D R A F T C O N F E R E N C E E D I T I O N PE 69640 OVERTY NVIRONMENT N EXUS Sustainable approaches to poverty reduction in Cambodia, Lao PDR and Vietnam THE WORLD BANK D R A F T C O N F E R E N C E E D I T I O N PE OVERTY NVIRONMENT N EXUS Sustainable approaches to poverty reduction in Cambodia, Lao PDR and Vietnam June 2006 THE WORLD BANK This publication is available online at <>. Front cover photos: Uwe Deichmann, World Bank Photo Library Cover design: Circle Graphics, Jostein Nygard Environment and Social Development Department East Asia and Paci�c Region The World Bank Washington, D.C. June, 2006 This volume is a product of an expert team from Cambodia, Lao PDR and Vietnam and the staff of the World Bank. The �ndings, interpretations, and conclusions expressed in this paper do not nec- essarily reflect the views of the Executive Directors of The World Bank or the governments they rep- resent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissem- ination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Of�ce of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522- 2422, e-mail pubrights@worldbank.org. Contents ACKNOWLEDGMENTS v ABBREVIATIONS AND ACRONYMS ix FOREWORD xiii PART I: STUDY CONTEXT AND METHODOLOGY 1 Introduction 3 1 Study Framework and Methodology 21 PART II: COUNTRY STUDY FINDINGS 41 2 Poverty-Environment Nexus in Cambodia 43 3 Poverty-Environment Nexus in Lao PDR 75 4 Poverty-Environment Nexus in Vietnam 123 PART III: SUB-REGIONAL PERSPECTIVES 175 5 Sub-regional PEN Findings and Conclusions 177 POVERTY AND ENVIRONMENT NEXUS iii Acknowledgements This report is the result of a collaborative In Lao PDR, an inter-ministerial project research effort by Laotian, Vietnamese, Cambo- team was assigned to work on the PEN study by dian and international staff. the Committee of Planning and Investment In Cambodia, Hang Chuon Naron and Ros (CPI) under the direction by Singkham Seilava from the Ministry of Economics and Khongsavanh. The work has been managed by Finance and Kim Saysamalen from the Ministry Monemany Nhoybouakong, National Project of Planning made the �rst recommendations for Director, and supported by Khampadith the scope of the PEN work. A case study on Khammounheuang from the Environment access to natural resources and poverty was lead Research Institute (ERI) under the Science, by Chea Sarin, World Bank consultant. Further Technology and Environment Agency (STEA). consultations have been held with Khieu Muth, Phouthone Sophatilath together with Soun- Pak Sokharavuth, Long Rithirak and Chuon thone Ketphan and Kamphay Manivong from Chanrithy from the Ministry of Environment, the National Agriculture and Forestry Research Ponn Narith, Peou Samy, Hang Samoeun, Khun Institute (NAFRI) under the Ministry of Agri- Sokha and Ross Sovann from the National Com- culture and Forestry (MAF) implemented the mittee for Disaster Management, Mao Saray, study on Non-Timber Forest Products (NTFP) Chan Darong and Bouy Kim Sreang from the and poverty; Bounmy Souksavath in coopera- Ministry of Rural Development, Susan Novak, tion with Pho Ngeun Soulsavath from the Phin Sopheap and Solieng Mak from Department of Roads under the Ministry of the Mekong River Commission, M S Shivaku- Communication, Transport, Post and Con- mar, advisor, Andrew Cock from the NGO struction (MCTPC) implemented the study on Forum of Cambodia, Sik Boreak from WFP Road and Poverty; Soulideth Souvannavong Phom Penh, Declan O’Leary from Danida, Sam- together with Lamphoukeo Kettavong from reth Sok Heng and Claire Liousse from the ERI implemented the study on Unexploded Urban Resource Centre Cambodia, Peter Feld- Ordinance (UXO) and poverty; Tayphasavan man from Partners for Development, Chan Fengthong with Onechanh Keosavanh and Sophal from Cambodia Development Resource Vilayvone Mangkhaseuam from the Environ- Institute, Ingrid Cyimana and Han Phoumin ment Hygiene unit of the Ministry of Health from UNDP Phnom Penh, and Anthony J. Jude, implemented the study on Rural Water and Sani- Chamroen Ouch and Paul Van Im from ADB. tation and poverty; and Chanthala Phimmachack POVERTY AND ENVIRONMENT NEXUS v ACKNOWLEDGEMENTS together with Bounno Fongkhamdeng and Division of MPI coordinated a case study on Vongsak Malivanh from the Urban Research multi-PEN subjects in the Cau River (sub-) Institute (URI) implemented the study on basin, while study implementation was led by Urban Water and Sanitation and poverty. Seng- Vu Tuan Anh from the Socio-Economic Devel- many Keolangsy and Somphath Souvannavong opment Centre (SEDEC) and also included from the National Statistical Centre assisted in Nguyen Xuan Mai from the Vietnamese Acad- case study design and national poverty and emy of Social Sciences, Nguyen Thi Bich Van environment analyses. from the Centre for Public Health and Devel- In Vietnam, TranThi Minh Ha, the Interna- opment in Hanoi, and Nguyen Duc Minh from tional Cooperation Department of the Ministry RCEE. Nguyen Van Pham from the General of Natural Resources and Environment Statistical Of�ce led a team that provided statis- (MoNRE), coordinated the overall PEN initia- tical updates for the study. tive with assistance from Nguyen Thi Thuy The World Bank team comprised mostly of Duong on land administration and poverty staff of the Environment and Social Develop- issues and Nguyen Viet Thang on industrial pol- ment Division of the East Asia and Paci�c lution and poverty. Implementation of the case Region, which is managed by Maria Teresa study on land administration and poverty was Serra. Speci�c directions were provided by led by Tran Nhu Trung from the Consultancy Magda Lovei, Sector Manager of the Environ- Services and Technology Development Com- ment Team. The World Bank team that sup- pany for Natural Resources and Environment ported the three national teams in their case (TECOS). In addition, the TECOS team study preparations, drafted the main report text included Tran Chuong Huyen, Dao Trung and managed databases. The team included Chinh, Nguyen Van Truong, Ho Thi Yen Thu, Jostein Nygard (task team leader), Nat Pinnoi, Ngo The An, Mai Van Trinh, Nguyen Thi Uwe Deichmann, Andrew Murray, Marija Kuz- Hong Nhung, Hoang Hong Hue, Nguyen manovic (Bank staff), Bjorn Larsen, Claude Thuy Phuong and Nguyen Anh Quan. Imple- Saint-Pierre, Bruce McKenney, AJ Bostian and mentation of the case study on industrial pollu- Pamila Aggarwal (consultants). Bjorn and Claude tion and poverty was led by Nguyen Van Tai drafted the initial PEN guidelines adopted in each from the Environment Department under country, assisted the local case study teams and MONRE. A team from the Research Center of performed report writing on the Lao PDR and Energy and Environment (RCEE), which Vietnam chapters. Bruce and Bjorn �nalized the included Nguyen Trung Thang, Nguyen Thi Cambodia chapter. Administrative assistance to Thuy Duong, Hoang Van, and Nguyen Duc the Bank team was provided by Indra Raja, Minh, implemented the study. The case study Natalya Gosteva, Giang Thanh Huong Le, Hoa on pesticide use and poverty in the Mekong Thi Phuong Kieu, Vatthana Singharaj and River Delta was lead by Nguyen Huu Dung Poonyanuch Chockanapitaksa. from the University of Ho Chi Minh City and Initially, the report built upon a Phase I PEN Khuc Xuyen from the Center of Occupational report �nalized by David Wheeler, Susmita Das- and Environmental Health. A study on environ- gupta, Uwe Deichmann and Craig Meisner mental health and poverty involved Truong Viet from the Bank’s Development Research Group Dzung and Tran Duc Thuan from the Depart- in 2002. Susmita and Craig have also coordi- ment of Science and Technology of the Ministry nated the pesticide–poverty case study in Viet- of Health together with Tran Van Dan, consul- nam referred to in this report. tant. Le Mihn Duc and Nguyen Tuan Anh from First drafts of the PEN case studies for each the Science, Tehcnogology and Environment country were presented and discussed at the vi POVERTY AND ENVIRONMENT NEXUS EXECUTIVE SUMMARY national consultation workshops in Hanoi in Shyamsundar and Jian Xie, all World Bank staff. June 2005, in Vientiane in August 2005 and in Elske van de Fliert from FAO-Hanoi, John Phnom Penh in December 2005. Conference Patterson Vietnam-Canada Environment Proj- and workshop information can be obtained ect; Duong Hong Dat Vietnam Plant Protec- on the following website: www.worldbank.org/ tion Association; Sarah Bales MoH in Vietnam, eapenvironment/news. Local PEN workshops Frank Radstake, consultant; Craig Leisher WWF, were held in several case study provinces in 2004 and Kiju Han Korea Institute for Industrial to �nalize PEN implementation guidelines and Economics and Trade all provided comments. discuss case study design. The report was edited by Robert Livernash, Peer reviewers included Jan Bojö, Benu consultant. Circle Graphics designed and man- Bidani, Robert Swinkels (World Bank staff) and aged the desktopping. Production was super- Berit Aasen (Norwegian Institute for Urban and vised by Jaime Alvarez. Photos provided by Regional Research). Additional inputs, com- Uwe Deichmann and the World Bank Photo ments and reviews were provided by Dan Biller, Library. Phillip Brylski, Rob Crooks, Giovanna Dore, Finally, we would like to express our gratitude Paavo Eliste, Patchamuthu Illangovan, Julien to the Government of Norway, which provided Labonne, Santanu Lahiri, Samuel Lieberman, the main trust funds (TFESSD) to carry out the Magda Lovei, Lars Lund, William Magrath, study and to the Government of the Netherlands Thomas Meadley, Stephen Mink, Glenn (BNPP) for providing funds for further publica- Morgan, Nalithone Phonyaphanh, Ian Porter, tions and dissemination of study results. The Kaspar Richter, Klaus Rohland, Jan Willem study was also supported by the World Bank’s Rosenboom, Steven Schonberger, Jitendra environmental mainstreaming funds. Shan, Susan Shen, Maria Teresa Serra, Priya POVERTY AND ENVIRONMENT NEXUS vii Abbreviations and Acronyms ADB Asian Development Bank ADI Acute Diarrheal Illness ARI Acute Respiratory Illness BOD Biological Oxygen Demand CAS Country Assistance Strategy CCB-NREM Commune and Community Based Natural Resource and Environment Management CEMMA Committee for Ethnic Minorities and Mountainous Areas (Vietnam) COD Chemical Oxygen Demand CPI Committee for Planning and Investment (Lao PDR) CPR Common Property Resources CPRG Comprehensive Poverty Reduction and Growth Strategy (Lao PDR) CSES Cambodia Socio-Economic Survey CV Craft Village DANIDA Danish International Development Agency DECRG Development Research Group (of the World Bank) DFID UK Department for International Development DHS Demographic and Health Survey DoSTE Department of Science and Technology Education (Vietnam) EASES Environment and Social Sector Unit of the East Asia and Paci�c Region (of the World Bank ERI Environmental Research Institute (Lao PDR) FA Forest Administration (Cambodia) FAO Food and Agriculture Organization GDP Gross Domestic Product GNI Gross National Income GPS Global Positioning System GSO General Statistical Of�ce (Vietnam) HH Household POVERTY AND ENVIRONMENT NEXUS ix ABBREVIATIONS AND ACRONYMS IBRD International Bank for Reconstruction and Development ICEM International Council for Educational Media IDA International Development Association IDS Institute for Development Studies IFPRI International Food Policy Research Institute IFSR Independent Forest Sector Review (Cambodia) INEST Institute of Environmental Science and Technology (Vietnam) IPM Integrated Pest Management ISO International Organization for Standardization IT Information Technology JICA Japan International Cooperation Agency KEI Korea Environment Institute KIET Korean Institute for Industrial Economics and Trade LECS Lao Expenditure and Consumption Survey (Lao PDR) LTC Land Tenure Certi�cate MAF Ministry of Agriculture and Forestry (Lao PDR) MARD Ministry of Agriculture and Rural Development (Vietnam) MCTPC Ministry of Communication, Transport, Post and Construction MDG Millennium Development Goals MoC Ministry of Construction (Vietnam) MOH Ministry of Health (Cambodia, Lao PDR, Vietnam) MoI Ministry of Industry (Vietnam) MoNRE Ministry of Natural Resources and Environment (Vietnam) MoST Ministry of Science and Technology (Vietnam) MPI Ministry of Planning and Investment MRC Mekong River Commission MRD Mekong River Delta NAFRI National Agriculture and Forestry Research Institute (Lao PDR) NDSP National Strategic Development Plan (Cambodia) NEAP National Environmental Action Plan (Lao PDR) NGO Non-Governmental Organization NGPES National Growth and Poverty Eradication Strategy (Lao PDR) NPEP National Poverty Eradication Program (Lao PDR) NPRS National Poverty Reduction Strategy (Cambodia) NSC National Statistic Center (Lao PDR) NSDP National Strategic Development Plan (Cambodia) NSEDP National Socio-Economic Development Plan (Lao PDR, Vietnam) NTFP Non-timber Forest Products PEN Poverty and Environment Nexus PPA Participatory Poverty Assessment PRSP Poverty Reduction Strategy Paper RCEE Research Center for Energy and Environment (Vietnam) RRD Red River Delta x POVERTY AND ENVIRONMENT NEXUS ABBREVIATIONS AND ACRONYMS SEDEC Socio-Economic Development Center (Vietnam) SFE State Forest Enterprise SIDA Swedish International Development Cooperation Agency SME Small and Medium Enterprise SPE Seriously Polluting Establishment STEA Science, Technology and Environment Agency (Lao PDR) TECOS Consultancy Service and Technology Development Company for Natural Resources and Environment (af�liated with MONRE) TVIE Township and Village Industrial Enterprise UNDP United Nations Development Program UNICEF United Nations Children’s Fund UNPF United Nations Population Fund URI Urban Research Institute (Lao PDR) UXO Unexploded Ordinance VDG Vietnam Development Goals VEPA Vietnam Environmental Protection Agency VND Vietnamese Currency (Dong) VNHS Vietnam National Health Survey VNLS Vietnam National Living Standard Survey WASA Water Supply Authority (Lao PDR) WBG World Bank Group’ WFP World Food Program WHO World Health Organization WSS Water Supply and Sanitation POVERTY AND ENVIRONMENT NEXUS xi Foreword to the Conference Edition This is a draft edition of the Poverty Environ- The material presented in this report is based ment Nexus (PEN) report for Cambodia, Lao upon comprehensive case studies as well as PDR and Vietnam, which will be presented at national analytical work performed in each the sub-regional PEN conference in Vientiane, country. We have omitted this background Lao DPR on June 21–22, 2006. material from this conference edition to avoid The purpose of this conference edition is to overwhelming the readers. This work will be present the �ndings from the studies that have included, however, in a CD-ROM attachment been undertaken in each country over the last in the �nal published version. three years as well as to obtain relevant com- As the team of authors is still working on the ments and feedback from the conference partic- overall conclusions and recommendations from ipants that could be included in the �nal edition the report, we have decided to postpone the of the report. inclusion of a comprehensive summary to the We realize that the material presented in this report until its �nal publication. report requires time to study and digest. Some of Wish you good reading of this edition and the linkages between poverty and environment looking forward to receiving your comments. are complex and dif�cult to capture. You, as the reader of this report, are, therefore, encouraged to give yourself suf�cient time to begin to dis- PEN Report Authors cern the main storylines. June 2006 POVERTY AND ENVIRONMENT NEXUS xiii I STUDY CONTEXT AND METHODOLOGY Introduction DEVELOPMENT OF THE PEN STUDY as village and commune data, however, con- strained the study. Moreover, the study found Past World Bank PEN Work that the PEN structures are often masked at in the Sub-region more aggregated levels and may be revealed only through study at more local levels. These In order to understand the extent and signi�cance findings and challenges were taken into con- of the PEN in the lower Mekong sub-region and sideration in the shaping of the second phase to expand the dialogue between the World Bank of the PEN study that is summarized in this and the three countries on the issue, the World report. Bank launched the �rst phase of the PEN study in 2001. Based upon secondary data sources, the study Motivation and Objectives examined seven main environmental indicators1 for the Study for mainly province-level correlations between environment (deforestation, fragile lands, indoor This second phase of the PEN study is designed air pollution, lack of access to water, lack of access to address several of the unsolved issues faced to sanitation, outdoor air pollution, and number by the first phase and within the existing PEN of diarrhea cases) and poverty and found that the literature on Cambodia, Lao PDR and Viet- nexus is differs between countries. nam. Since the earlier study found that the According to the study �ndings, the nexus in nexus may often be identified at disaggregated, Cambodia is con�ned primarily to ‘brown’ issues or local, levels, the analysis in the second phase such as indoor air pollution, water contamina- of the study was conducted both at the regional, tion and lack of adequate sanitation, in Lao PDR provincial and district levels for the national stud- it spans across all seven environmental issues, ies (macro-levels) and on the district, village, com- whereas in Vietnam, the study results were some- mune, and household levels in the (case-) studies what eclectic, suggesting a possible existence of a (micro-level). This combination of macro and nexus for fragile soils and indoor air pollution, but micro approaches results in more substantive relatively weak correlations for the remaining findings, which are elaborated in detail in the indicators (see Figure I.1). individual country chapters. The intention is that Data availability, particularly to the lack of these elaborations may then lead to a better district-level and more micro-level data, such scope for de�ning policy interventions. PEN STUDY IN THE LOWER MEKONG SUB-REGION 3 INTRODUCTION MAP I.1 Provinces and Socio-Economic Indicators for the PEN Study in the Sub-region Sources: World Development Report 2005; Poverty Assessment Papers for Cambodia, Lao PDR and Vietnam The study seeks to reflect the geographic tries. The gradual conversion of the joint strate- diversity of the poverty and environmental gies for environmental protection and poverty challenges, so it is sufficiently large in both reduction into actual operations is the main geographic and sectoral scope (see discussion in long-term development objective of this ana- the individual country chapters 2–4). It is also lytical work. designed to be policy- and response-driven, so In summary, the study had three main objec- it not only seeks to identify referred PEN hot- tives: (i) further identify and elaborate upon spots, where the poverty-environment dynam- PENs in the three countries, (ii) review policy ics is particularly strong, but also to provide options vis-à-vis these identi�ed PENs and (iii) basis for mainstreaming the PEN findings into establish links to ongoing PEN project initia- poverty reduction programs in the three coun- tives in the sub-region (see Table I.1). 4 PEN STUDY IN THE LOWER MEKONG SUB-REGION INTRODUCTION FIGURE I.1 Correlations Between Poverty and Different Environmental Indicators Found from Previous PEN Work in Cambodia, Lao PDR and Vietnam 1.00 Rank correlation with poverty 0.80 0.60 0.40 0.20 0.00 -0.20 Cambodia Lao PDR Vietnam -0.40 Deforestation Fragile Land Indoor Air No Access No Number of Outdoor Air Rate (Slope>16%) Pollution to Clean Sanitation; Diarrhea Pollution (wood/ Water Vietnam: Cases (number of charcoal use); notoilets deaths from Vietnam: ARI PM10) Indicator Regional and Country Context Based upon both the specific and general environmental and poverty features in the sub- Cambodia, Lao PDR and Vietnam face many region, several poverty—environment initiatives common challenges in their quest for economic have been generated by donors over the last years development and poverty reduction (see figure (see table I.1). While most of them originally I.2). An important shared feature is that poverty addressed poverty—natural resource management and environment issues in Lao PDR, Cambodia issues, newer initiatives have also increasingly and Vietnam fall into one of two broad cate- addressed poverty—environmental health issues. gories: environmental health and natural resource Some of these newer studies have also had multi- use. This feature is not speci�c only to these coun- sectoral PEN approaches. However, this study tries, as there is a body of research on poverty and may be among the �rst that tries to address environment linkages that gives this �nding global poverty environmental issues throughout the signi�cance.2 However, it is a particularly rele- natural resource management and environmen- vant classi�cation, as the national policy frame- tal health subjects at both the national and sub- works in the three countries already incorporate regional levels. As such, this study strives to both the issues of clean water and sanitation and make a contribution towards a comprehensive of sustainable use of natural resources into devel- understanding of the complexity of poverty— opment and poverty reduction strategies. environment linkages in the sub-region. PEN STUDY IN THE LOWER MEKONG SUB-REGION 5 INTRODUCTION FIGURE I.2 General Features of Environment and Poverty in the Lower Mekong Sub-region Specific to Vietnam Shared by Laos and Vietnam Specific to Lao PDR Absolute number of poor Upland poverty: Indochina Rural-to-rural concentrated in low lands. Cordillera migration at core of Strong government Poverty reduction policy Government policy institutions targeting the uplands Mainly planned economy in Widely diversified economy transition reflecting many sectors with PEN features. Shared by Vietnam & Specific to Cambodia Shared by Lao PDR Cambodia Post-war migration and Cambodia Urban poverty Flood regime in the central Weak institutional & region legal capacity Mainly market economy Shared by the three countries Access to natural resources critical to the poorest’s basic livelihood conditions Disproportionately low access to clean water and sanitation by the poor. Rural poverty traps in environmentally-fragile uplands and marginal areas Growing urban poverty with environmental health issues Migration What Laos, Cambodia and Vietnam also have In spite of the many common geo-political in common is fast-growing economies that allow and socio-economic circumstances, however, for rapid overall poverty reduction, but also it is important to acknowledge that Cambodia, “poverty traps� remaining in the rural sector. Lao PDR and Vietnam also confront unique These poorer areas are often, although not always, challenges in combating poverty and environ- environmentally sensitive areas. High economic mental problems. Their distinctive natural endow- growth rates occur jointly with the growth of ments and institutional and policy responses in urban centers3. Urban poverty rates are still low in their economic development, including the Vietnam and Laos and decreasing in Cambodia transition from socialism to more market- (in geographically comparable areas of Phnom oriented economies, have resulted in relatively Penh, poverty dropped from 11.4 percent to 4.6 distinct development paths, and consequently percent between 1993/94 and 2004), but the issue distinct priorities for environmental manage- of low-income groups is expected to become sub- ment and poverty reduction. The following sub- stantial in the future. The urban poor population sections deal with the speci�c national contexts is already facing speci�c environmental health of the three countries, which will contribute to issues in Vietnam, and so are the populations a better understanding of the differences in employed in heavy industries, but, in general, the scope, approaches and coverage of poverty and national policy frameworks appear to devote environment linkage areas between the three insuf�cient attention to this growing issue. countries. 