategies in the Agricultural Sector,” UNEP/Word Beauchemin, C. 2004. « Pour une relecture des tendances mi- Agro-forestry Centre (ICRAF) gratoires internes entre villes et campagnes: une etude com- pare Burkina Faso-Côte d’Ivoire » Cahiers Québécois de Démog- Kevane, M., Gray, L. C., 2001, “Evolving Tenure Rights and raphie, Vol. 33, no. 2, automne 2004, p. 167-199 Agricultural Intensification in Southwestern Burkina Faso,” World Development, Vol. 29, No.4, p.573-587, Elsevier Sci- Beauchemin C., Schoumaker B., 2005. “Migration to Cities in ence Ltd. Burkina Faso: Does the Level of Development in Sending Areas Matter?”, World Development, Vol. 33, No. 7, pp. 1129–1152. Kevane, M., Gray, L. C..1999. “A Woman’s Field Is Made At Night: Gendered Land Rights And Norms In Burkina Faso,” Benoit M., 1982. Oiseaux de Mil Les Mossi du Bwamu (Haute- Feminist Economics, Taylor And Francis Journals, Vol. 5(3), Volta), ORSTROM (IRD) p. 1-26. Black R., Ammassari S., Mouillesseaux S., Rajkotia R., 2004. Kondé, M., 2001. Bilan descriptif et analytique des cadres “Migration And Pro-Poor Policy In West Africa,” Working Pa- et outils méthodologiques de planification du développement per, C8, Sussex Centre For Migration Research. local et de lutte contre la désertification, Permanent Interstate Black, R.,, Sward, J., 2009. “Migration, Poverty Reduction Committee for Drought Control in the Sahel (CILSS) Strategies and Human Development,” Human Development Konseiga, A., 2007. “Household Migration Decisions As Research Paper 2009/38, UNDP Survival Strategy: The Case of Burkina Faso”, Journal of Af- Breusers, M., Nederlof, S., Van Rheenen, T., 1998. “Conflict or rican Economies, Centre for the Study of African Economies Symbiosis? Disentangling Farmers-Herdsmen Relations. Case (CSAE), Vol. 16(2), p. 198-233 of Moose and Fulbe of the Central Plateau, Burkina Faso.” The Lahmar, R., Bationo, A.B., Lamso, N.D., Guéro, Y., Tittonell, Journal Of Modern African Studies, 36 (3), 357-380. P., 2012. Tailoring Conservation Agriculture Technologies to Cambrézy L., Sangli G., 2011 (March). « Les effets West Africa Semi-Arid Zones: Building on Traditional Local géographiques de l’accroissement de la population en milieu Practices For Soil Restoration. Field Crop Res. 132, p. 158–167 rural africain : l’exemple du Sud-ouest du Burkina Faso », CFC, Lignon, M., Leclère, C., undated. « Développement Agricole N°207. Au Burkina Faso. » Dans le cadre de l’UV longue dirigée par Casenave A., Valentin C., 1992. A Runoff Capability Classifi- M. Dufumier, ‘Développement Agricoles dans les Pays en cation System Based on Surface Features criteria in Semi-Arid Voie de Développement’ Areas of West Africa. J. Hydrol., 130: 231-249. Lindqvist, S., Tengberg, A., 1993. “New Evidence Of Deserti- Cissé S., 1989. Pratiques De Sédentarité Et Nomadisme Au Mali. fication From Case Studies In Northern Burkina Faso,” Geo- Réalité Sociologique Ou Slogan Politique ?, Politique Africaine, grafiska Annaler. Series A, Physical Geography, Vol. 75, No. N°34 3, p. 127-135 De Bruijin M., Van Dijk H., 2003. “Changing Population Mo- Marchal, J.-Y., Quesnel, A., 1996. « Dans les vallées du bility In West Africa: Fulbe Pastoralists In Central And South Burkina Faso : l’installation de la mobilité », La ruralité dans Mali,” African Affairs, 102, 285-307, Royal African Society 7 les pays du Sud à la fin du XXème Siècle, ORSTOM Editions, De Haan A., Brock K., Coulibaly, N., 2002. “Migration, Liveli- Paris. hoods And Institutions: Contrasting Patterns Of Migration In Mathieu, P., Zongo, M., Paré, L., 2002 (December). “Mon- Mali”, Journal Of Development Studies, Volume 38, Issue 5, etary Land Transactions In Western Burkina Faso: Commodi- Pages 37-58 tisation, Papers And Ambiguities”, European Journal Of De- De Zeeuw, Fons, 1997. “Borrowing Of Land, Security Of Ten- velopment Research; Vol. 14 Issue 2, p. 109 ure And Sustainable Land Use In Burkina Faso,” Development Mortimore, M. J., Adams, W. M., 2000 (August). “Farmer And Change, Vol. 28 (1997), 583±595. Institute Of Social Adaptation, Change and & Crisis in the Sahel,” Global Envi- Studies, Blackwell Publishers Ltd ronmental Change, Vol. 11, p. 49-57

96 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Nelen, J., Traoré, N., Ouattara, M., 2004. “From Colonization Coulibaly S., Dallaire L., Gregory J. et. al, 1975. Les Migrations To Consultation: Regulating Use Of A Pastoral Zone In Sam- Voltaïques : Mesure De La Migration, 1969-73, National Insti- orogouan, Burkina Faso,” Issue Paper No. 129, International tute of Statistics and Demography (INSD), Ouagadougou. Institute For Environment and Development, IIED Denis, E., Moriconi-Ebrard, F., 2010. Africapolis : Etude De Ouedraogo, I., Ouattara, K., Kaboré, Sawadogo, S., Paré, S., L’urbanisation En Afrique De L’ouest, AFD/ SEDET Barron, J., 2011. “Permanent Internal Migration As Response De Schutter, Olivier, 2010. Report submitted by the Special Rap- To Food Shortage: Implication To Ecosystem Services In porteur on the right to food, Human Rights Council, UN Gen- Southern Burkina Faso,” Food Production – Approaches, Chal- eral Assembly lenges And Tasks, InTech, 41-55, Ed. Anna Aladjadjiyan Dabiré B., Kone H., Lougue S., 2006. « Theme 8 - RGPH- Ouedraogo, M., 2002. “Land tenure and rural development in 2006) », INSD, Ministry of Economy and Finance, Burkina Burkina Faso: Issues and Strategies,” Issue paper no. 112, IIED Faso, National Census Committee, Central Census Office Ouedraogo, T., Savadogo, Compaore, Oden O., 2010. “Land FAO/WFP, 2013. Sécurité Alimentaire Et Implications Humani- Cover Change And Its Relation With Population Dynamics In taires En Afrique De L’ouest Et Au Sahel, Décembre-Janvier 2013 Burkina Faso, West Africa,” Land Degradation & Development Vol. 21: 453–462 FAO, The Sahel Crisis 2012, Situation Update, 2013 (July 1). Ozer P., Gassani J., Hountondji Y-C, Niang A. J., Ambouta K., FAO, The Sahel Crisis 2012, Situation Update, 2012 (December 2007. « La désertification est-elle en recul au Sahel ? » Bois Et 4). Forêts Des Tropiques, N°293(3) FAO, 1987. Effets des migrations sur les structures agraires et Pearson, N., 2013. « Burkina Faso: The Effects of Environ- l’emploi rural au Burkina Faso: Rapport préparé par la société mental Migration on the Implementation of Agro-ecological africaine d’Etudes et de Développement methods , Exchanges and Cooperation, » Unpublished Master Government of Burkina Faso, 1999. « Programme d’Action Thesis, Sciences Po National de Lutte Contre la Désertification (PAN/LCD), » Piguet, E. et al. 2011, « Changements Climatiques Et Migra- Ministère de l’Environnement et de l’Eau Secrétariat Perma- tions : Quels Risques, Quelles Politiques ? », L’information Géo- nent du Conseil National Pour la Gestion de l’Environnement graphique, Ed. Armand Colin, Vol. 75(4), p. 86-109 Government of Burkina Faso, 2007. Programme d’action natio- Pretty et al., 2011, “Sustainable intensification in African agri- nal d’adaptation à la variabilité et aux changements climatiques culture,” International Journal of Agriculture, Vol. 9(1), p. 5-24 (Burkina Faso NAPA), Ministère de l’Environnement et du cadre de vie Reenberg A, Lund C., 1998. “Land Use And Land Right Dynam- ics - Determinants For Resource Management Options In East- Government of the Republic of Mali, 2011. Plan d’Action Natio- ern Burkina Faso.” Human Ecology, Vol. 26: 599-620. nal Climat du Mali, Ministère de l’environnement et de l’Assai- nissement et l’Agence de l’Environnement et du Développe- Sawadogo, H., 2011, “Using Soil and Water Conservation ment durable du Mali Techniques to Rehabilitate Degraded Lands in Northwestern Burkina Faso,” International Journal of Agricultural Sustain- Government of the Republic of Mali, 2005. Promotion De La ability Vol. 9(1), pp. 120-128 Mécanisation Agricole, Direction Nationale Du Génie Rurale, Ministry of Agriculture Sieveking N., 2009. « Dynamiques Migratoires, Mobilité Et Développement Au Mali », Étude Effectuée Pour Le Ministère Intergovernmental Panel on Climate Change (IPCC), 2007, Fédéral Allemand De La Coopération Economique Et Du Déve- Pachauri, R.K. and Reisinger, A. (Eds.), Contribution of Work- loppement, Centre On Migration, Citizenship And Develop- ing Groups I, II and III to the Fourth Assessment Report of the ment (COMCAD) Intergovernmental Panel on Climate Change, Synthesis Report, Geneva Stads, G.-J., Sawadogo Kaboré, S., 2010. Burkina Faso: Eva- luation de La Recherche Agricole, Institut de l’environnement Intergovernmental Panel on Climate Change (IPCC), 2008 et de recherche (INERA) / Indicateurs Relatifs aux Sciences et (June), Bates, B.C., Kundzewicz Z.W., Wu S., Palutikof J.P. Technologies Agricoles (ASTI). (Eds.), Climate Change and Water, Chapter 5: Analysing regional aspects of climate change and water resources, IPCC Technical Stringer, Lindsay C. Dyer , Jen C., Reed, Mark S., Dougill, An- Paper VI, Geneva drew J., Twyman, Chasca, Mkwambisi, David, 2009.“Adapta- tions To Climate Change, Drought And Desertification: Local International Monetary Fund, 2013 (May). “Mali: Poverty Re- Insights To Enhance Policy In Southern Africa,” Environmental duction Strategy Paper,” Country Report No. 13/111, Wash- Science & Policy Vol. 12 issue 7, p. 748-765 ington DC Thieba, D., 2003. Enjeux fonciers et développement durable International Organization for Migration (IOM), 2009. “Mi- dans les zones libérées de l’onchocerose au Burkina Faso, Groupe gration, Environment and Climate Change: Assessing The Evi- de Recherche et d’Action sur le Foncier (GRAF) dence,” p. 72-73. Retrieved May 2012 Wouterse F., Taylor J. E., 2008. “Migration And Income Diver- International Union For Conservation Of Nature, Change: Eco- sification: Evidence From Burkina Faso,” World Development. system-Based Adaptation And Lessons From The Field, Gland, Vol. 36, p. 625-640 Monographs, Reports SwitzerlandOCHA, 2013 (8 March). Bulletin Humanitaire Bahan, Dalomi, 2006, Monographie de la Région de l’Est, Re- Spécial, Crise Nutritionnelle Et Sécurité Alimentaire Au Sahel censement General de la Population et de l’Habitation de 2006, Kaboré, M., Ouedraogo, F.G., 2006, Monographie de la Re- National Institute of Statistics and Demography (INSD), Min- gion du Centre Ouest, RGPH-2006, INSD, Ministry of Eco- istry of Economy and Finance, National Census Committee, nomy and Finance, National Census Committee, Central Central Census Bureau Census, Bureau

IDDRI STUDY 09/2013 97 The State of Environmental Migration 2013

Permanent Interstate Committee for Drought Control in the Alliance for a Green Revolution in Africa (AGRA), Burkina Faso Sahel (CILSS), 2009. The Silent Transformation of Environ- Grant Page, http://www.agra.org/where-we-work/burkina-faso/ mental and Production Systems in the Sahel: Impacts of Pub- Web. 12 avril 2013 lic and Private Investments in Natural Resource Management, Almeria statement, II International symposium Desertification and Synthesis Report, Amsterdam University, CILSS-PRASA- migrations, (2006) http://www.sidym2006.com/eng/eng_ponen- LCD-POP Dev. cias_conclusiones.asp, Web. April 27, 2013. Plan Communal de Développement sectoriel en Assainisse- Dembele, D., « Retour à la campagne : le calvaire des aides ména- ment (PCD-AEPA) de Dandé 2011-2015, 2010, Retrieved at gères, » Le Journal du Mali.com, 10 Juillet 2010, http://www.jour- the Mayor’s Office of the village of Dandé, Haut Bassins, naldumali.com/article.php?aid=170324, Web. Avril 2013 Burkina Faso in November, 2012. FAO, FAO’s Response to the 2012 Sahel Crisis, http://www.fao.org/ Observatoire du Sahara et du Sahel (OSS), 2007. Adapta- crisis/sahel/en/ Web. April 2013. tion aux changements climatiques et lutte contre la désertifi- cation, Introductive Note n° 1, 2nd edition, GTZ, Tunis Government of the Republic of Mali, “Les Changements Climatiques au Mali,” Ministère de l’Environnement et de l’Assainissement, http:// OCHA, Mali, 2013 (27 March): Urgence Complexe, Rapport www.changementsclimatiques-mali.org/, Web. April 2, 2013. De Situation N°29 Konaté, Boukary, « Citoyen Rural Chôme Difficilement En Ville, » United Nations Convention to Combat Desertification (UN- L’info depuis les villages Maliens, http://villageinfos.mondoblog.org, CCD), 1999 (July). “Programme d’Action National de Lutte 20 March 2013, Web. April 7, 2013 Contre La Désertification (PAN-LCD),» Burkina Faso INSD Online Database, http://www.insd.bf/fr/ . Web. February, United Nations Environment Programme (UNEP), 2011. 2013. Livelihood Security: Climate Change, Migration And Conflict In The Sahel International Displacement Monitoring Centre, “Guiding Principles on Internal Displacement,” http://www.internal-displacement. United Nations Framework Convention on Climate Change org/, Web May 20, 2013 (UNFCCC), 2000 (September). Communication Initiale Du Mali International Organisation of Migration (IOM), “Migration for De- velopment in Africa MIDA,” http://www.iom.int/cms/mida, Web UNFCCC, 2011 (June). Seconde Communication Nationale April 7, 2013 Du Mali “Moving Stories: the Sahel,” UK Climate Change & Migration Coali- UNFCCC, 2001 (December). Communication initiale nation- tion, April 16, 2013, Web. http://climatemigration.org.uk/moving- ale du Burkina Faso stories-the-sahel/ UNCCD, 2004 (September). Rapport National Du Mali Office du Niger, http://www.office-du-niger.org.ml/internet/. U.S. Congress, Office of Technology Assessment (OTA), Web. April 2013, 1986, Continuing the Commitment: Agricultural Develop- Oxfam, Crise alimentaire dans le Sahel Cinq étapes pour rompre le ment in the Sahel—Special Report, OTA-F-308 (Washington, cycle de la faim en 2012, Note d’information inter agences, http:// DC: U.S. Government Printing Office. www.oxfam.org/sites/www.oxfam.org/files/ib-food-crisis-sahel_- World Food Programme (WFP)/UNICEF/Helen Keller In- 30052012-fr_0.pdf. Web. May 31, 2012 ternational, 2009. Etude De Base De La Sécurité Alimentaire Population Division of the Department of Economic and Social Af- Et De La Nutrition (EBSAN), Republic Of Mali fairs of the United Nations Secretariat, World Population Prospects: Legal Documents The 2010 Revision, Web, April 2013, http://esa.un.org African Union, 2009. Convention for the Protection and As- United Nations High Commissioner for Refugees, Barbara Hendricks sistance of Internally Displaced Persons in Africa (Kampala calls for aid during visit to Malian refugees in Burkina Faso, http:// Convention) www.unhcr.org/pages/49c3646c2.html, Web. July 9, 2012 ECOWAS, 1979. Protocol A/P.1/5/79 Relating to Free United Nations International Strategy for Disaster Reduc- Movement of Persons, Residence and Establishment tion (UNISDR), Burkina Faso National Platform, Hyogo Frame- UN General Assembly, 1990. International Convention on work for Action, http://www.unisdr.org/partners/countries/ the Protection of the Rights of All Migrant Workers and bfa, Web. June 30, 2013 Members of Their Families, adopted by resolution 45/158 United Nations Sahel Humanitarian Dashboard, Web. February 2013, http://www.unocha.org/crisis/sahel

D World Bank Burkina Faso Dashboard Overview, Web. April 2013, http://sdwebx.worldbank.org/climateportalb/home. INSD, 2006, Statistics derived from 2006 RGHP Census. cfm?page=country_profile&CCode=BFA Unpublished raw data from onsite database, Direction Ré- gionale de L’Est, BP 108 Fada N’Gourma, Burkina Faso, Re- I S trieved November 2012 Phone Interview With Mr. Manda Sadio Keita, Field Pro- O R gram Officer, And Mr. Modibo Touré, National Expert Sup- porting The Program, FAO Mali, March 29th, 2013 Aljazeera, Burkina Faso’s Sahel parched by drought, http://www. Interview With Mrs. Edwige Botoni, Director Of Natural Re- aljazeera.com/video/africa/2012/06/201261925515585267. sources Management, CILSS, Ouagadougou, Burkina Faso, html June 19, 2012 December 13th, 2012

98 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

IN FOCUS 2012 Floods in Bangladesh and food harvest, which occurs in September/October as well as in March/ April. Other contributing factors to food insecurity include a large insecurity and growing population, lack of assets, debt, social exclusion, low job opportunities, marginalization, and of course, natural disasters Emily Ferguson (ACF, 2013). In urban areas, food insecurity is mainly determined by unemployment or under-employment, and a lack of marketable skills On average, one-third of the country is flooded during the monsoon due to low levels of education (Ibid.). season, becoming a common fact of life in Bangladesh. However, Food security is not only about production stocks, but also about due to the importance of the agricultural sector in the country and the future food requirements, level of income and commodity prices its sensitivity to environmental disruptions, ensuring food security (Yu et al., 2010). Currently, the major challenge for the country is to -- both in the aftermath of a disaster and in the long run -- is a critical issue for the country. The great level of poverty in Bangladesh, keep up with the food demands of a fast-growing population (1.3% which currently stands at 40 per cent of the total population, further growth rate) with increasing incomes. On average, the dietary diver- exacerbates the threat to food security (GoB, 2010). sity is quite low; in 2008, the Bangladesh Institute of Development Studies (BIDS) reports that a full 76 percent of calories come from Floods and food insecurity cereals (mainly rice), with the remaining 17 percent coming from On the 25th and 26th, June 2012, the Chittagong division faced other plant and animal sources (primarily fish) (Rizvi, 2013). 65 cm of rainfall, one of the highest recorded rainfalls in the last two decades. The rain lasted for a month, inducing flash floods, and major rivers flooding. The conjunction of those events explains the severity of the damages. Despite under-coverage by local media, this was the worst flood and landslide to have hit the Chittagong region References in the last 60 years (WASH, Shelter and Early Recovery Cluster & Action Contre la Faim (ACF), “Bangladesh,” webpage http://www.actioncontrela- GoB, 2012). Official estimates found that 1,700 acres of crops had faim.org/en/content/bangladesh (last accessed 09/04/13) washed away; entire villages were submerged; and more than 200 Carlo del Ninno, Paul A. Dorosh, Lisa C. Smith, and Dilip K. Roy (2001) “The mud roads had been destroyed. As the affected regions struggled to 1998 Floods in Bangladesh: Disaster Impacts, Household Coping Strategies, and recover from the floods, thousands of people were left food insecure. Response,” Research Report 122, International Food Policy Research Institute, Poultry farms and fish hatcheries were also severely affected, both Washington, D.C. of which are important sources of protein in the Bangladeshi diet Government of the People’s Republic of Bangladesh, 2010. National Plan for Dis- (Ibid.). aster Management 2010- 2015. Disaster Management Bureau, Disaster Manage- Food security is threatened by natural disasters as these events ment and Relief Division, Ministry of Food and Disaster Management, 2010. cause food prices to increase and incomes to decrease, which Rizvi, Najma, 2013 “Enduring Misery,” D2+C, 22 22 March 2013 from: http://www.dandc.eu/ heightens a household’s vulnerability to future crises that threaten en/article/ bangladesh- does- not-guarantee-food-security-demanded-its-constitution food security (del Ninno et al., 2001). Improving the level of food WASH, Shelter and Early Recovery Cluster & Government of Bangladesh, 2012, security remains a key development goal for the country. Flooding in Southeast Bangladesh – Needs Assessment Report August 2012 (Chit- tagong, Bandarban, and Cox’s Bazar district) Understanding food insecurity in Bangladesh Yu, Winston; Alam, Mozaharul; Hassan, Ahmadul; Khan, Abu Saleh; Ruane, Alex; Under normal conditions, more than two-thirds of the rural poor Rosenzweig, Cynthia; Major, David; Thurlow, James. “Climate Change Risks and face food insecurity during the lean (monga) season before the rice Food Security in Bangladesh,” Earthscan Climate, Routledge: 2010. Print.