6 PEN STUDY IN THE LOWER MEKONG SUB-REGION TABLE I.1 Relevant PEN Studies in the Lower Mekong Sub-region, 2005 (to be completed) Study & Donor Initiative title methodologies Sector focus Country focus Key issues addressed Implementation International Organizations WB “Poverty Environment Case studies; map- Forestry, agricul- Mekong Linkages between 2003–2006 Nexus in Cambodia, ping PEN correla- ture, industry, Sub-region poverty and Lao PDR and Vietnam� tions; workshops pollution (Cambodia, access to natural management Lao PDR, resources; forest Vietnam) resource manage- ment; access to water and sanita- tion; pesticide use; and industrial pollution ADB/CIFOR “Mapping the Link Case studies; Forestry Mekong Forest & poverty link- 2004–2005 Between Poverty and mapping forest/ Sub-region ages; forest depen- Forest Dependence in poverty correla- (Cambodia, dence; innovative the Mekong Region tions; workshops Lao PDR, approaches to for Improved Pro-poor Vietnam) poverty alleviation Policies and Projects� through forestry ADB Poverty and air Ho Chi How to address poor 2005–2007 pollution Minh City, people dispropor- Vietnam tionately affected by Air Pollution UNDP/DFID “Harmonizing Poverty Case studies; ecology- Forestry, agricul- Vietnam Forestry and aqua- 2005–2009 Reduction and Envi- economy models; ture, industry, culture; water sup- ronmental Goals in analyzing, model- energy, pollution ply and sanitation; Policy and Planning ing and piloting management (rural) renewable for Sustainable best practices energy, sustain- Development� able energy use Bilateral Agreements DANIDA “Mapping of Monitoring Efforts on Poverty and Natural Resources Environmental PEN STUDY IN THE LOWER MEKONG SUB-REGION Management� SIDA *ADB—Asian Development Bank; CIFOR—Center for International Forestry Research; DANIDA—Danish International Development Assistance; DFID—Department for International Development (UK); UNDP—United Nations Development Program 7 INTRODUCTION INTRODUCTION Cambodia The 1990s witnessed a major recovery for the Cambodian economy. Vigorous economic Economic and social context growth of about 7.1 percent per year has resulted Cambodia is a small country in Southeast Asia in declining poverty rates and increasing real per encompassing a territory of about 181,000 km2 capita consumption. According to the latest and a population of 13.6 million in 2004. The World Bank’s Poverty Assessment draft report, country’s economic development in the past poverty incidence in Cambodia declined from thirty years has occurred against a backdrop of 47 percent in 1993–94 to 35 percent in 2004 in political instability. The rule of the Khmer Rouge geographically comparable areas, and the real per from 1975 to 1979 and the persistent civil strife capita household consumption rose about 32 per- following the regime’s decline have had lasting cent from 1993 to 2004. This economic growth consequences on the country’s economic infra- has also resulted in improvements in many structure and its human and natural capital. Even socio-economic indicators, such as school enroll- though the country is now enjoying peace and ment, housing quality, etc.4 greater political stability, it is still struggling to However, poverty reduction and consump- overcome its long legacy of conflict. tion increases were not uniform across the MAP I.2 Poverty and Population Density in the Sub-region Source: East Asia Poverty Reduction and Economic Management, World Bank 8 PEN STUDY IN THE LOWER MEKONG SUB-REGION INTRODUCTION country. Rural areas experienced significantly MAP I.3 Forest Cover in the Sub-region slower growth rates, even though they started off with the lowest average real consumption in 1993/94. This has resulted in a growing income disparity between the rural and urban areas. The Gini coefficient, which measures income inequality, is estimated to have risen from 0.35 to 0.40. Urban manufacturing, construction, and tourism have been the primary engines of growth in the past decade, and they have concentrated predominantly in Siem Reap and Phnom Penh. Since studies have also shown that the poor have bene�ted little from remittances,5 it is not sur- prising that this growth has not trickled down more towards the rural areas. Only about 13 per- cent of all households received such support, and it consisted only of about 8–10 percent of the total household consumption. The lack of strong development in rural areas is also attributable to the slow growth in agricul- tural production. Cambodia’s crop yields, labor productivity, and land productivity have been and still are among the lowest in Asia. Agricul- tural production is both dependent upon and signi�cantly impacted by the floods that are a unique feature of the Lower Mekong and Tonle Sap region. Drought occurrence in the country is also common. Vulnerability to natural disasters is, thus, another critical issue in poverty reduction efforts in Cambodia. Environmental trends The loss of forest cover, estimated at a rate of 70,000 to 100,000 ha per year, is one of the most Source: Global Land Cover 2000 database alarming environmental trends in Cambodia. The country’s forests comprise more than 50 per- ucts (NTFPs), which are used for subsistence and cent of the country’s territory and are, therefore, income generation. a major source of capital for the Cambodian Cambodia’s poor plateau areas have large areas economy and an important livelihood resource of fragile lands that are very susceptible to infer- for the poor (see Map I.3). In addition to timber, tility. Their sandy soils have little organic content which is the primary source of fuel for cooking and limited water retention capacity. Therefore, and heat (fuelwood and charcoal comprise about degradation resulting from inappropriate land use, 80 percent of the total energy supply), forests land clearings, and natural disasters further exac- provide the poor with non-timber forest prod- erbates the existing problem of infertile lands PEN STUDY IN THE LOWER MEKONG SUB-REGION 9 INTRODUCTION and maintains very low agricultural yields. As Access to safe drinking water and sanitation is the agricultural sector currently accounts about very limited. Only 17.7 percent of the total pop- 75–80 percent of the country’s population and ulation has access to sanitation6 (see Map I.4). accounts for about 40 percent of the national Limited capital prevents the establishment of GDP, such pressure on agricultural land resources infrastructures like sanitation and sewage systems, may have severe impacts on the overall agro-based and, if available, limited management prevents national economy. most poor people from access and connections. Cambodia is very rich in water resources, most Diarrhea and cholera have become endemic dis- important of which are the Mekong and Tonle eases in the country. Sap Rivers and Tonle Sap Lake. Even though these Air quality has been declining in Phnom Penh resources are critical to the country’s continued and other urban centers. Steady increases in sulfur economic growth, they are increasingly stressed dioxide and nitrogen dioxide have been observed by human activity. Increasing demand for water along with increasing incidence of respiratory combined with pollution from agricultural runoff, problems in children. The primary pollution industrial effluents, domestic sewage and solid sources are industrial emissions from fossil waste, are severely threatening the quantity and fuels, growing transportation fleet and waste quality of water supply. Groundwater is simi- incineration. larly impacted. A declining water table is an issue Growing solid and hazardous waste is also a in many areas. major issue. Urbanization, lack of integrated solid MAP I.4 Access to Clean Water in the Sub-region 10 PEN STUDY IN THE LOWER MEKONG SUB-REGION INTRODUCTION waste management system, absence of toxic and With the market-oriented reforms of the 1990s, hazardous waste disposal facilities and improper they saw increased economic growth and signif- management are the main culprits for these trou- icant poverty reduction. bling trends. On average, the country has achieved impor- tant strides in its �ght against poverty. Growth Lao PDR has averaged 6 percent per year between 1991 and 2003 and the incidence of poverty has fallen Economic and social context from 46 percent in 1992 to 33 percent in 2003. Lao PDR is a small, sparsely populated, land- However, as mentioned above, these bene�ts locked country with extensive natural resources. have not reached the poorest, most marginalized It covers an area of 236,800 km2, and its popu- populations. Many of these communities rely on lation of about 5.7 million people is growing agriculture as their main source of income, even rapidly. Similarly to Cambodia and Vietnam, though their land may not be �t for cultivation. the country’s economic development in the past Shifting cultivation remains an important land decades represents to some extent recovery from use system, and its application appears to have the destruction of past conflicts. Lao PDR suffered both its pros and cons as well as different views significant economic losses from the Indochina about its possible impact on the environment Wars, particularly the heavy US bombing of and poverty. As a matter of fact, most shifting the Ho Chi Minh trail. The consequences of the cultivation today is rotational agriculture and bombings are still visible today. Unexploded appears sustainable in districts where rural pop- ordinances (UXOs), for example, represent a ulation densities remain (e.g. below 20 inhabi- major issue in the eastern and southern part of tants/km2). However many districts are already the country in particular and were covered by beyond those levels, and land use systems pose the PEN study. issues of sustainability in terms of environment, Lao PDR still has limited infrastructure, such household incomes, or both. as roads, water, electricity and telecommunica- tions. However, it is well-endowed with natural Environmental trends resources, including forests, minerals and water resources. The loss of forest resources is one of the main Much of the country’s economic and social environmental problems that affect Lao PDR. In character has been defined by its topography the 1940s forest cover was estimated at about (see Map I.5). The upland areas, covering a 70 percent of the total land area, while today large proportion of the northern part of the that figure has decreased to about 40 percent, country and border areas in the southern part, which is, in fact, a comparatively high rate (see are characterized by low potential for sustain- Map I.3). A related and significant environ- able production of annual crops, lack of access mental challenge is the depletion of NTFPs, to roads (about 30 percent of the villages are which, like in Cambodia, are a widespread and not accessible during the wet season, and only significant income source in rural areas, includ- 20 percent can be reached by truck at any time ing for the poor. The habitat loss for many ani- of the year),7 high poverty incidence and dis- mal and plant species is a particularly worrisome proportionately high ethnic minority popula- trend, considering that Lao PDR is one of tion. The lowlands, on the other hand (especially the most biodiversity-rich countries in South- around the capital Vientiane), have benefited east Asia. Loss of topsoil due to erosion is greatly from their developed transportation another important issue potentially affecting infrastructure and their proximity to markets. all sloped land. PEN STUDY IN THE LOWER MEKONG SUB-REGION 11 INTRODUCTION Even though Lao PDR has a large per capita MAP I.5 Land Elevation in the volume of renewable water resources, the qual- Sub-region ity of both surface and groundwater is declin- ing. In addition, access to clean water and sanitation is a major problem. In rural areas, access to clean water is about 60 percent, whereas access to sanitation is estimated at about 36 percent (see Map I.4). There is a high incidence of diarrhea, dysentery and other waterborne dis- eases. Municipal solid waste is increasing rapidly, and disposal and pollution issues will feature prominently on the government’s environmen- tal agenda. Vietnam Economic and social context Vietnam is the largest and most populous coun- try in the Mekong sub-region. Its 82.2 million people live on an area of about 331,700 km2, which includes lowlands in the Mekong River delta in the South and the Red River delta in the North, and a large proportion of uplands. As in Lao PDR, the different topographical features of the country influence its social and economic context (see Map I.5). Even though Vietnam has also suffered sig- nificantly from its legacy of conflict, including (VDGs), which are the localized version of its involvement in all Indochina Wars, it has the Millenium Development Goals (MDGs), been moving much faster on the path of eco- demonstrate consistent improvements in school nomic recovery and consequently has the most enrollment, infant mortality rates, etc. developed and diversified economy in the sub- While poverty continues to decline, however, region. the pace of reduction has slowed, and there are With the second-fastest growing economy in increasing disparities between urban and rural the world after China and estimated growth rates areas. In rural areas, unemployment and under- of about 7 percent per year, Vietnam has man- employment are high, compelling many young aged to achieve impressive poverty reduction. A people to join the work force. Moreover, poverty decade ago, nearly 60 percent of the population reduction has been signi�cantly lower among lived below the poverty line. Today that �gure ethnic minorities compared to the rest of the has declined to less than 30 percent. Almost a population. third of the population—the equivalent of more The sustained economic expansion has pri- than 20 million people—was lifted out of marily been driven by the agricultural sector (the poverty in less than a decade. Mekong and Red River deltas are among the Progress has also been made in many social most productive and intensively cultivated areas indicators. The Vietnam Development Goals in Asia), including the expansion of cash crops 12 PEN STUDY IN THE LOWER MEKONG SUB-REGION INTRODUCTION particularly coffee and sugar cane. However, the extent of organic pollution in the rivers, exceed the country has also built up a relatively strong national water quality standards by several times. industrial base, which has created opportunities Lakes, small streams, and canals are also severely for many low-skilled workers in the cities. polluted. Urban air pollution is increasing due to a Environmental trends growing motor vehicle fleet, and growing indus- Unlike in Cambodia and Lao PDR, forest cover trial emissions. Fossil fuel consumption by power in Vietnam has been increasing in the past plants, which generate about 60 percent of the decade, a trend that is taking place together with country’s energy needs, emits high levels of a partial shift of the forestry sector towards envi- polluting gases such as sulphur dioxide, carbon ronmental objectives instead of production monoxide, nitrogen oxides and �ne particulates. objectives, with intensi�cation of farming sys- Other industrial sectors, such as metallurgy and tems and with land tenure on sloped land (see the chemical and cement industries, also gener- map I.3). At the same time, however, forest ate signi�cant levels of pollution. quality has continued to decline. Closed-canopy forest has dropped from almost 100 percent of PEN STUDY DEVELOPMENT forested areas a century ago to about 13 percent AND SUBJECTS today. Unsustainable use of fragile lands in Viet- Study Selection and Implementation nam is a major environmental threat, and steep slopes in upland areas make soils more suscep- The study selection process included not only tible to erosion when pressure from agricultural the selection of poverty and environment link- uses increases. Salinization in the Mekong delta age areas to be examined, but also the appropri- affects the quality of drinking water in the area. ate geographic locations for each study. The Groundwater contamination resulting from location of the studies was chosen in order to sat- the salt-water intrusion impacts drinking and isfy two major objectives. industrial water as well as ecosystems and agri- First, the PEN study sought to cover suffi- cultural production cient geographic area so as to adequately repre- As in Cambodia and Lao PDR, access to sent the geographic diversity in each country clean water is a major problem, even though the (map I.6). For example, studies were located in country is well-endowed with water resources upland, lowland and piedmont areas (Figure I.3), (see map I.4). In 2000, only 52.2 percent of the as well as in the northern, and southern areas in population had access to safe water. With water Lao PDR and northern, central and southern demand increasing rapidly due to growing areas in Vietnam, which have many geographic, urbanization and industrial demand, access to social (including ethnic), economic and politi- clean water can become an even more critical cal diversified features. issue in the future. Second, the study locations also sought to Industrial pollution affects water and air qual- explore the contrast between high- and low- ity, and thus has serious health implications for income provinces and between other high- and the population that may disproportionately affect low-income administrative units within the the poor. Pollution primarily affects surface provinces. In Lao PDR, the geographical focus of waters. Many of Vietnam’s rivers that flow the PEN studies was determined to a signi�cant through urban and industrial areas are heavily extent by the Laotian government’s selection polluted. Biochemical-oxygen demand (BOD) inputs, which reflected its own poverty priority and Ammonia-Nitrogen levels, which indicate locations (ex. 47 poverty priority districts). PEN STUDY IN THE LOWER MEKONG SUB-REGION 13 INTRODUCTION Access to Clean Water in the Sub-region MAP I.6 14 PEN STUDY IN THE LOWER MEKONG SUB-REGION INTRODUCTION FIGURE I.3 Elevation Distribution in Cambodia, Lao PDR and Vietnam Lao Cumulative share of land area by elevation zone Cambodia PDR Vietnam 100 < 100 61.5 1.4 35.5 90 101–500 33.1 36.5 31.3 80 501–1000 4.9 40.0 23.2 1001–1500 0.5 20.0 7.9 70 1501–2000 0.0 1.9 1.7 60 percent > 2000 0.0 0.1 0.4 50 Cambodia 40 Lao PDR 30 Vietnam 20 10 0 < 100 300 500 700 900 1100 1300 1500 1700 1900 2100 2300 2500 elevation Source: East Asia Update—Strong Fundamentls to the Fore—Regional Overview, April 2004, East Asia and Paci�c Region, The World Bank. In addition, in both Lao PDR and Vietnam, ii) the cost effectiveness (including administrative considerations were also made about addressing cost) of interventions. parts of the country That had not necessarily been Therefore, the �nal decision about study approached by international agencies (donors) areas drew a balance between analytical research before. In fact, it was recognized early on that the objectives as well as policy objectives by the host poorest parts of the countries had not necessary countries. been the main recipients of donor assistance.8 After careful consideration of the selection The choice of the poverty-environment link- criteria in a dialogue with multi-ministerial rep- age areas covered by the studies was subject to a resentatives in each country, the following envi- wider range of criteria. The environmental prior- ronment and poverty linkage areas were chosen ity issues identi�ed in the countries’ development for the studies: and poverty reduction strategic plans provided In Cambodia, the study examined the linkages critical guidance in the selection of the PEN stud- between poverty and access to natural resources, ies. In addition, a set of selection criteria were including agricultural land, forest products, graz- developed in the dialogue with the governments, ing land, �sheries and drinking water. which identified sectors affecting the highest In Lao PDR, the PEN studies examined link- number of poor people (table I.2). Other selec- ages between forest resource management (NTFP, tion criteria included: i) the severity of environ- roads and UXOs) and poverty, and between water mental impact (on income, health, vulnerability), supply and sanitation (rural WSS and urban WSS) whether the impact was alleviated by national and poverty. economic growth, the local institutional and The PEN studies in Vietnam included: land �nancial capacity to deal with the problem; and administration, environment and poverty; Cau PEN STUDY IN THE LOWER MEKONG SUB-REGION 15 16 INTRODUCTION TABLE I.2 Estimated Population Affected by Environmental/Health Indicators Cambodia Lao PDR Vietnam Estimated Number of People Affected Estimated Number of People Affected Estimated Number of People Affected Environment & % of non-poor % of poor % of non-poor % of poor % of non-poor % of poor Health indicator: population population population population population population Decline in Forest 66 percent 70 percent More data needed More data needed More data needed More data needed resources* Use of fragile land More data needed More data needed < 20percent < 45 percent More data needed More data needed Fuelwood/indoor 90 percent 99 percent 95 percent 98 percent 80 percent > 90 percent PEN STUDY IN THE LOWER MEKONG SUB-REGION air pollution Lack of water 16 percent 22 percent 20 percent 30 percent 15–20 percent 25–30 percent supply** Lack of sanitation 60 percent 80 percent 40 percent 70 percent 10–12 percent 30–35 percent Households 60 percent 35 percent 80 percent 60 percent always boiling drinking water Drought*** 12 percent 16 percent 0.2 percent 1.7 percent More data needed More data needed Flooding*** 24 percent 23 percent 5 percent 1 percent 1.3 percent 1.1 percent UXO/Mine�eld 19 percent 28 percent 20 percent 28 percent contamination**** * Cambodia: From 2005 survey on communes reporting status and trends in natural resources, combined with CSES 2004 consumption survey data. ** This is use of open, unprotected water sources (river, ponds and streams) in Cambodia and Lao PDR), and open, unprotected water sources and dug wells with nearby pollution sources in Vietnam. ***The data are not comparable across countries due to substantial differences in de�nitions. **** Estimated by combining data from household consump- tion surveys and UXO/mine�eld surveys. Cambodia: Mine�eld and Cluster bomb contamination from 2002 survey. Lao PDR: UXO contamination from 1997 survey. INTRODUCTION River basin, poverty and environment; pesticide the population in quintiles (strata) to distin- use and poverty, industrial pollution and poverty; guish between the different levels of access to and water supply and sanitation and poverty. The water, sanitation and health coverage. more diversi�ed PEN study structure for Vietnam is a result of this country’s more diversi�ed econ- Inclusive Approach and omy and the consequent greater variety of envi- Country Ownership ronmental issues with poverty implications. High country ownership was one of the most Policy Orientation important considerations in the design process for this study. The World Bank staff worked The expert team utilized the national �ve-year closely with local experts and government of�- plans, national poverty eradication strategies and cials from the initiation of the study to its com- poverty reduction strategy papers, which are pletion in order to ensure suf�cient input from jointly prepared by the national governments the local counterparts as well as shaping of the and the World Bank, as an important framework study and its outcomes to �t the government’s for the development and implementation of the objectives for poverty reduction. PEN study. They served not only as indicators of The local counterparts were involved signi�- government priorities for interventions in poverty cantly into the selection of the study locations and environment issues, but also as ways to assess and were responsible for implementation of local where the biggest gaps and shortcomings are in surveys.10 They remained involved throughout targeting and addressing poverty-environment the completion of the study through continued issues, in order to ensure that the outcomes of the dialogue and data sharing with the World Bank PEN study could be utilized to �ll those gaps (see staff as well as through consultation workshops Table I.3). and conferences that provided opportunities for These documents demonstrate that there is knowledge-sharing and enhancing the under- some recognition within the governments of standing of how the PEN �ndings �t into the Cambodia, Lao PDR and Vietnam of the poten- tial for jointly addressing poverty and environ- governments’ poverty reduction agendas. ment issues. However, there are many issues that deserve greater attention and more immediate Partnership and Cooperation interventions, and the PEN study seeks, among other things, to highlight those issues and bring The focus on country ownership provided basis them forward on the governments’ agendas. for continuous dialogue and cooperation between Moreover, the PEN work has already pro- the World Bank staff and the counterparts in vided some valuable inputs into these docu- Cambodia, Lao PDR and Vietnam throughout ments. The most substantive contributions were the completion of the study. Cooperation was made towards the Lao PDR National Poverty established with numerous government institu- Eradication Program and the Lao PDR Poverty tions and local non-governmental organizations Assessment. In both of these documents, a sep- (NGOs). arate section (NPEP) and a chapter (PA) were In Cambodia, the authors worked closely with devoted to findings from the PEN study. In the National Institute of Statistics in obtaining Vietnam, PEN approaches were also incorpo- data from the Population Census. In Lao P.D.R, rated in the Ministry of Planning and Invest- important poverty data was supplied by the ment work related to the Five Year Plan. Lastly, National Statistics Center. In Vietnam, the dis- the Poverty Assessment in Vietnam used the trict-level data were assembled by MoSTE, GSO initially-developed PEN approach of grouping and the Ministry of Health. PEN STUDY IN THE LOWER MEKONG SUB-REGION 17 18 INTRODUCTION TABLE I.3 Poverty-Environment Issues in Relevant Analytical and Planning Documents9 Key report statements/targets Relevant country documents Cambodia Lao PDR Vietnam Five-Year Plans National Strategic Development Five-Year Socio-Economic Develop- Five-Year Socio-Economic Develop- Plan 2006–2010 ment Plan 2006–2010 ment Plan 2006–2010 Main poverty-environment targets: Key environment-poverty targets: Main poverty-environment targets: increase access to safe water to increase in forest cover to 53–55%; increasing forest cover by 43%; 45% in rural and 85% in urban increase in the provision of clean increasing protected areas (esp. population; increase access to water to rural areas to about 65% of wetlands) by 11.2%; increasing sanitation to 22 percent of rural rural population; reduction in mor- access to clean water to 95% of and 67 percent of urban popula- tality rates of children under 5 to urban and 75% of rural popula- tion; increase forest cover to less than 30%; elimination of shift- tion; improving residential and 60 percent of the total area; and ing cultivation by 2009–2010; assign- industrial solid waste collection reduce fuel wood dependency in ment of land and forested areas to to 80–90%; improving industrial households from about 85.5 per- 3782 villages and 268,245 house- wastewater treatment and treat- PEN STUDY IN THE LOWER MEKONG SUB-REGION cent to about 61 percent of total holds between 2006 and 2010 ment of hazardous waste. population. Poverty Reduction National Poverty National Poverty Eradication The Comprehensive Poverty Strategy Papers Reduction Strategy Programme Reduction and Growth Strategy “Increasing opportunities for the “Poverty eradication activities should “Harmoniously combine socio- poor is the key element in pro- not degrade [the] quality and quan- economic development with tecting environment since natural tity of the environmental resources environmental protection and resources are eroded when the that the poor depend on, and (ii) improvement and ensure that all poor do not have alternative Improving environmental conditions people are able to live in a healthy means. Natural resource degrada- can lead to sustainable poverty erad- environment. Actively integrate tion is in part due to exploitation ication. Targeted poverty/environ- environment issues in plans, pro- by the rural poor, who are seeking ment linkage areas: deforestation grams, and projects and make it a to satisfy their basic needs. Reduc- and agricultural land; use of fragile criterion for evaluating develop- ing rural poverty is therefore land; water resources and environ- ment measures. Integrate fully the essential to achieving sustainable mental health; biodiversity; natural environment and natural resource development in Cambodia.� disaster vulnerability.� issues into the master plan for socio-economic development in provinces, districts, and ensure that the development is sustain- able and does not cause degrada- tion in natural resources.� Country Assistance Strategies The second objective of the Cambo- Under the �rst objective of the CAS, The draft version of the latest CAS dia CAS’s �rst pillar is to improve the Bank’s long-term and higher for Vietnam addresses both nat- natural resource management. order outcomes for Lao PDR include: ural resource management and Through lending and AAA, the a) improving governance and environ- environmental health issues. Pri- Bank expects to achieve the fol- mental management of natural mary focus areas in NRM include lowing outcomes: assets; forestry land, community-based a) improved provincial land use b) strengthening the collection and land management, and the con- management; use of revenues from natural version of agricultural land; alloca- b) accelerated land distribution & resources to help reduce poverty; tion of land titles, particularly in increased number of secure land c) maintaining productive forest cover uplands; biodiversity conservation. titles; as part of rural livelihoods; World Bank projects in environ- c) transparency in generation of d) reducing deforestation rates; mental health in urban areas will public & commercial bene�ts e) improving rural livelihoods. focus on waste water and waste from forests; management (PCB and hazardous d) expansion of community forestry; waste management, other indus- e) improved policies & institutions trial pollution) and carbon �nance. for protected areas management. Poverty Assesments The Vietnam Development Report “The Poverty Assessment highlights The latest poverty assesment for Lao 2004, which focused on poverty, trends in poverty in the past PDR highlights the strong poverty does not highlight environmental decade with some speci�c focus reduction in the country in the past on education, health, and agricul- decade, but also recognizes numer- issues as contributors to poverty, ture and rural development. How- ous poverty-related areas, where but does discuss environmental ever, a separate chapter devoted more work is needed. It highlights health issues, such as access to to poverty-environment linkages— the strong geographic features of clean water and sanitation, as similar to the one prepared in the poverty in the country, the dispro- important aspects of poverty and Lao PDR Poverty Assessment—has portionately higher poverty among inequality in the country. The not been included. The intention ethnic minorities as well as the slow newest update on the Poverty of Chapter 2 of this report is to progress in improving many social update in Vietnam (from 2006) provide such information� indicators. The report states that shows further reductions in poverty is linked to environmental poverty, even if they are still conditions and dedicates a chapter to relatively unequally distributed the environment, which is mostly between the different regions. based on �ndings from this PEN The statistics show a reduction in study. The NRM section of the chap- the urban-rural ethnic divide, but ter covers forest resources, land the ethnic minority population resources, UXOs and natural disas- lags farther and farther behind in ters. The environmental health sec- poverty reduction. Ethnic minori- PEN STUDY IN THE LOWER MEKONG SUB-REGION tion focuses on access to clean water ties comprise 39% of the total and sanitation as well as indoor air number of poor. pollution. 19 INTRODUCTION INTRODUCTION Cambodia’s Ministry of Environment and the II consists of the three country chapters, which Ministry of Planning in Lao are expected to play present the �ndings from the studies in each a key role in the mainstreaming the PEN �ndings country. The sub-regional �ndings and conclu- in their respective National Poverty Reduction sions are presented in Part III of the paper. Strategy (NPRS) processes. In Lao PDR, the The separate studies reflecting �ndings from Science Technology and Environment Agency each of the study areas (map I.6), national PEN (STEA) and the Committee for Planning and analyses and additional supporting material is Cooperation (CPC) will be the key partners provided in the CD-ROM attachment. in integrating the PEN results in the National Poverty Eradication Program (NPEP). The Endnote/Reference: recently set up Ministry of Natural Resources and Environment (MoNRE) and the Ministry To follow: 1. The study itself included 5 environmental indicators as of Planning and Investment in Vietnam will access to water, sanitation and health impacts were provide support in incorporating the PEN �nd- clustered into one indicator. ings in the CPRGs. 2. Examples are the jointly DFID, EC, UNDP and World To sustain the dialogue and cooperation Bank-prepared “Linking Poverty Reduction and Environ- mental Management—Policy challenges and Opportuni- throughout the process of completion of the ties—A contribution to the World Summit on Sustainable study and to share �ndings and knowledge from Development Process,� the World Development Report the three countries’ PEN work, the World Bank 2003 “Sustainable Development in a Dynamic World and its partners organized a series of sub-regional Transforming Institutions, Growth, and Quality of Life,� and Poverty—Environment material issued in the workshops (see table I.3). World Bank Environmental Economics Series. In completing the study, the World Bank staff 3. There are 4 large urban centers in the sub-region: Hanoi, also worked closely with international organiza- Ho Chi Minh City, Phnom Penh and Vientiane. tions such as the World Food Program (WFP), 4. Cambodia Poverty Assessment Draft Report. the Mekong River Commission (MRC), the 5. Cambodia: Poverty Assessment, presentation of pre- liminary �ndings, October 13, 2005. World Health Organization (WHO), the United 6. The World Bank’s Cambodia Environmental Monitor Nations Population Fund (UNPF), the United for 2003. Nations Children’s Fund (UNICEF), and the 7. NSC-Lao PDR 1999 (Social Atlas of the Lower United Nations Development Program UNDP). Mekong Basin). 8. This fact appears also to be the case for Cambodia— refer to draft Poverty Assessment. 9. The PEN study contributed to the following docu- IV. Structure of the Report ments: Lao PDR PRSP and Poverty Assessment and The paper is organized as follows. Part I consists Vietnam’s Five-Year Plan and Poverty Assessment. The Vietnam PRSP was prepared prior to the PEN study’s of one chapter that provides a detailed descrip- completion. tion of the methodology and data sources 10. Refer to comprehensive reports for local surveys in the applied to the PEN studies in each country. Part CD-ROM attached to the report. 20 PEN STUDY IN THE LOWER MEKONG SUB-REGION 1 Study Framework and Methodology INTRODUCTION for faster growth and poverty reduction. Greater wealth will then enable a society to restore envi- There is broad agreement in the development ronmental quality later. Empirical evidence to literature that there is a close link between support the existence of these dynamics and thus poverty—and human welfare more generally— the benefits of such strategies has been mixed and environmental quality. The existence of (Ansuategi and others 1998; Dasgupta and others such a link is intuitively credible, given that 2002; Perman and Stern 2003). degradation of natural systems adversely affects In fact, much of the work on the poverty- the livelihoods of people who depend on these environment nexus shows that the dual develop- resources for agriculture, forestry, and �sheries. ment objectives of maintaining environmental Similarly, environmental degradation in the form quality and increasing living standards do not of air, land, and water pollution is often higher in have to be contradictory, even at earlier stages of marginal settlements in undesirable areas. These tend to be areas with a high concentration of development. The full costs of environmental poor residents. Contaminated air and water leads degradation include loss of productivity and to health problems, which in turn reduce the increased demands on social systems due to the ability to pursue income-earning or livelihood- environmental health burden. A potential long- enhancing activities. term stream of future bene�ts from sustainably The strength and direction of the relationships managed natural resources remains unrealized between poverty and environmental conditions when such resources are irreparably damaged. continue to be ill-de�ned. It has been dif�cult to Typically, poor people, especially in rural areas, con�rm the existence and nature of these links in depend directly on the natural resource base and empirical studies. In a review of forestry and therefore carry the largest burden from pollution poverty, Angelsen and Wunder (2003) call such and degradation of natural resources (Ekbom relationships a “controversial link.� At a macro and Bojo 1999). Deterioration or destruction of level, some researchers believe that in early stages natural resources tends to be triggered by demand of development some environmental degradation or pro�t-seeking from wealthier people through is unavoidable—a necessary trade-off that will be large-scale logging, depletion of water resources, reversed with increasing wealth. The implication soil degradation, land degradation due to over- is that countries should accept increased pollution grazing, or pollution from agro-chemicals. The and environmental degradation as the price to pay poor often fail to obtain a fair share of the bene- PEN STUDY IN THE LOWER MEKONG SUB-REGION 21 STUDY FRAMEWORK AND METHODOLOGY �ts of rapid industrialization and natural resource environment analysis—especially as these relate exploitation. Where the poor contribute to land to geographic aspects. Some of the common degradation, it is often due to population pres- elements in the PEN methodology used in the sure that forces overexploitation of a shrinking national and regional analyses are highlighted in resource base, frequently in already marginal the �nal section. The country chapters contain areas. In many areas, rising populations are due to additional information on data collection and voluntary or prescribed in-migration rather than implementation issues that are specific to each natural growth of population. Such situations of the case studies. often coincide with a breakdown of traditional resource management systems that helped main- Poverty and Environment Concepts tain an environmental balance. While develop- ment in the form of higher living standards in Poverty the long run can help reduce direct pressure on Poverty is a multi-dimensional concept and there natural systems, the short-run objective is to �nd is continuing debate about how poverty should new regimes that regulate access to resources for be de�ned and measured. A comprehensive sustainable use or effective protection. Such sys- overview of poverty concepts is provided in the tems need to be developed by or evolve in close World Bank’s World Development Report on collaboration with local communities. It is the Attacking Poverty (2001a). This report stresses quality of this response that determines whether that poverty is not just a matter of material depri- population growth and economic dynamics lead vation, but also relates to broader notions of risk, to further degradation or whether the outcome vulnerability, social inclusion and opportunities. combines improved living standards with better Poverty encompasses all forms of deprivation natural resource management. that prevent a person from achieving their aspi- In the context of the World Bank poverty rations. This wider range of poverty dimensions reduction and environmental strategies, a main also has policy implications as it acknowledges objective is therefore to ensure that poverty allevi- the mutually bene�cial interactions from inter- ation activities should not damage the environ- ventions in a broad set of welfare dimensions. In ment on which poor people depend, as this would practice, it is the nature of the problem that is essentially trade off gains in one area for losses in studied that will determine which welfare mea- another (World Bank 2005a). More pro-actively, sure is most appropriate. This section will briefly poverty reduction strategies should also aim at review a number of poverty concepts that are rel- improving environmental conditions that can help evant for applied policy research on poverty-envi- reduce poverty (Bojö and others 2001; 2004). In ronment linkages. this context, the Southeast Asia PEN study aims to The concept of poverty that most closely contribute to the debate by providing empirical matches most people’s notions is absolute evidence on relevant linkages in a variety of envi- poverty—the ability of people to satisfy their ronmental sub-sectors and geographic settings. basic needs in terms of nutritional intake, shelter, The overall goal is to help identify strategies and basic amenities, and the ability to take advantage interventions that improve people’s lives and of opportunities such as education. Living stan- maintain environmental quality. dards are determined as the consumption expen- This chapter is composed of three main sec- diture equivalent of market and non-market tions. First, it provides a brief review of poverty goods that are required to satisfy these basic needs. concepts and main environmental issues in South- This level is called the poverty line. Poor house- east Asia. The second section describes conceptual holds are those whose consumption falls under and methodological issues pertinent to poverty- the level that is considered suf�cient to maintain 22 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY basic needs. The degree to which the household is more easily designed to match a particular issue falling short is called the poverty gap, the sum of under study. The problem with service access, which for the entire population is an estimate of asset, or wealth indexes is that there is no the monetary resources required to lift everyone universal methodology for creating them. For out of poverty. A further commonly used mea- instance, individual components of the index sure, the squared poverty gap or poverty severity, could be weighted by perceived importance or gives extra weight to people who are far below the simply added up. This can create a high degree poverty line. This measure therefore better incor- of subjectivity and limits interpretation and porates inequality among the poor and within comparability. the population as a whole. Consumption ex- Broader concepts of poverty have been pro- pressed in monetary terms provides perhaps the posed to more closely address the issue of most objective way of comparing living standards, opportunity. Rather than looking at income or and methods for data collection and analysis are consumption itself, the sustainable livelihoods well developed (Ravallion 1994; Deaton 1997). approach, for instance, focuses on the types of Most comprehensive poverty assessments resources that determine a household’s standard employ this approach, often in combination with of living. These are summarized by �ve types of other methods. Alternative indicators can provide capital: natural, social, human, physical, and more direct information on welfare outcomes. �nancial.1 In this framework, the environment is For instance, nutritional status as indicated by seen as the sum of the resource stock accessible stunting or wasting is used in Demographic to households; that is, its natural capital. Other and Health Surveys and in food security assess- forms of capital come into play by substituting ments. Similarly, one could define education or modifying the relationship with the environ- poverty on the basis of literacy rates or test ment. These livelihood concepts have the advan- score achievements. tage that they explicitly consider access to assets In contrast to absolute measures of poverty, such as resource stocks in determining poverty relative poverty focuses on inequality across a pop- levels. But they are often more dif�cult to oper- ulation in the distribution of consumption, assets, ationalize than standard poverty measures, since education, or any other welfare measure. This the components do not always correspond to is relevant for targeting interventions to the pop- easily measurable and comparable indicators. As ulation groups at the bottom of the welfare distri- such, they are most useful as a conceptual model bution. But it is also relevant in its own right, that can guide more informal and qualitative because severe inequality is considered detrimen- poverty analyses, like those in several of the PEN tal to the overall development prospects of a coun- case studies. try (World Bank 2005). A further aspect of well-being is vulnerability, Where detailed poverty data are unavailable, which captures the risk of being poor or falling policy researchers sometimes use a simple asset- into poverty at some point in the future. Vul- based indicator of human welfare. This may be nerability strongly affects choices and options, an index of ownership or availability of certain for instance with respect to food production or goods or services such as a vehicle, good quality investment in productive activities. For example, housing, agricultural land holdings, or connec- farmers close to the subsistence minimum are tion to utilities (Filmer and Pritchett 2001). In rarely able to risk adoption of unproven but the absence of detailed consumption data, these potentially superior new varieties or techniques. asset-based indicators can reflect the nature of Coping capacity in the presence of fluctuating the problems faced by the communities, espe- incomes also affects the subjective perception of cially in rural areas. An asset index can also be a person’s well-being. The ability to deal with PEN STUDY IN THE LOWER MEKONG SUB-REGION 23 STUDY FRAMEWORK AND METHODOLOGY sudden shocks and the existence of safety nets tors, but covers all households in the country. By provide a feeling of security, which influences combining these two data sources using statistical people’s actions. These concepts are highly rele- techniques, reliable estimates of poverty and vant to the analysis of human welfare, but in inequality can be derived for much smaller practice they are dif�cult to measure systemati- administrative units than the province—such as cally. Likewise, vulnerability is also relevant when districts or even communes. Such “poverty applied to issues such as health. Poorer persons maps� have become an invaluable tool for visu- with less access to healthcare will be more vul- alization and analysis of poverty patterns and for nerable to environmental and other health prob- targeting of interventions. lems than people who have access to safety nets The �rst phase of the PEN study made use of or other coping mechanisms. poverty maps that were created for Vietnam The various PEN studies described in the fol- (Inter-Ministerial Poverty Mapping Task Force lowing chapters employ poverty concepts that 2003), Cambodia (MoP and WFP 2002) and are most appropriate in the speci�c context. The Lao PDR (WFP 2001). The more recent analy- regional analysis in Phase I largely followed the ses took advantage of revised poverty maps and poverty de�nitions used in national poverty assess- new data. These are based on large sample national ments. The core indicator is typically the propor- household surveys and the latest available census tion of individuals or households that fall short of information.3 Both surveys and censuses are com- consumption expenditures that support a food plex and expensive undertakings. National statis- intake of more the 2,000 calories per person, plus tical and census of�ces therefore carry out such minimum non-food expenditures. This indicator enumerations only every few years (10 years in is called the poverty headcount index and the the case of censuses). Some of the data available threshold level, the poverty line, is de�ned some- for the three countries are therefore seven to what differently in each country.2 The national eight years old. While the broad patterns are analyses in the second phase of the PEN study unlikely to change, much of the region is expe- used the same concepts where data were available, riencing rapid economic growth that triggers but augmented the set of indicators with speci�c dynamic migration patterns and poverty reduc- outcome measures such as health status. A key tion in some regions. Wherever possible, the component of the spatial analysis of PEN issues at PEN study therefore augmented available sta- the regional and national level were the available tistical and census data sets with other infor- poverty maps that show welfare indicators at a �ne mation from smaller surveys, district records, level of geographic disaggregation. or field data collection. A typical living standard survey collects com- Table 1 summarizes different sources of socio- prehensive information about a relatively small, economic data that may be suitable for poverty- but nationally representative sample of residents— environment studies. Each has advantages and usually several thousand households. This allows disadvantages, but all are useful for policy analy- estimation of statistically reliable poverty rates and sis, although researchers and decision makers other indicators for a small number of regions or, need to be aware of the limitations in using any at best, at the province level. World Bank of these data sources. researchers and others have recently developed a methodology to combine the rich detail of wel- Environment fare surveys with information from population The three countries included in this study face the and household censuses (Hentschel and others challenge of pursuing economic growth to gener- 2000). A census collects a limited set of indica- ate adequate living standards for a rapidly grow- 24 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY TABLE 1.1 Socioeconomic Data Sources Data source Pros Cons Population and Can provide reliable data for very Typically only available every 10 years; household census small areas; sometimes includes small number of basic indicators; information on household assets often, only aggregate data are that reflects welfare status. published (province or district level); often a long lag between census-taking and dissemination. Comprehensive, Often the main source of reliable Sample sizes are too small to yield multi-topic data on poverty and other aspects geographically detailed indicators. household survey of household welfare; usually col- (LECS, CSES, lects a very comprehensive set of VNLSS, DHS) individual, household, and com- munity-level indicators. Short, sector-speci�c Fairly low cost and quick turn- Often not designed to meet statistical surveys (e.g., village around; can be used for monitor- requirements (sampling, etc.); small survey) ing through repeated surveys sample sizes; focused on limited (e.g., annual). number of issues, making more comprehensive analysis dif�cult; do not typically cover large areas or entire country. Facility and service Provides information on the supply- Frequently include quantitative provider surveys side of service delivery that can indicators only (e.g., number of otherwise be dif�cult to monitor doctors), which do not necessarily (e.g., health, education); can be reflect quality of service; user combined with user surveys at surveys at facilities leave out peo- facility. ple who do not have access for geographic, social, or economic reasons. Participatory data Involves citizens; low cost; provides Dif�cult to maintain data quality collection and rich contextual information. and objectivity; not possible to rapid appraisal extrapolate to other parts of the (focus group country or to country as a whole. interviews) Transactions data Information base is always up-to- Limited number of indicators; com- (e.g., birth/ date; serves many administrative plex and expensive to maintain death register, functions. electronically. cadastre, etc.) ing population. This process requires a large Environmental Health and expansion of public services—such as water sup- Quality of Life ply and health care—and puts increasing strains Environmental health, according to the World on natural resources. The PEN study distinguishes Health Organization, broadly between environmental conditions that directly affect human health and quality of life “comprises those aspects of human health, versus those related to the maintenance of natural including quality of life, that are determined by resources that support fundamental biological sys- physical, chemical, biological, social, and psy- tems and rural livelihoods. chosocial factors in the environment. It also PEN STUDY IN THE LOWER MEKONG SUB-REGION 25 STUDY FRAMEWORK AND METHODOLOGY of disease, and contaminates drinking water; if FIGURE 1.1 Historical and Projected Population Growth in Lao PDR, Cambodia untreated sewage is released into streams, it can and Vietnam also affect natural systems. The impact on infants and small children is particularly severe, but ill- 120 ness among adults also contributes to a large 110 social and economic burden due to infectious 100 diseases from inadequate water supply, sanita- Vietnam 90 tion, and hygiene. Although many countries have Population (million) 80 made progress in extending safe drinking water 70 supplies to an increasing share of their popula- 60 tion, coverage in many rural areas is still low. In part, this is because the population is widely scat- 50 tered and the cost of supplying infrastructure is 40 high, although other issues such as a lack of voice 30 Cambodia among rural populations are equally important 20 Lao PDR factors to consider. In towns and cities, problems 10 are often localized in low-income neighborhoods, 0 where residents may not be able to afford con- nection fees. 50 60 70 80 90 00 10 20 30 40 50 19 19 19 19 19 20 20 20 20 20 20 Water supply is also threatened by direct indus- Source: UN World Population Prospects 2004, medium variant, UN Popu- trial and municipal pollution of freshwater lation Division, New York. resources. Facing limited regulations and enforce- ment, many factories release toxic waste directly into streams and rivers, a problem particularly refers to the theory and practice of assessing, cor- acute in the densely populated booming industrial recting, controlling, and preventing those factors in the environment that can potentially affect centers in Vietnam. Such pollution originates adversely the health of present and future gener- from large, formal industrial plants, but also from ations.� (World Health Organization; draft smaller cottage industries, many of which are definition developed at a WHO consultation located in smaller towns or villages. Effluent pol- in Sofia, Bulgaria, 1993). lution problems are not limited to manufacturing. In the context of the PEN study, the most In the agricultural sector, farmers have greatly important aspects of environmental health are increased production, in part with the help of the effects of inadequate water supply and sani- agro-chemicals such as fertilizers and pesticides. tation in rural and urban areas; air and water pol- The use of agricultural inputs is not strongly con- lution from industrial activities in cities, towns, trolled; inappropriate application of chemicals, and villages; indoor air pollution, especially in some of which are banned elsewhere but imported the upland areas of Lao PDR and Cambodia; illegally, pollute land and water, and pose a signif- and pesticide use in agriculture. icant health risk to farm workers who apply them. Water-related problems enter the poverty- Air pollution is largely a problem in urban environment nexus through basic public health areas where industrial �rms and power plants issues. Lack of access to clean water among the often operate with higher pollution intensities poor is one of the primary health problems, (pollution per unit of output) than modern largely through coliform bacteria in drinking plants that employ state-of-the-art technology. water that cause diarrhea. Lack of proper sanita- Accelerating motorization and the popularity of tion affects hygiene conditions, increases the risk motorcycles with two-stroke engines also leads 26 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY to high concentrations of particulates that are bene�t from logging beyond occasional employ- the main cause of pollution-related respiratory ment opportunities, local economic stimulus from problems. Whether air pollution affects the poor timber operations, and possible pro-poor govern- more severely than the non-poor depends on the ment programs funded through logging proceeds. type and location of pollution sources. While The habitat destruction from large-scale logging, vehicle emissions affect everyone in a city equally, however, can damage the natural resource base industrial plants tend to be located in undesir- on which poor farmers, hunters, and gatherers able parts of a city that also have low-income depend. neighborhoods. Low-skilled industrial workers Logging operations are often poorly regulated, also tend to be located in more urban areas and and illegal logging of high-value timber is a severe may face a high occupational health risk due to problem in the region. Much of the demand pollution. Other types of pollution such as comes from neighboring countries, especially untreated effluents are also a bigger problem in after the logging ban in China in 1998 led traders poor neighborhoods. there to look elsewhere in the region for timber In contrast to outdoor air pollution, indoor supplies. Overall, deforestation processes in the air pollution is a signi�cant health problem in region reflect the problems globally. The drivers rural areas, where traditional cooking methods include a range of mutually reinforcing factors with wood or charcoal in poorly ventilated houses such as economic and institutional conditions, generate high particulate concentrations. Espe- national policies and enforcements, and proximate cially in the cooler upland regions of Vietnam and and remote influences that encourage agricultural Lao PDR, indoor air pollution is thought to have expansion, wood extraction, and infrastructure considerable effects on the health of women and extension (Geist and Lambin 2002). children and may be most severe in poor house- Remaining forest areas in Southeast Asia con- holds. Much of the evidence in this area is from tain an extraordinary degree of biodiversity, par- indirect health studies of the incidence of respi- ticularly in the remaining montane forest regions. ratory disease, since systematic pollution moni- One way for the poor to bene�t from this species toring programs and large-scale epidemiological richness is through non-timber forest products studies have been scarce. (NTFPs). NTFPs can be important for the poor as a livelihood strategy and safety net. A meta analysis of empirical studies worldwide showed Natural Resource Management that forest environmental income provides an Natural resource management issues have great average of 22 percent of household income, about relevance in the subregion. All three countries half of which is cash income (Vedeld and others contain forest areas with high economic and eco- 2004). But few such products have the potential logical value. About 40 percent of the land area of for sustainable scaling up of collection and com- Lao PDR, for instance, is covered by forests. Esti- mercialization, and therefore have limited poten- mates of the forest sector’s contribution to GDP tial for supporting socioeconomic advancement. vary between 