IDDRI STUDY 09/2013 99

The State of Environmental Migration 2013

DROUGHT AND INTERNAL DISPLACEMENTS OF PASTORALISTS IN NORTHERN KENYA IN 2012: AN ASSESSMENT PIERRE BONNEAU

INTRODUCTION pastoralists that could be considered as internally displaced persons. Factors underpinning In March , the county of Isiolo was crip- pastoralists internal displacement were seen as pled by inter-ethnic clashes among Borana, very diverse and often multicausal, according Turkana and Somali pastoralists that triggered to the report. The report mainly focused on is- displacement of approximately , Turkana sues of displacement related to conflict – the pastoralists. Reasons behind the conflict were a biggest issue in Kenya. Climatic triggers of dis- complex mix of political and economic tensions placement such as drought were not addressed in addition to competition over access to natural in depth. Nevertheless, issues of drought and re- resources (grazing lands and water), accen- source depletion in northern Kenya are of utter tuated by recent migration flows of pastoral- importance and are connected, to some extent, ists from drought-affected areas. To quell the with flows of displacement—as the Isiolo con- violence, the provincial administration ordered flict shows. Indeed, there are many connections the pastoralists from neighbouring county of between the conflicts in the Northern regions of Wajir and Mandera who had recently arrived in Kenya and stress factors linked to droughts, in- Isiolo to return to their areas of origin, where cluding resource depletion of pasture and water, pasture had replenished (IRIN, b). livestock loss. This event encapsulates the complexity of The chapter assesses therefore drought-re- drought-related pastoralist displacement in lated internal displacements of pastoralists in Northern Kenya. Droughts trigger urban migra- northern Kenya as of  and questions to what tion flows of pastoralists, using displacement as extent droughts may be direct and indirect ma- a coping strategy. The arrival of new comers to jor factor for displacement. It interrogates the existing communities can reignite long-standing responses at the national and international level disagreements and conflicts over use of natural and presents the challenges and gaps to be ad- resources with other communities. Furthermore, dressed. The object of study is complex as it fo- local politicians exploit latent and long-standing cuses on different categories of pastoralist dis- community hostilities to fuel conflicts to affect placement, from flows of pastoralists displaced voting patterns and win office. Theses overlap- in  to communities of pastoralists that can ping environmental, economic and political con- still be considered in a state of protracted dis- ditions lead to deaths and displacement of pasto- placement as of . ralist communities. The pastoralist lifestyle relies on livestock . According to the UN Guidelines on Internally Displaced production in arid and semi-arid land through Persons, internally displaced persons are “persons or extensive mobility patterns and use of natural groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual resi- forage and water points. As such, it is intrinsi- dence, in particular as a result of or in order to avoid the cally linked with mobility and migration and in effects of armed conflict, situations of generalized vio- the popular perception of pastoralism, speaking lence, violations of human rights or natural or human- made disasters, and who have not crossed an interna- of pastoralists’ displacement would be tautologi- tionally recognized State border.” (UN, ) cal. Contrary to this perception, a report of IDMC . Secondary sources of information along with grey lit- and ISS (Sheekh et al., ) concluded that erature (media reports, articles, humanitarian reports) as of the end of  there were over , have been used for this paper.

IDDRI STUDY 09/2013 101 The State of Environmental Migration 2013

It describes firstly how drought in context in cycle in , compared with  per cent nationally northern Kenya contributes, along with other fac- (Fitzgibbon, ). The populations of the north- tors, to the slow-onset erosion of pastoralist live- ern areas are also extremely vulnerable to food lihood systems. It assesses secondly the situation insecurity: as of  all the northern areas were of drought-related pastoralists’ internal displace- classified as stressed or in crisis in terms of food ment in , by proposing a qualitative typology security (USAID, ). and some elements of quantification. It will then This poor level of development relates to a high- analyses the attention and assistance given at the ly insecure context. State security services lack national and international level, focusing on the any meaningful presence in the regions. Since the policies targeting IDPs, disaster-risk reduction, s, firearms have proliferated in the region. development of ASALs and security. Finally, it Inter-communal conflicts are frequent in these will conclude and propose recommendations for districts, killing  people in  (OCHA, b). a better understanding of drought-related migra- They are often based on long-standing, resource- tion and displacement flows within existing policy based issues (grazing rights, access to water and frameworks. land), ethnic and/or political hostilities. The practice of cattle raiding, traditionally used in a rather non-violent way to capture livestock, has become 1. DROUGHTS IN KENYA: BETWEEN more violent in the last decade because of the pre- SUDDEN DISASTER AND SLOW- viously mentioned factors, as well as the increased possibility of commercialization of the stolen live- ONSET EROSION OF PASTORALIST stock (Sheekh et al., ; Kaimba et al. ). LIVELIHOOD SYSTEMS The term “pastoralism” covers a wide variety of societies, cultures, livelihood strategies, mobility .. Pastoralism in Arid patterns and geographic repartition. Their com- and Semi-Arid Lands mon feature is to rely in majority, on livestock (cat- northern Kenya regions tle, sheep, goats and camels) for food and income, and on natural sources of forage. As presented in the introduction, this article In Kenya, “census surveys of pastoralists are con- focuses on the northern arid and semi-arid lands sidered inaccurate and available data on pastoral- districts of Kenya that are home to most of the ism are largely inconsistent and unreliable, even pastoralists in Kenya, represented in the map though their regional presence is significant: it is below. believed pastoralists occupy  per cent of the na- ASALs cover about  per cent of Kenya’s land- tional land mass in Kenya” (Sheekh et al., ). mass and support about one-third of the country’s Pastoralism contributes to around  per cent and human population and  per cent of the national  per cent to the country’s gross domestic product livestock herd. The economy of the arid districts (GDP), with the livestock sector providing an esti- is marked by pastoralism, while the better-watered mated  per cent of all employment opportunities semi-arid district features more diversified sources and more than  per cent of household incomes of income, such as agro-pastoralism, mixed farm- in ASALs (Schilling et al., ). Those statistics ing and rain-fed agriculture (Fitzgibbon, ). might not even reflect the total economic value Arid and semi-arid districts feature a very low of indirect benefits of pastoralism to the country level of development and a very high level of pov- (IIED, ). erty. Poverty rates in arid provinces are approxi- The single term “pastoralist” is misleading: mately  per cent in , whereas the national pastoralists are not a single homogeneous group. average is of . per cent (Fitzgibbon, ). In the Their mobility patterns differ, from wholly mobile northern districts of Marsabit, Turkana, Wajir and to semi-nomadic with seasonal migration (tran- Mandera, between  per cent and  per cent of shumant) or nearly sedentary with short-term and the population live below the absolute poverty line closed migrations. Far from being unchanging, (Government of Kenya, ). Decades of political pastoralist lifestyles are also evolving at a rapid and economic marginalization have contributed to pace through increasing diversification and edu- this low level of development. Infrastructure, fa- cation. Many pastoralists rely on other economic cilities and basic services are weakly developed: activities such as crop cultivation (agro-pastoral- access to water is problematic, the road network ism, predominantly in semi-arid districts in Kenya) is poor, health facilities are unevenly distributed and only one district – Isiolo – is connected to the . In order to restock, assert domination on a territory, national electricity grid. Only . per cent of stu- acquire honour, pay the dowry for a marriage, or other dents in the north completed their primary school cultural reasons.

102 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Map 1. The Arid and Semi-Arid Land Districts in Kenya

Source: Arid Land Resource Management Project

IDDRI STUDY 09/2013 103 The State of Environmental Migration 2013

Figure 1. Seasonal calendar and critical events timeline

Source: FEWSNET, 2012 or petty trade (Fitzgibbon, ). Also, because of to food insecurity vulnerability of approximately education many children are attending sedentary . million people (OHCHR, ; IRIN, b). schools while other parts of the family, mostly Adverse extreme environmental events affected men, are migrating seasonally with the herds. The different parts of northern Kenya in . Certain role of women, previously in charge of the domes- areas of Kenya where affected by an early drought tic chores and of childcare, is also changing to- prompted by erratic and short rainfalls. wards more involvement in diversification labour In July , parts of the north-eastern Kenya, through urban migration (IRIN, a). Pastoral- especially Wajir, were reported to suffer early ists populations are also characterized by impor- drought because the March-April-May (long-rain tant inequality of wealth and assets. season) rainfalls were delayed and “at a depressed volume, erratic, and unevenly distributed across . Droughts and vulnerability the northern, northeastern and southeastern pas- of pastoralists: repeated shocks toral areas” (FEWSNET, ). Water reservoirs and erosion of livelihood systems only received up to  per cent of expected levels, and pasture resources were depleted. The drought Importance and impact of drought in especially affected northern and north-eastern northern Kenya districts Turkana, Wajir, Marsabit, Mandera, Moy- ASALs areas in Kenya are characterized by low ale, Tana River, where . million people were con- and irregular rainfall (mm to mm annu- sidered as food insecure, either in “Stressed” or in ally) as well as periodic droughts. The pastoralist “Crisis” situations (FEWSNET, ). This early lifestyle is supposed to be adapted to drought by droughts triggered concerned of humanitarian ac- relying on extensive mobility to access pastureland tors that populations strongly affected by drought and water sources. However, for the past decades, of  were likely to fall again into a food-securi- droughts have had increasing impacts on popula- ty emergency. Indeed, most of the food insecure tion. Indeed, in  approximately , persons population in  consisted of pastoralists that were reported as affected by drought and requiring were already strongly affected by the droughts of food aid. Two decades years later in - this  – if not before — and who had not fully recov- figure reached . million, and an estimated . ered (KRCS, ). Hence, drought, it is very acute million during - (OHCHR, ). The form, can be classified as a sudden-onset event, al-  drought was one of the most severe in  years though it should not shadow the gradual erosion to impact the Horn of Africa. The impacts were of livelihood systems of pastoralists. multi-faceted. Because of reduced availability of water source and of pasture land, livestock deaths Droughts as an external livelihood shock and/or diseases were widespread. Crop failure, and the erosion of traditional coping resource depletion, rising food prices, decreasing strategies livestock prices and health and patchy livestock Over the last decade, droughts have been perceived buy-back programmes (IRIN, b) led to a loss as hotter and drier than before and it has been of livestock, a loss of purchasing power and hence attributed to climate change, even though there are uncertainties on the causality (IRIN, d). But drought in itself is not the main problem because . Interview with Justin Ginetti, IDMC,  April  for centuries pastoralists have developed coping

104 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

strategies to survive in naturally arid environ- cattle-raiding or resource-based ones (Sheekh et ments. Drought becomes a problem when coupled al.,  ; Opiyo et al., ). The discussion over to the vicious circle of vulnerability and poverty this link is of particular importance in the assess- of pastoralists in northern Kenya, demographic ment of drought-related displacement, often pressure, the erosion of traditional livelihoods, the categorized as conflict-induced displacement absence of service provision, political marginaliza- (OHCHR, :). tion, and the already significant food and human The advocates of the positive drought-conflict insecurity (Overseas Development Institute, link generally state that because of drought- ). Indeed, the drought starts a chain of events induced resource scarcity, pastoralists’ coping that can result in changing mobility patterns, over- strategies involve extending of mobility patterns, grazing, resource depletion, livestock disease and which increases the risk of competition for the death and destitution of communities. This mix of same resources and border conflicts on territory. factors may contribute to forced or obliged migra- Advocates refer to numbers to assert that there tion flows of pastoralists, or internal displacement, is a correlation between deaths led to clashes especially if coping strategies are unsuccessful. over resource and the presence of above-normal Droughts increasingly challenge pastoralists’ dry conditions. For example,  conflict-related coping strategies for livelihood shocks. One of the deaths were reported in northeastern Kenya in most common strategies is changing the mobility , mostly concerning pastoralists fighting over patterns of livestock to access farther pastures or resource, whereas only  deaths were reported water sources (Opiyo et al., ). Other strate- in  (IRIN, d). This approach of a positive gies involve diversification of livestock species and drought-conflict link is also framed by many me- economic diversification towards non-pastoralist dia reports (IRIN, a ; b ; c). activities such as crop farming, petty trade, rural- Others advocate on the contrary for a negative rural or rural-urban migration or charcoal burn- link between droughts and conflict by showing ing. However land fragmentation and the increase that cattle-raiding is more developed during the of insecurity and resource-related conflicts affect wet seasons and the short-rains period (Opiyo et changing mobility patterns. Land fragmentation al., ). This theory bases itself on cultural rea- is the result of a trend of privatization and enclo- sons, this season being the time where pastoralists sure of land for diverse purpose such as farming, restock their herds, implement rituals of passage ranching, conservancies, and touristic reserves. (that give importance to cattle raiding). Opportu- As a consequence, traditional resources and mi- nities to sell livestock in good condition and at a gration routes are not available any more, which good price are also more important during this pe- leads pastoralists to extend their mobility patterns riod. For Opiyo et al. (), this “suggests that the farther to access natural resources in territories two contesting resources-based theories apply not whose use is not previously agreed upon through as discrete scenarios but in a ’resource abundance- shared resource agreements. Hence, the extension resource scarcity’ continuum, thereby creating of mobility increases the risk of border conflicts non-deterministic relationship between resource with other pastoralist communities competing availability and pastoral conflicts.” Moreover, this for the same resource (Opiyo et al., ). Land argument for a non-deterministic relationship is fragmentation, combined with conflict and per- strengthened by the importance of external factors ceptions of insecurity (Opiyo et al. ) can also and stakeholders in conflict motives and incen- led to concentration of herding in certain areas, tives. In some cases, political incitements are key which leads to over-grazing and further depletes triggers of conflict in pastoralist communities (see natural resources. Because of the erosion of tradi- figure ) (IRIN, c ; g ; a). tional coping strategies, the impacts of droughts Empirical evidence from a survey amongst the are more significant than before as are livestock conflicting Turkana and Pokot communities at the deaths and conflicts. Turkana-West Pokot border, highlight how the motives for cattle-raiding differ between communi- The complex link between drought and ties: “on the Turkana side, drought-related hunger, conflicts poverty and lack of pasture are the central conflict Besides the impacts on natural resources, live- stimuli, while on the Pokot side the accumulation stock and food security, questions were raised on of wealth, payment of dowry and the expansion the links between droughts and conflicts, be they of territory are the main motives behind raiding” (Opiyo et al., ). These mixed analyses on the links between . The latter diversification being considered as negative because increasing resource depletion of scrubs and droughts and conflicts make necessary to ana- trees. lyse conflicts and associated displacements on a

IDDRI STUDY 09/2013 105 The State of Environmental Migration 2013

case-by-case basis without drawing early and sim- themselves identify drought as a major cause of plistic conclusions. At the same time, politically- their displacements. But drought is generally constructed conflicts would not happen if the is- not responsible alone, as our first section widely sues at stake, such as access to natural resources emphasised. It is sometimes difficult to identify and their depletion, were not felt as very problem- the first factor that contributed to displacement atic by the communities involved. It follows that and/or loss of livestock because conflict or cattle- displacement flows in  in northern Kenya raiding can precede or follow drought-related should be analysed in the light of this complex cattle deaths. Even in discussion with pastoralist nexus of drought, erosion of livelihood systems communities, it appears to be difficult to identify and conflicts. the first factor that led to displacement. On the whole, cattle-raiding, resource-based conflict and loss of livestock (triggered by drought, conflicts, or 2. A TYPOLOGY OF DROUGHT-RELATED both) appear to be the main factors contributing to PASTORALIST DISPLACEMENT IN 2012 drought-related displacement. Thirdly, due to lack of data on pastoralists communities, it is very difficult to identify precisely .. Qualitative and quantitative which communities are most likely to be dis- methodological challenges placed. However, nomadic communities that are extremely reliant on livestock and less diversified A first set of challenges come from the displace- are more likely to be strongly impacted by drought ment itself and the type of legal categorization and the loss of livestock and be subsequently dis- it leads to. By definition, an IDP is a person that placed. Finally, pastoralist communities’ reac- was “forced” to move. Yet, the pastoralist lifestyle tion to environmental drivers and drought may is based on mobility, adapting to rainfalls and vary greatly hence complicated the possibilities of resource availability, which are sometimes scarce systemizing their responses. Indeed, they will be or far away, leaving them “no alternative” but different according to their internal social organi- extending their patterns to access the resource. As zation, their integration and access to state and a consequence it is not easy to differentiate their non-governmental organization (NGO) services as regular forms of movements and forced displace- well as the economic resources and asset base they ments, and monitoring flows is problematic can tap into. (Sheekh et al., ). The same difficulty applies for pastoralists moving to urban centres or refugee .. A typology of pastoralists camps during droughts after having lost all or forced migration large part of their livestock, or before drought in anticipation. Secondly, it is difficult to distin- Providing with a clear typology of drought-related guish whether displacement are temporary or not, displacement is hence a main challenge. Yet, this because of the lack of access to information on article try here to assess the variations amongst the situation of IDPs once they are displaced. The incentives for migrations and migrants’ profile. few assessments realized by OCHA and the IDMC The first main aspect of distinction is the direct provide for aggregated numbers but generally do or indirect impact of drought on the decision to not precise the situation of IDPs, the assistance migration. given to them, their patterns of return. Finally, some pastoralist found in Northern Kenya may ) Direct drought-related displacement: Interest- be of bordering States, especially Somalia. They ingly, the drought as a push factor as been little ad- cannot therefore be considered as IDPs, although dressed in our context. This type of displacement belonging to cross-border pastoralist communi- is reported to be rare: out of  communities sur- ties. The legal category of IDPs therefore finds its veyed for a research purpose, only one declared limits in the light of the human realities on the to have been displaced only by drought, following ground. Additionally, when displaced, pastoralists total loss of livestock. generally do not identify and consider themselves as IDPs but rather as pastoralists without livestock . Interview with Martina Caterina,, IDMC, th April . – which is extremely grave in pastoralists’ culture. . Idem. A second row of challenges concerns the trig- . Idem. gers of drought-related displacement. Pastoralists . Interview with Malika Peyraut, th March . . Date of redaction of the article. . Interview with Martina Caterina, th April . Further . Interview with Justin Ginetti, th April  ; information can be accessed in the report of IDMC on

106 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Yet, the drought of  has probably led many and visibility on the issue (as the Isiolo conflict il- Kenyan Somali pastoralists of north-eastern Ken- lustrates (IRIN, b). ya to flee to refugee camps of Dadaab, claiming The latter category of urban migrant pastoralists to be Somali refugees, and register as prima facie that do not come back to their traditional lifestyle refugees (with reduced individual procedures to is generally referred as “pastoralist drop-outs” determine the refugee status, usually applied in (IOM, ). Even though the term is well known, situations of mass movements). This phenomenon there is still little research on the issue. A base- is thought to happen by many researchers, with line study by IOM () in Garissa showed that a high uncertainty on its extent because of absence common feature of all the dropouts was the com- of any monitoring. This phenomenon has grave plete loss of livestock. Most of them were strongly consequence in terms of human rights because vulnerable in terms of food security, shelter and once registered as refugees, pastoralists lose their access to water and generally lived off of wages rights as Kenyan citizens. Another practice of cer- from day-to-day informal odds job. IOM consid- tain Kenyan pastoralists would be to go into the ers that the situation of pastoralist drop-outs is refugee camps temporarily in order to access ser- comparable to internally displaced persons (IOM, vices (food, shelter, health, education) and then ). However, dropouts are not a homogeneous leave. This is further corroborated by members of group because “they range from destitute persons pastoralists communities themselves stating that to those who have adapted other livelihoods and “some families are disguising themselves as So- may be thriving within them” (HPG, ). While mali refugees to get food and medical assistance, destitute dropouts could be considered as IDPs for not because they are corrupt but because they are the purpose of assistance, how “thriving” dropouts desperate” (IRIN, c). These migration flows to should be categorized, is another issue. Another is- refugee camps in seek of service provision can be sue concerns the categorization of these dropouts seen at the same time as a voluntary migration as for the purpose of aid and whether they should be a coping strategy, or as an obligation for survival, entitled to “pastoralists policies” and assistance, technically classified as internal displacement. including food aid, or considered only as urban Indeed, the same applies for migration in urban population (HPG, ). settlements. According to OHCHR (:), “in- These complex urban migration flows are blur- creasingly severe and more frequent droughts {...} ring the distinction between coping adaptative have forced many to search for new forms of live- strategies and internal, rather forced, displace- lihoods, including in urban areas. However, there ment. It is likely that they happen chronically, in- has been a tendency in Kenya to consider that cluding in , but there is no information on the these groups are not displaced, since they are by scale of it, which precludes further understanding definition mobile”. The drought of  has likely of means of prevention, assistance, and durable driven many pastoralists into urban and peri-urban solutions for these specific displaced persons. areas. There, the situations might have differed significantly between individuals and communi- ) Indirect drought-related conflict displacement: ties. Either the whole family could have migrated in , five main areas of conflict provoked flows in urban areas, or some members of the family. of IDPs – or were hosts to enduring internally Moreover, some could have used this urban migra- displaced persons. UNOCHA (b) reported tion to access new economic resources and come , displaced and  killed in . back to their former lifestyle few months after or The following table reviews the most important support it, while others could have stayed indefi- to assess the extent to which conflict can be linked nitely in urban areas and permanently abandoned to drought-related factors, such as issues concern- the pastoralist way of life. It practically means ing livestock, resource access and resource deple- that pastoralists that moved to urban areas in  tion. The core of the conflict dynamics is linked to might still be there in , with little recognition cattle-raiding/livestock issues, or grazing land/ water issues, and therefore has environmental dimensions. It cannot be concluded that drought- displaced pastoralists published at the end of . related stress factors on the environment are the . Interview with Justin Ginetti, th April  ; Interview primary factors for conflict and subsequent death with Malika Peyraut, th March  ; Interview with Nuur Sheekh, th April . and displacement. On the other hand it cannot be . Interview with Malika Peyraut, th March . contested that the resource depletion associated . Interview with Justin Ginetti, th April . . The assessment of factors behind this choice of tempo- . The term encapsulates all the pastoralists that have rary or permanent urban migration is beyond the scope definitely left the pastoralism lifestyle – not necessarily of this study. urban migrant.