4.5 percent—if only of�cial com- Consequently, NTFPs are often a symptom of mercial bene�ts are included—to 7 to 10 percent poverty rather than a solution. Yet, especially in if the contribution of forest resources to sub- Lao PDR, NTFP dependence by the rural poor sistence households and illegal logging opera- may be higher than in many other countries. tions are added (World Bank 2001). Commercial Some forest products—such as sticlac, car- timber extraction is capital- and skill-intensive, damom, bamboo, benzoin, rattan, and pine requires large economies of scale, and relies on resin—are legally harvested NTFPs with com- specialized marketing channels.4 The poor rarely mercialization potential (WFP 2004). A more PEN STUDY IN THE LOWER MEKONG SUB-REGION 27 STUDY FRAMEWORK AND METHODOLOGY sinister form of NTFP utilization is the illegal land, but can also contribute to large-scale removal trade in threatened plant and animal species. of forest cover and negative impacts on other Trade in wildlife that originates in or passes bene�ciaries of forest resources and on environ- through Vietnam, for instance, is well docu- mental functions. Due to its high labor require- mented (Song 2003; World Bank 2005c). Most ments and low productivity, shifting cultivation of the demand is from urban areas in the region is often done by poor farmers. It is a sustainable and from China. The sources are mostly the pro- use of forest areas when suf�cient fallow periods tected areas of Vietnam, Lao PDR, Cambodia, follow land conversion (Fox and others 2000). and Myanmar, where the last remaining popula- But various factors may lead to reduction of fal- tions of many threatened species live. With rising low periods to unsustainable levels. Population living standards in the region, demand for wild growth—often through in-migration rather than animal and plant products will increase further. natural growth—increases the number of farm- This is likely to lead to the extinction of many ers dependent on a given land area, and large-scale unique species. This kind of trade is dif�cult to logging or setting aside of forest areas for other stop given the scarce resources of wildlife agencies uses further contribute to a shrinking resource and the vast pro�t margins available to traders. base available to the poor for forest-based agricul- In the upland areas of Lao PDR and Vietnam ture. With shorter rotation cycles, fragile soils— and the lower lying regions of Cambodia, forests especially in steep upland areas—lose their ability provide vital ecological services to rural commu- to support traditional agricultural practices. nities as well as to downstream towns and cities. Where fallow periods are too short for systems to Ecological services such as regulation of hydro- recover, sustainable shifting cultivation turns logical functions, carbon storage, and preserva- into unsustainable slash-and-burn agriculture that tion of genetic diversity are forest functions for may lead to complete depletion of soil fertility which compensation or bene�t-sharing schemes (Casetella and Quang 2002; Palm and others could be designed. Direct payment schemes for 2005). This process can be reversed, but requires such services have been tested in various parts of considerable investment in restoration of plant the world, but it is no trivial matter to implement nutrients, improved land management, or plant- ecological service payment systems that bene�t ing of trees. Such long-term investments will the poor (Pagiola and others 2005; Landell-Mills make sense to farmers only if they have suf�ciently and Porras 2002). Still, successful experiences secure tenure to their land. with payments for environmental service schemes Intensified permanent agriculture—using especially in Latin America may hold promise for modern farming techniques adapted to the ter- similar projects in Southeast Asia. Indirect bene- rain and climate—can be successful in providing �ts from ecological endowments such as tourism a livelihood base to farmers. But in areas of rapid are increasingly important in the economies of the population growth, rural strategies based on three countries. Future expansion of those rev- introduction of modern technologies alone are enues is threatened by large-scale deterioration unlikely to be successful in absorbing surplus labor of the most valuable regional ecosystems. without simultaneous expansion of cropping areas Perhaps the most important role that forests in previously forested regions. However, scaling play in providing livelihoods for the poor is as a up of technology transfer and extension services, land reserve that can be converted to agricultural combined with provision of education and other uses on a temporary (rotating or shifting cultiva- social services, infrastructure and effective regu- tion) or permanent basis (Angelsen and Wunder lation, can have important spillovers for job cre- 2003). For the poor, this may often be the most ation in the rural non-farm sector, where most economically bene�cial use of previously forested of the future rural employment growth needs to 28 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY BOX 1.1 Sources of Environmental Information Poverty-environment linkages are varied and complex. Consequently, the range of environmental indicators that need to be evaluated to study various aspects of the nexus is also very broad. Some relevant information can be extracted from socioeconomic data sources (reviewed earlier in this chapter). Examples are health outcomes that may have environmental causes or household behav- ior with respect to the use of natural resources. For many other aspects, however, information needs to be collected using specialized tools. Remote sensing—capturing images of features on the ground from space-based systems or airplanes—is one of the most useful tools for collecting natural resources data for large areas at regular intervals. Optical remote sensing systems such as Landsat or Spot have been invaluable in national or regional assessments of land cover, including forests, wetlands, and agricultural areas (see Figure XX). Some of these systems have been in operation for several decades, so archived images can provide historic views for change analysis. Remotely sensed data is usually the method of choice for forest cover inventories and monitoring because they provide data collected using a consistent methodology for large areas. This circumvents the problem of subjective evaluation in compiling ground-reported information from local observers. Other remote sensing platforms generate more specialized information, such as topography (elevation and slope) from radar- or laser-based systems, climatic and atmospheric data, and even particulate pollution concentration. Lately, commercially available high resolution (60cm–1m) satellite data have become available. These products can be interpreted visually, reducing the need for complex image analysis proce- dures that are required for the coarser resolution data (15m–1km). However, the cost of covering large areas and the time required for visual interpretation mean that high-resolution images have so far not been used extensively in natural resources applications. Not all environmental information can be captured remotely. Obtaining high-quality informa- tion on water and air pollution, flood levels, rainfall, or temperature requires a network of ground- based monitoring stations. Those recording stations can be permanent or mobile. Best known are weather stations that report continuously and yield long-term climatic averages. Unfortunately, for many environmental indicators permanent monitoring networks are very sparse and recording periods are often interrupted. An example is the scarcity of river gauge data that severely hampers the analysis of flooding that might be linked to land use changes upstream. With prices falling for many kinds of monitoring devices, targeted data collection campaigns are increasingly feasible. For instance, Dasgupta and others (2004) deployed air pollution monitoring devices in a large number of households for 24-hour periods to study indoor air pollution exposure in Bangladesh. Such data also add context and substantive evidence to the analysis of household survey data on health out- comes, especially if it includes medical information such as lung capacity measurement. The bene�ts from combining different, multidisciplinary data sets in analyzing human-environment interactions are clearly very large. Collection of environmental data is most useful if it is part of a long-term information manage- ment strategy and if the outputs are widely disseminated to policy makers, researchers, and the pub- lic. Since ecological processes do not follow political boundaries, comprehensive regional approaches should play an essential role in improving natural resource management. Examples are the Brazilian initiatives on rainforest monitoring, which are based on remote sensing and include public disclosure of real-time information on forest cover changes. An institutional template for such efforts in South- east Asia is the Mekong River Commission, but any effort focused on terrestrial ecosystems should also include the parts of mainland Southeast Asia that are not part of the Mekong watershed. occur (Müller and Zeller 2002; Cuong and usually concentrated in urban centers and within others 2003). parts of larger cities, so the burden of air, land, and water pollution will also vary by geographic location. Countries also often show large varia- GEOGRAPHIC SCALE AND SCOPE tions in living standards, even when adjusting for OF THE PEN ANALYSES cost-of-living differences. Apart from this geo- Natural resources are distributed unequally graphic heterogeneity, the three countries also dis- within a country. Similarly, industrial activity is play great complexity in the nature and severity of PEN STUDY IN THE LOWER MEKONG SUB-REGION 29 STUDY FRAMEWORK AND METHODOLOGY FIGURE 1.2 Land Cover in Northern Lao PDR Source: NASA Landsat Geocover 2000. Vegetation in green, bare soil/agriculture in red/purple. environmental and social problems. While there Multi-Scale Analysis— are many similarities, such as the natural resource from Regional to Local management issues affecting the poor in the uplands of Vietnam and Lao PDR, each country The study applies a multi-scale approach com- also has its speci�c set of problems. The booming bining analysis of aggregate data at national cities of Vietnam face pollution problems that and regional levels with very detailed case stud- have not reached the much smaller urban centers ies that reach down to the village and individ- of Lao PDR. Indoor air pollution is a problem in ual household level. The �rst phase of the project the uplands of these two countries, but less so in took a regional perspective by analyzing similar Cambodia where lower elevations and different variables for the three countries, typically at the cultural preferences mean that most cooking is level of provinces (Dasgupta and others 2005). done outside. The PEN study accounts for these What emerged was that for many environmen- complexities by considering a very broad (though tal and welfare-related questions, the province by no means exhaustive) range of poverty-envi- level provides insufficient detail to yield use- ronment issues. It also accounts for geographic ful answers. Within-province variation is often heterogeneity explicitly by analyzing problems at greater than that between provinces. In the sec- multiple scales and by employing spatially refer- ond phase, national-level analysis, data were enced data whenever possible. collected and analyzed for all parts of the coun- 30 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY try, and therefore relied as much as possible on district or subdistrict level. Similarly, forestry district-level information. Finally, core contri- information is sometimes collected using satel- butions of the PEN study are the topic- and lite remote sensing with a resolution of 500m place-specific case studies that support the or 1km. This is too coarse to detect deforestation local analysis. Here, information is typically due to shifting cultivation or selective logging. collected at the subdistrict, village, and house- Even where reliable data by administrative units hold level. are available, the analyst often faces challenges Any speci�c poverty-environment issue should in interpreting the information (see box 1). be analyzed at the scale that is most relevant to the characteristics of the topic. For instance, air Use of Spatial Data to Account for pollution from cement mills may only be dan- Geographic Heterogeneity gerous in the urban area where the facility is located, while acid rain may affect natural sys- In addition to looking at PEN issues across multi- tems far away from the pollution source. In the ple scales, the PEN study relied as much as possi- Cau River Basin case study in Vietnam, pollu- ble on geographically referenced information. It tion from riverside industrial facilities has an consists of statistical information and derived indi- impact downstream, so a focus only on the imme- cators that can be linked, for instance, to a digital diate pollution source would miss the wider map of administrative units, as well as inherently context of the nexus. Other applications—for spatial data such as digital forest cover, slope, or instance, using province-level poverty data— elevation maps. These maps are used to display the provide many insights into the spatial distribu- geographic distribution of the data, combine or tion of welfare levels when looking at the overlay different maps to generate “spatial cross- country as a whole. Investigation of specific tabulations,� and to extract new variables that are regional linkages between human welfare and used in descriptive and statistical analysis. environmental processes, however, requires For instance, among the most dominant geo- “zooming in� to the district level or below in graphic patterns in the subregion are large dif- order to capture variations in patterns. For ferences between the lowland and upland areas, instance, most behavioral aspects of poverty- especially in Lao PDR and Vietnam. More than environment interactions need to be assessed at 60 percent of the Lao PDR land area is above the household level whenever possible. An 500 meters, and more than 20 percent above example is hygiene behavior, which often varies 1,000 meters. For Vietnam, the corresponding �g- between economic or ethnic groups and has ures are 30 percent and 10 percent. Lowland areas strong relations to health outcomes in areas of along the major river systems in the region—the poor access to water and sanitation. However, Mekong and the Red River—were traditionally when looking at water supply issues, the most more suitable for agriculture and have long sus- suitable level of analysis may be the commu- tained higher population densities than the often nity or village as the basis for provision of steeply sloped upland areas. Higher elevation networked services, or even the district where areas, in contrast, contain much of the remain- infrastructure investment targeting decisions ing forest cover and a high degree of biodiversity. are made. The nature of the poverty-environment nexus is In practice, a major barrier to multi-scale therefore very different in the two areas. With the analysis is the availability of data of sufficient exception of the plantation agriculture in the Cen- quality. Apart from census information, most tral Highlands in Vietnam, most intensive com- socioeconomic indicators are derived from sam- mercial agriculture is in the lowlands, where issues ple surveys, which are usually not reliable at the such as overuse of pesticides are thus most severe. PEN STUDY IN THE LOWER MEKONG SUB-REGION 31 STUDY FRAMEWORK AND METHODOLOGY BOX 1.2 Working with Data Recorded by Administrative Units Many socioeconomic indicators and some environmental data are available for administrative units such as districts or provinces. Analyzing such data geographically or statistically often poses major challenges. The most common of these is when data are only available at fairly coarse levels of aggregation. Data at the province or district level may hide considerable heterogeneity that is visible at the subdistrict level (see Fig XX below). For instance, relatively rich provinces might contain pockets of poverty that require urgent policy intervention. Two related problems that are frequently encountered are that the units do not match the phenomenon being studied and that boundary changes over time make it difficult to perform change analysis. The first of these problems can occur in poverty-environment studies when administrative units were drawn with no consideration of underlying geographic features such as river basins, areas with rural versus urban characteristics, or the distribution of ethnic groups. Depending on how these reporting units are drawn, the summary values of socioeconomic or bio- physical indicators may change signi�cantly. In principle, just by re-aggregating existing data, one may �nd completely different results in correlation studies. The larger the units, the bigger this problem will be. A related issue is that different levels of administration may have very different influence on policies relevant to the topic studied. An example is Cambodia, where elections have been held only at national and commune level. The intermediate province and district levels there- fore have much less relevance for policy implementation. Consequently, the PEN study found that research at the district level in Lao PDR provides useful insights into poverty-environment inter- actions, while such patterns are nonexistent at the province level. For Vietnam, both levels are useful, while for Cambodia it is necessary to go to the subdistrict level to analyze these topics. The second problem occurs when boundaries change signi�cantly. This tends to happen before each major data collection effort, such as a census. It has also been observed as a consequence of decentralization, where units might be created largely as a way to increase the number of political appointees that might bene�t from newfound decision-making power (Fitrani and others 2005). Analysis of possible causal relationships requires observations of changes over time, where changes in one variable during a certain period of time have an impact on changes in another variable dur- ing a subsequent period. But often it is impossible to reconcile data sets collected in different peri- ods. To address both problems, an important task for data collection agencies is to maintain and disseminate information at the lowest possible level of aggregation that does not compromise data con�dentiality concerns—for example, at the enumeration area level, which will include a few hun- dred households. This allows for aggregation of indicators to any set of units that are appropriate for the problem being studied. Increasingly, census and statistical agencies in many countries are moving in this direction. Similarly, severe industrial pollution is mostly exceptions, poverty is most severe in the upland concentrated in the densely populated urban areas, where the districts with the highest propor- areas of Hanoi and Ho Chi Minh City. In the tion of poor are located. These areas, however, uplands, on the other hand, forest-related issues have low population densities, so the absolute are very signi�cant. This includes maintenance number of poor is relatively small. Lowland areas of ecological functions, the use of non-timber have seen the fastest decline in poverty rates. But forest products, and intensi�cation of tradition- their much higher population size means that they ally sustainable forms of mountain agriculture continue to contain by far the largest share of the such as shifting cultivation. poor, especially in Vietnam. Fig. 4 The contrast between the natural resource The upland-lowland dichotomy is also relevant management issues in the uplands and pollution- in the context of public infrastructure provision. related environmental problems in the lowlands Provision of networked infrastructure such as is mirrored by a distinct difference in the nature electricity and water, but also roads, is much more of poverty between the two regions. With very few costly in sparsely settled rural areas. In the case of 32 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY FIGURE 1.3 Poverty Rates for Provinces, Districts, and Communes Data source: MoP and WFP (2002) the upland regions, the terrain also contributes to PDR, the sector-speci�c studies were very closely high costs in relation to the size of the popula- coordinated and implemented by local expert tion served, and local governments may have less teams that interacted frequently among them- capacity for effective management of investment selves and with a group of advisers. Consequently, programs. In more densely settled urban or high- all �ve case studies followed a very similar path. In intensity agricultural areas, in contrast, there are Cambodia, the comprehensive natural resource large economies of scale in service provision, which management case study was implemented by a is reflected in much higher coverage rates. local policy research group, and in Vietnam sec- toral policy research was conducted by different teams of national and international experts. Analytical Methods Used in the PEN Study Tight Linking of National/ Local Analysis The PEN analyses in the three countries followed slightly different approaches, since the nature of For all countries, the sector-speci�c, local studies speci�c issues and data availability vary to some are embedded in a comprehensive analysis at the extent. However, for all three countries there is an national level. This analysis relied on a variety of overall unifying framework anchored in a tight information sources, including published reports linking of a national analysis using secondary data and statistical data sets. For each country, the with locally relevant in-depth studies. In Lao national population and household census, and to PEN STUDY IN THE LOWER MEKONG SUB-REGION 33 STUDY FRAMEWORK AND METHODOLOGY FIGURE 1.4 Relation Between Poverty and Elevation in Vietnam a lesser extent the agricultural and economic cen- welfare indicators. While correlation does not sus, provided a core set of indicators at all levels necessarily imply causality, the resulting infer- of the administrative hierarchy. The census yields ences provide some indication of possible rela- demographic information, as well as some indica- tionships, which in turn suggest appropriate tion of access to water and sanitation and available interventions. More formal testing of causal assets. As described previously, each country also links is not possible in most cases, because most has implemented comprehensive household sur- of the required information is available for only veys with the objective of collecting detailed infor- one time period. Fundamental questions—such mation on general living conditions or on health as whether poor people are mostly victims of status for a large sample of the population. While environmental degradation, or whether poverty these samples are not large enough to yield reliable itself causes a decline in the natural resource data at the district level, analysis of the individual- base—cannot be answered with statistical cer- level micro data provides valuable insights, espe- tainty without fairly extensive time series data. cially on environmental health issues. Wherever Furthermore, in attempting to identify causal possible, indicators at various levels of aggregation links, there is also a problem of intervening vari- were presented in map form for exploratory visu- ables that affect the poverty-environment inter- alization and presentation of key results. actions. For instance, children in poor Statistical analysis of secondary data reveals households are more likely to suffer from water- initial correlations between environmental and borne diseases, but there is a signi�cant influence 34 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY from factors such as the education of the – Estimates of the number of poor and non- mother, which can be both a cause and conse- poor affected by environmental issues. quence of poverty. For these reasons, the studies presented in this report do not claim to show sta- For Lao PDR, three studies relate to natural tistically sound proof of speci�c relationships. resource management issues: use of non-timber forest products by upland communities; investi- Rather they are designed to provide solid empir- gation of the link between road access, poverty, ical evidence of relevant interactions and corre- and environmental quality; and the impact of lations that point to an urgent need for policy unexploded ordnance on poverty and natural intervention. resource management. Two closely related stud- In the first phase of the PEN study, the ies, one in urban centers and the other in rural national-level analysis assigned districts or prov- villages, focused on access among the poor to inces into four categories in each of a number water supply and sanitation and the sustainability of dimensions of the poverty and environment of water and sanitation services in poor villages. nexus: poverty incidence, deforestation rate, steep- In each of these, a subset of districts was chosen ness of slope as an indicator of soil vulnerability, based on criteria that ensured variations in condi- wood/charcoal use, unsafe water source, and tions across the sample. For each selected district, prevalence of childhood diarrhea. The compiled the study teams assembled a detailed statistical information was presented as a color-coded table profile by combining nationally available indi- that clearly highlights coincidence of high or low cators with information collected locally from rankings (Fig. XX). district of�cials. Further sampling proceeded by selecting villages within each district and house- Selection of Case Studies holds within each village. The selection aimed The national analysis in the �rst phase was exten- at achieving a pre-defined stratification—for sively updated and expanded in the second phase. instance, based on distance from roads or degree This process helped re�ne the development of a of in-migration. This approach is not intended to matrix of PEN issues and focused the identi�ca- produce a scienti�cally accurate sample, as is the tion of key knowledge gaps to be considered case in a large comprehensive household survey. in the in-depth analyses. In discussions with Rather it supports more qualitative information national counterparts, the potential focal studies collection through community questionnaires were then narrowed down to a subset for each and interviews of focus groups and key stake- country, as discussed below. The geographical holders. The outcome of this process in each selection of these studies was guided by the fol- in-depth study is a contextually rich package lowing main criteria: of information that helps identify the complex interactions between human and natural systems – Poverty incidence, geographic distribution of in the focus regions. poverty, and poverty persistence (so-called The Cambodia study focuses on the status, “poverty traps�). trends, and impacts of access to natural resources – Government priority regions, provinces and among poor and non-poor households. The districts for poverty reduction, economic devel- study selected two provinces, each of which had opment, and environmental management. a protected area, different types of land access – Representation of the distinct socioeconomic, situations, and high migration and poverty rates. environmental, and renewable resource use Households in six villages were selected for a patterns in the countries (e.g., upland/lowland, structured interview. Additional information industrial concentration and “hot spots�, pes- was collected from key informants, such as gov- ticide intensity, etc.). ernment of�cials and focus group interviews. As PEN STUDY IN THE LOWER MEKONG SUB-REGION 35 STUDY FRAMEWORK AND METHODOLOGY FIGURE 1.5 Poverty and Environmental Problems in Lao PDR Provinces Poverty Deforest Sloped Wood/ Unsafe Child Province rate -ation land Charcoal Water Diarrhea Luangphrabang Huaphanh Oudomxay Savannakhet Champasack Xiengkhuang Phongsaly Saravane Khammuane Xayabouri Vientiane Municipality Vientiane Luangnamtha Bokeo Attapeu Borikhamxay Sekong Xaysomboon 1 – worst 2 3 4 – best Source: PEN I final report Source: PEN I �nal report in the Lao PDR case studies, poor and non-poor which life depends—is reflected in the in-depth households were identi�ed using a simple asset- study of the Cau River Basin. The Cau River based indicator approach. experiences signi�cant pollution from industrial The in-depth studies carried out in Vietnam and municipal sources. The study aimed at iden- present a more varied mix of PEN situations, tifying the distribution of the impacts and �nd- reflecting the more complex set of problems faced ing possible interventions for remediation. In in a country with a large population and rapidly contrast to most of the other studies, the geo- growing economy. This balancing act of creating graphic unit of analysis is the watershed, which a dynamic economy that contributes to poverty does not completely match the administrative reduction—while preserving natural systems on hierarchy for which most socioeconomic data 36 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY area available. The area of interest falls into two contamination of shallow dug wells, on which provinces, in which structured interviews with the poor rely disproportionately. The same sur- 700 households were conducted. This informa- vey also yields insights at the national level on tion was augmented by direct observation and public health concerns related to substandard interviews with decision makers and stakehold- water supply and sanitation, as well as possible ers. The poverty measure employed is based on effects of pesticides in agriculture. income data from the questionnaire, which facil- Environmental health implications of pesti- itates ranking of households by economic status. cide use are the topic of a case study in the Industry is also the focus of another in-depth Mekong Delta, which is the largest region of study that analyzed the spatial relation between intensive rice cultivation in the country located in poverty, industrial concentration, air and water a fragile ecosystem. A team of medical researchers pollution, and environmental health risk. This implemented a detailed survey of about 600 farm- analysis was accompanied by rapid local-level ers in 10 districts that collected socioeconomic surveys of a number of urban hot spots and craft information as well as medical tests for the expo- industry villages, highlighting that the pollution sure to toxic chemicals. The land policy case problem is not restricted to large industrial estates study, finally, combined an analysis of a land- but also affects life in smaller towns and even related database of 610 districts in Vietnam with villages. The comprehensive Vietnam National a survey in three provinces in the upland regions. Health Survey of 36,000 households provided Within these, districts with contrasting condi- additional insights on possible outcomes of expo- tions were selected. Asset-based poverty indi- sure to pollution such as respiratory illness and cators aided the household stratification for TABLE 1.2 Data Sources Used at Different Administrative Levels Nation Province District Commune Village Household Cambodia Census of Population 1998 Cambodia Socioeconomic Survey (CSES) 1997 Case studies Lao PDR Census of Population 1995 Lao Expenditure and Consumption Surveys (LECS) II-1997/98, III-2002/03 Case studies Vietnam Population and housing census 1999 Vietnam Living Standards Surveys (VNLSS 1998, 2002); Vietnam National Health Survey (VNHS) 20XX Nationwide land inventories INEST/MOST (CV) Case studies PEN STUDY IN THE LOWER MEKONG SUB-REGION 37 STUDY FRAMEWORK AND METHODOLOGY analysis of structured interviews and focus Ansuategi, A., E. B. Barbier, and C. A. Perrings. 1998. “The group discussions. environmental Kuznets curve.� In J. C. J. M. van den Bergh and M. W. Hofkes, eds. Theory and Implementa- In summary, the process of analysis and tion of Economic Models for Sustainable Development. consultation resulted in the following thematic Dordrecht, NL: Kluwer. stratification of in-depth studies of poverty- Bojö, J, K. Green, S. Kishore, S. Pilapityia, and R. Reddy. environment linkages in the three countries: (2004). “Environment in PRSPs and PRSCs.� World Bank Environment Department Paper No. 102. Wash- – PENs with a potentially high number of poor ington, DC: World Bank. being affected by renewable natural resource Bojö, Jan, Julia Bucknall, Kirk Hamilton, Nalin Kishor, Christiane Kraus, and Poonam Pillai. 2001. Poverty issues and environmental health concerns: Reduction Strategy Papers Source Book. (Chapter 11, rural water supply and sanitation in Lao PDR Environment). Washington, DC: World Bank. and Vietnam, NTFPs and road development Casetella, Jean-Christophe, and Dang Dinh Quang, eds. in Lao PDR, land use in Vietnam, and natural 2002. “Doi Moi in the Mountains. Land use changes resource management in Cambodia. and farmers’ livelihood strategies in Bac Kan Province, Vietnam.� Hanoi: The Agricultural Publishing House. – Development trends and emerging PENs in Chantirath, K. 1998. “Forestry resources and the underlying urban areas: small town water supply and san- causes of deforestation and forest degradation in Lao itation in Lao PDR, industrial pollution in P.D.R.� In IGES Forest Conservation Project. “A Step rapidly growing cities in Vietnam. toward Forest Conservation Strategy.� Interim Report. – Previously unexplored PEN issues where sig- www.iges.or.jp/en/fc/phase1/interim-contents.htm. Cord, L., H. Lopez, and J. Page. 2003. “When I Use a ni�cant knowledge gaps exist: UXOs in Lao Word . . . Pro-poor Growth and Poverty Reduction.� PDR, pesticides in Vietnam, indoor air pollu- Processed. Washington, DC: World Bank. tion in Vietnam. Cuong, Pham Manh, Manfred Zeller, Regina Birner, and – Geographic priorities and cross-sectoral regional Daniel Mueller. 2003. “Improvement of access to avail- issues: Son Cau River basin and pollution issues able agricultural technologies: did it cause land-use changes and poverty reduction in the North-western in craft villages in Vietnam. upland of Vietnam?� Paper presented at the Conference on International Agricultural Research for Development, Endnotes Göttingen, Germany, October 8–10. Dasgupta, S., B. Laplante, H. Wang, and D. Wheeler. 2002. 1. http://www.undp.org/pei/pdfs/PEIPhase1Summary- “Confronting the environmental Kuznets curve.� Jour- Paper2.pdf nal of Economic Perspectives 16: 147–168. 2. This information also forms the basis for estimation of Dasgupta, Susmita, Mainul Huq, M. Khaliquzzaman, Kiran cross-nationally comparable welfare measures, such as Pandey, and David Wheeler. 2004. “Indoor Air Quality the proportion of the population living on less than $1 or $2 a day. for Poor Families: New Evidence from Bangladesh.� Pol- 3. Vietnam National Living Standards Surveys (VNLSS icy Research Working Paper 3393, Development 1997/98); Cambodia Socioeconomic Survey (CSES Research Group. Washington, DC: World Bank. 1997); Lao PDR Expenditure and Consumption Survey Dasgupta, Susmita, Uwe Deichmann, Craig Meisner, and (LECS II 1997/1998; LECS III data from 2002/03 were David Wheeler. 2005. “ThePoverty-Environment also utilized in Chapter 3). Census enumerations in the Nexus in Cambodia, Lao PDR, and Vietnam.� World 2000 census round were carried out in Vietnam in 1999, Development 33 (4): 617–638. Cambodia in 1998, and Lao PDR in 1995. Data from the Deaton, A. 1997. The analysis of household surveys. Balti- 2005 census in Lao PDR were not yet available. more, Maryland: Johns Hopkins University Press. 4. See Angelsen and Wunder (2003) for a discussion of DFID, EC, UNDP, and World Bank. 2002. Linking Poverty poverty-forest linkages on which this section draws. Reduction and Environmental Management: Policy Chal- lenges and Opportunities. Washington, DC: World Bank. Duraiappah, Anantha K. 1998. “Poverty and Environmen- References tal Degradation: A Review and Analysis of the Nexus.� Angelsen, A., and S. Wunder. 2003. “Exploring the forest- World Development 26 (12): 2169–2179. poverty link: Key concepts, issues and research impli- Ekbom, Anders, and Jan Bojo. 1999. “Poverty and Environ- cations.� CIFOR Occasional Paper No.40. Bogor, ment: Evidence of Links and Integration into the Coun- Indonesia: Center for International Forestry Research. try Assistance Strategy Process.� Discussion Paper No. 4, 38 PEN STUDY IN THE LOWER MEKONG SUB-REGION STUDY FRAMEWORK AND METHODOLOGY Environment Group, Africa Region. Washington, DC: Perman, R., and D. I. Stern. 2003. “SP3Evidence from panel World Bank. unit root and cointegration tests that the environmental Filmer, D., and L. Pritchett. 2001. “Estimating wealth Kuznets curve does not exist.: Australian Journal of Agri- effects without expenditure data–or tears: An applica- cultural and Resource Economics 47 (3):325–347. tion of educational enrollment in states of India.� Ravallion, M. 1994. Poverty comparisons. Chur, CH: Har- Demography 38(1):115–132. wood Academic Publishers. Fitrani, F., B. Hofman and K. Kaiser (2005), Unity in diver- Song, Nguyen Van. 2003. “Wildlife Trading in Vietnam: sity? The creation of new local governments in a decen- Why it Flourishes.� EEPSEA Research Report 2003 - tralizing Indonesia, Bulletin of Indonesian Economic RR6. Singapore: Economy and Environment Program Studies, 41, 1:57–79. for Southeast Asia. (http://www.idrc.ca/uploads/ Fox, Jefferson, Dao Minh Truong, A. Terry Rambo, Nghiem user-S/10705050241SongRRlayout.doc, last accessed Phuong Tuy En, Le Trong Cuc, and Stephen Leisz. 2/13/06). 2000. “Shifting Cultivation: A New Old Paradigm for UNDP & EC. 2000. “Attacking poverty while improving Managing Tropical Forests.� BioScience 50 (6): 521–528. the environment: Towards win-win policy options.� Geist, Helmut J., and Eric F. Lambin. 2002. “Proximate Causes and Underlying Driving Forces of Tropical Poverty and Environment Initiative. New York: UNDP Deforestation.� BioScience 52 (2): 143–150. and European Commission. Hentschel, J., P. Lanjouw, J. Lanjouw, and J. Poggi. 2000. Vedeld, P., A. Angelsen, E. Sjaastad and G. Kobugabe-Berg. “Combining Household Data with Census Data to 2004. “Counting on the Environment: Forest Incomes Construct a Disaggregated Poverty Map: A Case Study and the Rural Poor.� Environment Department Paper of Ecuador.� World Bank Economic Review, 14 (1): No. 98. Washington, DC: World Bank. 147–166. World Food Programme (WFP). 2001. “Lao PDR Poverty Inter-Ministerial Poverty Mapping Task Force. 2003. Analysis.� Asian Development Bank, Lao National Sta- Poverty and inequality in Vietnam. Spatial patterns and tistics Center, United Nations World Food Programme. geographic determinants. Hanoi: Ministry of Agriculture Analysis and report prepared for the Asia and Paci�c and Rural Development, International Food Policy Forum on Poverty, Manila. Research Institute, Institute of Development Studies. World Food Programme (WFP). 2004. “Analysis of Vul- Landell-Mills, N., and I. Porras. 2002. Silver Bullet or Fool’s nerability to Food Insecurity District Level, Vulnerabil- Gold? A global review of markets for forest environmental ity Analysis and Mapping.� Vientianne: United Nations services and their impacts on the poor. London: Interna- World Food Programme. tional Institute for Environment and Development. World Bank. 2001. Lao PDR Production Forestry Strategy. MoP and WFP. 2002. Estimation of poverty rates at the com- 2 vols. Washington, DC: World Bank; SIDA; Ministry mune level in Cambodia. Phnom Penh: Ministry of Plan- of Foreign Affairs, Government of Finland. ning and UN World Food Program. World Bank. 2003. World Development Report 2003: Sus- Müller, D., and M. Zeller. 2002. “Land use dynamics in the tainable Development. Washington, DC: World Bank. central highlands of Vietnam: a spatial model combin- World Bank. 2005a. Environment strategy for the World ing village survey data with satellite imagery interpreta- tion.� Agricultural Economics 27: 333–354. Bank in the East Asia and Paci�c Region. Washington, Pagiola, S., A. Arcenas, and G. Platais. 2005. “Can Payment DC: World Bank. for Environmental Services Help Reduce Poverty? An World Bank. 2005b. World Developing Report 2005: Equity Exploration of the Issues and the Evidence to Date from and Development. Washington, DC: World Bank. Latin America.� World Development 33 (2): 237–253. World Bank. 2005c. “Going, going, gone . . . The illegal trade Palm, Cheryl A., Stephen A. Vosti, Pedro A. Sanchez, and in wildlife in East and Southeast Asia.� Environment and Polly J. Ericksen. 2005. Slash-and-burn agriculture. The Social Development East Asia and Paci�c Region Discus- search for alternatives. New York: Columbia University sion Paper. Washington, DC: World Bank. Press. PEN STUDY IN THE LOWER MEKONG SUB-REGION 39 II COUNTRY STUDY FINDINGS 2 Poverty and Environment in Cambodia This chapter examines poverty-environment Cambodia’s population is located in rural relationships in Cambodia through an analysis areas— 11 million out of a total population of of national data and case study �ndings. We 13 million—poverty remains largely a rural begin with an overview of the national context problem. Indeed, of the total population living and trends for poverty and natural resources, below the poverty line in Cambodia, more providing a synthesis of existing studies and infor- than 90 percent are located in rural areas, equal mation across Cambodia’s regions and provinces to more than 4 million people. (Map 2.1). Next, we conduct a range of new In addition to urban-rural differences in pov- analysis to identify factors associated with poverty, erty, poverty rates are quite uneven across Cam- and to investigate speci�c poverty-environment bodia’s regions and provinces (Map 2.2). The relationships pertaining to natural resource depen- percentage of the population living below the dence, drinking water sources, sanitation, nat- poverty line in coastal areas is 27 percent, but ural disasters, and mine/UXO contamination. this rises to 52 percent in the plateau/mountains To complement national analysis of poverty and region. Although poverty severity is highest in the natural resource relationships, we then present rural plateau/mountains region, this region is less a case study on poverty and access to natural populated than the plains and Tonle Sap regions, resources conducted in two provinces. Finally, which together account for about 75 percent of we provide a summary of key �ndings and dis- the population in poverty. cuss policy implications. Trend estimates across geographically compa- rable areas from CSES 1993–94 to 2004 indicate poverty has declined in all regions except for the POVERTY: CONTEXT AND TRENDS rural plateau/mountains. Overall, most provinces Poverty incidence in Cambodia fell from about made progress in reducing poverty over the 47 percent in 1993–94 to 35 percent in 2004, past decade. Poverty rates only increased in two according to �ndings from the 2004 Cambodia provinces—Kampong Thom and Kampong Socio-Economic Survey (CSES).1 However, Speu, though these increases are within the statis- rural poverty incidence (37.8 percent) remains tical margin of error. Poverty reduction was con- considerably higher than urban areas (17.6 per- siderably less on a percentage basis in the rural cent) (Table 2.1). Since about 85 percent of areas during this period than in the urban and PEN STUDY IN THE LOWER MEKONG SUB-REGION 43 POVERTY AND ENVIRONMENT IN CAMBODIA MAP 2.1 Regions in Cambodia (as applied in CSES Survey 2004) Source: World Bank 2005b. semi-urban areas. The main reason is because inequality in the distribution of per capita con- sumption increased signi�cantly.2 In decompos- ing inequality, approximately 86 percent of the TABLE 2.1 Poverty Estimates by Region, Urban and total increase can be attributed to a rise in inequal- Rural, 2004 ity within rural regions, while only 14 percent Percentage of Population can be attributed to an increase in rural-urban population below distribution of poverty inequality (World Bank 2005a). poverty line (% of all poor) Nationwide, poverty rates of households headed by ethnic minorities (defined as non- Region Urban Rural Total Urban Rural Total Khmer) are only slightly higher (35.8 percent) than Khmer-headed households (34.6 per- Phnom Penh 1.1 8.9 4.6 0.2 1.0 1.1 Plains 13.7 32.9 32.1 0.7 39.0 39.7 cent). However, because Chinese and Vietnamese Tonle Sap 28.2 45.4 42.8 3.7 33.4 37.0 households are typically wealthier than Khmer Coastal 20.4 30.1 26.8 1.6 4.6 6.1 households, these minority groups may offset Plateau/ 32.6 56.3 52.0 1.8 14.2 16.1 mountains higher poverty rates among smaller ethnic groups Cambodia 17.6 37.8 34.7 7.9 92.1 100.0 concentrated in the upland areas of the north- eastern provinces—about 17 ethnic groups com- Source: CSES 2004. prising over 100,000 people (Map 2.3). 44 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA MAP 2.2 Poverty Incidence by Region & Rural Poverty Incidence by Province Source: CSES 2004. NATURAL RESOURCES: Fisheries CONTEXT AND TRENDS The national catch has been estimated in the range Most of Cambodia’s rural population depends of 300,000 to 400,000 metric tons (360,000 met- on one rice crop per year, �sh and other aquatic ric tons in 2003), though the challenges of resources, forest products, and wage labor. CSES collecting catch data suggest a wide range of (2004) data indicate the income of the average uncertainty about this estimate. About 80 per- rural household is comprised of crop cultivation cent of the catch is from inland fisheries, most (30 percent); livestock (10 percent); common from the Tonle Sap Lake, as well as the Mekong property resources, especially �sh and forest prod- and Tonle Bassac River areas. Marine fisheries ucts (25 percent); and wage labor, nonagricultural along Cambodia’s 435-km coastline provide the activities, other (35 percent). Although rice farm- other 20 percent of the catch, important to coastal ing is the dominant livelihood activity, often areas of Kampong Som, Kampot, and Koh Kong households are active in all of these income- provinces. generating pursuits, diversifying in order to opti- Although fish catch and composition data mize labor resources during different seasons, and are lacking, there is a strong public perception safeguard against the risks of agricultural failure that fisheries resources are becoming scarcer (McKenney and Prom 2002; Chan and Acharya due to over�shing and various destructive �shing 2002; Prom and Ballard 2005). approaches. In a recent opinion poll of commune Cambodia enjoys one of the highest natural leaders in Cambodia’s rural areas (1,500 com- resource endowments per capita in the region. munes), 86 percent reported that the volume of This includes considerable unused land, relatively �sh catch has decreased compared to �ve years abundant forest resources, and highly productive ago (Seila and Danida 2005). A major concern is freshwater fisheries. A lack of (credible) data, that ongoing habitat destruction and �shing pres- however, makes it dif�cult to characterize the cur- sure will reduce �sh stocks and diversity to the rent status of natural resources with precision. point where �sh production consists of a limited PEN STUDY IN THE LOWER MEKONG SUB-REGION 45 POVERTY AND ENVIRONMENT IN CAMBODIA MAP 2.3 Minority Ethnic Population & Uplands/Lowland in Cambodia Source: Population Census 1998. number of small, low-value species, while larger, them more vulnerable to fishing pressure. The higher-value �sh species become a diminishing resulting change in the species composition proportion of the catch. This occurs because reduces catch values per unit of fishing effort larger fish species typically require more than a (Degen and others 2000). one-year cycle for reproduction, which makes Forests MAP 2.4 Forest Cover of Cambodia, 2002 The Forest Administration estimates that forest cover increased from 58 percent of the country in 1997 to 61 percent in 2002 (Map 2.4). This estimate stands in stark contrast to numerous other forest sector studies, which indicate that logging and clearing of forests have been rampant during this period (Independent Forest Sector Review 2004; Fraser Thomas and others 2000; www.globalwitness.org). For example, the Inde- pendent Forest Sector Review (IFSR 2004) esti- mates 55 percent of forest area has been disturbed or heavily degraded (Map 2.5). The IFSR also highlights the problem of uncontrolled clearing of forests, which has been occurring at an accelerat- ing pace over the past decade. Whereas 0.3 mil- lion ha of forest were cleared from 1993 to 1997, 1.3 million ha were cleared from 1997 to 2002 Source: Forest Administration 2003. (Map 2.6). This �nding is consistent with other 46 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA estimates indicating that the rate of deforestation MAP 2.5 Forest Disturbance by Roads and Villages increased from an average annual rate of about in Cambodia, 2002 0.5 percent prior to the mid–1980s, to about 1 percent from the mid–1980s to mid–1990s, and then to 1.0–1.7 percent from the mid–1990s to 2000 (IFSR 2004; McKenney and others 2004). The most recent indication of forest decline comes from a national opinion poll of commune leaders, which �nds that 72 percent of commune leaders in rural areas report forest cov- erage has declined compared to �ve years ago (Seila and Danida 2005). Land Approximately 40 percent of the rural popula- tion lives off of less than 0.5 hectares of agricul- tural land, an area too small to yield suf�cient rice to meet the needs of the average rural household. Source: IFSR 2004. Landlessness appears to be on the rise, a result of population pressure, land sales under duress (for example, health emergencies), and specula- tive land purchases (usually by Phnom Penh res- MAP 2.6 Forest Cover Clearance in Cambodia, idents). About 13 percent of the rural population 1993–1997 and 1997–2002 was estimated to be landless in 1997 (CSES 1997), rising to 16 percent in 1999 (CSES 1999) and to 20 percent in 2004 (CSES 2004). The chief means of acquiring new land for cultivation is forest clearance. This, along with uncontrolled logging, has resulted in rapid rates of forest loss in recent years. But with relatively large areas of forest remaining, it is not surpris- ing that only 28 percent of commune leaders reported less access to land for cultivation com- pared to �ve years ago. Along with uncontrolled forest clearing, a key management concern is the quality of cleared land for agriculture over the long term. If cleared land does not support pro- ductive agriculture due to poor soil fertility and water availability (and much of Cambodia’s land has low/medium soil productivity, Map 2.7), forest clearance could result in an ongoing cycle Source: IFSR 2004. of poverty and environmental loss. Indeed, cur- rent estimates are that more than half the land cleared from 1997 to 2002 had low-to-medium PEN STUDY IN THE LOWER MEKONG SUB-REGION 47 POVERTY AND ENVIRONMENT IN CAMBODIA 2001 have penetrated much more deeply into MAP 2.7 Soil Productivity of Cambodia forest areas (Map 2.9). Water With an estimated annual runoff of 475 billion cubic meters from the Mekong system draining over 85 percent of the country, Cambodia’s water resources are plentiful (World Bank 2005b). But rainfall is concentrated during a six-month wet season, with little infrastructure for capturing, storing, and regulating this runoff. Flooding is an annual occurrence, damaging infrastructure and crops, but also providing benefits to the floodplain by bringing soil nutrients, water, and creating areas for �sh spawning/habitat. During the dry season, river levels drop substantially, and surface water resources are largely limited to rivers (Mekong, Sap, and Bassac) and the Tonle Source: Forestry Administration 2003. Sap Lake. Groundwater resources are largely untapped; they appear to be uneven geographi- cally (most in close proximity to rivers), but more soil fertility (Dümmer 2004). Agricultural yields study is needed. Irrigation coverage in Cambodia in these areas are likely to be poor. In turn, this (7 percent) is far less than the neighboring coun- may lead to greater pressure on natural resources tries of Vietnam (34 percent), Thailand (26 per- and more land clearing in these areas—creating cent), and Laos (18 percent) (World Bank 2005a). an ongoing cycle of forest clearing, poor (and As a result, most agriculture is highly dependent declining) agricultural yields, greater pressure on natural resources, and more forest clearing. on the timing and levels of rainfall. Establishing land use management regimes to bring logging and forest conversion for agricul- The Management Vacuum ture under control will be a signi�cant challenge. Approximately one-third of the Cambodian pop- Over the past decade, the Cambodian govern- ulation (about 4 million people) live within 5 km ment’s chief approach to natural resource man- of forest, and about 10–15 percent of the popula- agement has been to contract out large areas to tion (about 1.5 million people) live within 5 km Cambodian and foreign investors as forest, �sh- of high-value forest (evergreen/semi-evergreen), eries, and agricultural/economic concessions. This which provide richer timber resources and other has proven to be a highly ineffective management forest products (Map 2.8). Moreover, the com- approach. It has failed to galvanize economic bination of in-migration and population growth growth in rural areas; generated only limited levels is resulting in a rapid increase in new villages of government revenue through taxes and fees; being established in rural forested/cleared areas. and often resulted in conflict between local com- Whereas the establishment of new villages largely munities and concession operators when commu- occurred along the agriculture-forest frontier nities have lost access to resources (McKenney prior to 1998, villages established from 1998 to 2002; World Bank 2005b). 