IDDRI STUDY 09/2013 107 The State of Environmental Migration 2013

Map 2. Inter-communal conflicts by district (21 November 2012)

Source: OCHA, 2012.

108 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Table 1. Conflict-induced displacements in 2012 and assessment of their linkage with drought-related issues Area Number of IDPs Qualitative assessment Linkage with drought-related issues as of 2012 Turkana 0 for OCHA Cattle-raiding between Pokot and Turkana communities Conflicts perceived to be politically connected, (2012b), over Attacks by the Pokot were reported in March 2012 and explained by but at the same time aggravated by drought 3000 persons political dynamics for contest of border territories. It would have led impacts on resource depletion. for other sources to 3000 IDPs in neighbouring areas. (Business Today, Attacks of the same kind (also between Uganda border Dodoth or 27th March 2012) Toposa and Turkana) are regularly occurring every year and tend to increase during drought period. (IRIN, 2011e) Samburu Between 5000 Cattle-raiding based conflict between Samburu and Turkana It has been linked with climate change and (OCHA, 2012b) communities increasing drought (ISS, 2012), with politics, And 11000 The Samburu massacre is closely linked to inter-community hostile as well as opposition to a disarmament (IDMC, 2012) dynamics between Turkana and Samburu since the 1990s linked programme. with access to resource and cattle raiding. The displacement was triggered by a military deployment following the massacre of 42 police officers during an operation to take back stolen livestock in November. Isiolo 9,575 (OCHA, Conflict between Borana, Turkana, Somalis A conflict “about political numbers, not 2021b) Presented as cattle-rustling but thought to be politically-instigated resources because civilians, including women in order to reach ethnically-based voting patterns before the 2013 and children are being killed and nothing election. stolen” (IRIN, 2011g). Also linked to claims over grazing land/rights by different On the other hand, the conflict was linked the pastoralists communities. presence in 2012 of Wajir/Mandera pastoralists that had previously migrated in the town during a drought period (IRIN, 2012b). Moyale 50,592 (OCHA, Conflicts between Borana and Gabra “A complex interplay between ecology, politics 2012b) Moyale clashes were a tribal civil unrest due to historical lands/ and ethnicity” (KCRS, 2012) territorial disputes, the upcoming general elections and historical disagreements/injustices (KRCS, 2012).

Tana River 30,000 (OCHA, Conflict between semi-nomadic Orma and farmers Pokomo Said to be politically linked, but at the same YEAR) Revenge conflicts, fuelled by a history of farmers-herders time long-standing resource base. conflicts over grazing lands, pasture and water (IRIN, 2012c). The 2011 drought has been described as a Many displaced families seeking refuge in nearby forests, host potential stress conflict factor in the area communities and camps. Kenya Red Cross assisted with food and (IRIN, 2011a). non-food items as well as medical care (IRIN, 2012e) Source: author.

Figure 2. Direct and indirect drought-related displacement dynamics in pastoralist communities

Source: author

IDDRI STUDY 09/2013 109 The State of Environmental Migration 2013

with repeated drought and the exhaustion of cop- but started recently to formally recognize and ing strategies are not important factors in conflict address the problem during the post-election triggers, along with political, historical and eco- violence of - that displaced approxi- nomic factors, and in the resulting migration mately , persons. This was one of the only patterns. cases where the Kenyan government implemented large-scale assistance for and monitoring of IDPs. .. A simplified model of the On the whole, the monitoring, registration, assis- relation between drought tance have remained ad hoc, weak and/or poorly and displacement factors coordinated, especially regarding pastoralist displacement. This subsection assesses the imple- The dynamics induced by the external shocks of mentation of the IDP frameworks in Kenya in this drought in the livelihood systems of pastoral- very specific context, to allow for a comparative ists are summarized in the figure below. Bidirec- analysis and subsequent assessment of the situa- tional arrows correspond to feedback interaction tion of drought induced IDPs. between factors. The words written in italic appear to be among the most important factors leading Monitoring and profiling of IDPs to pastoralist displacement, respectively cattle- The monitoring of internal displacement, through raiding, resource-based conflicts and loss of live- identification of IDPs, their locations, patterns stock. The two boxes in orange symbolize the lack of mobility, vulnerabilities and needs remains of conceptual understanding so far concerning “extremely weak” (Metcalfe, ). In , the the previously mentioned categories of pastoralist Kenyan Government, along with stakeholders in displacement. civil society and UN agencies, created the National Protection Working Group on Internal Displace- ment (PWGID), “with the objective of enhancing 3. POLICY RESPONSES the capacity of the Government and its overall AND CHALLENGES response to internal displacement in the country” (OHCHR, ). However, official presence in The policy and legislative responses to the specific northern Kenya is limited and the Government issue of drought-related pastoralist displacement has not undertaken any monitoring exercise for is nearly absent in Kenya. The issue of pastoralist displaced pastoralists (Sheekh et al., ). displacement in northern areas itself is not really acknowledged by policymakers, and suffer inex- Registration istent or incomplete data (Sheekh et al., ). Registration of IDPs is lacking coherence: except This section will assess the policy responses given during the post-election violence period in  through different bias, responding to specific and , no displaced have been profiled or regis- aspect of displacement: IDP protection frame- tered on the national database (IDMC, ). Other works, development in ASAL areas, disaster- non-registered IDPs include: post-election violence management policies and security policies. (PEV) displaced persons in host communities and urban settings labelled as “integrated IDPs” .. National responses by government (amounting to around , and frameworks persons), people displaced by natural disasters (flood, droughts), development or environmental Assistance and protection of internal project (IDMC, ), inter-communal conflict displaced (resource-based, cattle raiding, for other motives), Kenya recognizes the UN Guidelines for Inter- and pastoralists migrating in urban and peri-urban nally Displaced Persons, which explicitly recall settings following drought (OHCHR, ). state’s responsibility in its Principle : “states are under a particular obligation to protect against Assistance the displacement of indigenous peoples, minori- Registered PEV IDPs have been provided with ties, peasants, pastoralists and other groups with protection and assistance, though inadequate special dependency on and an attachment to their because of the lack of sufficient or nutritious food; lands” (UN, ). Yet, until recently, Kenya lacked access to shelter, water and sanitation facilities; frameworks to address pastoralist displacement. and other services such as education and health Kenya has a long history of internal displacement care (OHCHR, ). Also, most of the assistance has been provided to landowner IDPs (IDMC, . Especially issues related to land claims that sometimes ), and fraud has also been reported (Kenya date back to colonialism. National Assembly, b). For other IDPs, the lack

110 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

of registration entails a lack of visibility and recog- ). Also, IDPs could in theory seek “legal re- nition. Hence, assistance and protection, if any, course for compensation for life and property lost” are generally limited to food aid (OHCHR, ). because the state now is legally responsible to pre- According to IDMC, “many IDPs are displaced in vent displacement and can hence be held account- areas of the country that are environmentally and able (IRIN, f). However, the implementation, economically vulnerable, and as such they enjoy especially in northern areas where resources and fewer opportunities for integration and develop- capacity-building are scarce, will be a tremendous ment. This in turn increases the likelihood of their challenge. The data collection system has not yet living in situations of prolonged displacement” started to function, and it is unlikely that local (IDMC, : ) This also applies for drought- administrations will be provided with enough re- related pastoralist displacement. sources to do so on the short-term. Moreover, it remains to be seen if pastoralists migrating to ur- Return and durable solutions ban areas during drought or for drought-related A report of the Parliamentary Committee on the reasons will also be considered as IDPs and be the resettlement of the IDPs in Kenya criticized the granted adequate assistance and protection. return policies for registered PEV IDPs as being badly managed and flawed (Kenya National Assembly, Enhancing resilience and adaptive capacities b). Concerning displaced pastoralists, no such assistance seems to have taken place. According to Disaster-management policies Sheekh et al. () “as the displacement of pasto- Kenya has adopted several frameworks governing ralists is little understood either qualitatively or disaster management such as the draft National quantitatively, they face a greater risk or protracted Disaster Management Policy developed since displacement. There is currently no conceptual , a National Disaster Response Plan, and understanding of durable solutions specific to the Climate change response strategy of  Kenya’s displaced nomadic population, and as a (IDMC, ). However, these frameworks lack result no response that aims to achieve them”.  enforcement, adequate infrastructures and human resources, especially in northern areas (OHCHR Possible impacts of the development of a new ; IRIN, a, b). Additionally, the Kenyan IDP policy National Assembly created in November , the In October  a National Policy on the Preven- National Drought Management Authority to serve tion of Internal Displacement and the Protection as the focal point for information-sharing, early- and Assistance to Internally Displaced Persons has warning, rapid reaction and coordination of poli- been adopted after three years of drafting. The cies and measures targeting drought. text is largely based on international and regional The successes of these institutions and policies instruments such as the Great Lakes Protocol on remain however very doubtful. Communicat- IDPs and the African Union Kampala Convention ing early warning systems towards pastoralists – though Kenya has ratified none of them both. has to take their culture, beliefs language and lo- It provides a comprehensive, detailed and indis- cal knowledge into account. Beyond information criminate approach for management and protec- diffusion, infrastructural problems are also im- tion of IDP. For example, the definition of IDPs portant as pastoralists are sometimes not able to includes persons internally displaced by “natural access markets to trade their stock to anticipate disasters whether or not triggered by the change of drought hardships. Indeed the main challenge climate”, “politically instigated or inter-communal for disaster-management policies is the absence hostilities such as competition over lands or other of local capacity-building actors and access to re- resources” (Kenya National Assembly, a). As a sources and funds to implement prevention, miti- consequence, pastoralists’ internal displacement gation and relief actions. The existing contingency should finally be recognized. funds lack financial capacity. The long-term objec- Another central aspect is the provision for the tive, according to Mohamed Elmi, the Minister of set up of a data collection, registration and profil- ASALs, is to access Adaptation Fund money of the ing system on IDPs, while taking into account right UNFCCC for long-term interventions (IRIN, f). to privacy. The profiling is supposed to be carried Beyond the State, another key actor in natural within  days of an internal displacement (IDMC, disaster response is the Kenya Red Cross Society, one of the most specialized humanitarian actors in responses to displacement, notably consequent to . The report on pastoralists displacement published by IDMC at the end of  aims at providing a detailed review of displacement flows as well as a conceptual . Interview with Justin Ginetti, th April ; Interview understanding of durable solutions. with Nuur Sheekh, th April .

IDDRI STUDY 09/2013 111 The State of Environmental Migration 2013

natural disasters. Through its “tracing” approach, Security policies it provides much needed information on migra- Disaster-management and development policies tion flows of IDPs and refugees. It has provided will most likely remain ineffective if the secu- assistance to floods-induced displaced people in rity situation remained bleak in northern areas.  (IDMC, ). Quite interestingly, the KRCS Government policies on this matter have generally considers droughts, floods and inter-communal been criticized for their inadequacy, if not their conflicts as part of one single “complex emer- adverse effects. gency” calling for unified funding, preparedness The Kenyan State has tried to decrease insecuri- and response (KRCS, ). However there is no ty with strong disarmament programmes. Several information on their assistance to drought-related operations seem to have fuelled inter-ethnic clash- pastoralists displaced or pastoralist drop-outs. es and displacement, because the programmes were felt to be biased and heavy-handed (Sheekh Development policies for Arid and Semi-Arid et al., ). The proliferation of firearms from the Lands Sudan and Somalia has severely undermined dis- Development policies for Arid and Semi-Arid armament operations. Lands have been limited during a long-time. A Conversely, the Kenyan State has also carried traditional approach of ASALs by government out community armament programmes to out- agencies as well as humanitarian actors has been source security provision by turning pastoralist short-term assistance, notably in terms of food. communities themselves into home guards and There was a common perception of ASALs as a reservists. But many of the distributed arms ended “food aid sink hole” with little recognition of the up traded or used for personal security and even economic potential of these areas (IRIN, f). criminal attacks, fuelling the conflicts (IRIN, However, there seems to be a recent evolution in e). This short-term security approach, cou- the past few years. pled to the short-term food aid approach, has been In , the Ministry of State for the Develop- interpreted by certain communities that end up ment of Northern Kenya and Other Arid Lands was demanding “bullets and food” from their govern- created – the only of its kind in the whole Horn ment (IRIN, th July b). of Africa. Moreover, the draft policy on the sus- Long-term policies of conflict-management, tainable development of ASALs was approved in peace-building to try to end cattle raiding and re- February  after nine years of waiting (Kenya source-based conflicts have been left to NGOs sup- Rural Development Programme, ). It provides porting informal resource-sharing agreements set for investment in security, infrastructure, job crea- up by communities themselves. But those agree- tion, adaptation to climate change, management ments often break up as soon as the rain comes of drought, job creation and emphasizes the need back and replenishes the resources. More, the defor a “special treatment” of ASALs. But quite increase of the strength of traditional peacekeeping terestingly, none of these policies and approaches institutions, especially the role of elders in sanc- refers to the terms “displacement” and the term tioning conflict, is an issue upon which it is diffi- “drop-out” is referred only one time, without stra- cult to act. The issue of reconciliation for conflict- tegic approach. induced IDPs also plays a crucial role, but is rarely Concerning on-going implementation, a live- addressed on a large-scale and is poorly funded by stock marketing board is in working process to donors (IDMC, ). provide support for the pastoralist livestock value- chain. Indeed, while Kenya is a net importer of .. International responses meat (Government of Kenya, ), the livestock and frameworks value-chain of pastoralists in ASALs has never been strongly supported by government (IRIN, f). Donor countries and international organizations Finally, the new Constitution of Kenya adopted support different measures targeting IDPs (though in  is supposed to provide opportunities for not pastoralist) and pastoralist development, from securing tenure rights and rights to natural re- food assistance to livelihood transition. sources of pastoralists communities, which could reduce the process of land fragmentation (around Responding to internal displacement  per cent of the land in semi-arid areas is now Donor countries and international organizations used as national park or reserves) (IRIN, f). have supported the post-election process in  and  and assisted the Government and other actors to monitor and assist PEV IDPs, notably . https://www.kenyaredcross.org/, accessed the th of with interventions to provide water, sanitation July . and shelter as well as technical support to develop

112 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

the IDP policy through the setting up of a Protec- Migration has recently opened a pastoralist dis- tion Working Group (IDMC, ). pensary in Turkana. The IIED and the Climate However the same level of interest and assis- Adaptation Fund are funding early warning pro- tance has not been dedicated to displaced pasto- jects tapping into pastoralist use of radio and ralists and other non–PEV IDPs. For example, no mobile phones. USAID and UNICEF are devel- organization has undertaken comprehensive sur- oping mobile schools projects. Livelihood di- veys of the migration flows of pastoralists towards versification is also promoted and supported by urban centres during the drought of , as well international actors, for example through training as their situation and needs in  and . Aside in small-scale farming for pastoralist women and from the KRCS, the organizations that have paid targets pastoralist dropouts that have lost their theoretical attention to the issue are UNOCHA, the livestock and need to start a new livelihood (IRIN, IOM (through the focus on drop-outs) the Internal a). Finally, projects of livelihood transition are Displacement Monitoring Centre (IDMC) and the promoted, notably by Oxfam, to support pastoral- Institute for Security Studies (ISS).  ists in developing new livelihoods such as small businesses. Notwithstanding, these long-term Enhancing resilience and adaptive capacities approaches are not yet developed and coordinated The response of the international community on the same scale as food assistance. to pastoralist situation in northern areas is a continuum from short-term food assistance .. Policy recommendations (historically the most important), to livelihood support, livelihood diversification and finally The previous analyses call for several non-exhaus- livelihood transition (the three latter remaining tive policy recommendations in order to improve largely project-based). the situation of displaced pastoralists. Responses to drought emergencies have often Develop a better knowledge on the nature, been focused on short-term reactive food aid, be- scale and reasons of migratory flows as well as cause of the absence of long-term projects focus- of the specific needs of displaced pastoralists. ing on livelihood resilience and supposed lack of First, the Government of Kenya, international or- mechanisms to fund pre-emptive action. For ex- ganizations, humanitarian actors, researchers and ample massive humanitarian funding for food as- donor countries must work together to develop sistance during the  drought was only released adequate assessment methodologies that fully when UN officially declared state of famine in July grab the complexity and multi-causality of such , whereas early warnings of poor rainfall were flows. This is a necessary step towards develop- noted since May  (Fitzgibbon,  ). During ing a stronger conceptual understanding of these the last droughts of / and , the hu- displacements, of the possibility of prevention, manitarian communities also fell short of under- and of durable solutions. In Kenya, the Ministry standing the utter importance of focusing on live- of State for Special Programmes and the Ministry stock as key assets of the pastoral communities, for the development of Arid and Semi-Arid Lands even though things seem to evolve with the de- could work together to develop new assessment velopment of the Livestock Emergency Guidelines tools and methodologies. Second, this develop- and Standards (LEGS) (IRIN, c) and the un- ment of methodologies would adequately support derstanding of the importance of timely restock- the data collection and profiling system that will be ing programmes to supplement communities that set up by the new IDP policy of Kenya. However, due have lost their livestock with new ones. Other projects focus on livelihood support, ser- . http://www.iom.int/cms/en/sites/iom/home/news- vice provision (such as health and education), and-views/press-briefing-notes/pbn-/pbn-listing/ livestock assistance, information sharing, risk pastoralist-dispensary-opens-in.html (Accessed the th management and adaption to climate change. of April ) For example, the International Organization of . http://reliefweb.int/report/kenya/kenya-information- strengthens-pastoralists-resilience (Accessed the th of April ) ; http://reliefweb.int/report/kenya/cell- . UNOCHA has realized surveys and assessment of pas- phones-revolutionizing-kenya%E%%s-livestock- toralists displaced by floods and/or by inter-communal sector (Accessed the th of April ) conflicts in . The Internal Displacement Monitoring . http://kenya.usaid.gov/programs/education-and- Centre has realized one scoping study to cast light upon youth/ ; http://reliefweb.int/report/kenya/ the issue (see Sheekh et al., ), and will also publish kenya-educating-nomadic-pastoralist-children-low-cost in October  an extensive in-depth qualitative and (Accessed the th of April ) quantitative assessment of the scale and factors of pasto- . http://blogs.oxfam.org/en/blogs/---helping- ralists displacement in Northern Kenya. pastoralists-cope-changing-way-life-kenya (Accessed the . Interview with Nuur Sheekh, th April . th of April )