48 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA Recent policy interventions suggest the gov- MAP 2.8 Villages within 5 km of Evergreen/ ernment recognizes some of these problems Semi-Evergreen Forests in Cambodia, 2002 and is seeking alternative approaches to natural resource management. Support for “social con- cessions�—a mechanism that could promote land distribution—is one acknowledgment by the government that a focus on large land- holdings as “economic concessions� has, with few exceptions, attracted little investment (World Bank 2004). In the forest sector, several conces- sions have been canceled, and remaining conces- sionaires are being required to submit strategic forest management plans while a ban on log- ging and log transport is in effect. However, it remains unclear whether this process will result in tangible improvements in forest manage- ment, or to what extent alternative management approaches will be promoted. A number of for- Source: Adaptation of data from IFSR 2004. est sector studies call for a broader shift in focus from industrial approaches to the exploration of various commercially oriented community forestry approaches (IFSR 2004; McKenney and others 2004). In the �sheries sector, the govern- MAP 2.9 Cambodian Village Locations 1998, and New Villages 1998–2001 ment acted to improve access to fisheries in 2001 by canceling commercial �shing lots cov- ering more than half the �shing concession area (McKenney and Prom 2002). However, many of the remaining �shing lots control the most productive �sheries. As a result of the changing management ap- proaches, much of Cambodia’s natural resources (perhaps 30 to 40 percent of total area, Figure 2.1) are now under no clear management regime (World Bank 2005b). Canceled forest conces- sions have reverted to the “forest reserve,� but management arrangements have not been de�ned (Table 2.2). Likewise, the roughly 500,000 ha of �sheries concessions released in 2001 have not been brought under clear management regimes. Finally, an estimated 1.75 million ha of “unused� land has not yet been declared to be under any Source: IFSR 2004. specific ownership, use, or control. Until these management vacuums are addressed, there will be the potential for misappropriation and over- exploitation of resources (World Bank 2005b). PEN STUDY IN THE LOWER MEKONG SUB-REGION 49 POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.1 Managed/ TABLE 2.2 Cambodia Land Tenure/ Unmanaged Areas Use Estimates, 2004 Towns, Area infrastructure, Category (million ha) % Share land mines 7% Forests not under 4.7 26.1 concessions No clear Protected areas/forests 3.3 18.3 management Cultivated areas 2.7 15.0 38% Forest concessions 3.1 17.2 Scrub land, nonwooded 1.7 9.4 land Towns, infrastructure 1.1 6.1 Protection, Released �shing 0.5 2.8 concession, cultivation concession lots 55% Agricultural concessions 0.8 4.4 Land mines contami- 0.1 0.6 nated areas Total 18.0 100.0 Source: World Bank 2005b. Source: World Bank 2005b. Natural Resources and Moving Out of Poverty restricted or prohibited; others located in the With much of Cambodia’s natural resources Mekong and Tonle Sap areas have seen a sharp under no clear management, and existing man- decline in �sh catch. agement regimes often weakly enforced, natural • CPR played little or no role for households resources are frequently accessed and used by that moved out of poverty. households as common property resources (CPR). • Total income for the “chronic poor� group has The importance of these resources as a safety net become more dependent on labor (up from 28 for rural households has been established in a to 38 percent), and less dependent on agricul- number of studies (Prom and Ballard 2005; CSES ture (33 to 23 percent) and CPR (28 to 18 per- 2004; Chan and Acharya 2002; McKenney and cent). This reflects agricultural problems with Prom 2002). Indeed, the majority of rural poor floods and droughts, as well as declining access households depend on CPR for some amount of and availability of CPR. consumption and income generation. But a recent • CPR access still played a crucial role as a safety study entitled Moving Out of Poverty Study (Cam- net in sustaining rural poor households cop- bodia Development Resource Institute 2006) ing with shocks (floods, droughts, and health attempts to take the analysis a step further by emergencies). investigating the extent to which CPR may help households move out of poverty over time.3 The FINDINGS: ANALYSIS OF POVERTY- study’s �ndings suggest the role of CPR has been ENVIRONMENT RELATIONSHIPS limited. Key conclusions include the following: In this section we present �ndings from our analy- • CPR household income declined for all mobil- sis of poverty-environment relationships, drawing ity categories due to declining availability and on a number of recently developed datasets.4 We access. For some villages, forest access has been identify major factors associated with poverty and 50 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA investigate speci�c poverty-environment relation- TABLE 2.3 Factors with Statistically Signi�cant ships pertaining to natural resource dependence, Relationship to Poverty drinking water sources, sanitation, natural disas- ters, and mine/UXO contamination. Factors associated with Factors associated with To set the stage for this analysis, Table 2.3 pre- poor households non-poor households sents major factors with a statistically signi�cant Household in Pleateau/Mountain Household in urban area relationship to poor and non-poor households in or Tonle Sap region Cambodia. While most of these relationships are Provincial capital not in close Household in Phnom Penh well-established, some are surprising (floods asso- proximity Household size All-weather road in close ciated with non-poor households). Likewise, proximity Table 2.4 summarizes the disparity between poor Female-headed household Head of household has and non-poor households for a number of key some schooling Drought-affected in past Head of household age poverty and environment indicators. �ve years (older) Mines severely affect Flood-affected in past amount of available �ve years Natural Resource Dependence agricultural land Cluster bombs dropped in/ Nationwide, a substantial proportion of house- near village holds are engaged in collecting fuelwood and other wood forest products (72 percent), collect- Source: Estimated from CSES 2004 and National Level 1 Survey 2002. ing non-wood forest products (21 percent), and catching �sh/seafood (53 percent) (Figure 2.2). Not surprisingly, it is much more common for TABLE 2.4 Poverty and Environment Indicators and rural households to be engaged in these activities Population Affected than urban households. Focusing on rural areas, the relationship between poverty and engaging Estimated number of in these natural resource-dependent activities is people affected clear—more than twice as many households in % of non-poor % of poor the poorest quintile engage in these activities Poverty & environment indicator: population population compared to the richest quintile (Figure 2.2). This relationship was noted in a previous analy- Lack of water supply (using open 16 22 sis of CSES data by Knowles (2005): water sources)* Using unprotected dug wells* 8 16 “Two variables referring to sources of house- Lack of sanitation (no latrine)* 60 80 hold income are positively and signi�cantly Households never boil drinking 24 48 water* related to food poverty, i.e., the “household Mine �eld and Cluster bomb 19 28 catches �sh, shrimp, crabs or oysters� and the Contamination (2002)** “household collects �rewood, charcoal, timber Household use of Solid fuels/ 90 99 or other forest products.� The fact that these indoor air pollution* variables are positively related to food poverty Flood (≥3 years in last 5 years)* 24 23 Drought (≥3 years in last 5 years)* 12 16 does not mean that they cause food poverty. Urban environment*** <20 <10 Instead, the causal link most likely runs in the No All-weather Road* 19 28 opposite direction, i.e., food-poor households are more likely to use these activities to obtain *Estimated from CSES 2004. **Estimated from a combination of CSES 2004 additional income. This result underlines the and the National Level 1 survey of mine�elds. ***This is simply the percent- importance of access to common resources as a age of non-poor and poor living in urban areas, and therefore potentially coping mechanism for the poor.� affected by urban environmental issues such as outdoor air pollution. PEN STUDY IN THE LOWER MEKONG SUB-REGION 51 POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.2 Households Engaged in Natural Resource-Dependent Activities (left), by Household Consumption Quintile in Rural Areas (right), 2004 90.0% Rural 80.0% 100% 70.0% 90% 60.0% 80% 70% 50.0% 60% 40.0% 50% 30.0% 40% 30% 20.0% 20% 10.0% 10% 0.0% 0% National Rural Urban Poorest 2 3 4 Richest Collects fuelwood/forest products Collects NWFPs Catches fish/seafood Collects fuelwood/forest products Collects NWFPs Catches fish/seafood Source: Estimated from CSES 2004. To take this analysis a step further, we assess the (Seila and Danida 2005).5 Speci�cally, we focus potential relationship between poverty, house- on responses to four poll questions (Table 2.5). holds engaged in natural resource-dependent Map 2.10 indicates where these problems may activities, and areas where resources may be in be most concentrated, showing the percentage of decline. First, to understand trends in the nat- communes within each province that responded ural resource base, we draw on recent opinion “decrease� for at least three of the four responses. poll data from a survey of rural commune leaders Problems of natural resource decline are most acute in the northern most provinces, with about 60 percent of the communes reporting natural resource decline in Ratanakiri, Stung Treng, and TABLE 2.5 Commune Opinion Poll Responses: Oddar Meanchey. Regionally, the relationship Natural Resource Decline between natural resource decline and poverty appears strongest in the plateau/mountain region Communes responding (Figure 2.3). Poll question decrease The poor are likely to be disproportionately Volume of �sh catch compared 86 percent affected by the negative impacts of natural to 5 years ago? resource decline (Figure 2.4). By examining only Forest coverage compared to 72 percent the households engaged in natural resource- 5 years ago? Number of people with access 28 percent dependent activities in communes reporting to land for cultivation compared resource declines, it becomes clear that the poor to 5 years ago? are considerably more dependent on these dete- Number of people able to secure 19 percent livelihood from natural resources riorating resource bases than the non-poor. in 2010? Indeed, about twice as many households in the poorest quintile depend on these activities com- Source: Seila and Danida 2005. pared to the richest quintile. 52 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA Drinking Water Sources MAP 2.10 Percentage of Communes within a Province Reporting Decreases in Natural Resources The proportion of Cambodian households with (3 out of 4 responses of decrease—�sh, access to safe or improved drinking water forest cover, land access, and long-term sources ranges from about 65 percent (dry sea- sustainability) son) to 75 percent (wet season) (Figure 2.5). For this analysis, sources generally considered safe or improved include water piped in dwelling or on the premises, tube/piped wells or boreholes, protected dug wells, and rainwater. Water from tanker trucks and vendors may also be safe, but as this market is unregulated it is difficult to confirm. Unsafe sources include unprotected dug wells and ponds, rivers, and streams. Over 20 percent of the population uses rainwater for drinking water during the wet season, switch- ing to this source mainly from rivers, ponds, and streams, tanker trucks and vendors, and tube/piped wells. Drinking water sources: non-poor and poor households The proportion of non-poor Cambodian house- Source: Estimated from Seila and Danida 2005. holds with access to safe/improved drinking water sources ranges from about 70 percent (dry season) to 80 percent (wet season), compared to 55 per- cent (dry season) to 65 percent (wet season) for FIGURE 2.3 Natural Resource Decline and Rural poor households (Figure 2.6). Water piped in Poverty Incidence by Zone dwelling is much more common for the non-poor (17 percent) than the poor (2 percent). And Plateau/Mountain almost one-quarter of non-poor households use rainwater as drinking water during the wet Coastal season compared to only 19 percent of poor households. Tonle Sap A strong relationship between poverty and unsafe drinking water sources (unprotected dug Plains wells and ponds, rivers, and streams) is further revealed when examining across household con- Phnom Penh sumption quintiles (Figure 2.7). Nationally, 0% 10% 20% 30% 40% 50% 60% households in the poorest quintile are more than Rural Poverty Incidence % Communes Reporting NR Decline (3 of 4 responses) three times as likely to use unsafe drinking water sources as the richest quintile. This disparity is Source: Estimated from CSES 2004 and Seila and Danida 2005. lower in rural areas, where the poorest quintile is only twice as likely to use unsafe water sources. But in urban areas, households in the poorest PEN STUDY IN THE LOWER MEKONG SUB-REGION 53 POVERTY AND ENVIRONMENT IN CAMBODIA 60 percent (dry season) to 70 percent (wet season), FIGURE 2.4 Households Engaged in Natural Resource-Dependent Activities in compared to over 85 percent for urban households Communes Reporting Resource Decline, (Figure 2.8). Looking at speci�c water sources, By Household Consumption Quintile water piped in dwelling is far more prevalent for urban households (over 40 percent) compared 80% 70% rural households (4 percent). Use of rainwater 60% during the wet season is more common for rural 50% households (25 percent) than urban households 40% 30% (13 percent). 20% Analyzing rural and urban households accord- 10% 0% ing to non-poor or poor status sheds further light Poorest 2 3 4 Richest on relationships between poverty and drinking Fuel wood/wood products collection and decline in resource water sources (Figure 2.9). First, while rural non- NWFP collection and decline in resource poor households have somewhat greater access Fish/seafood catching and decline in resource to safe/improved drinking water sources than rural poor households, this is largely due to more purchases from tanker trucks and vendors during Source: Estimated from CSES 2004 and Seila and Danida 2005. the dry season and greater use of rainwater during the wet season. The disparity is greater in urban quintile are 8–10 times more likely to use unsafe areas, where about 90 percent of non-poor house- water sources compared to the richest quintile. holds have access to safe/improved drinking water, compared to only about 65–70 percent Drinking water sources: of poor households. Moreover, non-poor house- rural and urban households holds in urban areas are far more likely to have The proportion of rural households with access the convenience of water piped in dwelling to safe/improved drinking water ranges from (50 percent) than poor households in urban areas (7 percent). FIGURE 2.5 Drinking Water Sources, Dry and Wet Season, 2004 Factors associated with accessing drinking water sources: regression results National Several variables show a statistically significant relationship with access to drinking water sources. 100% First, access to unsafe drinking water sources is 90% Pond, river or stream associated with households that are poor, located 80% Unprotected dug well in the coastal or plateau/mountain regions, and 70% Tanker truck, vendor located in rural areas far from district and provin- 60% cial capitals and all-weather roads. Access to 50% 40% Rainwater unsafe water is also associated with male heads 30% Tube/piped well/borehole or of households who have no schooling (female- 20% protected dug well headed households are associated with using 10% Piped in dwelling or on safe/improved drinking water sources at a statisti- premises 0% cally signi�cant level). Dry season Wet season During the wet season, approximately one- quarter of households nationwide switch to the Source: Estimated from CSES 2004. use of rainwater. Variables associated with using 54 PEN STUDY IN THE LOWER MEKONG SUB-REGION FIGURE 2.6 Drinking Water Sources, Non-Poor and Poor, Dry and Wet Season, 2004 Dry Season Wet Season 100% Pond, river or stream Pond, river or stream 100% 90% 80% Unprotected dug well Unprotected dug well 80% 70% 60% Tanker truck, vendor Tanker truck, vendor 60% 50% Rainwater Rainwater 40% 40% 30% Tube/piped Tube/piped 20% 20% well/borehole or well/borehole or protected dug well 10% protected dug well Piped in dwelling or Piped in dwelling or 0% 0% on premises on premises non-poor poor non-poor poor Source: Estimated from CSES 2004. FIGURE 2.7 Unsafe Drinking Water Sources, by Household Consumption Quintiles, 2004 National 50% 40% 30% 20% 10% 0% Poorest 2 3 4 Richest Wet Season Dry Season Rural Urban 50% 40% 35% 40% 30% 30% 25% 20% 20% 15% 10% 10% 5% 0% 0% Poorest 2 3 4 Richest Poorest 2 3 4 Richest Wet Season Dry Season Wet Season Dry Season Source: Estimated from CSES 2004. PEN STUDY IN THE LOWER MEKONG SUB-REGION 55 POVERTY AND ENVIRONMENT IN CAMBODIA quintile boiling water compared to the poorest FIGURE 2.8 Drinking Water Sources, Rural and Urban, Dry and Wet Season, 2004 quintile. In rural areas, boiling drinking water by non-poor households (71 percent) and poor Rural & Urban households (51 percent) tracks rates at the national level. But in urban areas, boiling drinking water Pond, river or stream 100% is much more common, especially by non-poor households (nearly 90 percent). Note that boiling 80% Unprotected dug well is by far the most common method for making drinking water safe; few households treat their 60% Tanker truck, vendor drinking water by other methods. Analyzing by drinking water source reveals 40% that those households accessing unsafe water Rainwater sources are the least likely to be boiling their water 20% (Figure 2.11). That is, the households that most Tube/piped well/borehole or need to boil their water are the least likely to do 0% Rural & Rural & Urban & Urban & protected dug well so. This is especially true for poor households, Piped in dwelling or Dry Wet Dry Wet on premises where boiling is done by less than 25 percent of Season Season Season Season households using water from unprotected wells and only about 40 percent of households using Source: Estimated from CSES 2004. water from ponds, rivers, and streams. But poverty is not a suf�cient explanation, as only a modest proportion of non-poor households using these rainwater include non-poor households, female- unsafe drinking water sources boil their water— headed households, and older heads of house- 40 percent using water from unprotected wells holds with some schooling. Rainwater use is also and 57 percent using water from ponds, rivers, associated with households located in the coastal and streams. or plateau/mountain regions, in rural areas with Regression analysis indicates several variables all-weather roads, and in areas close to the provin- associated at a statistically signi�cant level with cial capital. Similarly, variables associated with the households boiling drinking water. These vari- use of drinking water from tanker trucks and ables include households being non-poor, located vendors include non-poor households, female- in urban areas, and located in the plains or plateau/ headed households, and heads of households with mountain regions. Boiling drinking water is also some schooling, as well as households located in associated with older heads of households who closer proximity to district and provincial capitals. have had some schooling, as well as with house- holds that have separate kitchens and use clean fuels for cooking. Boiling or otherwise treating drinking water Sanitation Nationwide, approximately 75 percent of non- poor households boil or otherwise treat their Nationwide, only about 30 percent of Cambo- drinking water compared to only about 52 per- dian households have access to basic sanitation— cent of poor households (Figure 2.10). This rela- they are connected to sewerage, have a septic tank, tionship between poverty and boiling drinking pit latrine, or other facility without septic tank, water is illustrated even more clearly when look- or use a public or shared toilet (Figure 2.12). The ing across household consumption quintiles, with rest of the population (70 percent) has no access about twice as many households in the richest to sanitation services. 56 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.9 Drinking Water Sources, Rural and Urban, Non-Poor and Poor, Dry and Wet Season, 2004 Dry Season & Rural Wet Season & Rural 100% Pond, river or stream 100% Pond, river or stream 90% 90% 80% Unprotected dug well 80% Unprotected dug well 70% 70% 60% Tanker truck, vendor 60% Tanker truck, vendor 50% 50% Rainwater Rainwater 40% 40% 30% 30% Tube/piped Tube/piped 20% 20% well/borehole or well/borehole or 10% protected dug well 10% protected dug well Piped in dwelling or Piped in dwelling or 0% 0% on premises on premises Non-Poor Poor Non-Poor Poor Dry Season & Urban Wet Season & Urban 100% Pond, river or stream 100% Pond, river or stream 90% 90% 80% Unprotected dug well 80% Unprotected dug well 70% 70% 60% Tanker truck, vendor 60% Tanker truck, vendor 50% 50% Rainwater Rainwater 40% 40% 30% 30% Tube/piped Tube/piped 20% 20% well/borehole or well/borehole or 10% protected dug well 10% protected dug well Piped in dwelling or Piped in dwelling or 0% 0% on premises on premises Non-Poor Poor Non-Poor Poor Source: Estimated from CSES 2004. Basic Sanitation: non-poor and sumption quintiles, with about three times as poor pouseholds many households in the poorest quintile lacking Nearly 40 percent of non-poor Cambodian house- basic sanitation compared to the richest quintile. holds have access to basic sanitation (primarily connection to sewerage and septic tanks) com- Basic sanitation: rural and pared to only about 8 percent of poor households urban households (Figure 2.13). This relationship between poverty Access to basic sanitation is far more common in and lack of basic sanitation is illustrated even urban areas (65 percent of households) than rural more clearly when looking across household con- areas (20 percent of households) (Figure 2.14). PEN STUDY IN THE LOWER MEKONG SUB-REGION 57 FIGURE 2.10 Boiling/Treating Drinking Water, Non-Poor and Poor, by National (top-left), Household Consumption Quintile (top-right), Rural (bottom-left) and Urban (bottom-right), 2004 National Boiling by Household Consumption Quintile 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% No, never 50% 50% Sometimes 40% 40% Yes, always 30% 30% 20% 20% 10% 0% 10% Poorest 2 3 4 Richest 0% Non-Poor Poor Always Sometimes Rural Urban 100% 100% 90% 90% 80% 80% 70% 70% 60% No, never 60% No, never 50% Sometimes 50% Sometimes 40% Yes, always 40% Yes, always 30% 30% 20% 20% 10% 10% 0% 0% Non-Poor Poor Non-Poor Poor Source: Estimated from CSES 2004. FIGURE 2.11 Boiling or Otherwise Treating Drinking Water by Source, Non-Poor FIGURE 2.12 Access to Basic and Poor, 2004 Sanitation, 2004 100% National 90% 80% 100% 70% 90% None 60% 80% Open land 50% 70% 40% 60% Public or shared toilet 30% 50% Pit latrine or other w/out 20% 40% septic tank 30% Septic tank 10% 0% 20% Connected to sewerage 10% am r er er l e l el do el l ho at at w w tre en nw w 0% d d e ,s /v te te or d ai ck pe 1 ec ec er l/b R ru Pi iv t ot el ro t ,r Pr w np er nd be nk U Po Tu Ta Source: Estimated from CSES 2004. Non-Poor Poor Source: Estimated from CSES 2004. 58 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.13 Sanitation by Non-Poor and Poor (left) and Household Consumption Quintile (right), 2004 Non-Poor & Poor Lack of Basic Sanitation by Household Consumption Quintile 100% 100% 90% None 90% 80% 80% Open land 70% 70% 60% Public or shared toilet 60% 50% Pit latrine or other w/out 50% 40% septic tank 40% 30% Septic tank 30% 20% Connected to sewerage 20% 10% 10% 0% 0% Non-Poor Poor Poorest 2 3 4 Richest Source: Estimated from CSES 2004. Connection to sewerage—30 percent of urban cent of rural households lacking sanitation in all households compared to less than 2 percent of but �ve provinces. rural households—accounts for much of the Analyzing rural and urban households accord- difference. A closer examination of the lack of ing to non-poor or poor status sheds further light sanitation for rural households shows how the on relationships between poverty and sanitation. problem is nationwide, with more than 80 per- Only 7 percent of rural poor households have FIGURE 2.14 Sanitation by Rural and Urban (left); Lack of Sanitation for Rural Households by Province (right), 2004 Rural & Urban 100% 90% None 80% Open land 70% 60% Public or shared toilet 50% Pit latrine or other w/out 40% septic tank 30% Septic tank 20% Connected to sewerage 10% 0% Rural Urban Source: Estimated from CSES 2004. PEN STUDY IN THE LOWER MEKONG SUB-REGION 59 POVERTY AND ENVIRONMENT IN CAMBODIA access to basic sanitation, compared to 26 per- on CSES 2004, which provides data on house- cent of rural non-poor households (Figure 2.15). holds affected by floods and droughts between Greater use of septic tanks by the rural non-poor 1999 to 2003. Over this �ve-year period, 53 per- accounts for much of this difference. In urban cent of households report being affected by flood areas, about 20 percent of poor households have and 63 percent report being affected by drought access to sanitation, compared to 73 percent of in at least one year. To isolate the households non-poor households. Here, the urban non-poor experiencing more severe problems with these households are 10 times as likely as urban poor natural disasters, we identi�ed households that households to be connected to sewerage. had been affected in three of more of the five years, either by flood (23 percent of households) Factors associated with access to or drought (14 percent of households), for fur- sanitation: regression results ther analysis. For these flood-affected households, the poor- Several variables show a statistically significant est quintile was less often affected by floods than relationship with access to basic sanitation. These richer households (Figure 2.16). While it may variables include households that are non-poor, seem counterintuitive that richer households located in urban areas, with access to all-weather appear located in more flood-prone areas than roads and in close proximity to district and provin- poor households, there may be two explanations cial capitals. Access to sanitation is also associated for this �nding. First, for rice farming purposes with older heads of households who have had there are both “good� floods (providing water some schooling. Female-headed households are and replenishing soil nutrients) and “ bad� floods more likely to lack access to sanitation. (large unmanageable amounts of water). It is not clear whether the survey (and respondents) made Natural Disasters this distinction. Second, the fact that richer households are more affected by floods may sim- Analysis of the relationships between poverty and ply reflect their ownership of higher quality rice households affected by natural disasters is based land nearer to water sources. For households affected by drought in three or more of the five years, poorer households are more often affected than richer households FIGURE 2.15 Sanitation, Rural and Urban, Non-Poor and (Figure 2.16). From 1999 to 2003, about 17 per- Poor, 2004 cent of households in the poorest quintile were affected by droughts in three or more of the 100% five years. This percentage declines as house- 80% None hold consumption increases, with only about Open land 13 percent of the richest quintile affected by 60% Public or shared toilet drought. 40% Pit latrine or other w/out septic tank Controlling for other factors, both types of Septic tank natural disasters show a statistically signi�cant Connected to sewerage 20% relationship to poverty. For households affected 0% in three or more of the �ve years, drought is asso- Urban & Non-Poor Urban & Poor Rural & Non- Rural & Poor Poor ciated with poor households, whereas flood is associated with non-poor households (Table 2.3 Source: Estimated from CSES 2004. above). 60 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.16 Housholds Affected by Flood (left) or Drought (right) in Three or More of the Five Years from 1999 to 2003, by Household Consumption Quintile Affected by Flood in 3 or more of past 5 years Affected by Drought in 3 or more of past 5 years 30% 18% 16% 25% 14% 20% 12% 10% 15% 8% 10% 6% 5% 4% 2% 0% 0% Poorest 2 3 4 Richest Poorest 2 3 4 Richest Source: Estimated from CSES 2004. Mine and Unexploded Ordinance nation limiting access to agricultural land shows (UXO) Contamination a statistically significant relationship with being poor. Analysis of the relationship between poverty and mine/UXO contamination is based on CSES poverty data (2004) and the National Level 1 Sur- CASE STUDY FINDINGS: vey (2002) conducted in nearly 14,000 villages to ACCESS TO NATURAL RESOURCES identify mine/UXO contamination. The National AND POVERTY6 Level 1 Survey distinguishes between villages The PEN II case study was undertaken to inves- affected by mines, cluster bombs (a class of UXO tigate relationships between access to natural that may affect villages in a similar way as mines), resources and poverty (Box 2.1: Case Study and spot UXO. Villages are more likely to expe- Methodology). Household survey �ndings indi- rience socio-economic impacts, such as reduced cate that across the six villages studied the per- access to agricultural land, from mine and clus- centage of poor households range from 30 to ter bomb contamination. 60 percent. More than half of poor households Analysis of affected households shows a experience signi�cant annual rice de�cits—they strong relationship between mine and cluster- must buy rice for more than half of the year to bomb contamination and poverty (Figure 2.17). support consumption. The average amount of The poorest quintile is almost four times as agricultural land for poor households is only likely to be affected by mine contamination, 1.1 ha compared to 2.4 ha for non-poor house- and twice as likely to be affected by cluster bomb holds. Overall, the poor are more likely to live in contamination, as the richest quintile. Like- a thatch (rather than wooden) house, own fewer wise, mine contamination is much more likely and lower-value livestock, and have less involve- to reduce access to agricultural land where ment in petty trade and small business activity. households are poor. Indeed, controlling for While the proportion of poor and non-poor other factors, the presence of mine contami- households hiring out their labor is similar, a PEN STUDY IN THE LOWER MEKONG SUB-REGION 61 POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.17 Housholds Affected by Mines (top left), Cluster Bombs (top right), Mines/Cluster Bombs (middle left), Spot UXO (middle right), and Mines Limiting Access to Agricultural Land (bottom left), by Household Consumption Quintile Mine Contamination - National 14% 12% 10% 8% 6% 4% 2% 0% Poorest 2 3 4 Richest Mine and/or Cluster Bomb Contamination - National Spot UXO Contamination - National 35% 50% 45% 30% 40% 25% 35% 20% 30% 25% 15% 20% 10% 15% 10% 5% 5% 0% 0% Poorest 2 3 4 Richest Poorest 2 3 4 Richest Mine Contamination Severely Affecting Access to Agricultural Land Cluster Bomb Contamination - National 7% 25% 6% 20% 5% 4% 15% 3% 10% 2% 5% 1% 0% 0% Poorest 2 3 4 Richest Poorest 2 3 4 Richest Source: Estimated from CSES 2004 and National Level 1 Survey 2002. higher percentage of the poor hire out labor to what more households in the study areas col- clear degraded forests, while non-poor tend to be lect forest products and somewhat fewer raise more involved in timber extraction. livestock. Nearly all rural households across the study At the commune and village level, officials areas, whether poor or better-off, depend on nat- report a number of changes in natural resources ural resources for their livelihoods. In most cases, over the past decade. First, significant forest households are cultivating rice (83 percent), rais- clearing for rice and other agricultural crops has ing livestock (68 percent), collecting forest prod- occurred, consistent with national trends. At the ucts (98 percent), and �shing (75 percent). The same time, progress in improving agricultural proportion of households engaged in these liveli- productivity has been mixed, with some areas hood activities is quite similar to rural households reporting increases, some decreases, and some across Cambodia (CSES 2004), though some- no change. Across the study areas, households 62 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA BOX 2.1 Case Study Methodology In the design of PEN phase II for Cambodia, interviews were conducted with a range of key stake- holders to identify poverty-environment issues of national priority. This included consultations with a number of government partners, including the Ministry of Environment, Ministry of Economy and Finance, and Ministry of Planning. These interviews revealed several recent and ongoing studies addressing critical issues associated with poverty and natural disasters, land, water, and health and sanitation. In addition, a number of studies were focused on poverty issues in floodplain areas around the Tonle Sap Lake and in the Mekong Delta region (Prom and Ballard 2005; So and Ballard 2005). However, one issue of increasing concern for Cambodia—the relationship between access to natural resources and poverty—was not the explicit focus of any of these studies. To address this concern, PEN II examines access and poverty issues for a variety of natural resources, including agricultural land, forest products, grazing land, �sheries, and drinking water. As the study sought to avoid floodplain areas (in order to avoid repeating the work of other studies), �sheries concerns are not covered in depth, whereas forest and land issues receive more comprehen- sive coverage. For the purposes of this study, access to natural resources is de�ned as households having the ability to collect/obtain the speci�ed resource or product. In practice, it was found that all households with access to resources had also made use of this access in the past year.1 Lack of access may be attributed to a lack of means (transportation, resource extraction equipment), physical barri- ers such as poor road conditions, and/or legal and institutional barriers such as restrictions on access to land or speci�ed areas/resources. The case study is guided by the following research questions: • To what extent do the poor differ from the non-poor in terms of current access to natural resources and environment in Cambodia? • What trends in access to natural resources and environment can be observed, and do these trends differ between the poor and the non-poor? • What are the environmental impacts from different types of natural resources and environment utilization by the poor and non-poor? What are the key determinants of the environmental impacts? To what extent can these environmental impacts be economically evaluated? • What are the speci�c and practical policy options leading to poverty reduction and sustainable utilization of natural resources and environment? PEN II case studies are intended to deepen current knowledge about poverty-environment relation- ships and to support broader investigation of potential poverty-environment linkages in national data. In Cambodia, two provinces were selected for study of natural resource access issues— Kampong Speu and Kampong Thom. These were chosen because each ful�lls the criteria of having a “protected area,� diverse types of concessions, ongoing in-migration, and a substantial share of poor districts/communes. A total of six villages (three in Aoral district of Kampong Speu province and three in Santuk district of Kampong Thom province) were then selected for more focused study based on their proximity to protected areas, proximity to roads, and population densities. Data were collected through (a) semi-structured interviews with of�cials at the provincial, district, commune, and village levels; (b) a household survey of 120 households (20 households randomly selected in each of the six villages); and (c) focus group discussions to capture additional information on the potential effects of wealth differences and gender. Data analysis required classifying house- holds into “poor� and “non-poor� categories. This study relied on methods developed and employed by GTZ in its rural development programs (including in the villages in Kampong Thom). Key criteria in establishing poor and non-poor status included amount of residential land, amount of productive land, sources of family income, livestock, housing, means of transport, electric appliances, and period of time in which the household lacks food within a year. It should be noted that problems were encountered in applying these selection criteria. First, proximity to a protected area was revealed to be unimportant, as parks tend to lack enforcement. Proximity to (quality) forest resources is a more meaningful variable to consider. Second, proximity to roads was not as meaningful as proximity to markets. For example, in Kampong Thom, while the three selected villages have differing levels of road access, all are less than one hour from a marketplace. Likewise, the three selected villages in Kampong Speu have differing road access, but all are very far from a marketplace. Finally, although provinces were selected according to their (continued ) PEN STUDY IN THE LOWER MEKONG SUB-REGION 63 POVERTY AND ENVIRONMENT IN CAMBODIA BOX 2.1 Case Study Methodology (Continued ) low and high population density, at the village level it was found that all potential rural villages had a population density well below the national average. A number of limitations of the PEN II study require noting here. First, the sample size (20 house- holds per village for six villages) for the case study is small; the intention was for the survey to provide on-the-ground insights regarding the poverty and environment nexus, but not necessar- ily to provide data for analysis of statistical signi�cance. Second, as noted above, the classi�cation of “poor� households is not based on a robust poverty line. This may lead to a small margin of error in the classi�cation of “poor� and “non-poor� households, and in the comparison of case study �ndings to national data. Third, at the time of writing, PEN II had only limited access to data and �ndings from the CSES, which suggests potential scope for identifying additional relationships between case study �ndings and the national context and trends beyond those discussed in this chapter. Finally, analysis of spatial associations between poverty and environment was hampered because most credible data (such as CSES) are only presently available at the provincial level, with some data also available at the commune level (for example, Seila/Ministry of Planning commune database 2003). For analysis in Cambodia, Vietnam, and Laos, PEN II has targeted poverty-environment spatial analysis at the district level. While this may be an appropriate administrative level for analysis in Vietnam and Laos, it has a number of shortcomings in Cambodia. First, Cambodia has only had elec- tions at the national and commune level.1 As a result, most policy (and donor) support targeting local issues focuses on communes and their elected Commune Council. Policy development is less driven at the district and provincial levels because they are administered by unelected of�cials. Second, from a geographical standpoint, aggregating data at the district level has not been found to be highly relevant because most rural districts include semi-urban areas (district markets). This mix of rural and urban areas makes district-level data less meaningful for study of poverty than more disaggregated units. Moreover, districts may include floodplain and upland areas,2 and thus villages dependent on different types of natural resources (e.g., �sheries, forestry) for their liveli- hoods, all of which complicate assessments of poverty-environment relationships at the district level. To achieve more meaningful results, with greater opportunities for policy mainstreaming at the local level, future analysis of poverty-environment issues in Cambodia would be better targeted at the commune level. 1When the term “access� is translated into Khmer, it implies the household has the internal ability to obtain a resource. Households with the ability to obtain resources also have experience doing so, given that these resources are relied on and often constitute a major part of livelihoods. 2Cambodia’s governing levels include national, provincial, district, commune, and village. 3For instance, many districts surrounding the Tonle Sap Lake also include forested areas far from the lake. and of�cials report that availability of natural and better-off “newcomers�). Newcomers often resources (especially forest and �sheries resources) hire local people to clear land, usually for agri- has decreased substantially over the past decade, culture but also for speculative purposes. In with the sharpest declines occurring in the past some areas, there is plenty of access to land avail- �ve years. Officials blame the declines on in- able, but the land is unproductive. For instance, creased demand, overexploitation, and uncon- migrants were drawn to one village by the promise trolled activity. of three ha of land. While this was provided, they have found yields to be very low due to poor soils and a lack of rain. Most are now trying to Access to Agricultural Land make a living in charcoal production. Gaining access and claim to land in the study Approximately 25 percent of the households areas involves little more than clearing it. As a surveyed report being landless. The landless result, in some villages there has been signi�cant households are almost all poor, and most are in-migration of landseekers (including landless highly dependent on access to natural resources 64 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA (especially forest products) for their livelihoods. very unjustly and cruelly take a share from the Many of these households express concern that products collected by the innocent people. For increasing overexploitation of resources may instance, when people transport the products they worked very hard to collect to the market or home, make it dif�cult for them to sustain their liveli- these people charge very high informal fees from hoods in the future. In addition to landlessness, people without shame or con�scate their transport there are indications of increasing concentration means, making people lose their last resort of liveli- of land ownership. For example, the average hood. The community can only educate people to non-poor landholding household in Santuk owns love nature, but cannot protect ourselves from such 2.77 ha, compared to 0.73 ha for poor land- abuses. (A village chief in Santuk district) holding households. Given the restrictions, some households have found creative ways in which to collect timber Access to Forest Resources while avoiding fees and con�scation. For instance, villagers often cut up large logs (worth consider- Availability of forest products has decreased able sums) into low-value firewood to avoid markedly over the past several years in all study restrictions. Likewise, villagers may cut timber for areas. Timber, resins (which are produced by house construction and wait for a year or more for large trees), and wild animals are among the the wood to become “old� looking before trans- products that have declined most dramatically. porting it. This way they appear to be moving Wood for charcoal production is also on a rapid their house, rather than construction materi- declining trend. Decreasing at a somewhat slower als, and can avoid paying of�cials on the road. rate are wild vegetables, vinery, and other minor Such delays and inef�ciencies caused by the cur- products. As low-value products, these resources rent enforcement regime represent a sub-optimal are likely under less extractive pressure than tim- use of forest resources, but villagers see no other ber and other wood resources. options, given the current restrictions. Access to forest resources in the study areas is In Aoral district, many households are involved limited by legal restrictions, taxes, and Forest in charcoal production. Although access to for- Administration (FA) control of forest concessions est resources is open, a fee of R10,000 ($2.50) (in Santuk district, where three forest concessions must be paid to authorities for each unit of char- operated until recent suspension of activities). coal produced (1 meter by 1 meter). Charcoal Households in Santuk district recall that when producers consider this to be a high fee, as it is concessionaires tightened forest boundaries and equal to their labor costs per unit. Authorities placed armed guards to restrict access several enforce against timber extraction selectively, stop- years ago, they became poorer. The hardship con- ping villagers from logging, but also allowing tinued during the years of concession operation, bands of loggers to operate as long as they pay and has become worse as the forest resources have fees twice per month. Households in the study come under FA management. The FA not only areas complain about what they perceive to be restricts forest access but also either charges fees to unjust restrictions and selective enforcement. villagers collecting forest products, or con�scates They believe the presence of enforcement author- the products altogether. The restrictive regime ities, or designation of protected areas, has little appears to be selectively enforced, with more non- effect if outside logging interests are involved. poor households gaining access to timber (46 per- “The whole village is in protected area but log- cent) than poor households (11 percent). ging is happening every day.� (A village chief in Aoral district) At present people have the right to access the forests to collect forest by-products. However, the author- As charcoal production has been ongoing for sev- ities (forest wardens, soldiers, and police) wait to eral years in Aoral district, the forest surrounding PEN STUDY IN THE LOWER MEKONG SUB-REGION 65 POVERTY AND ENVIRONMENT IN CAMBODIA the study villages has all been cut. The forest is lagers �sh during the wet season and early dry now 6–10 km away, with the frontier moving season, when there is enough water in these areas farther away each year. Early in the morning, to support more abundant aquatic life. Catches these villagers depart on koyun7 to go to the for- are primarily for household consumption. est to collect �rewood for charcoal and timber There are no restrictions on access to �shing for fences. Most villagers use axes; chainsaws are areas for the study villages. Although roughly the only owned by powerful people in the area. In the same proportion of poor and non-poor house- current forest area, villagers are in the process of holds �sh, non-poor households catch twice as making a “second cut� (the best wood has already much �sh as poor households. This occurs because been collected). While villagers express concern the non-poor can afford better �shing equipment. about the future, they also say they will make a Households report a decline in their overall catch, third cut (to extract roots) if they have no other especially over the past �ve years. They attribute option. this decline to overexploitation, including the illegal use of destructive �shing equipment and Access to Grazing Land methods. Access to grazing land is important to rural liveli- hoods because raising cattle and buffalo provides Access to Drinking Water draft animals for agriculture as well as supple- Access to drinking water is similar for poor and mental income if sold. On average, 65 percent of non-poor households in the study areas. How- the households in the study villages raise cows or ever, all villages complain about the quality of buffaloes, with poor households owning fewer water and the dif�culty of collecting it. Drinking animals than non-poor households. Likewise, the average value of animals owned by poor house- water in Santuk district is mainly drawn from holds is only half that of non-poor households. wells, but many households note that the water Where grazing areas are available, access is smells and tastes bad. They believe there is lime generally open to all. Most households report a in the well water, and therefore they prefer other decrease in grazing areas compared to 10 years sources when available. Households in Aoral dis- ago due to conversion of these areas into crop- trict use a variety of sources, including rain water, land. However, some households (usually non- streams, and hand pumps. Although households poor) no longer require grazing land, as they have understand that water from streams is generally opted for machinery (koyun) to replace draft ani- not clean, they often use it. Where possible, house- mals, so they have less need to raise cattle and holds in both study areas use water from streams buffalo. for irrigating vegetables, bathing, washing clothes, and feeding animals. Less than one-quarter of households boil Access to Fisheries water before drinking. Villagers note that it is Cambodia’s �sheries play a major role in rural quicker and easier to drink un-boiled water. livelihoods, supporting consumption and pro- Most households report that they do not wash viding the chief source of protein in the rural their hands before having meals and there is a diet. Even though the six surveyed villages are lack of other hygienic practices. However, there primarily forest-dependent, 75 percent of house- is a degree of awareness among most households holds bene�t from catching a small amount of that hygienic practices and drinking boiled water �sh, frogs, and other aquatic resources. Most vil- help reduce health problems and illness. 66 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA Summary of Case Study Findings portation and equipment for extraction, sup- porting even more extraction, while the poor Restrictions on timber access appear to bene�t non- continue at a level of subsistence. poor households; no restrictions were identi�ed for Poverty reduction associated with opening of other natural resources. Although this case study access tends to be short-lived, likely to be followed by did not �nd restrictions on access to most natural increasing poverty over the long term. Although resources, access to timber was restricted for for- both non-poor and poor households may initially est areas currently managed by the Forest Admin- bene�t from the opening of access to natural istration in Kampong Thom. This restriction resources (for example, cancellation of forest con- appears to be selectively enforced, with more cessions or �shing lots), this poverty reduction is non-poor households gaining access to timber likely to be short-lived. Opening access—with no (46 percent) than poor households (11 percent). clear management structure to �ll the vacuum— In Aoral, charcoal producers pay an informal tax results in a period of overexploitation, often to authorities on each unit of charcoal produced; involving destructive practices, that will likely be in effect, the tax doubles their cost of production. followed by increasing poverty over the long term For other natural resources studied (agricultural as the resource base declines (Figure 2.18). land, grazing areas, �sheries, and drinking water), For example, rapid exploitation of charcoal no notable differences in the proportion of poor resources by villages in Aoral district appears to and non-poor households with access were iden- have led to short-term poverty reduction. Many ti�ed. This appears to be the case because nearly households used initial pro�ts from charcoal to all households in the six villages studied, whether make home improvements and invest in koyun very poor or better off, depend on access to these transportation (to support farming and the trans- resources. port of charcoal). It is only more recently that, as Non-poor households make more effective use of the forest frontier has moved farther away, char- open access. The study did �nd differences in how coal is becoming less pro�table and charcoal pro- effectively (and pro�tably) non-poor households ducers are worried about the future. Due to poor make use of their access to natural resources farming conditions (drought and low soil fertil- compared to poor households. This appears to ity), they do not see options other than continu- occur for two reasons. First, non-poor house- ing charcoal production. Thus, it can be expected holds tend to have better means of transporta- that poverty will increase as resources become tion and equipment for exploitation of forest scarcer and farther away, unless other opportu- and fisheries resources. This allows them to nities arise for the community members to make collect more, over greater distances. Second, their living. because the non-poor are able to collect prod- Although Figure 2.18 reflects this case study’s ucts more efficiently, they are better able to �nding for villages in forest areas, the same appears cover the fixed costs associated with travel to true for �shing villages in Cambodia. Non-poor resource-rich areas, pay fees to officials, and �shing households tend to have better boats and still make modest earnings. Poorer households gear, providing them a greater ability to bene�t may require longer travel times, collect fewer from open access �sheries. As noted in the interim products for their labor output, and have dif- findings of the Tonle Sap Participatory Poverty ficulty covering fee charges. These differences Assessment: in the productive use of access tend to com- “[Fishing] villagers de�ned three different cat- pound over time, as non-poor households use egories of well-being: medium, poor, and des- some of their earnings to invest in better trans- titute. Medium households are differentiated PEN STUDY IN THE LOWER MEKONG SUB-REGION 67 POVERTY AND ENVIRONMENT IN CAMBODIA FIGURE 2.18 Poor vs. Non-Poor—Open Access and Natural Resource Dependent Livelihoods Subsist Poor Poverty Increase Invest in Resource income increased depletion resource extraction Non-Poor Other options or migration Time Poor: subsistence or some short-term gains Poor: continuing poverty, return to poverty, or alternatives Non-poor: larger short-term gains Non-poor: alternatives or possible move into poverty from poor and destitute households based on ous rural livelihood studies (Oxfam GB 2000; ownership of different �shing assets, both in Chan and Acharya 2002; McKenney and Prom terms of quantity and quality. The medium 2002; IFSR 2004; World Bank 2005b; www. households tend to have larger boats with larger ngoforum.org.kh). This case study found that engines, which enables them to go further out in the lake to �sh. These same households also stricter controls on forest product collection tend to have much better quality �shing gear imposed by concessionaires, and then the Forestry than do the poor/destitute households� (Prom Administration (FA), have caused considerable and Ballard 2005). hardship to local communities in Santuk dis- trict. These measures reduce the income of local Like the overexploitation of open access to for- households, to the bene�t of concessionaires, FA, est resources, open-access �shing leads to a rapid and a few non-poor households (who still have decline of �sheries resources. A study of �shing the means to harvest timber and simply bribe lots released in 2001 for open access �shing of�cials and guards when they are caught). More- notes: “Initial positive impacts from improved over, selective enforcement of restrictions (for access to �sheries are now perceived to be declin- example, outsiders are allowed access while ing due to unregulated competition for resources locals are not) likely intensi�es the scramble for and widespread use of illegal �shing gear� (Com- resources, as local people act to bene�t from munity Fisheries Department Of�ce 2004). rapidly declining resources in any manner they Strict controls on previously accessed resources can. For instance, in Santuk district, logging by can increase poverty and cause sub-optimal use of outsiders continues despite restrictions affecting resources. This impact has been noted in numer- local households. At the same time, locals have 68 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA identi�ed a number of informal ways of bene�t- SUMMARY AND ing from timber resources as well, but in a highly POLICY IMPLICATIONS inef�cient manner. Natural resource dependence and poverty The abundance and quality of natural resources in the areas surrounding the six villages is in 1. The poor are disproportionately dependent significant decline, suggesting the potential for on natural resources for their livelihoods. increasing poverty in the future. This is especially And this holds true in areas reporting the case for forest and �sheries resources. Vil- declines in the resource base. Problems lagers indicate that the decline has worsened in of poverty and resource decline appear recent years. While resource decline has nega- to be most pronounced in the plateau/ tively affected both poor and non-poor house- mountain region. In analyzing households holds, it was not possible to distinguish if one that are both engaged in natural resource- group has been disproportionately impacted. dependent activities and located in com- Compared to poor households, non-poor house- munes reporting resource declines, we find holds tend to collect more resources, often of twice as many households in the poorest higher value. As a result, a decline in resources quintile depend on these activities com- would likely reduce the total income of non- pared to the richest quintile. This suggests poor households by a greater amount than poor that if these resource bases continue to dete- households. But this may not move these house- riorate, the poor will bear a disproportion- holds into poverty, as they may still earn an ade- ate burden. quate income from natural resources or be able Policy implication: Targeting support of to redirect their labor to alternative income-gen- local resource management, land use plan- erating activities. Poor households earn less from ning, and agricultural and off-farm assis- natural resources than non-poor households, but tance should consider communes reporting these earnings are often a substantial proportion declines in forest cover, fisheries, and land of their total income. As