IDDRI STUDY 09/2013 113 The State of Environmental Migration 2013

assistance and capacity-building for implementa- cyclical drought affects more and more citizens, tion and profiling will be needed, especially at local triggering vicious circles of vulnerability in already level – otherwise the policy will remain mere words. marginalized areas. It follows that the impact of Strengthen the use of climatic early-warning drought on displacement is mostly related with systems at all levels, especially at local and pas- loss of livestock, erosion of livelihood systems toralist community’s level. The communication of and/or increase of the resource-based conflicts as information towards pastoralist communities also well as cattle-rustling – along with other political, needs to be done by taking into account their specific economic and historical factors. Drought-related cultural values, beliefs and perceptions to trigger ad- displacement can also be linked to urban migra- equate responses (IRIN, b). tion that can be seen either as diversification Develop proactive policies to improve liveli- strategies or as the result of the destitution of hood opportunities of pastoralists in northern pastoralists “drop-outs”, that can be compared to Kenya as part of the long-term answer to current internally displaced persons: any generalization is issues. The Government of Kenya should realize the difficult as there is little research and assessment importance of livestock value-chain, and adequately on these migration flows. Pastoralist communities develop markets and livestock value-chain support with eroding livelihood can also end up migrating system. This should be done by paying attention to to refugee camps. In , the main displace- the fact that for certain pastoralists communities, in- ment cause amongst pastoralist communities was tegration in the market economy and livestock sell- conflict. In most case, drought-related issues, i.e. ing for cash is not a natural reasoning (The Guardian, resource depletion and livestock, were closely ). Livestock assistance preparedness needs to be related to conflict drivers – along with many other improved to be able to anticipate droughts and food factors, composing a complex nexus of drought – security emergencies. livelihood erosion – conflict leading to death and Promote conflict prevention, resolution, as well displacement. as reconciliation The question of resource-based Attention and assistance given to environmental conflicts, cattle-rustling and underlying ethnical displacement of pastoralists in Kenya are scarce hostilities fuelled by political incitements and other and incomplete. At the national level the only IDPs economic incentives, is one of the most important is- to be considered and assisted so far by the govern- sues in Northern Kenya nowadays: it is at the same ment are the post-election violence IDPs, the other time the result of several decades of marginalization, being non-registered. Prevention of displacement, resource depletion and poverty, and a cause of liveli- in the form of sound policies for ASALs develop- hood erosion and displacement. While solutions for ment, disaster-risk reduction and security are still this problem are specific for every counties and com- lagging behind. The government has been promis- munities, the involvement of elders and religious and ing new IDP laws, drought-management and ASAL traditional leaders, support for community-based development policies, but implementation will resource-sharing agreements (Opiyo et al. ), be challenging. Security concerns remain highly development of early-warning systems could be pos- problematic and so far not really addressed. At the sible means of advancing conflict prevention and international level, few organizations have paid resolution. The issue of reconciliation is also key and attention to pastoralists’ displacement, and even will need to be addressed in order to achieve peaceful less to drought-related displacement and dropouts. coexistence and resource-sharing of communities. However, it seems that the approach towards pas- Consult and associate affected pastoralist com- toralism is slowly shifting from food assistance to munities Finally, a transversal recommendation is to long-term support, diversification and/or transi- consult and associate affected pastoralists communi- tion of livelihoods. Donor countries and humanities in the development of the previously mentioned tarian actors were instrumental in assisting PEV measures. Realizing adequate profiling and qualita- IDPs, and should now assist Kenya in implementa- tive assessments of their needs would be a first step tion of its new policy, with due capacity-building. for such integration. Finally, the review of secondary sources of information of drought-related and conflict-related factors in displacement flows points to differing per- CONCLUSION ceptions and narratives among stakeholders – such as the media, local communities, policymakers and Drought-related internal displacements of pasto- aid workers – with respects to the relative impor- ralists in northern Kenya are multi-causal and tance of drought-related and conflict-related fac- complex. The resource base and the coping strat- tors, which calls for further field research to ques- egies of the communities are slowly eroding as tion these perceptions.

114 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Sheekh N. et al., 2012, Kenya’s neglected IDPs: internal dis- BIBLIOGRAPHY placement and vulnerabilities of pastoralist communities in Northern Kenya, Institute for Security Studies and Internal A   Displacement Monitoring Centre

FEWSNET, 2012, Kenya Food Security Outlook June-De- cember 2012. N  FEWSNET, 2012, Potential impact of El Nino event in Business Today, 27th March 2012, http://www.businesstoday. 2012/2013. co.ke/news/2012/03/27/3000-displaced-clashes-turkana Fitzgibbon Catherine, 2012, Economics of resilience study – (Accessed the 30th of April 2013) Kenya Country Report IRIN, 2nd February 2009a, http://www.irinnews.org/Re- Humanitarian Policy Group, 2010, Pastoralism demograph- port/82683/KENYA-Drought-exacerbating-conflict-among- ics, settlement and service provision in the Horn and East pastoralists (Accessed the 30th of April 2013) Africa Transformation and opportunities, REGLAP Project. IRIN, 13th July 2009b, http://www.irinnews.org/Re- Humanitarian Policy Group, 2009, Pastoralism, policies and port/85252/KENYA-The-dangers-of-pastoralism (Accessed the practice in the Horn and East Africa: A review of current 30th of April 2013) trends IRIN, 18th December 2009c, http://www.irinnews.org/Re- port/87450/KENYA-What-drives-conflict-in-northern-Kenya IIED, 2009, Arid waste? Reassessing the value of dryland (Accessed the 30th of April 2013) pastoralism, Policy Brief IRIN, 11th February 2010a http://www.irinnews.org/Re- IDMC, 2012, Annual Review of Internal Displacement in port/88067/KENYA-Plugging-the-gaps-in-disaster-prepared- Kenya ness (Accessed the 30th of April 2013) IOM, 2010, Rapid baseline assessment with exclusive focus IRIN, 15th April 2010b, Extreme weather tests pastoralists per- on pastoralist drop-outs (Garissa municipality) ceptions http://www.irinnews.org/Report/88815/KENYA- ISS, November 2012, Violence in Kenya: a climate change Extreme-weather-tests-pastoralist-perceptions (Accessed the dimension, http://www.issafrica.org/publications/pub- 30th of April 2013) lication-archives/cpra-daily-briefing/violence-in-kenya- IRIN, 20th August 2010c, KENYA: Mohamed Barre Ali, “Our land a-climate-change-dimension (Accessed the 30th of April has become a warzone », http://www.irinnews.org/HOVRe- 2013) port.aspx?ReportId=90230 (Accessed the 30th of April 2013) Kaimba et al., Effects of cattle rustling and household char- IRIN, 19th January 2011a, http://www.irinnews.org/Re- acteristics on migration decisions and herd size amongst port/91666/KENYA-Act-now-to-mitigate-drought-effects-say- pastoralists in Baringo District, Kenya, Pastoralism: Re- agencies (Accessed the 30th of April 2013) search, Policy and Practice 2011, 1:18 IRIN, 16th June 2011b, http://www.irinnews.org/Re- Kenya Red Cross Society, 2012, presentation, http:// port/92997/KENYA-Severe-drought-high-food-prices-hit-pas- www.iom.int/jahia/webdav/shared/shared/mainsite/ toralists (Accessed the 30th of April 2013) microsites/IDM/workshops/moving-to-safety-complex-cri- th ses-2012/speeches-presentations/Session-3-Abbas-Gullet. IRIN, 4 July 2011c, http://www.irinnews.org/Report/93142/ pdf KENYA-SOMALIA-Drought-decimates-livestock-hits-incomes (Accessed the 30th of April 2013) Kenya Red Cross Society, June 2012, Complex emergency th appeal (floods, drought, civil unrest) IRIN, 12 July 2011d, http://www.irinnews.org/Re- port/93204/EASTERN-AFRICA-Too-soon-to-blame-climate- Metcalfe et al., 2011, Sanctuary in the city? Urban displace- change-for-drought (Accessed the 30th of April 2013) ment and vulnerability in Nairobi, Humanitarian Policy th Group Working Paper IRIN, 29 July 2011e, http://www.irinnews.org/Re- port/93363/KENYA-Drought-exacerbates-conflict-in-Turkana OHCHR, 2012, Rapport of the Special Rapporteur on Rights (Accessed the 30th of April 2013) of Internally Displaced Persons, Human Rights Council IRIN, 21st September 2011f, http://www.irinnews.org/Re- OCHA, 2012a, reports on inter-communal conflicts and port/93786/KENYA-Push-to-put-pastoralists-on-the-map (Ac- floods. cessed the 30th of April 2013) OCHA, 2012b, Inter-communal conflict by district as of No- IRIN, 25th November 2011g, A boomtown powder keg (Acccess vember 2012 the 30th of April 2013) Opiyo F. et al, Resource-based conflicts in drought-prone IRIN, 26 March 2012a, The changing face of pastoralism (Acc- Northwestern Kenya: The drivers and mitigation mecha- cess the 30th of April 2013) nisms, Wudpecker Journal of Agricultural Research Vol. IRIN, 2nd of April 2012b, Several thousand displaced after fresh 1(11), pp. 442 - 453, December 2012 clashes in Isiolo http://www.irinnews.org/Report/95219/ Overseas Development Institute, 2009, Pastoralists vulner- KENYA-Several-thousand-displaced-after-fresh-clashes-in-Isio- ability in the Horn of Africa lo (Accessed the 30th of April 2013) Schilling et al., 2012, Raiding pastoral livelihoods: motives IRIN, 22nd August 2012c, http://www.irinnews.org/Re- and effects of violent conflict in north-western Kenya Pasto- port/96153/KENYA-Dozens-killed-in-Tana-River-clashes (Ac- ralism: Research, Policy and Practice, 2:25 cessed the 30th of April 2013)

IDDRI STUDY 09/2013 115 The State of Environmental Migration 2013

IRIN, 29th August 2012d, Politics, pastureland and conflict Government of Kenya, 2009, Releasing our Full Potential – http://www.irinnews.org/Report/96201/KENYA-Politics- Draft Sessional Paper on National Policy for the Sustainable pastureland-and-conflict (Accessed the 30th of April 2013) Development of Northern Kenya and other Arid Lands IRIN, 17th September 2012e, Needs rise among Tana River Kenya Rural Development Programme, 11th February 2013, violence survivors, displaced http://www.irinnews.org/ Government launches ASAL policy, http://www.dmikenya. Report/96327/KENYA-Needs-rise-among-Tana-River-vio- or.ke/component/k2/item/13-government-launches-asal- lence-survivors-displaced (Accessed the 30th of April 2013) policy.html (Accessed the 7th of August 2013) IRIN, 31st October 2012f, Pastoralists, too, can be displaced, Kenya National Assembly, 2012a, Internally Displaced Per- http://www.irinnews.org/report/96669/KENYA-Pastoral- sons Act ists-too-can-be-displaced (Accessed the 30th of April 2013). Kenya National Assembly, 2012b, Report of the parliamen- IRIN, 18th February 2013a, Rising insecurity in Northern tary select committee on the resettlement of the internally displaced persons in Kenya Kenya http://www.irinnews.org/Report/97499/Rising- insecurity-in-northern-Kenya (Accessed the 30th of April United Nations, 11 February 1998, Guiding Principles on 2013) Internal Displacement, E/CN.4/1998/53/Add.2 IRIN, 24th April 2013b, http://www.irinnews.org/Re- port/97916/Floods-highlight-disaster-management-chal- I lenges-in-Kenya (Accessed the 30th of April 2013) Malika Peyraut, former consultant in Dadaab, 14th March The Guardian, 22nd October 2012, Turkana pastoralists en- 2013. couraged to innovate to build resilience to drought http:// www.guardian.co.uk/global-development/2012/oct/22/ Martina Caterina, Assistant country analyst for Africa re- turkana-pastoralists-innovate-resilience-drought gion at IDMC, 10th April 2013. Justin Ginetti, Advisor for natural disasters at IDMC, 5th April 2013. P    Mohamed Nuur Sheekh, Former consultant at IDMC, and Government of Kenya, 2004, Draft National Policy on Arid board member of IDPAC, 30th April 2013. and Semi-Arid Lands, http://www.theredddesk.org/sites/ default/files/kenya_draft_land_policy_asals_0.doc (Ac- cessed the 30th of April 2013)

116 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

THE FUKUSHIMA EVACUEES’ RETURN CLOÉ DEVLIN, LUCILE LECLAIR AND TIMOTHÉE DIDIER-BANDOU

INTRODUCTION promised compensation contributed to evacuees favouring one strategy over the other. The paper On  March , Japan’s Tokohu region was focuses on the nuclear evacuation as it remains devastated by a triple disaster that had a the most political and controversial consequence “profound and transformative effect on Japa- of the / catastrophe. nese society” (Hasegawa, ). The destruction caused by the earthquake, tsunami and nuclear meltdown exceeded the authorities’ expectations 1. THE POLICY FOR RETURN and had caused one of the worst nuclear disasters in history (ibid): in total, , people died After the / disaster, a “return to normalcy” was and hundreds of thousands were displaced. As of at the center of the Japanese society’s recovery mid-, a large part of those evacuated are still objectives and priorities. However, this urge for unable to return to their homes. normality resulted in the obliteration of the events This article interrogates why the Japanese gov- and their consequences rather than in develop- ernment’s decision to encourage return has not ment of new coping strategies, based on the very yielded the intended results and why migration acknowledgement of the events themselves. The has not been considered as an adaptation strat- policy of return can therefore be understood as an egy. It therefore explores the present situation of illustration of this desire to return to the previous the evacuees: while some have indeed returned status quo. to their communities as a result of the governmental support upon population’s return, the ma- .. The tsunami evacuees: jority have not. Fears over radiation levels and return and resettlement financial concerns are especially prevalent. In light of the alarming radiation levels detected in Out of the , people displaced in the after- Fukushima and the surrounding prefectures, re- math of the / disaster, about , people settlement could have been considered a vi- were taken off the list of evacuees in  as they able policy option for the Japanese government. returned to their homeland or resettled elsewhere. Despite an initial effort to provide assistance to Those who returned quickly were almost entirely those who wanted to resettle, political discourse tsunami evacuees who, despite facing many remained predominantly orientated towards an administrative and financial problems, received ambiguous policy of return. Most of the evacuees some support from the Japanese Government. lost their livelihoods and were in very difficult fi- Indeed, in , Japanese authorities pursued the nancial situations: return would have been a real- reconstruction efforts whose cost Reconstruction ist possibility for them as long as the State would Agency Minister Takumi Nemoto was reappraised have financed the reconstruction. Along with the by “ trillion to  trillion yen” (EUR  million to evacuees’ perception of displacement, their cur- EUR  million) (Nemoto, ). At the end of rent considerations on their future prospects as , the grand majority of infrastructures were well as the policy responses formulated by the rebuilt and rendered operational. Electricity, gas, Japanese government, the financial implications water, phone lines, roads, harbours and railways of return versus resettlement are also key in the were restored back to  per cent of their capacity article: we will try to assess to what extent the on average (Reconstruction Agency, ).

IDDRI STUDY 09/2013 117 The State of Environmental Migration 2013

However, such a percentage can be misleading: The analysis of how the reconstruction and if it indicates that facilities were largely restored, it resettlement policies have been carried out ena- also tends to hide the remaining difficulties faced bles to identify the underlying attitudes around by the population as private structures, which the / catastrophe. Earthquake and tsunami were left to the evacuees to rebuild, did not have being natural catastrophes, the government-led such high percentage of restoration. A reason for response did not acknowledge any form of re- this lies in the administrative and financial prob- sponsibility for their grave consequences. It para- lems that prevented evacuees from reconstructing doxically facilitated the government’s response, the structures necessary to their reestablishment. considered as their normal duty to protect its pop- The EUR ,-worth financial support and the ulation. Overcoming the tsunami evacuee crisis EUR , worth low-interest housing loans has been more about coping with financial and were not enough to give the evacuees sufficient administrative problems than about facing criti- resources to undertake the reconstruction projects cisms of risk prevention and management, much and compensate the financial losses they under- more highly politically sensitive in the context of went - often jobless and facing a ‘double loans’ the nuclear disaster. (situation of those paying loans both for their former house – destroyed by the catastrophe – and for .. Nuclear evacuees: their new one) (Hasegawa, ) From forced displacement The resettlement policy faced additional chal- to forced returns lenges. The need for new land raised both financial and spaces challenges as evacuees were con- The difficulty of returning fronted with the difficulty of finding available Unlike some of the tsunami evacuees, practically space for reconstruction because of the geographi- none of the nuclear evacuees have been able to cal characteristics of the Japan, and the sacred return. They represent , of the , character of ancestral lands. The little coordina- people still evacuated in  (Bøhmer, ), of tion between the state and the local authorities – whom , are found within the Fukushima in terms of the grant of financial support, for ex- Prefecture and , outside (Fukushima on the ample - added obstacles to the implementation of Globe, ). Most of the zone within a radius of  the resettlement schemes. kilometres from the nuclear plant is still completely Because it creates new alternatives for the evacu- forbidden because of radioactive contamination ees, the resettlement policy is not per se an instru- and nuclear evacuees still faced arrest or fines if ment for maintaining the pre-disaster status quo. they tried to go back within the forbidden areas However, the desire to return to ‘how things were (McCurry, ). As of , Fukushima locals before’ was omnipresent in the way the policy was have been allowed to visit their properties during conducted. Indeed, in order not only to rebuild the day but still cannot stay the night nor culti- homes but also to rebuild the previous communities vate their land (Bøhmer, ). This impossibility ties, the government laid down the condition that contrasts with the evacuees’ expectations as, one resettlement should be carried out collectively. year after the disaster, a majority still wishes to To be resettled, eac h family was supposed to return (Imai, ). Those who have returned are form a group of families (at least four) with whom to a great extent the elderly since “they do not they would collectively decide of the resettlement have that much time or energy left to rebuild” or details. Only on this condition would they receive resettle (Aizu-Wakamatsu, ). However, for the government assistance. If this policy was intended majority of the evacuees, returning remains diffi- to respond to the population’s desire to rebuild cult. Fear of the consequences of radioactivity was the community ties shattered by the / disaster omnipresent in the evacuees’ testimonies, which (Hasegawa, ), it created more problems than expressed high mistrust towards the Japanese it offered solutions, as it was often very difficult for government’s declarations on the gravity of situa- evacuees to get in touch with friends and former tion (Myles, ). In addition to reappraising the neighbors who were spread around the region dur- ‘acceptable’ rate from  to  millisieverts per year, ing the evacuation. Accordingly, the government’s the Government was accused by Greenpeace of will to encourage the reconstruction of traditional lowering the actual figures in order to conceal the community dynamics hindered the possibility for risks of radioactivity (ibid). The mistrust towards many evacuees to move forward. the government raised the concerns on the actual safety of going back home, an option that was . Interview with researcher Reiko Hasegawa, March , gradually dismissed: between June  and IDDRI Paris March , the proportion of evacuees wishing . Ibid. to return dropped from  per cent to  per cent

118 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

(Hasegawa, ). However, this evolution was compensated – a little more than half a per cent not taken into account by the government officials of all evacuees (Le Figaro, ). Many evacuees who deployed an active policy in order to push for blamed the long and complicated administrative return. procedure imposed by the operator of the power plant, Tepco, who provided a -page explicative The pressure upon return document to help the evacuees fill out the necessary forms (Ibid). After public and government Financial pressures pressure, Tepco simplified the procedure and In March , the Japanese Government hired , people to work on analyzing individu- launched a plan to organize return to the evacu- al files, promising to hold numerous information- ated areas according to their radioactivity level. sharing sessions and to open specific help desks Areas with less than mSv/year were consid- for the evacuees (Ibid). ered as “areas for which evacuation orders are However, a schism emerged between the Gov- ready to be lifted” (Hasegawa, ). In the areas ernment and Tepco on the nature and responsi- in which there was between mSv and mSv/ bility of compensation. Presented by the Gov- year, evacuees were expected to be able to return ernment as the main responsible party for the within two to three years (Ibid). Finally, areas nuclear catastrophe, Tepco refused to assume with more than mSv/year were considered to an expansive interpretation of the compensa- be inhabitable again after five years (Ibid). This tion policy as deplored lawyer Shigeo Takanashi: plan was rolled out unilaterally by the Recon- “Tepco believes that the responsibility for decon- struction Agency that, without consulting the tamination belongs to the State, not to the firm, population, made it the base for the continua- which is questionable” (Le Figaro, ). But even tion of financial support (Greenpeace, ): at the individual level, Tepco demonstrated some those who would not return to an area officially reluctance to hand out compensation, as it mostly presented as ‘safe’ would be considered as volun- handed out “temporary compensation” which tary departures and therefore would not be enti- victims of the meltdown were expected to repay tled to government assistance anymore. There- (Willacy, ). Yukiko Kameya, a - year old fore, from a situation of forced displacement, the nuclear evacuee, currently living in a “tiny public victims of Fukushima increasingly suffered from housing apartment in Tokyo’s (…) urban sprawl” forced return. An evacuee from Naraha expressed (Willacy, ), was initially given EUR , by the frustrating situation of many of the displaced Tepco, of which EUR , was deemed “tempo- people: “The government forced us to evacuate in rary” (Ibid), an amount that the evacuee could the first place. Now it’s trying to force us to return have to reimburse once a final settlement was de- without much information.” (Hasegawa, ). cided. Such uncertainty concerning financial re- Financial pressures for return can also be found courses represents a strong obstacle in the search within the compensation policy. Initially, each of for life alternatives. the , Fukushima evacuees was promised The importance of the pressure for evacuees to EUR , in compensation for material dam- return is relevant. In contrast to the earthquake ages, costs of evacuation and possible trauma (Le and tsunami, the responsibility for the nuclear Figaro, ). Compensation was regarded as an accident is partly attributable to the Japanese essential mechanism to help people financially State and its ‘zero risk’ guarantees regarding nu- recover from the disaster and to invest in the re- clear power. The issue around the nuclear evacu- construction of the devastated areas. However, ees’ situation is therefore much more political the compensation effort was quickly redirected than for tsunami evacuees. The “forced march” from public to private entities, namely Tepco the (Linton, ) towards status quo launched by main responsible party. A Greenpeace report, re- the Government through its different post-disas- leased in February , pointed out that many ter policies could be seen as a way to avoid the firms who helped design and build the Fuku- debate raised around Japan’s energy policy. Re- shima reactors were not held legally responsible ducing the visible consequences of the nuclear and were not required to pay compensation to the accident by favoring evacuees’ return may have victims (Greenpeace, ). As a result, at the eve been a way to reduce the controversies on nu- of , only one thousand individuals had been clear power. Nuclear safety being a condition for return, lowering radioactivity levels has become . “mSv” stands for millisiverts, the unit used to measure a huge challenge for the government to be able the impact of radiation on human beings. to carry out a return policy. Decontamination was . Interview with researcher Reiko Hasegawa, March , to be the key policy for return and, as such, many IDDRI Paris efforts were put into it.

IDDRI STUDY 09/2013 119 The State of Environmental Migration 2013

The decontamination policy (Ibid) was even more controversial as it was not In , decontamination was at the core of the followed by the expected decrease in radioactive Government’s policy for return and was, as such, levels. carried out doggedly. The objective was to reduce Although current decontaminating methods can radioactivity to below  mSv/year by  be very effective in the short term (in some case, ra- (Tabuchi, ) on a territory of about ,km dioactivity was divided by five after a decontami- (Linton, ). To achieve this goal, around EUR nation operation (Linton, )), some note global billion were mobilized on a period of three years ineffectiveness in the long term. With rainfalls, (Linton, ). The Government presented this radioactivity comes back sometimes to reach even immense decontamination effort as the only solu- higher levels than before decontamination. Green- tion Japan had in order to overcome the evacuees’ peace nuclear expert Heinz Smital made statements crisis. Making the comparison with the Cher- on the impossibility to decontaminate efficiently: nobyl nuclear accident, a representative of the “It’s not possible to decontaminate whole swathes Japan Atomic Energy Agency, Shinichi Nakayama, of land, mountains, rivers and riverbanks. You explained: “Conversely to the very vast USSR, Japan can’t get rid of that contamination” (Smital, ). cannot afford to abandon a part of its territory” Despite the inefficiency of the decontamination (Linton, ). Indeed, with a population density of policy – which has led the government to recog-  inhabitants/km, giving up even a small part of nize that there was little hope of recuperating areas the territory would entail huge extra demographic where radioactivity exceeds mSv/year (Linton, pressures on the other regions. Decontaminating ) - few other alternatives have been explored. was therefore considered as necessary. Nonethe- According to Kobe University professor Tomoya less, deep concerns have arisen about the current Yamauchi, the Japanese Government should have decontamination policy, particularly in relation to prioritized evacuation to frenzied decontamina- its environmental impact. In order to lower radio- tion: “For very affected cities, we can of course activity,  to  million cubic metres of soil are to be choose to destroy and rebuild everything, which is scraped off the ground’s top layer (The Guardian, very complicated. But we can also evacuate, at least ); trees and grass have been cut down (Linton, pregnant women and children” (Linton, ). ); dead leaves and forests’ natural compost The Government’s determination in decontami- have been removed (World Nuclear News, ); nating reveals the logic of its response to the / all of this being stocked in millions of plastic bags disaster, considering population displacement as or dumped into rivers (Tabuchi, ). Although it a temporary “crisis” response rather than a vi- is still too early to know exactly what will be the able long-term adaptation strategy. As such, the long-term environmental consequences of decon- post-catastrophe policies tended to privilege return, tamination, some impacts have already been noted contain population movement, and aim for status and are foreseeable(The Guardian, ). First, the quo. However, ‘going back to before’ is not only not removal of soil and trees increases the predisposi- always the best solution to overcome a crisis, it is tion to floods in the region that would be devas- often impossible. tated again in the case of a new tsunami. Second, in addition to this, the removal of dead leaves and natural compost will probably durably affect soil 2. THE LIVING CONDITIONS quality, which hinders the reestablishment of a OF THOSE DISPLACED ‘normal’ natural and agricultural activity. Third, the release of radioactivity in sea waters might Face to the impossibility to return home, Fuku- durably affect marine life in the Pacific (The shima evacuees were left with few alternatives: Guardian, ). The chemist Elizabeth Grossman stay (where they were evacuated or in their early reported that nuclear molecules were found contaminated homelands) or leave elsewhere in higher and higher stages of the local food chain ‘voluntarily’. The latter option was the least imple- (The Guardian, ). In light of all this, radiation mented and most of them chose to stay, despite the expert Tomova Yamauchi criticized the decon- difficulties it supposed. tamination policy as “absolutely irresponsible” (Tabuchi, ) as “Fukushima nuclear cleanup .. Sheltering in precarious could create its own environmental disaster” (The evacuation structures Guardian, ). This “environmental disaster” Out of the , persons initially displaced, about , are still evacuated in tempo- . As of . World Bank database. http://data.worldbank. org/indicator/EN.POP.DNST (consulted on August th rary housing developments in the prefecture or ) in subsidized apartments where rent is free but

120 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

utilities are not. Face to the impossibility to return been exposed to the extremely high radiation home because of the many obstacles cited earlier, levels since then. After being forced to leave by many feel “stuck between past and future” (Green- Japanese soldiers, he was rejected by his family peace, ). “I don’t have any sense of progress or because of their fear of him being contaminated. goal. I’m just living day to day” said -years-old When he was turned out a second time by the Itsuko Suzuki, an evacuee who has been living in a over-crowded refugee camp, he decided to go back temporary house for two years (Bird, ). As of into the forbidden zone, where he used to live , her situation is an illustration of the one lived (Arte Journal, ). As the only current inhabit- by the evacuee who are also face deep uncertain- ant of the prohibited zone, Naoto Matsumura sur- ties about their future. Recognizing the difficulties vives thanks to the food and water that admiring of immediate return, the Japanese Government compatriots send, as they see him like a ‘heroe’: planned to extend the period during which evac- “Naoto Matsumura is an original person and uees could remain in the temporary houses to an what he is doing is admirable” declared one of his additional two years (until ) (Ibid). former neighbours (Ibid). Although living in temporary houses represented a considerable qualitative decrease in many Maintaining Fukushima’s agriculture evacuees’ standard of living, it was generally Naoto Matsumura’s situation is not unique, and preferred to returning home. Chizu Matsumoto, a significant portion of the Japanese population a young mother living in temporary facilities be- is also living in high radioactivity levels without cause her house is so contaminated that the gov- receiving such publicity. In the regions where ernment did not plan to clean it up, highlights the radioactivity has spread, many farmers are still relief that seeing her boys is good health and far living and cultivating. A farmer from Nihonmatsu, from radioactivity is (Bird, ). Health preoccu- interviewed in the documentary Japan: dirty land pations have generally overridden the evacuees’ (Robin, ), explained: “Japanese people have urge to return. As result, many decided to stay always cultivated in this region and they have got away from their contaminated homes and keep liv- roots in here since many generations. This is why I ing ‘temporarily’ in secured areas. This option was cannot abandon my land”. however not feasible for everyone and many kept A -year-old female farmer from Nihonmatsu, on living in their contaminated homelands. added: “We need at least thirty years to decontaminate our land. If we leave, we will not be able to .. Staying in convey the story of this land to young people that contaminated homes build the future.” (ibid) A reason farmers continue to cultivate is that Fukushima’s “last man” (Pagnotta, ) they are backed by Japanese State, which hands After the / accident, the  km radius-zone out financial compensations provided Fukushima around the nuclear power was entirely evacuated farmers continue to produce. According to the and entering it is still strictly forbidden. For some, State, there is no immediate danger for culti- evacuating was more complicated. In particular, vating the land and no problem for eating food among the farmers, the evacuation has left a of the region of Fukushima. Kiyoshi Fujimoto, rep- long-term trauma because of the strong attach- resentative of the Agriculture Minister, said that ment they had to their land. Some even committed the radioactivity of the land of Nihonmatsu is “ab- suicide when forced to destroy their contaminated normally high, but considering the characteristics harvests (Robin, ). of the absorption of caesium by plants, cultivation For many farmers who were evacuated, it was is not a major problem” (Arte Journal, ). In morally impossible to abandon their animals. spite of the reassurance of the State, some farmers However, no evacuation program was organized organized in order to decontaminate the zone. For for animals, who were slaughtered or abandoned. instance, some started growing colza, famous for Facing this situation, one farmer, Naoto Matsumu- its capacity to absorb radioactivity. Some coop- ra, decided to stay in the forbidden km-radius eratives were created where customers can verify zone and took care of these animals. According themselves the level of radioactivity in the prod- to Antonio Pagnotta, Naoto Matsumura lives his ucts. Despite these reassurance actions, various struggle as a demonstration against Tepco and the scandals have alerted consumers, and it became Government (Pagnotta, ). Naoto Matsumura, very difficult to sell products from Fukushima. , was in his farm on the  March  and has Shinichi Ouchi, a farmer from Nihonmatsu, explained that since the accident, farmers have lost . Interview with researcher Reiko Hasegawa, March ,  per cent of their business. According to him, IDDRI Paris “All the product from Fukushima cannot be sold.

IDDRI STUDY 09/2013 121 The State of Environmental Migration 2013

The biggest difficulty is the damage done by have even been asked to give a medical certificate the rumours: even though the radiation level indicating their caesium levels (the caesium rate is very low, products are not sold” (Belogolova, indicates the presence of radioactivity) (Wallace, ). Japanese citizens mistrust information ). given by the State and they organize collectives to measure food safety. As Mitsuhiro Fukao, an New divisions among the society economics professor at Keio University in To- As researcher Reiko Hasegawa highlighted, fear kyo, says: “Many Japanese feel they’ve been lied of contamination created may divisions within to by their government” (Glionna, ). Thus the the Japanese society. Among families, separations only possibility for farmers in Fukushima region to between married couples became frequent as the continue to their livelihood is to receive compen- result of strong disagreements about the dangers sation from the State. A signif icant portion of radioactivity. The divorce rate has increased of their production is thrown away as it does not and this phenomenon led to a new expression in reach/contain the acceptable level of radioactiv- Japanese language: “genpatsu rikon”, meaning ity. Therefore, although radioactivity in the Fuku- “atomic divorce”. Noriko Kubota, a professor of shima region is “abnormally high” (Arte Journal, clinical psychology at Iwaki Meisei University, also ), financial assistance from the state tended to noticed that families are abnormally stressed and encourage the farmer’s immobility. that “suicides, alcoholism, gambling and domestic violence across the area” have increased (Haworth, Discriminations towards the ‘contaminated ). The “disaster honeymoon period” (Ibid) – i.e. people’ the great cooperation set up just after the accident Another factor discouraging the mobility of Fuku- – ended and people are facing long-term trauma. shima residents lies in the discriminations they Parents feel guilty to stay in Fukushima because face in the rest of Japanese society. Researcher they know that radioactivity would have a bad Reiko Hasegawa also explains that there is a strong effect discrimination against people that come from on their children’s health. In an article published Fukushima. They have no possibility to marry a in The Guardian (Haworth, ), an interview of person that comes from another region, for the a Japanese couple illustrated the dilemma faced assumption that they have bad health: a survey by Fukushima parents. Kenji Nomura, living with published in February  by the World Health his wife Aiko in Fukushima Prefecture with their Organization (Larramée de Tannenberg, ) said daughters, explained their situation: “We would that the risk of developing a thyroid cancer for girls rather move away from here altogether, but we from Fukushima Prefecture is  per cent higher can’t afford it” because “I would have to give up my than the risk for Japanese population on average. job” and “it is hard to find a new work in the cur- In September , the chairman of Ecosystem rent economy” (Haworth, ). Furthermore, it Conservation Society Japan Hobun Ikeya, who is socially hard to leave one’s job as the colleagues campaigned against nuclear power, said in a public consider it as “desertion” (Ibid). meeting, that “People from Fukushima should not Family pressures are indeed generally coupled marry because the deformity rate of their babies with social pressures. As Aiko Nomura explains, will skyrocket” (Haworth, ). families feel guilty when abandoning their re- Representing the fear of nuclear power, former gion when there is a strong need of rebuilding it Fukushima residents that live in other areas of Ja- (Ibid). They feel responsible for the disaster, since pan are discriminated in their everyday life. Chil- they voted for the nuclear plant to be created and dren at school, for instance, are seen as dangerous they feel that they are part of the nuclear disas- by other children (Wallace, ). According to The ter (Ibid). In addition, several campaigns came to Los Angeles Times (), “Apartment dwellers urge Fukushima residents not to leave the region have complained of cooking smells or noises that with slogans such as “Without the revitalization were unusual only in that they were produced by of Fukushima, there is no revitalization of Japan” former Fukushima residents” (Hays, ). We can and “Don’t give up Fukushima!” (Hasegawa, ). also mention the case of a transportation company This campaigning is relevant of a more global at- in Iwaki (located in Fukushima prefecture) that mosphere in which reluctance to live in the con- was asked by its clients “not to use trucks with Iwa- taminated areas was quickly described as anti-pat- ki license plates” (Hays, ). Discrimination for riotism or treason. work application is also frequent as some people . Interview with Reiko Hasegawa, March , IDDRI . Interview with Reiko Hasegawa, March , IDDRI, Paris. Paris . Ibid

122 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Therefore, because of the insufficiencies of the and the conditions imposed by the government to state responses to the / catastrophe (in terms favour return are not working as planned. of return, immobility or departure) displaced the Tepco’s reluctance to fully engage in the re- issue from the institutional level to the individual construction process contributed to complicat- level, thus creating social conflict and divisions. ing the evacuees’ short and long term adaptation strategies. A particular understanding of the term “responsibility” led to the State shouldering CONCLUSION the vast majority of reconstruction financing. The nuclear meltdowns were, in part, an industrial In , the nuclear evacuees’ situation can still disaster and a consequence of the “zero risk” be characterized by Reiko Hasegawa’s assess- myth that had bypassed basic safety concerns. ment: “evacuees continue to suffer from uncer- As a result, the company should have been held tainties about their immediate future and the accountable proportionally to their responsibil- affected communities are at risk of disintegra- ity in advocating the “zero risk” myth. As ob- tion. Reconstruction is still a distant process for served in the Greenpeace report, the Japanese the nuclear evacuees” (Hasegawa, ). The government is leading the reconstruction effort difficult and perhaps counter-productive decon- with little assistance from the nuclear industry. tamination efforts have not reassured the evac- Nevertheless, the policy of return is largely failing uees or the general population that the radi- in large part due to its rigidity and the govern- ated areas are safer than they were two years ment’s negligence in considering migration a vi- ago. Financial compensation remains insufficient able adaption strategy.

IDDRI STUDY 09/2013 123 The State of Environmental Migration 2013

McCann Christine. 2013. « Fukushima Nuclear Crisis Update BIBLIOGRAPHY for February 22nd to February 25th, 2013 ». In Greenpeace International (on line) http://www.greenpeace.org/inter- A national/en/news/Blogs/nuclear- reaction/fukushima-nuclear-crisis-update-for-july-19th/blog/46031/ (consulted Aizu-Wakamatsu. 2012. “Hopes of Home Fade Among Ja- on July 26th 2013) pan’s Displaced”. In The New York Times (on line) http:// McCurry Justin. 2011. « Fukushima evacuees face arrest if www.nytimes.com/2012/11/26/world/asia/hopes- of- they return home ». In The Guardian (on line) http://www. home-fade-among-japans-displaced.html?_r=0 (consulted guardian.co.uk/world/2011/apr/21/japan-declares- fuku- on April 26th 2013) shima-no-go-zone (consulted on March 25th 2013) Bird Winfried. 2013. “Two years on, Fukushima evacuees Myles Robert. 2013. “Fukushima two years on – the plight seek justice and a normal life”. In Japan Times (on line) of 300,000 displaced Japanese”. DigitalJournal (on line) http://www.japantimes.co.jp/life/2013/03/10/people/ http://digitaljournal.com/article/345181 (consulted on two-years-on-fukushima- evacuees-seek-justice-and-a-nor- June 29th 2013) mal-life/#.UfJ-oG0_8e- (consulted on July 24th 2013) Pagnotta Antonio. 2013. Le dernier homme de Fukushima. Bøhmer, Nils. 2013. « Analysis: Fukushima two years later Ed Don Quichotte. ». In Bellona (on line) http://www.bellona.org/articles/ar- ticles_2013/Nils_fukushima_comment (consulted on April Tabuchi, Hiroko. 2013. “In Japan, a Painful Slow Sweep”. 28th 2013) In The New York Times (online) http://www.nytimes. com/2013/01/08/business/japans-cleanup-after-a- nu- Belogolova Olga. 2013. “Rumours and Prejudice Plague clear-accident-is-denounced.html?pagewanted=all&_r=0 Fukushima Farmers Two Years Later”. In The Atlantic (consulted on April 27th 2013) (on line) http://www.theatlantic.com/international/ar- chive/2013/03/rumors-and- Wallace Rick. 2011. “Discrimination increases torment of Fukushima”. In The Australian (on line). http:// prejudice-plague-fukushima-farmers-two-years-lat- www.theaustralian.com.au/news/world/discrimina- er/274283/ (consulted on July 26th 2013) tion- increases-torment-of-fukushima/story-e6frg6so-%20 Digges, Charles. 2013. “Displaced Fukushima survivors 1226073275610 (consulted on July 25th 2013) come to terms with never going home again”. In Bellona (on line) http://www.bellona.org/articles/articles_2013/ Willacy Mark. 2013. “Residents paying back Fukushima fukushima_series2 (consulted on April, 27th 2013) compensation”. In ABC News (on line) http://www.abc. net.au/news/2013-02-19/companies-not-paying- com- Glionna John. 2012. “A Year After Tsunami, a Cloud of pensation-at-fukushima/4528150 (consulted on July 26th Distrust Hangs Over Japan”. In Los Angeles Times (on 2013) line) http://articles.latimes.com/2012/mar/11/world/ la- fg-japan-quake-trust-20120311 (consulted on July 26th 2013) W  Hasegawa, Reiko. 2013. “Disaster Evacuation from Japan’s Fukushima on the globe, http://fukushimaontheglobe.com 2011 Tsunami Disaster and the Fukushima Nuclear Acci- dent”, Study N° 05/13 April 2013. IDDRI. (consulted on April 28th 2013)Le Figaro. 2011. “Les évacués de Fukushima toujours pas dédommagés”, http://www.lefi- Haworth Abigail. 2013. “After Fukushima: families on the garo.fr/international/2011/11/21/01003-20111121ART- edge of meltdown”. In The Guardian (on line) http:// FIG00479-les-evacues-de-fukushima-toujours-pas-dedom- www.guardian.co.uk/environment/2013/feb/24/di- mages.php (consulted on August 19th 2013) vorce- after-fukushima-nuclear-disaster (consulted on July 25th 2013) Reconstruction Agency, http://www.reconstruction.go.jp/ english/ (consulted on April 27th, 2013). Hays Jeffrey. 2013. “Facts and Details: Fukushima Evacuees”. In Facts and Details (on line) http://factsanddetails.com/ja- The Guardian. 2012. “Fukushima nuclear cleanup could pan.php?itemid=2288 (consulted on July 26th 2013) create its own environmental disaster” http://www.guardian.co.uk/environment/2012/jan/09/fukushima-cleanup- Imai Akira. 2012. The Third Resident Survey. Fukushima environmental-disaster (consulted on April 29th 2013) University. Used in Hasegawa, Reiko. 2013. “Disaster Evac- uation from Japan’s 2011 Tsunami Disaster and the Fukush- World Nuclear News. 2011. « Clean-up in Fukushima ». ima Nuclear Accident”, Study N° 05/13 April 2013. IDDRI. http://www.world-nuclear- news.org (consulted on April 29th 2013) Larramée de Tannenberg Valéry. 2013. « Fukushima : l’OMS se veut rassurante ». In Journal de l’environnement (on line) http://www.journaldelenvironnement.net/arti- R cle/fukushima-l-oms-se-veut- rassurante,33384 (consulted on July 26th 2013) Greenpeace International. 2013. « Fukushima deux ans Linton, Marie. 2012. « A Fukushima, une décontamina- après: des victimes livrées à elles-mêmes » (on line) http:// tion à marche forcée ». In Sciences et Avenir. http:// www.greenpeace.org/france/PageFiles/300718/190213_ sciencesetavenir.nouvelobs.com/crise-nucleaire-au- Fukushima_2ans_in demnisations.pdf (consulted on July japon/20120307.OBS3181/a-fukushima-une-decontam- 24th 2013) ination-a-marche-forcee.html (consulted on March 30th 2013)

124 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

A-  I

Arte Journal. Broadcasted on March 11th 2013. Direct interview with Reiko Hasegawa, in March 2013, in Robin Marie-Monique, Théron Roland, Boulège Françoise. IDDRI, Paris 2012. “Japon: terres souillées”. In Arte television. Broad- Smital Heinz, interviewed by Hao, Gui. 2013. “Greenpeace: casted on December 1st 2012. Fukushima victims are victims once again”. http://www.dw.de (consulted on April S 29th, 2013)

Nemoto, Takumi. 2013. Speech for the Second Anniversary of the 3.11 Earthquake.

IDDRI STUDY 09/2013 125 The State of Environmental Migration 2013

IN FOCUS The New Towns of L’Aquila: allowed for public tenders and requests for proposals to be carried out considerably raised the whole cost of the project (40% higher A successful post-quake rehousing than the average market price of the square meter in the country), which amounted to a total of 809 million3. project or a wrong long-term Secondly, the amount of money poured into the project and the reconstruction policy? wide publicity made for and around it in the media contrasted sharply with the ruins of the city centre. This fact actively contributed from Lorenzo Pomarico fuel the feeling that C.A.S.E was mainly a strategy to divert attention to the fundamental issue of a durable reconstruction of the city On 6 April 2009 at 3.32 a.m. an earthquake of magnitude 5.8 struck centre (whose cost was estimated to be of EUR 10 billion) and raise the city of L’Aquila, located 7 km away from the epicentre, as well as electoral consensus desperately needed by the ruling government at 124 surrounding towns. The death toll was 309 people (Il Secolo XIX, the time and its Prime Minister Silvio Berlusconi. Furthermore, the 2013), while 1500 people were injured and about 50000 left home- local authorities were completely marginalized and excluded from less in the immediate aftermath of the quake (Repubblica, 2009; Il the decision-making phase, which led to a feeling of disempower- Corriere, 2009), Most of the people affected came from the city cen- ment of the inhabitants and the victims of the quake. tre. The management of the crisis underscores interesting aspects of Finally, it emerges from media and interviews carried out by the reconstruction and rehousing measures. author4 that the quality of life of the people resettled is actually very Interestingly, the government decided to defer the reconstruction poor. The 19 areas of the project are far from the city centre while of the historical centre of the city and gave instead priority to the poor transportation has been set up. No services or social structures C.A.S.E1 project was comprised of a total of 185 buildings designed have been included in the project, and the New Towns resemble dor- to host approximately C.A.S.E, project that was comprised 15,000 mitories rather than the lively neighbourhoods they were meant to be. people2. The project aimed at relocating the displaced population In light of these elements, we may ask ourselves whether the New living in the tent camps into newly built anti-seismic apartment Towns are just the incongruous outcome of an ineffective long-term blocks before the winter: the so-called New Towns. Four years after reconstruction policy or yet an incredibly rapid and effective project the quake, the city is still in ruin, the situation raises a wide range aimed at rehousing in few months thousands of displaced families of criticism with regard to the effectiveness of the measures imple- while permanently rendering urban areas vulnerable to earthquakes mented in the aftermath of the quake, their durability, cost and more secure. While the debate still rages in the country and among impact on the quality of life of the people relocated. the public opinion on the final assessment of C.A.S.E project. The Firstly, the flats were not meant to be provisional accommoda- most important question has not yet found an answer: will the dis- tions, but were instead conceived as high-quality flats that would placed be able to return? generate a profit on the market in the long term for all the actors involved in their construction. On the other hand, the anti-seismic platforms upon which the buildings were erected turned out to be very expensive devices. These facts coupled with the short time 3. Court of Audit of the EU, The response of the SFEU to the 2009 Abruzzo earthquake: pertinence and costs of the operations (in Italian), Special Report n. 24, 2012 4. Phone interview with Carla Tatone, Professor at L’Aquila University and 1. The acronym forms the word “houses” in Italian. New Town resident, conducted in Paris on April 28, 2013; Phone interview 2. Protezione Civile, Numeri del Progetto Case: http://www.protezionecivile. with Giovanna di Carlo, former resident of the city center, conducted in gov.it/cms/view.php?dir_pk=395&cms_pk=15963 ; Accessed on April Paris on April 7, 2013. See also Sabina Guzzanti, Draquila: l’Italia che 26th, 2013 trema, 2010 (documentary)

126 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

DAM-INDUCED MIGRATION IN THE MEKONG REGION MARIE LE TEXIER

INTRODUCTION that is essential, both at the national and regional levels, if policies are to be adapted to protect these All four countries comprising the Lower Mekong environmental migrants. Basin – Laos , Cambodia, Viet Nam and Thailand – agree that “the social and political economy of resettlement […] is one of the most serious stra- 1. THE ENVIRONMENTAL CRISIS: tegic trans-boundary issues facing hydropower HYDROPOWER SURGE IN THE development in the area” (ICEM, b: ). Dam building on the Mekong is indeed at the heart of MEKONG RIVER BASIN discussions at the Mekong River Commission (MRC) – the River Basin Organization institution- .. The Mekong River Basin alized by the  Mekong Agreement to foster cooperation over the shared river of the Mekong— The Mekong River, which is the tenth largest river level at present. This is due to the controversial in the world and the longest of Southeast Asia, decision of Laos in November  to give the is an international river originating in China and go-ahead to the construction of the Xayaburi spanning all of the countries of mainland South- dam, against the will of the three downstream east Asia except mainland Malaysia (MRC, ; riparian States. This dam, which would be the Cronin and Hamlin, ). It is commonly divided first to be built on the mainstream of the LMB, is between the Lower and the Upper Mekong Basins indeed expected to bring important transnational (LMB & UMB), the latter being comprised of socio-environmental consequences, among which the Chinese and the Burmese parts of the River. important migration flows. This division makes sense both from an hydro- Studies of environmental migration in the Me- logical point of view since the Upper Mekong kong region so far have focused on the linkages Basin comprises most of the river’s descent from between climate change and human migration over , metres above sea-level (Baker, ), patterns in the Mekong Delta (Dun, ; ICEM, and from an institutional point of view as neither ; IOM, UNDP and CTU, ). On the other Myanmar nor China have ever accepted to be part hand, hydropower developments in the region of the MRC regime. The present paper focuses on have been extensively studied but not with regards the LMB only, as this is the area of the Basin covered to environmental migration. Rather, the issues by the MRC mandate, where cooperation and poli- covered included security (Goh, ; Baker ), cies strategies may therefore be planned realisti- political-economy (Matthews, ), institutions cally. As far as hydrological cycle is concerned, the (Ratner ; Hirsch et al., ), hydro-politics most important characteristic to point out, aside (Keskinen et al., ; Schmeier, ), common- from the Southwest monsoon affecting the region, pool resources (Fox and Sneddon, ) and im- is the special functioning of the Tonle Sap Lake pact assessment (MRC studies principally). This in Cambodia, which fills up and empties in the article is therefore an attempt to bring together Mekong according to the seasons. This accounts these two bodies of literature so as to highlight the for most of the extraordinary biodiversity of the key features of dam-induced migration in the re- river, which is considered to be the second most gion. It is, above all, an attempt to raise awareness biologically diverse river following the Amazon. about this under-researched issue, an awareness It is, subsequently, essential to the food security

IDDRI STUDY 09/2013 127 The State of Environmental Migration 2013

of an estimated  million people, whose liveli- Theun  Purchasing Power Agreement (PPA) in hood is mostly based on fisheries and agriculture.  that foreign hydropower developers rushed More precisely, it has been estimated that more back to Laos. This marked the beginning of a new than  per cent of the economically-active popu- power surge in Laos: it has been estimated that lation have a job that is vulnerable to changes in there are currently  dams on the Mekong tribu- the river’s hydrology (MRC, ). For a long time, taries with nine under construction, another  at the Mekong remained a worldwide example of an the planning stage and  at the feasibility stage, untapped, free-flowing river, as no dams had been as shown on Figure  (Ministry of Energy and built on its mainstream (Fox and Sneddon, ). Mines ). Christopher G. Baker goes so far as to speak of a “hydropower gold rush” since regional .. Projected dams on the state-owned enterprises from China, Laos and Viet mainstream of the Mekong Nam are competing with foreign investors from France, Korea, Japan, and Norway for the hydro- It was not until  that China completed its power resources of the basin (Baker, : ). first dam on the Upper Mekong mainstream. The , megawatt Manwan dam was in fact the first .. The environmental crisis of a series of eight planned mainstream dams (cf. Figure ), known as the Lancang cascaнорǧȘde It is generally acknowledged that a dam has impor- (Goh, ). The total planned installed capacity tant consequences both upstream of the reservoir, is of about , megawatts and the cascade is by flooding an important area of the basin, and to capture an estimated  billion cubic metres of downstream, by modifying the hydro-morpholog- water from the river (about the same amount as ical regime of the river (both the hydrology and the Three Gorges Dam) (Baker, ). So far, four the transport of sediments). The latter type of of these mega-sized projects have already been effects brings about important modifications in constructed (Cronin and Hamlin, ). However, the ecosystems of the river, which in turn affect none of the  planned mainstream LMB dams fisheries and agricultural patterns of the basin. In has been completed yet. Xayaburi, as explained in the case of the Mekong, both of these activities are introduction, is the only project under construc- essential to the livelihoods of millions of people tion at present. The planned cascade includes in the basin. This is especially exacerbated in two eight more dams in Laos and two additional hydro- key zones: () the Tonle Sap Lake (cf. part .); () power stations in Cambodia. If these were to be and the Mekong Delta, which is also known as the built,  per cent of the river between Chiang Saen, “rice bowl of Southeast Asia” and is responsible Thailand and Kratie, Cambodia would become a for half the national rice production of Viet Nam reservoir (Baker, ). While these plans on the (Lisandre, ). LMB mainstream are quite recent, damming the Upstream consequences bring what will hereaf- tributaries of the river, however, is not new on the ter be called direct consequences in terms of mi- agendas of the MRC member countries. gration patterns in the region, i.e. the resettlement of entire villages out of the flooded areas. These .. Building dams on the population movements are planned, short term tributaries of the Mekong and local. Indeed, construction companies, before – power surge in Laos the actual construction phase of dams, plan and organize the resettlement of the flooded house- “Becoming the ‘battery of Southeast Asia’ through holds in areas close to their original villages. On exploiting its hydropower potential has been a the other hand, downstream consequences are longtime dream of the Government of Laos (GoL) expected to generate long term indirect migration and its backers” (Lawrence, : ). This is why flows, due to the above mentioned modifications the present paper focuses on Laos when it comes in terms of river-based activities. These are expect- to the study of tributary dams. Laos’s “hydro- ed to be especially important due to the cumula- boom” (Ibid.) started in the s with the signing tive effects of all tributary and mainstream dams. of numerous Memorandums of Understanding They are trans-boundary by nature and difficult to (MoUs), both with foreign hydropower developers predict precisely. That is why the adequate scale ( MoUs signed with Korean, Australian, Euro- of study for such effects is regional and long-term. pean and North American corporations) and with the Thai and Viet Namese governments (to export a total of  MW by  ) (Ibid.). The Asian financial crisis of  ended this race to hydropower, and it was not until the signing of the Nam

128 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

2. KEY DAM-INDUCED MIGRATION governmental support, due to repeated floods in FLOWS the lowlands (ICEM, a). The main issue resulting from these repeated relocations is the impov- erishment of these communities, who have been .. Direct consequences of dams identified as being among the poorest segments of on migration flows: resettlement society (ICEM, b). The ICEM team even stated patterns in the Mekong region that frequent forced hydropower-related displacement is “one of the most impoverishing acts that Direct dam-induced migration flows may be can occur to communities” (ICEM, a: ). divided up into two broad categories: () out- displacement induced by forced resettlement in the Laos’ planned and constructed tributary area flooded by a reservoir; and () in-migration dams due to direct job creation in the dam area. In the According to the Hydropower Database of the case of the LMB, it has been evaluated that direct MRC, the total number of people displaced by Laos’ job creation resulting from the  planned main- tributary dams is ,, with an average of  stream dams should generate an estimated USD persons displaced per dam. Again these figures . billion in wages (ICEM, : ). Moreover, are probably an underestimation of the reality, as it is interesting to note that “much of the labour there are currently no data available for  of the (especially for skilled and semi-skilled jobs) is Laotian tributary dams. The precise numbers of likely to be imported from surrounding countries people displaced per dam are presented in Figure other than the host countries (especially Viet Nam  below. and China)” (Ibid.). This articles focuses however Concerning the evaluation of the resettlement on population movements that followed the reset- process in Laos, International Rivers, an NGO that tlement process. Resettlement being a local issue, protects rivers and defends the rights of communi- specific to each dam, the same distinction as in the ties that depend on them, highlighted some ma- preceding part, between mainstream and tributary jor concerns in an extensive study conducted in dams, is made here. . Among these are the facts that lots of ethnic minorities are being resettled and some of them Planned mainstream dams have been resettled more than once already (see Regarding the planned mainstream dams of the double jeopardy issue explained above). There is LMB first of all, a strategic environmental assess- overall a lack of information disclosure regard- ment (SEA) of hydropower on the Mekong main- ing the process and a lack of opportunity for the stream was carried out in  by the Interna- displaced populations to voice their concerns; and tional Centre for Environmental Management the quality of resettlement programs is rather low (ICEM) for the MRC. The SEA team reviewed (Lawrence, ). The micro-level case study of existing secondary data sources in the region and the Theun Hinboun Expansion Project presented estimated to , the total number of people below helps to get a more in-depth understanding that would be displaced, should the cascade of  of the process. mainstream dams be constructed in the LMB. The This is a particularly interesting project as it was dams that are expected to account for most of the funded by a national agency (the Theun Hinboun resettlement are Luang Prabang; Stung Treng and Hydropower Company, which is a consortium of Sambor, as shown in Table  below. These figures companies in which Electricité du Laos owns % are probably an underestimate and are expected to of the share) and therefore shows the normal re- rise given more precise data from the developers settlement process in Laos. The project involved (ICEM, a). building a storage dam on the Nam Gnouang One of the most problematic issues at stake here River and doubling the capacity of the existing is the one of “double jeopardy” (ICEM, a: ), Theun-Hinboun power plant (Lawrence, : which designates the fact that some of the villages ). It required the resettlement of approximately displaced by mainstream projects have already , people (THPC, ). The resettlement pro- been forcibly displaced once or twice within the gramme has been evaluated as effective overall past decade. This is the case of some of the house- (ICEM, ), especially since the Theun Hinboun holds of Ban Houay Xong, in Nan district, one of Hydropower Company gave special attention to the potential districts affected by the Xayaburi livelihood development, public involvement, pub- dam, for instance: displaced from the uplands lic health and education, and ethnic minorities to the lowlands in the mid-s, these house- (THPC, ). Despite these efforts, some salient holds were forced to relocate themselves seven issues have been highlighted, among which: (i) the years after the first displacement, without any absence of commitment to provide land-for-land

IDDRI STUDY 09/2013 129 The State of Environmental Migration 2013

Map 1. Map of the tributary hydropower dams (circles) either built (red) or under construction (yellow); or licensed (green); or planned (black); and of the mainstream dams (triangles).

Source: MRC Hydropower Database.

130 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Figure 1. Number of people resettled by Laos tributary dams

9 000 8 000 7 000 6 000 5 000 4 000 3 000 2 000 1 000 0 Tha 1

Xeset 1 Xeset 2 Theun1

Xe Neua Xekatam Nam Pha Nam Ou 1 Nam Ou 7 Nam Leuk Nam Ou 4 Number of persons resettled Nam Ngay Nam Nam Ngieu Nam San 2 Dak E Mule Xe Lanong 1 Nam Mang 1 Nam Xe Kong 3up Nam Ngum 1 Nam Ngum 5 Xe Kaman 2B Xe Nam Noy 5 Nam Ngum 4B Nam Feuang 3 Theun-Hinboun Nam Hinboun 1 Xe Bang Hieng 2 Tributary dams in Laos

Note : Only the dams where figures were available have been included (13 additional tributary dams have been constructed). Source: author; with data of the MRC Hydro- power database. compensation (Lawrence, ; Matsumotos, in the future. The results are shown in Figure . It is ; Imhof, ); (ii) a lack of productivity of important to note that the diagram only accounts the replacement land for the resettled (Lawrence for the operating phase of the dams. The construc- ; Imhof ). This is particularly clear in a tion phase is bound to have some migratory effects study of the MK project in Phoumarkneng village, as well, but these are likely to disappear after the a relocation site where five villages merged into - years of construction. It is also essential to one due to the building of the dam (Bénézit et al., reaffirm the fact that both the tributary and the ). It demonstrated that the resettlement site mainstream dams, constructed and projected, was subjected to an overexploitation of non-tim- are taken into account in this part since indirect ber forest products resources and of fish resources, migration flows are most likely to occur due to the while relocated people were generally not self-suf- cumulative effects of dams. ficient in rice and vegetable production. The main From the diagram, it appears that the two key explanation accounting for these findings is that push factors of forced migration are the changes the land provided for compensation was of lower in agriculture and in water-related activities, es- quality than the former one. pecially in fisheries. Both are directly related to One of the fiercest criticisms against the THXP food security, the main triggering factor of dam- resettlement process is the lack of consideration induced displacements for Baker (: ), along for the downstream communities also impacted with water security issues: “Continued LMB hydro- (ICEM, ; FIVAS, ). Indeed, downstream development will result in strong push factors that impacts of dams may induce further forced dis- come from the problems associated with food and placement, often not taken into account by the water scarcity issues”. Water security also appears hydropower firms. These impacts are bound to be in our diagram, under the more general category even more important and to bring even more mi- of human health and security issues, as having gration flows, if considered cumulatively. a significant impact on the decision of affected households to leave. .. Indirect consequences As for pull factors, the main one is the increase of dams on migration in electricity provision, should the electricity pro- patterns: long-term duced by each dam be directly connected to the ecosystem modifications dam area. In the case of the main regional dam builder – namely Laos – though, most of the elec- Causal linkages between dams and long- tricity produced is sold to Viet Nam, Thailand term forced migration and China (Hribernik, ). This may therefore The MRC has undertaken extensive work on the change electricity generation into a push factor in future socio-economic impacts of dams on the favor of rural exodus. Still, it has to be noted that Mekong (ICEM, a,  b; MRC,  ). Our electricity exports to neighboring countries will contribution to this work is to reconstruct the are bound to bring economic development to Laos. causal links potentially leading to forced migration Depending on the management of such economic

IDDRI STUDY 09/2013 131 The State of Environmental Migration 2013

development, especially whether it is used or not development plans. How can this classification be towards dams’ areas, the situation might be revert- justified? ed again and electricity generation turned once The study of the four respective national eco- more into a pull factor. Assuming that push factors nomic contexts provides for a first baseline vul- are more important in amplitude than pull factors, nerability assessment of the region. Indeed, it has the rest of this section focuses on out-migration been proved that “resource users living in support- patterns rather than on those related to new mi- ive environments (with strong economic links, gration into the area. well-developed infrastructure and social services) A key question to answer to, beforehand, is would be less vulnerable to changes in water re- whether or not further migration is likely to occur sources availability than those living in less sup- in the region. The answer seems to be affirmative portive environments” (Hall and Bouapao, : since increasing proportions of the LMB popula- ). Laos and Cambodia both face multiple chal- tion have been shown to move across national and lenges in providing a supportive environment for international boundaries (ICEM a). their respective populations, contrary to Thailand and Viet Nam which both offer satisfying eco- Identification of dam-induced-migration nomic growth and access to basic services to their hotspots populations (Hall and Bouapao ; MRC ). Intuitively, the two zones which are bound to Therefore, the Tonle Sap and the Siphandone ar- be the most strongly impacted by the different eas are likely to be more vulnerable to changes in development projects in terms of migration are water resources than the Mekong Delta and the the Tonle Sap Lake and the Mekong Delta, as Chiang Rai area the primary fisheries and rice production areas The concepts of dependency and resilience, respectively (Baker, ). In order to make a more taken together as they are closely related, allow precise assessment of the situation, the assump- to go one step further in the classification. The tion is made that the most vulnerable sites are results of interviews carried out during the MRC the ones where displacements are the most likely vulnerability study (Hall and Bouapao, ) in to occur in the long run. Vulnerability, which is the four above mentioned areas show that nearly defined as “the degree to which a social group is  per cent of the Cambodian households, when likely to experience negative effects due to expo- asked what their second most important liveli- sure to changes in its environment” (MRC, : hood occupation was, answered that they had ), is a function of “exposure”, “sensitivity” and none, contrary to Laotians households,  per cent “resilience to change” (MRC, : ). Exposure of whom answered “fishing”. Full-time fishers are is essentially measured by the location of people rare in Laos, where the primary economic activity relatively to the environmental changes, while is agriculture (MRC, ). These findings show a resilience may be assessed via the existence of very high level of vulnerability of the Cambodian alternative livelihoods for affected households. households near the Tonle Sap Lake as they have The concept of sensitivity is closely related to the no livelihood alternative, should their primary oc- one of dependency on natural resources affected cupation – fishing – be threatened. The Tonle Sap by the changes (Ibid.). For the sake of simplicity, area is thus more vulnerable than the Siphandone dependency will be considered as a good proxy for zone. The dependency of households on fish, as sensitivity. shown in Table  below, further confirms this pat- The concept of exposure helps to identify, on top tern. It also enables to classify Thailand and Viet of the Tonle Sap Lake and of the Mekong Delta, Nam’s studied areas: with twice as many people two additional highly vulnerable sites (Hall and being highly dependent on fish and other aquatic Bouapao, ; MRC, ): (i) the Siphandone animals as compared to Chang Rai, the Mekong stretch of the Mekong (Southern Laos), in an Delta is more vulnerable to any change in the river. important fishing migration area prone to extensive dam building; and (ii) the Chiang Rai area, given that in the northern Thailand all the rivers are still connected to the Mekong and therefore likely to be highly impacted by any dam-induced changes. The MRC vulnerability assessment of the basin (Hall and Bouapao, ) then goes one step further and shows that the Tonle Sap area is the most vulnerable zone, followed by the Siphandone zone. The Thai side of the mainstream and the Me- kong Delta to a lesser extent are less vulnerable to

132 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Figure 2. Causal linkages between dams and human migration flows.

Note: If there is no “+” red sign indicated, the impacts are assumed to be are negative. Source: Author.

IDDRI STUDY 09/2013 133 The State of Environmental Migration 2013

Table 1. Assessment of household’s dependence on fish Vientiane, Luangnamtha and Bokeo provinces and other aquatic animals by study site (Kabmanivanh Phouxay, ). It appears, howev- Level of Cambodia Lao PDR Thailand Viet Nam er, that most of the people arriving to these places dependence study sites study study sites study sites come from northern provinces, while the majority sites of the migrants originating from southern prov- Low 21% 17% 45% 59% inces are more likely to move to Thailand than to Medium 42% 68% 45% 29% Vientiane (Ibid.). Thus the majority of migrants High 22% 14% 8% 9% coming from the Siphandone area would likely go Very high 15% 1% 2% 4% to Thailand receiving urban centers. Finally, when Source: Hall and Bouapao, 2010: 126 it comes to Cambodia, it has been assessed that migration flows are essentially rural-to-rural (Malto- Concerning the distributional impacts of dams, ni, ). Rural-urban migration, even though less or what may be called the social hotspots of migra- important, still exists with the main urban centers tion, most studies agree that the impacts “related of in-migration being Phnom Penh, Koh Kong, to the loss of agricultural land (…), off-shore fish- Battambang, Siem Reap, and Banteay Meachey eries and flood plain agriculture (…), are likely to (MRC, ). Moreover, internal migration is rath- fall more heavily on poorer groups” (ICEM, a er short-range (Maltoni, ), and when inter- : ). This finding is further confirmed by the fact province migration occurs, this is from “the more that the poorest segment of the population in the densely populated provinces in the south and west basin is much more dependent on fisheries for to the more sparsely populated provinces in the their livelihoods than higher income groups: in north-east” (MRC, : ). All in all, it seems Cambodia for instance, “ per cent of the poorest that the population from villages around the Tonle quintile catch fish (and seafood) as part of their Sap would migrate primarily towards other vil- livelihood compared to only  per cent of the rich- lages in close and sparsely populated provinces, est segment” (MRC, : ). before moving to the above-mentioned urban destinations. Main destinations for the migrants Concerning the amplitude of these potential Two main types of migration in the LMB have been migration flows, Baker () identifies three po- identified (ICEM, a; Baker, ): (i) rural to tential scenarios (cf. Figure ). The policy implica- urban migration; and (ii) migration to Thailand tions of these different scenarios are incremental from neighboring countries, especially from Laos and vary from the mere need of policies tackling and Cambodia. Consequently, a probable desti- unemployment and inflation increase in receiving nation for the farmers and fishermen from the areas in the first scenario to the necessary resort to Tonle Sap Lake and the Siphandone area might international aid in the third case (Ibid.). be the main migrants’ receiving cities in Thailand, namely Bangkok and to a lesser extent Chang Mai (Promburom and Sakdapolrak, ; Baker, ). 3. MAIN GAPS IN POLICY Nonetheless, international migration is not the RESPONSE BOTH AT THE NATIONAL migrants’ first choice, and “rural-urban migrants usually move within their own country before AND REGIONAL LEVELS moving across borders” (Baker, : ). Within Viet Nam, first of all, there has been an .. Lack of protection of increase in rural-urban migration since the mid- resettled people in the national s (Viet Khoa et al., ). Among the main legal and policy frameworks in-migration cities are: Ho Chi Minh City, Ha Noi, Hai Phong and Da Nang (Ibid.). Regarding the Preliminary assessment of the LMB Mekong Delta more specifically, “trends continue countries’ respective resettlement to show movement of migrants from the rural Me- frameworks kong Delta towards urban centres, industrial parks “Numerous gaps remain in land acquisition and and rapidly developing high-growth provinces in compensation policy and procedure compared to the Southeast and Central Highlands of Viet Nam” international best practice for all LMB countries” (Dun, : ). As far as Thailand is concerned, (ICEM b: ). Lack of political will and human rural-urban migration has increased since the capacities have been identified as other obstacles s, with Bangkok being the main destination to a socially equitable resettlement practice (Ibid.). (Promburom and Sakdapolrak, ) along with The main issues include: Chiang Mai. As for Laos, the main receiving area m Inadequacy of the overall Environmental and is Vientiane Capital followed by Borikhamxay, Social Impact Assessment (EIA) process (ICEM,

134 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Figure 3. Three potential migration scenarios

Moderate, ordered Heavy, unstable Severe, chaotic migration migration migration

If dams impacts If the dams result If massive are minor and in more serious livelihoods localized impacts such as destructions Regular, economic hunger and loss of (especilly in the migration will many livelihoods, Delta and the occur. Irregular migration Tonle Sap Lake) could increase Refugee like flows significantly. of people may arise.

a: ): in Cambodia for instance, a clear participation of local populations and livelihood “lack of political commitment for the role of improvement for resettled communities – rather EIAs in development decision-making” has been than mere monetary compensation – with special identified, along with a lack of capacity within attention being given to ethnic minorities (Nam the Ministry of Environment to critically review Theun  Project, ). In practice however, these the EIA findings (Suhardiman, de Silva and provisions are either not enforced or implemented Carew-Reid, : ); on an ad hoc basis (Lawrence ). m Limits of national capacities to implement the A thorough analysis of the legal and institutional respective national social and environmental frameworks of the different sectors related to hy- guidelines, especially on hydropower planning dropower – namely the water, land, environment, and monitoring (ICEM, a); energy and resettlement sectors – is necessary to m Insufficiency of budget funds allocated by hy- provide an explanation for such a gap between dropower developers to social and environmen- these legal and institutional frameworks and the tal safeguards (ICEM a); reality on the ground. Such an analysis has been m Barrier of land management to a sound resettle- completed by Suhardiman, de Silva and Carew- ment management: as most rural lands do not Reid (), the results of which are summarized have any legal title, rural communities are of- in Chart  below. In a nutshell, it appears that ten in a weak bargaining position vis-à-vis hy- the current legal framework is giving too much dropower developers’ interest in terms of land responsibility to project owners regarding the concession (Ibid.; ICEM, a: ); resettlement process, while the main regulatory m “Legal pluralism” (Suhardiman, de Silva and bodies over land and water resources lack author- Carew-Reid, : ), whereby hydropower ity to effectively monitor the hydropower projects. development is managed by different govern- This is explained by the underlying rationale of the ment agencies operating in more than one legal GoL who is encouraging private investment in the order (Ibid); country to boost its economy. m Lack of public participation in hydropower deci- As far as the hydropower decision making pro- sion-making: public participation, if any, mainly cess is concerned, it seems that the Ministry of occurs at the project level regarding mere pro- Environment and Mines (MEM) bases its decision cedural matters (Suhardiman, de Silva and to undertake the building of a dam essentially on Carew-Reid, ). the project’s economic feasibility. The EIA has to be approved as well but only by the Water Resourc- Focus on Laos’s hydropower legal and es and Environmental Administration (WREA), institutional framework which in practice has less bureaucratic power than An extensive legal and policy framework in Laos the MEM. Another issue is the one of the discrep- has been developed, especially over the last decade, ancy between the respective positions of national on environmental and water management. More- level and provincial level authorities regarding re- over, “in the lead-up to Nam Theun ’s approval, settlement and compensation issues. This is an im- the World Bank, the ADB and other donors worked portant obstacle to a sound resettlement practice with the GoL to establish social and environmental since each power company, after having obtained laws and policies to guide hydropower develop- the national authorities’ approval for their respec- ment in the country” (Lawrence, : ). These tive projects, has to negotiate again with provin- different documents include provisions to ensure cial authorities about the resettlement issue at the

IDDRI STUDY 09/2013 135 The State of Environmental Migration 2013

implementation phase of the project. The main strategy towards migration flows. In a sense, it explanation for such an inconsistency between is already doing it indirectly, through the miti- national and provincial level authorities is the ab- gation of other effects of dams: by limiting the sence of a “connected multi-level regulatory sys- impacts of dams on fish migration and sediment tem” (Suhardiman, de Silva and Carew-Reid, : transport for instance, the MRC is indirectly ) in the hydropower development sector. helping in diminishing human migration flows. Another key area where the MRC could poten- .. Lack of awareness tially act is by extending its data and knowledge and authority to deal on the subject. In reality, “the use of knowledge with indirect long-term to make better decisions is at the heart of the migratory consequences of MRC’s governance role (…). Producing scientific dams at the MRC level knowledge (…) is both its unique strength and an essential foundation for providing disinterested Contrary to the issue of resettlement in the flooded scientific advice to governments, international in- area, which is local, indirect migration flows stitutions, project developers and (…) other river arise from the cumulative effect of all tributary Basin stakeholders” (Hirsch et al., : -). and mainstream dams. As such, they are a trans- The MRC programme the most closely related to boundary issue by nature, which should therefore social sciences is the Basin Development Plan, be addressed at the MRC level. Currently, there is which aims at ensuring that the use of the Ba- no clear framework for dealing with such trans- sin’s natural resources contributes to sustainable boundary issues, as “they raise issues of account- economic development. Currently dealing with ability that none of the LMB countries have previ- issues such as employment and food security, it ously had to address” (ICEM, b: ). might easily encompass migration as well. The real impact of such actions remains to be A lack of awareness? seen, as the MRC has no supra-national authority The main obstacle to the establishment of such over its member States. It is and remains an inter- a framework is the lack of awareness of such State body with no supra-national authority (Cro- issues on the part of the MRC. Although the MRC nin and Hamlin, ), a governed organization framework also presents the issue of social conse- rather than a governing one (Hirsch et al., ). quences of dams in terms of direct and indirect Many observers – especially civil society groups impacts – referring respectively to losses of live- – consider this as one of the main MRC’s weak- lihood resources and to the remaining impacts nesses and expect it to “intervene in its own right such as new employment opportunities, cumu- to address concerns or resolve grievances” (Lee lative loss of livelihood related activities, health and Scurrah, : ). Furthermore, the MRC risks and water pollution – it does not encom- also lacks the power to influence the four LMB pass the issue of population movements (ICEM, countries’ respective policies: considering the a). sensitivity of water-related issues, there is little But rather than a lack of awareness as such, it if no chance that the MRC may influence in any would be more accurate to speak of a restricted way the respective national migration policies of mandate of the MRC. Indeed, employees of the its member States or the signing of a multilateral MRC Secretariat are well-aware of the migra- migration agreement between them. tion issue. The main reason why the latter is not Even though the MRC did have such a power, discussed officially is that it is not part of the of- the outcomes of such discussions would likely be ficial MRC agenda. The MRC is indeed focusing threatened by two external actors. Firstly, China, on water-related issues in a narrower sense. For a non-member of the MRC, is building eight main- the issue of migration to be included in the MRC stream dams on the upper reach of the Mekong mandate, there needs to be a consensus and a col- where the MRC has no authority whatsoever. As lective request from the four member States. This one of the main investors in the new waves of dams is unlikely to happen given the current regional in Laos, the country is also pushing for rapid hydro- setting, but even if it were the case, the question power development in the LMB with little concern remains of what the MRC could potentially do. over socio-environmental issues. The prospects of improvement of the situation are rather pessi- A lack of supra-national power mistic since China, as both an upstream riparian Given the current practices of the MRC when (able to reap all the benefits of increased hydro- faced with other social effects of dams in the power and export most of the social and environ- region (Hall and Bouapao, ), it is probable mental consequences) and the economic leader of that the institution would adopt a mitigation the region, is not likely to accept to become part

136 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

Table 2. Summary of the assessment of the existing legal and institutional hydropower framework in Lao PDR. Sector Main legal texts Main issues impeding sound resettlement practices Resettlement Legal and Decree on Compensation and Resettlement of This legal framework gives project owners the full responsibility Policy framework People Affected by Development Projects (STEA, to conduct the overall resettlement process. The main problem is 2006); therefore one of conflict of interests. This is especially problematic Regulations for Implementing Decree on given that there is currently no clear definition of any mechanism in Compensation and Resettlement of People Affected place to monitor the projects’ owners conduct. by Development Projects (2006) Energy Legal and Policy Electricity Law (1997, 2010); The rationale of the GoL is to encourage private sector investment framework Power System Development Plan (2004); in hydropower development to promote economic growth. This is National Policy on Sustainable Hydropower (2006); the main reason why the rationale to regulate or plan is most often Renewable Energy ; side-lined, and why the resettlement framework gives so much Development Strategy (2010) responsibility to the project’s owner.

Water management Water and Water Resources Law (1996); The analysis of this sector may provide an explanation for the Legal and Policy National Water Resources Profile (2008) difficulty to monitor hydropower projects. Indeed, the different framework Draft National Water Resources Policy (2010) texts define a regulatory body (WREA) in charge of regulating the different water related activities. Yet, it is also clear from these texts that this body lacks the authority to effectively monitor the different hydropower projects. In practice, there is a clear lack of inter-ministerial/cross-sectoral coordination, resulting in unclear operational boundaries and as a consequence to a “institutional and responsibility vacuum”. Land management Land Law (2003); The same is true for the land sector where the main regulatory Legal and Policy Decree on state land lease or concession (2009); body (NLMA) has no authority over the different ministries. Another framework Instruction as regards the implementation of important issue which may impede sound resettlement is the fact decree on state owned land approval for lease or that “the law does not oblige land registration for smallscale land concession (2010) use (article 45). Being not formally registered as official land users, they might receive lower compensation and lack any legal back up to negotiate about their lost opportunities in land use with the respective project staff” (p. 24). Environment protection Environment Protection Law (1999); The main issue outlined by the analysis of these texts is the one of Legal and Policy Decree on the agreement and endorsement of the public participation: mechanisms for public participation remain framework National Strategy on Environment years 2020 and limited, especially in the EIA process. Action Plan for the years 2006-2010 (2004); Decree on Environmental Impact Assessment (2010)

Source: Author, with data from Suhardiman, de Silva and CarewReid (2011).

of the cooperation regime (Menniken, ). The A need to transfer the resolution of dam- second external actor potentially at odds with any induced migration issues to other regional cooperation improvements in trans-boundary mi- bodies gration management is the Greater Mekong Sub- To avoid a potential political dead-end, a real- region (GMS) programme, initiated in  by the istic option may be to transfer the resolution of ADB. “The GMS Program focuses on economic dam-induced migration issues to other regional and infrastructure development, (…) and largely bodies, such as the ASEAN and the ADB’s GMS. ignore(s) the Mekong River and aquatic biodiver- The ASEAN, firstly, represents the right political sity aspects” (Keskinen et al,. : ). Part of its level of discussion. Regional migration policies portfolio of activities is the GMS Energy Roadmap, are already being discussed at this level, especially which explicitly aims at implementing an integrat- regarding the implementation of a common visa ed approach to deliver sustainable energy, in par- for ASEAN citizens similarly to the Schengen visa ticular through hydropower development (ADB in Europe. Moreover, some observers have advo- ). If the MRC and the GMS continue to com- cated for the incorporation of security-oriented pete rather than to cooperate (Keskinen, ), regional water issues to the higher political level this may directly go against any policy improve- of ASEAN: “A shifting of the politicized and ment on the part of the four member States. securitized part of the Mekong issue to genuine

IDDRI STUDY 09/2013 137 The State of Environmental Migration 2013

political bodies such as ASEAN, or to economically of South-East Asia”, the existing legal and institu- powerful institutions such as the ADB and GMS, in tional framework presents significant weaknesses order to combine water-related with other issues in protection resettled populations. As for indi- could help to outweigh China’s ‘Rambo’ position rect consequences, they are trans-boundary by in the water field” (Menniken, : ). The nature, and as such should first be addressed at GMS, on the other hand, might be the adequate the regional level. If the MRC is the most evident body for the funding of such regional migration arena for such an issue, it has been shown that the mechanisms. There is a lot to be done before organization might gain momentum in seeking getting there however, for the MRC has not taken alliances with the ADB’s GMS programme as well part in the “extraordinary momentum towards as with the ASEAN forum. More precisely, our regionalism that swirls around it” so far (Hirsch main recommendations would be: et al : ). m Governments should “treat resettlement not as If dam-induced displacements could be intro- a problem but as an opportunity. One key ele- duced on the ASEAN agenda, what should be the ment in this is to mix land-based and diversified position advocated by the MRC? Should it stick to strategies, not just to restore people’s incomes a mitigation point of view, as mentioned earlier? but to improve them” (World Bank, ). Or should it rather adopt, along with mitigation, m More resources ought to be invested in the an adaptation vision of migration? Given the un- implementation of sound resettlement policies, avoidable impacts of dams already constructed or especially by international funding agencies about to be, the latter option seems to be the more such as the World Bank or the ADB. appropriate one. It is essential to consider migra- m The MRC should adopt a radical change of par- tion as part of the solution to the problem. In the adigm and include adaptation options along same way as foreign migrants are considered as with mitigation strategies of dams’ social and cheap workforce for the construction of dams, environmental effects. dam-induced migrants may become resource- m The migration issue should be included to the ful agents in the receiving areas, if migration is MRC mandate. This would enable the exten- well-managed. Contrary to Baker’s vision on the sion of data and knowledge on water-related security implications of migration, migration may migration through the BDP programme of the well represent an improvement of human security MRC. rather than a threat to it. m Cooperation between the MRC and the other powerful regional bodies should be enhanced: the ASEAN may be an indirect way for the MRC 4. CONCLUSION AND POLICY to tackle the issue of dam-induced migration RECOMMENDATIONS while the ADB and its GMS could provide a non-negligible source of funding of such mi- Dam building on the Mekong brings both direct gratory governance mechanisms. migratory consequences, by resettling important Only a few authors have already acknowledged numbers of people located in the areas flooded by the true scale of the dam-induced migration is- the reservoir of each dam, and indirect migration sue in the Mekong region. Baker (: ) is one flows due to the cumulative effects of tributary of them: “impacts [of the Mekong dams] on food and mainstream dams on water-related activi- security and livelihoods threaten to create a situ- ties – especially fisheries and agriculture. The ation where migration […] leads to even greater former type of impacts is insufficiently addressed pressures on populations and states”. This article at the policy level of each LMB country. In Laos attempts to swirl around this momentum and, in particular, which aims to become the “battery hopefully, to further it.

138 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

MRC. 2010. Assessment of basin-wide development scenarios BIBLIOGRAPHY – Technical Note 12: Social Assessment. Vientiane: Mekong River Commission. A  Phouxay, Kabmanivanh. 2010. Patterns of Migration and Baker, Christopher G. 2012. Dams, Power and Security in the Socio-Economic Change in Lao PDR. Umea: UMEA Mekong: A Non-Traditional Security University. Assessment of Hydro-Development in the Mekong River Basin. Promburom, Panomsak and Sakdapolrak, Patrick. 2012. Singapore RSIS Centre for Non- “Where the Rain Falls” project. Case Study: Traditional Security (NTS) Studies for NTS-Asia. Thailand. Results from Thung Hua Chang District, Northern Cronin, Richard and Hamlin, Timothy. 2012. Mekong Turn- Thailand. Bonn: UNU-EHS. ing Point: Shared River for a Shared Future. Suhardiman, Diana; de Silva, Sanjiv; and Carew-Reid, Jer- Washington DC: Stimson Center. emy. 2011. Policy Review and Institutional Analysis Dun, Olivia. 2009. Linkages between flooding, migration and of the Hydropower Sector in Lao PDR, Cambodia, and Viet- resettlement - EACH-FOR project Vietnam Case nam. CPWF Mekong. Study Report. UNU-EHS. THPC. 2011. Theun Hinboun Expansion Project, Social and Dun, Olivia. 2011. Migration and Displacement Triggered Environmental Division: From Inception to 2010. by Floods in the Mekong Delta. International Vientiane: THPC. Migration 49 (S1): e201-e223. Viet Khoa, Nguyen; Cong Thao, Nguyen; and Van der Geest, Goh, Evelyn. 2004. China in the Mekong River Basin: The re- Kees. 2012. “Where the Rain Falls” project. Case gional security implications of resource Study: . Results from Dong Thap Province, Thap Muoi Dis- development on the Lancang Jiang. Singapore: Institute of trict. Bonn: UNU-EHS. Defence and Strategic Studies. Hall, David and Bouapao, Lilao. 2010. Social Impact Moni- toring and Vulnerability Assessment: Regional I  

Report. Vientiane: Mekong River Commission. Bénézit, Maud; Bouahom, Bounthanom; Croutaix, Orianne; Hirsch, Philipp and Morck Jensen, Kurt. 2006. National In- Souny, Phomdouangsy; and Santisouk, Lengsay. terests and Transboundary Water Governance in 2013. Presentation at the Environmental Mekong Symposium the Mekong. Copenhagen: Australian Mekong Resource 2013: Exploring the feasibility of potential livelihood opportu- Centre, The University of Sydney, in nities in a relocation area, in the Theun Hinboune Expansion collaboration with Danish International Development As- Project site, Lao PDR. Ho Chi Minh City: WISDOM & DLR. sistance. ICEM. 2010a. Strategic Environmental Assessment of Hydro- W power on the Mekong Mainstream – Final Report. Ministry of Energy and Mines, Hanoi : Mekong River Commission. ICEM. 2010b. Impacts assessment (Opportunities and Risks) http://www.poweringprogress.org/index. Volume II: Main report. Hanoi: Mekong River php?option=com_content&view=article&id=272&Item id=2 (consulted on 15 April 2013). Commission. Nam Theun 2 website, http://www.namtheun2.com/ (con- ICEM. 2011. Theun Hinboun Expansion Hydropower project – sulted on 20 April 2013). review of ESIA. Hanoi: ICEM. Keskinen, Marko; Mehtonen, Katri; and Varis, Olli. 2008. Transboundary cooperation vs. internal ambitions: I The role of China and Cambodia in the Mekong region. Tokyo: Interview with Anonymous, engineer specialized in hydro- United Nations University Press. electricity, conducted on 21 March 2013 by Lawrence, Shannon. 2008. Power Surge: The Impacts of Rap- id Dam Development in Laos. Berkeley: International Rivers. .

Maltoni, Bruno. 2007. Migration in Cambodia: Internal vs. Interview with Olivia Dun, PhD Candidate at The University External Flows. Fuzhou: ARPMN of Sydney, conducted on 17 March 2013 by Menniken, Timo. 2007. China’s Performance in Internation- al Resource Politics: Lessons from the Mekong. . Contemporary Southeast Asia 29 (1): 97-120. MRC. 2005. Overview of the hydrology of the Mekong Basin. Interview with Anonymous, a Mekong policy specialist, con- Vientiane: Mekong River Commission. ducted on 19 April 2013 by Skype.

IDDRI STUDY 09/2013 139

The State of Environmental Migration 2013

EDITORS

F G P B D I François Ge- Pauline Brücker Dina Ionesco is menne teaches is a Migration Migration Policy the course on Research Fellow Officer for the “ E n v i r o n m e n t at the Institute International Or- and Migration” at for Sustainable ganization for Sciences Po’s Par- Development and Migration (IOM) is School of International Affairs International Relations (IDDRI). in Geneva (Switzerland). She (PSIA). A specialist of environ- She is also a PhD student at the is responsible for the «Migra- mental changes and migration at Centre for International Relations tion, Environment and Climate the universities of Liège (ULg) and Study (CERI/Sciences Po) and the Change» portfolio. Dina joined Versailles (UVSQ), he has con- CEDEJ (Cairo/Khartoum). Gradu- IOM in  to work on migration ducted field studies in the United ated from a Research Master in and development, particularly on States (New Orleans, Louisiana), Political Sociology, and a Master policies towards diaspora. As a re- Tuvalu, China, Central Asia, Mau- in International Public Law, both searcher she has developed many ritius, D.R. Congo and Japan, in at Sciences Po Paris, she works on projects and trainings as well as the aftermath of the Fukushima the sociology of migration in the coordinated and written several disaster. He has been involved in Middle East and the Horn of Afri- publications by the IOM. Dina was different international research ca and the processes of migrants’ also Administrator for the Organi- projects, and has consulted for categorization. Her researches fo- zation for Economic Cooperation the Asian Development Bank, the cuses also on a critical approach and Development (OECD) from International Organization for to environmental migration as  to , on issues of local Migration and the Cartier Foun- a socio-political object, through development, entrepreneurship, dation for Contemporary Art. He examination of normative protec- employment and gender. She also has published books and articles tion gaps of environmentally dis- worked for an organization for the on environmental migration, ad- placed persons and the protection defense of human rights, as teach- aptation and vulnerability. He is of human rights in times of natu- ing assistant and as an independ- also the editor of the ‘Sustainable ral disasters. She also pursued ent consultant. Dina was trained Development’ series at the Presses field researches on environmental in political science, international de Sciences Po, a leading French migration related policy and legal relations and economics at the In- academic publisher. A political challenges in Bangladesh. She is stitut d’études politiques in Paris scientist by training, he was first a member of the Management (France), at Sussex University and educated at the Univesity of Liège Committee of the European Coop- the London School of Economics (Belgium) and holds a joint PhD eration for Scientific and Techni- (United Kingdom). from Sciences Po and the Univer- cal research (COST) on “Climate sity of Liège, as well as a Master of Change and Migration”. Research in Political Science from M K the London School of Economics Mark Koski is in and a Master of Sciences in Envi- his final year as a ronmental Studies from the Uni- Master’s of Public versity of Louvain. Policy candidate at the Gerald R. Ford School of Public Policy at the University of Michigan. He is focusing on international public policy and human rights. He interned with the International Organiza- tion for Migration in the Migration, Environment and Climate Change team in  where he helped in the editing process of the State of Envi- ronmental Migration. Before coming to the Ford School, Mark graduated from the University of Florida with Bachelor of Arts in History.

IDDRI STUDY 09/2013 141 The State of Environmental Migration 2013

CONTRIBUTORS

P B J C F D Pierre Bonneau Juliette Cody is a Fannie Delavelle is holds a Bachelor’s Franco-Irish stu- a political scientist, degree of Litera- dent currently en- currently complet- ture from the Uni- rolled in a Master’s ing a joint Master’s versity of Poitiers, a programme in En- in Political Econo- Master’s Degree in vironmental Policy my at the London Environmental Policy from the Paris at the Paris School of International School of Economics and in Interna- School of Internal Affairs (PSIA/Sci- Affairs (PSIA/Sciences Po). She tional Relations at Sciences Po Paris. ences Po), and is currently studying completed her Bachelor’s degree at She has worked in the private and in the Master’s Degree Dynamiques Sciences Po and studied for a year at public sectors in the United States, Comparées des Développements at the University of Toronto, Canada. conducting research on climate AgroParisTech where he specializes Her main interests lay in sustainable change and migration at Columbia in rural and agricultural develop- development policies as well as cli- University, in Latin America,within ment. He is interested in issues re- mate change mitigation and adapta- the International Relations Depart- lated to implementation of sustain- tion strategies. She is currently in- ment of the Government of Buenos able development – especially in the terning with the UN Office for REDD Aires, and in Europe, analyzing eco- field of food and agriculture — and + Coordination in Indonesia and nomic shocks and food security in engaged in youth civil society move- focusing on policies linked to the the Middle East for the UN World ments, at different levels. He was a green economy and green growth. Food Programme. Her career objec- French Youth Delegate at the United She is also involved in CliMates, a tive is to contribute through policy- Nations Rio+ Conference and is a global student network researching making to the cooperation between member of the French Student Net- and advocating for innovative solu- the USA and the EU, particularly work for Sustainable Development. tions to climate change. She hopes on security and global governance Through his internships at the GRET to enhance her experience in the issues. and at the Department of Agricul- field of climate change related poli- ture of Western Cape, he has come cies through further professional ex- C D to understand the importance of the periences in France or abroad. Cloé Devlin is a territorial levels of action. French-Australian S C Master’s student T B Sabira Coelho in Paris School of Tania Boulot is a recently gradu- International Af- French-German ated from the Paris fairs (PSIA/Sci- student holding a School of Inter- ences Po). After several months of Bachelor’s degree national Affairs volunteering in the Immigration De- from the European (PSIA/Sciences partment of the Pro-Human Rights campus of Sciences Po) with a Master’s degree in Inter- Association of Andalucía (APDHA) Po in Nancy, and is currently at- national Development. As a Scienc- in Granada, Spain, she joined the tending her first year of the Human es Po student, Sabira was engaged Human Rights and Humanitarian Rights and Humanitarian Actions in raising awareness amongst the Action Master’s program and spe- Master’s program at the Paris School youth on environmental issues and cialized in Migration Studies. Her of International Affairs (PSIA/Sci- the Rio+ conference through a concern about the social and struc- ences Po), where she is specializing project called “MyCity+”. She is tural marginalization of migrants in in Migration Studies as well as Afri- also extremely interested in the hu- modern societies led her to orien- can Studies. She has completed an man aspect of climate change, which tate her career towards encourag- internship at Human Rights Watch was triggered off by her internship ing the integration of migrants and in Berlin, Germany. So far her focus at IOM’s Asia-Pacific headquarters refugees. She is currently working as has been on human rights, thus her in Bangkok, Thailand. Following her an intern for an NGO in the Canary fields of interests also include envi- Master’s degree, Sabira will return Islands providing accommodation ronmental equality and social jus- to IOM in Bangkok as a consultant, as well as legal, social and profes- tice. She is working towards a better carrying out research on the climate sional guidance to first-generation understanding and further experi- change and migration nexus in Asia. migrants. Cloé intends to keep on ence in these issues. focusing on the southern border of Europe and work on the relation

142 STUDY 09/2013 IDDRI The State of Environmental Migration 2013

between peripheries - be it its geo- S K C N graphical, social or legal aspects – Sora Kim Sora Kim After entering Sci- and exclusion. is from South Ko- ences Po Paris in rea. She is current- , Camille trav- T D-B ly in a Master’s in elled to Burkina Development Prac- Faso and Mali, L LC tice at Science Po where she worked Lucile integrated (PSIA/Sciences Po). Before entering in the fields of microfinance and Sciences Po Paris her Master’s program, she earned a women’s rights. She then studied in . She ended Bachelor’s degree in French Litera- Environment, Sustainable Develop- up her bachelor ture at Hankuk University of Foreign ment and Risks during her Master’s with an exchange Studies. In , Sora also worked as programme at the Paris School of year in Ecuador, an intern for in Korean NGO help- International Affairs (PSIA/Scienc- where she learnt a lot from Latin ing African asylum seekers obtain es Po). She also started a Research American culture. In parallel to her refugee status, shelter and consulta- Master’s in Medieval African History university exchange, she made an tion. She followed this experience at the Sorbonne University in Paris internship in an agricultural organ- up with an internship at Amnesty and created her own small trans- isation (Federación de Centros Agrí- International Korea in . Begin- lation and writing business. She colas del Litoral del Ecuador) where ning August , she began an in- worked for the Water and Sanitation she worked with farmers on sustain- ternship working for the UNHCR Programme in President Jacques able agriculture. Back to France, she office in Burundi. She would like to Chirac’s Foundation in France, vol- chose the Master “International Af- work to improve the conditions of unteered for the Earth Charter Inter- fairs- Environmental Politics” and refugees and immigrants after com- national Secretariat in Costa Rica, interned in the pedagogic farm of pleting her Master’s in , with a and currently is the Communication the French association Protection special focus on forced environmen- and Partnership Deputy Director in of farm animals and in the National tal migrants. the international student-led think- Federation of Organic Agriculture. tank CliMates. She has started an in- A M ternship at Ifri (French Institute for M L T Anne Meldau International Relations) in Paris, in Marie Le Texier graduated from the Sub-Saharan Africa Programme. is a young engi- Sciences Po Paris neer in water sci- with a Master’s de- N P ` ences, recently gree in “Environ- A degree-holder graduated with mental Policy” in from Paris School a master degree June . She also holds a bachelor of International in environmental policies from the in Political Science from Mannheim Affairs-Sciences Paris School of International Af- University. Anne is particularly inter- Po and the College fairs (PSIA, Sciences Po). Passionate ested in climate change mitigation of St. Benedict’s in about international water-manage- and adaptation with a special focus Minnesota, USA, Nakia is special- ment issues, she participated in the on the challenges faced by Small Is- ized in sustainable development and th World Water Forum in Marseille land Developing States (SIDS). She migration. She has worked and vol- in  and has just been selected as has inter alia worked for the German unteered for various organizations a Junior Rapporteur for the World carbon-offsetting start up ecogood, in Japan, China, India and Cam- Water Week , held in Stock- the Permanent Mission of Tuvalu to bodia. In her native Bahamas, she holm on September -. She joined the United Nations in New York and worked as a Development Associate the water geopolitics department of the UNESCO Intersectoral Platform at the College of The Bahamas. An the think tank Youth Diplomacy in for SIDS in Paris. Currently, Anne is avid environmentalist, she traveled August , while she was complet- a Carlo Schmid Programme scholar- by bicycle from Beijing to Paris from ing a six-month internship with the ship holder at the UNESCO Regional - (website: www.fueled- United Nations in Laos. Since then, Office for the Pacific States in Apia, byrice.org). In , she interned at Marie has been pursuing her own Samoa, where she works on various the Permanent Interstate Commit- personal research on the Mekong environmental projects. tee for Drought Control in the Sahel region. Research is in fact an activity (CILSS) in Burkina Faso, where she she came to appreciate during her conducted research on the agricul- two six-month internships with the tural practices of migrants escaping International Research Institute for droughts. Her academic interests Development in Peru and Brazil. include sustainable agriculture,

IDDRI STUDY 09/2013 143 The State of Environmental Migration 2013

slow-onset climate migration, and N S G Z female migration, particularly in Af- Nicole is a Ger- Geraldine Zam- rica and the Caribbean. man native who brana is from Bo- was enrolled at livia. After a Fine L P Sciences Po Paris Arts school in Bo- Lorenzo Pomarico for a yearlong ex- livia, she did her Lorenzo Pomarico change during her B.A. at Sciences Po was born and Master’s programme at Heidelberg Poitiers campus and specialized in raised in Italy. He University. While studying in the Latin America She then joined the attended the Euro- USA at the University of Georgia in Master’s in Human Rights and Hu- pean School of Va- /, the earthquake in Haiti manitarian Action programme and rese, where he successfully obtained and the floods in Pakistan led her to specialized in Project Management a European Baccalaureate and dis- become more involved with natural and Migration (PSIA/Sciences Po). covered an interest for international disaster management and climate So far, her work experiences in the affairs. He obtained a Bachelor in change issues. She wrote her Bach- field of development include Public Political Sciences and Eastern Eu- elor’s thesis on relief aid efforts ex- Partnerships Assistant in a micro- ropean Studies at Sciences Po Paris, amining a variety of factors that cor- finance NGO in Paris; Public Rela- at the European Campus of Dijon. responded positively to the amount tions Manager in a cultural associa- He is currently in his second year of donations given to countries af- tion in Johannesburg; and Project of the Human Rights and Humani- fected by natural catastrophes. In Manager at a shelter for children in tarian Action Master Degree at the Heidelberg, she has also been part New Delhi. She aims to keep devel- Paris School of International Affairs of the interdisciplinary project, “The oping her skills in the Project Man- (PSIA/Sciences Po) and interning at Global Governance of Climate Engi- agement field. the Protection Department of UN- neering”, and will critically investi- HCR at the Impfondo Field Office, gate this topic in her Master’s thesis. Republic of Congo. After graduation, expected in June , he is U J W looking forward to continue work- Ulrike Julia Wendt has recently coming with refugees and specialize in pleted a Master’s in their protection at an international Development Prac- level, as well as getting some work tice at Sciences Po experience in the NGOs sector. in Paris and prior to that, a Bache- M R lor’s in Internation- Born in Bordeaux, al Development at La Trobe Univer- Maëlle Robert is a sity in Australia with an academic Master’s student at excellence recognition. Her research Sciences Po Paris. interest focuses on development, Earlier, she obtained humanitarian and children’s issues a literary baccalau- in Africa, Asia and South America. reate and studied social sciences in Julia has gained work experiences Nancy, at the French-German campus in various national and interna- of Sciences Po, where she contributed tional organizations in developed to a weekly radio broadcast. Interested and developing countries. Her work in journalism, she spent then one year experience include UNICEF’s child in Hamburg, working for the daily protection unit in Burkina Faso, and newspaper, Hamburger Abendblatt, Oxfam’s Youth Program in monitor- and the radio station or programme? ing and evaluation, the Oxfam Inter- TIDE. She is currently in the Inter- national Youth Partnerships (OIYP), national Development Master’s pro- in communication, and the refugee gramme at the Paris School of Inter- organization, RISE, in advocacy. Pri- national Affairs (PSIA/Sciences Po) or to that, she worked on children’s where she studies and researches how rights in cooperation with a local to reconcile growth, wellbeing, justice NGO in Nicaragua. and sustainability. In the fall semester of , she will intern with a newspaper in Ouagadougou, Burkina Faso.

144 STUDY 09/2013 IDDRI

The State of Environmental Migration 2013 A REVIEW OF 2012

Edited by François Gemenne Pauline Brücker Dina Ionesco

his volume is the third of an annual series, which aims to provide the reader with regularly-updated assessments on the changing nature and dynamics of environmental migration throughout the world. The idea for Tit stemmed from the course “Environment and Migra- tion”, taught at the Paris School of International Affairs (PSIA) of Sciences Po. The course, which is thought to be the first of its kind in the world, examines the complex relationship between environmental change and migration flows. The best of these papers have been selected and edited, and are presented in this volume. Most of them constitute the first detailed analyses of the migration flows that were induced by some of the most dramatic events of 2012, paving the way for future scholarly works.

With the support of