a result, any reduction access, and concerns about future resource- in earnings from natural resources can cause based livelihoods, as well as poverty levels severe hardship, especially since poor households and other indicators (such as soil quality, often lack alternative income-generating options. water resources). BOX 2.2 Potential Policy Links: Ongoing Initiatives on Natural Resources • Policies to promote institutionalizing land and natural resource management at the commune level (e.g., Seila Program and Commune, Community Based Natural Resource and Environment Management (CCB-NREM)). • Initiatives aimed at improved productivity and diversi�cation of agriculture, off-farm opportuni- ties, and land, forest, and �sheries reform under the Rectangular Strategy and the National Strate- gic Development Plan (NSDP). The NSDP consolidates a number of past planning processes and documents, including the National Poverty Reduction Strategy (World Bank), Socio-Economic Development Plan (Asian Development Bank), and the UN Millennium Development Goals. • Reform efforts aimed at improving land, forestry, and �sheries laws; regulations; and enforcement mechanisms. Of particular importance are policy interventions for forest and �sheries concessions (especially addressing possible reform/termination, monitoring, and enforcement) and land distribution mechanisms (for example, social concessions). • Programs using an array of criteria for identifying target areas and communes (Seila, CCB-NREM, and various other rural development and conservation NGOs). PEN STUDY IN THE LOWER MEKONG SUB-REGION 69 POVERTY AND ENVIRONMENT IN CAMBODIA 2. Natural resources provide an important local management regimes, consider target- safety net for the poor, but not often a path- ing interventions where rapid resource extrac- way out of poverty. The poor are dispropor- tion is either beginning, and/or the resource tionately dependent on natural resources for base is not yet highly degraded, and encour- income/consumption, and these resources play age a combination of sustainable resource a critical role in helping households cope with management and investment of extraction crises, such as floods and droughts. Despite the earnings in agricultural and off-farm oppor- important role of natural resources in rural tunities (rather than even more extraction). livelihoods, income generation activities (�sh- 3. Natural resource management at the “ex- ing and forest product collection) appear to tremes� is not pro-poor; both highly restric- play little or no role in moving households tive regimes and open access management out of poverty. Indeed, �ndings from our case appear to result in negative outcomes for study in Kampong Speu on charcoal produc- the poor compared to the non-poor. Clearly, tion indicate households often use initial earn- restricting the rural population’s access to nat- ings from resource extraction to invest in more ural resources (as has been the case for many ef�cient extraction (equipment, transporta- forest, �shing, and economic/land concession tion), which in turn can lead to an acceleration areas) takes away productive resources that in the decline of the resource base. Both non- support rural livelihoods. While this can make poor and poor households may bene�t from the surrounding rural population worse off, this period of rapid extraction, though non- the impact can be especially severe for food- poor households tend to be better positioned poor households, which tend to be more to do so. As resources are depleted, however, dependent on these resources. many households may return to poverty, espe- But the converse to restrictive regimes— cially where other livelihood options are lack- “open access� (or a management vacuum)— ing, as is the case in Kampong Speu. The fact may not be much of an improvement for the that natural resource-dependent activities play poor over the long run. First, the poor tend to little role in moving households out of poverty be less able to take advantage of open access to is unlikely to change, as income from these exploit resources as pro�tably as non-poor activities is falling as a proportion of house- actors, because they lack the capital means hold income, and this is occurring across all (equipment, transportation). Second, “open households, rich and poor. The fall in income access� generally is not fully open, but rather reflects an overall decline in the resource involves paying a range of informal fees for base, and in some areas increasing control access—a disproportionate burden for the over resources by private concessions and poor. Lastly, the poor tend to be more depen- powerful interests (as illustrated by our case dent on natural resources and have fewer alter- study site in Kampong Thom). native options for income generation. As a Policy implication: Natural resource man- result, when open access exploitation leads to agement should be focused on developing resource decline, the impact on livelihoods is local management regimes suitable for main- again felt disproportionately by the poor. taining resources and providing access for the Policy implication: The management poor/vulnerable, not aimed at industrial- vacuum created by past cancellations of for- level extraction, which has a history of dismal est concessions and fishing lots needs to be results for the environment, local communi- addressed with greater support for natural ties, and government revenue. In establishing resource assessments, setting priorities for 70 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA management (for example, rural poor, conser- holds using water from ponds, rivers, and vation), and promoting locally appropriate streams. However, only a modest proportion management regimes. Remaining concessions of non-poor households using these unsafe in Cambodia still claim a large proportion of sources boil their water— 40 percent using the richest resource areas. These (selectively) water from unprotected wells and 57 per- restrictive management regimes should be cent using water from ponds, rivers, and reformed to improve appropriate access for streams. local communities (with ongoing indepen- Policy implication: Consider education and dent monitoring), or the concessions should awareness programs to encourage the boiling face termination. It is critical, however, that of drinking water, especially for households reforms and terminations do not lead to man- using unsafe sources. Although poor house- agement vacuums (as has been the case in the holds using unsafe sources are more at risk past), which invite ad hoc control by powerful due to their lower levels of boiling, even a interests. There must be support for resource considerable proportion of non-poor users of assessments, consultation and priority-setting, unsafe sources do not boil. This suggests and establishment of fair and enforceable poverty itself is not a suf�cient explanation management. for the lack of boiling, and there may be a potential for signi�cantly reducing drinking Drinking water wources and poverty water-related health impacts through greater 4. The poor are disproportionately depen- education and awareness programs targeting dent on the use of unsafe water sources. this segment of the population. Our analysis indicates households in the poorest consumption quintile are three Sanitation and poverty times as likely to use unsafe water sources 6. Access to basic sanitation is lacking across (unprotected dug wells and ponds, rivers, all rural areas. In urban areas, 65 percent of and streams) as households in the richest households in urban areas have access to basic quintile. sanitation, with 30 percent connected to sew- Policy implication: In targeting service erage. But such sanitation infrastructure/ provision for drinking water, consider factors services are dearly lacking in rural areas, where identi�ed in this analysis as associated (at sta- only 20 percent of households have access to tistically signi�cant levels) with the use of sanitation and less than two percent are con- unsafe drinking water sources. These include nected to sewerage. The problem is even worse households located in the coastal or plateau/ for the rural poor—only seven percent have mountain regions, as well as rural areas far access to basic sanitation and less than 0.5% from district and provincial capitals and all- are connected to sewerage. weather roads. Additional factors include Policy implication: With so much of the households that are poor and have heads of rural (poor) population lacking access to households with no schooling. basic sanitation, targeting based on poverty 5. Households accessing unsafe water sources is unlikely to be meaningful. To make are the least likely to be boiling their water. progress at such a broad scale, improving This is especially true for poor households, basic sanitation in rural areas needs to be where boiling is done by less than 25 percent elevated as a national policy priority, with of households using water from unprotected appropriate technical and resource support wells and only about 40 percent of house- for action. PEN STUDY IN THE LOWER MEKONG SUB-REGION 71 POVERTY AND ENVIRONMENT IN CAMBODIA Natural disasters and poverty areas have higher poverty rates, demining these 7. Both droughts and floods show a statisti- areas should be given higher priority. cally signi�cant relationship to poverty. But whereas droughts are associated with Endnotes poverty, floods are associated with not 1. The CSES is the largest and most extensive multi-objec- being poor. Droughts and floods are a com- tive survey undertaken in Cambodia, involving a sample mon problem for Cambodian households. of 15,000 households. Cambodia’s baseline poverty line For the �ve-year period from 1999 to 2003, consists of a single national food poverty line and three regional nonfood allowances. Poverty estimates are based 63 percent of households report being on per capita consumption. affected by a drought and 53 percent report 2. A sharp and statistically signi�cant increase in the Gini being affected by a flood in at least one year. coef�cient occurred in rural areas—from 0.27 in 1993– Whereas poorer households tend to be more 94 to 0.33 in 2004. 3. The study assesses how poverty has changed for often affected by drought than richer house- 1,000 households in nine villages between 2001 and holds, the reverse is true for households 2004. The villages were selected to represent Cambo- affected by floods. This may reflect the chal- dia’s different eco-regions. To capture mobility, the lenges of de�ning “flood� (which depending study groups households into seven categories: (1) com- on its nature may be viewed as good or bad fortably rich, (2) climbing into wealth, (3) escaping poverty, (4) static middle, (5) chronic poor, (6) deepen- for rice farming), as well as the likelihood ing poverty, and (7) falling into poverty. that richer households own higher quality 4. Key datasets include CSES 2004, Seila and Danida rice land (in flood-prone areas). 2005, and National Level 1 Survey 2002. Policy implication: In targeting problems 5. Rural communes account for nearly 1,500 out of just over 1,600 communes in Cambodia. affecting the poor, developing appropriate 6. This section summarizes key case study �ndings; more disaster responses to drought problems should detailed case study �ndings, data, tables, and informa- be elevated to a higher priority. This is not to tion can be found in the Annex. diminish the importance of addressing floods, 7. “Koyun� is a Khmer word referring to a machine intended to be used for pulling a plow, but it is also com- but only to recognize that droughts dispro- monly used for transportation. To a degree, it is a mech- portionately affect the poor and therefore anized substitute for draft animals and an oxcart. deserve due attention in poverty reduction strategies. References Andersen, Henny. 2003. “Poverty and Natural Resources Mine/UXO Contamination and Poverty and the Environment in Cambodia.� Phnom Penh: Royal Danish Embassy (Danida). 8. Mine and cluster bomb contamination Asian Development Bank. 2001. Participatory Poverty show a strong relationship with poverty. Assessment in Cambodia. Manila: ADB. The poorest household consumption quin- Cambodia Development Resource Institute. 2006. “Moving tile is almost four times as likely to be affected Out of Poverty Study: Cambodia National Synthesis Report.� Phnom Penh: CDRI. by mine contamination, and twice as likely to Chan, Sophal, and Sarthi Acharya. 2002. Facing the Challenge be affected by cluster bomb contamination, as of Rural Livelihoods: A Perspective from Nine Villages in the richest quintile. Cambodia. Working Paper 25. Phnom Penh: Cambodia Policy implication: In targeting areas for Development Resource Institute (CDRI). demining, village poverty rates (as identi�ed in Chan, Sophal, Tep Saravy, and Sarthi Acharya. 2001. Land Tenure in Cambodia: A Data Update. Working Paper 19. the CSES 2004 data) should be considered as Phnom Penh: CDRI. an additional element of the targeting criteria Community Fisheries Department Of�ce. 2004. “Impacts currently used. Where villages in contaminated of Fisheries Policy Reforms: Kampong Cham, Pursat 72 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN CAMBODIA and Takeo Provinces.� Phnom Penh: DFID and McKenney, Bruce, and Prom Tola. 2002. Natural Resources IMM, Ltd. and Rural Livelihoods in Cambodia: A Baseline Assess- Conway, Tim. 2005. “Identi�cation and Targeting of the ment. Working Paper 23. Phnom Penh: CDRI. Poor—Basic Principles.� Presentation to the National McKenney, Bruce, Yim Chea, Prom Tola, and Tom Evans. Forum on Identi�cation of Poor Households, Phnom 2004. Focusing on Cambodia’s High Value Forests: Liveli- Penh, February 15. hoods and Management. Phnom Penh: CDRI and Council for Social Development. 2005. “Report on Pro- Wildilfe Conservation Society. ceedings and Results of the National Forum on Identi- Ministry of Planning. 2005. National Strategic Development �cation of Poor Households.� Draft (March 4, 2005). Plan for Poverty Reduction 2006–2010: Process for Prepa- Phnom Penh: Ministry of Planning. ration. Phnom Penh: Ministry of Planning. Dasgupta, Susmita, Uwe Deichmann, Craig Meisner, and Murshid, K. A. S. 1998. Food Security in an Asian Transi- David Wheeler. 2002. The Poverty-Environment Nexus tional Economy: The Cambodian Experience. Working in Cambodia, Lao PDR, and Vietnam. Washington, DC: Paper 6. Phnom Penh: CDRI. World Bank. NGO forum. www.ngoforum.org.kh. Degen, P. F., N. Van Acker, N. van Zalinge, N. Thuok, and OXFAM Great Britain. 2000. The Cambodia Land Study D. Loeung. 2000. “Taken for granted: Conflicts over Project (Vols. 1–3, 1999; Vol. 4, 2000). Phnom Penh: Oxfam Great Britain. Cambodia’s freshwater �sh resources.� Paper presented Prom, Tola. and Brett M. Ballard. 2005. “Interim Report 1 at 8th IASCP Conference, Bloomington, Indiana, May for the Tonle Sap Participatory Poverty Assessment 31 to June 4. Project.� Phnom Penh: CDRI. Dümmer, Ignas. 2005. “Information and Knowledge Based Royal Government of Cambodia. 2004. “The Rectangular System for Land Suitability Analysis in Cambodia.� Pre- Strategy for Growth, Employment, Equity and Ef�- sented at the Land Resource Assessment Forum, Phnom ciency in Cambodia.� Paper presented by Prime Minis- Penh, September 14–17. ter Samdech HUN SEN and the Third Legislature of Fraser, Thomas, GFA-AGRAR, and ANZDEC. 2000. the National Assembly at the Of�ce of the Council of “Cambodia Forest Concessionaire Review Report.� Ministers, Phnom Penh, July 16. Phnom Penh: Sustainable Forest Management Project, Seila and Danida. 2005. A National Opinion Poll: Commune Asian Development Bank. council’s perception of its natural resource base and liveli- Global Witness. www.globalwitness.org. hood options. Phnom Penh: Danida. Independent Forest Sector Review. 2004. The Forestry World Bank. 2004. Cambodia at the Crossroads: Strengthening Sector in Cambodia. Phnom Penh: Joint Coordinating Accountability to Reduce Poverty. Report No. 30636-KH. Committee—Donor Working Group on Natural East Asia and the Paci�c Region: World Bank. Resources and the Forestry Administration. World Bank. 2005a. From peace to prosperity: an assessment Knowles, James C. 2005. A New Set of Poverty Estimates of poverty in Cambodia. Phnom Penh: World Bank. for Cambodia, 1993/4 to 2004. Report to the EAS World Bank. 2005b. Cambodia Rural Sector Strategy Note: Country Units of the World Bank. Washington, DC: Towards a Strategy for Rural Growth and Poverty Reduc- World Bank. tion. Phnom Penh: World Bank. McKenney, Bruce. 2002. “Questioning Sustainable Con- Zohir, Sajjad. 2005. Report on CMDGR 2005 Update Exer- cession Forestry in Cambodia.� Cambodia Development cise: Undertaken during 25 May–15 June 2005. Phnom Review 6 (1): 1–7. Penh: UNDP. PEN STUDY IN THE LOWER MEKONG SUB-REGION 73 3 Poverty and Environment in Lao PDR GEOGRAPHIC DIMENSIONS and clothing, permanent housing, access to OF POVERTY transportation, and inability to meet health and education expenses. In this context the Govern- Over the last decade, Lao PDR’s economic ment identified 46 districts of first priority for growth averaged nearly 6 percent per year and poverty reduction.2 Many of these districts are was accompanied by signi�cant poverty reduc- also heavily contaminated by unexploded ordi- tion. Poverty incidence declined from 46 percent nance (UXOs) from the Vietnam War (Map 3.1). in 1993 to 33 percent in 2003.1 At the same time, In the eyes of the poor themselves, poverty there was a substantial narrowing in poverty inci- includes rice insuf�ciency, lack of large animals, dence among provinces and the highest provin- and vulnerability to ill health (ADB 2001). cial poverty incidence came down from over From these broader perspectives, improving our 70 percent in 1993 and 1998 to less than 55 per- understanding of poverty-environment linkages cent in 2003 (Map 3.1). is vital for the development of more sustainable Lao PDR is still, however, one of the poorer livelihoods and an improved quality of life, both countries in Asia, with a GDP per capita of $425 in terms of natural resource management and in 2004 (World Bank, 2005). Substantial urban- environmental health risks. rural divides remain, and there are large geo- The poorest districts in Lao PDR are char- graphic variations in poverty and development. acterized by very sloped land, relatively low More than 85 percent of the poor live in rural areas population density, and—particularly in the with rural poverty incidence at 38 percent in 2003, South—by ethnic minorities. Many communi- in contrast to 20 percent in urban areas (LECS III). ties in these districts are small and remote, with Poverty, as discussed above, is de�ned as con- limited access to roads and markets and improved sumption or expenditure below a minimal level. water supply and sanitation, and a high reliance This provides a quanti�able measure useful for on natural resources for their survival. In the monitoring progress over time and for identifying central and southern part of Lao PDR, it is the geographic areas with persistent poverty in need poorest districts that also have the highest for- of focused attention and interventions. Poverty est cover (Map 3.2). can also be de�ned in broader perspectives. The The National Growth and Poverty Eradica- Government of Lao PDR de�nes poverty as the tion Strategy (NGPES) of the Government of lack of basic necessities—such as inadequate food Lao PDR establishes a link between poor districts PEN STUDY IN THE LOWER MEKONG SUB-REGION 75 POVERTY AND ENVIRONMENT IN LAO PDR MAP 3.1 Poverty Incidence and Government Priority Districts for Poverty Eradication Provincial poverty District poverty Government priority Percent of villages Incidence 2003 incidence 1998 districts with high UXOs Sources: and upland districts, but there is no of�cial de�- 200 m, but only 28 percent of the 46 priority nition of upland districts in Lao PDR and no poor districts are among them. Slope, however, is direct correlation between elevation and poverty. a variable that may describe poor upland districts Some 55 percent of the 142 districts in Lao PDR better since about half of the districts with higher have a maximum elevation between 500 m and poverty incidence than the national average have MAP 3.2 Geographic and Demographic Characteristics of Lao PDR Sources: 76 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR more than two-thirds of their land area with slopes Phongsaly, Houaphanh, and Saravane—and higher than 16 percent (Maps 3.1–3.2). along the border with Vietnam in the South Grouping the districts in Lao PDR by poverty (Map 3.1). A comparison of these districts to the incidence provides some perspectives on poverty- 46 districts with highest poverty incidence in environment linkages (Table 3.1). The two groups 1998 is revealing.8 The government’s priority dis- of districts with the highest poverty incidence are tricts have (a) the lowest rate of water supply cov- less populous, but have the largest number of erage and highest reliance on open unprotected poor.3 In these districts, forest cover was lower and surface water; (b) the highest reliance on fuel deforestation higher during the 1993–97 period.4 wood; (c) the highest share of sloped land (slope Per capita agricultural land and rice production greater than 16 percent); (d) the highest crude was substantially lower, and signi�cantly less flat mortality rate; (e) the lowest literacy rates; and land is available for cultivation.5 The districts’ (f) the lowest rice production per capita. The share of population with sanitation facilities was only environmental indicator that is higher in substantially lower, and reliance on open unpro- 46 poorest districts is the deforestation that tected surface water was higher.6 Crude mortality occurred from 1993 to 1997. rates were also higher, and literacy rates consider- The priority districts show high consistency ably lower. Access to roads was also lower in these with the government’s de�nition of poverty and poor districts.7 overlap closely with many environment-poverty The government of Lao PDR has identi�ed priorities. NGPES, which initiated the �rst 46 priority districts—or nearly one-third of the poverty reduction program with a geographical country’s districts—for poverty reduction and focus on 46 priority poverty districts, identi�es development (NGPES). Many of these dis- the environment as one of the key inter-sectoral tricts are in the poorest provinces—Oudomxay, priorities in the �ght against poverty. For forest TABLE 3.1 District-Level Poverty Pro�le District poverty incidence (1998) 0–25% 25–41% 41–64% 64–96% All Lao PDR* Number of districts 32 32 32 32 128 Population (million) 1995 Census 1.4 1.2 0.9 0.9 4.4 Population (million) 2000 Estimate 1.7 1.4 1.0 0.9 5.1 Population in poverty 1998 (million) 0.27 0.47 0.52 0.70 1.96 Percent of total area with slope >16% 41.9 50.5 66.0 70.3 59.5 Rice production per capita (kg/person) 280.4 264.5 229.4 188.1 247.3 Large animals per household 1.84 2.75 2.52 2.59 2.36 Forested land % of total land 1997 39.4 43.8 37.7 34.6 38.6 Deforestation rate of forested land 1993–97 2.9 2.4 2.7 3.2 2.8 Agricultural land % of total land 1997 12.2 8.3 3.2 1.7 5.5 Percent of households relying on open 24.6 53.8 63.4 74.2 49.9 unprotected surface water in 1995 Percent of households without toilet 52.0 76.3 76.6 84.3 69.7 facilities in 1995 Crude death rate in 1995 6.8 10.6 12.1 14.7 10.5 (per 1000 population) Percent of population with education 72.9 58.3 49.7 40.3 57.7 in 1995 Sources: *ADB (2001) only presents poverty estimates for 128 districts (of a total of 142 districts). PEN STUDY IN THE LOWER MEKONG SUB-REGION 77 POVERTY AND ENVIRONMENT IN LAO PDR resource degradation and issues related to agri- and urban air pollution remains moderate and cultural land, however, it is necessary to con- nowhere close to the levels in many large cities in sider a broader focus than the priority group of South-East Asia. Access to a road, an important districts. determinant of, for instance, environmental ser- vices, such as water supply and sanitation, is sig- ni�cantly lower among the poor population. POVERTY-ENVIRONMENT ISSUES Suf�cient data is not available to provide an A set of environmental issues in relation to estimate of the number of poor and non-poor poverty can be estimated from available data, per- affected by environmental issues related to nat- taining to environmental health, natural disas- ural resources such as forests, non-timber forest ters, and the urban environment (e.g. outdoor air products, or land. In agriculture, however, rice pollution). The disparity between poor and non- suf�ciency and livestock holdings are de�ned by poor households is particularly substantial with the poor themselves as principal indicators of respect to lack of potable water supply, basic san- poverty in Lao PDR (ADB 2001). Illness and itation, and UXO contamination. While a lesser lack of health services were also raised by the number of people are affected by malaria and poor as central issues. drought, the poor are also disproportionately In the context of this study, poverty dimen- affected by these problems.9 Flooding seems to sions are those factors that affect rice suf�ciency affect a larger share of the non-poor population and livestock holdings, which include, but are than the poor. Solid fuels are the main sources of not limited to, quality and quantity of natural household energy for cooking both among the resources. Indicators for groups of districts by poor and non-poor population. As more than poverty incidence con�rm the importance of 85 percent of the poor reside in rural areas, and rice suf�ciency (Table 3.1). However, the analy- urban poverty incidence is “only� 20 percent, sis does not apply to animal numbers: poorer urban environmental problems, such as outdoor districts have higher overall livestock numbers, air pollution, are more an issue for the non-poor even though the poor within these districts may population. Urbanization is still low in Lao PDR, often be those with fewer or no animals. TABLE 3.2 Number of Poor and Non-Poor Affected Estimated number of people affected Poverty & environment indicator: % of non-poor population % of poor population Lack of water supply (using open water sources)* 20 30 Lack of sanitation (no latrine)* 40 70 UXO Contamination (1997) 20 28 Household use of Solid fuels/indoor air pollution* 95 98 Malaria Incidence (con�rmed cases 1999–2001)** 0.2–0.4 1.0–2.0 Flood (1995–97 and 2002)*** 5 1 Drought (1995–97)*** 0.2 1.7 Urban environment na <10–15 No Access to Road* 15 30 Source: Estimated using a variety of statistical techniques from the following data: *LECS III 2002/03, **WHO data, ***Natural Disaster Management Of�ce data. ****Less than 15 percent of the poor reside in urban areas (LECS III 2002/03). 78 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR PEN Study Focus Areas The PEN focus areas cover three of the four NGPES policy sectors: transport, health, and The PEN study set out to deepen the under- agriculture and forestry. None of the focus areas standing of poverty and environment linkages in dealt directly with education. However, it became Lao PDR by examining their relationship at the evident during the course of the PEN study that province, district, and household levels and to education—and more broadly information- and mainstream the �ndings into the national policy awareness-building—is an essential ingredient agenda. The scope of the study was designed to �t in all �ve PEN focus areas. One of the focus into the national policy framework for poverty reduction. This includes the National Growth areas relates to the Lao UXO program, which is and Poverty Eradication Strategy (NGPES) and one of three poverty-related national programs the National Poverty Reduction Strategy Paper included in NGPES. (PRSP), published in 2003. A two-pronged approached was followed in The main national partners of the PEN study the study of the PEN focus areas. A nationwide in Lao PDR were (a) the Committee for Planning analysis was carried out using secondary data at and Investment (CPI); and (b) the Environmental province, district, village and household level, Research Institute (ERI) of STEA (Science, Tech- and �eld surveys were undertaken in four dis- nology & Environment Agency), which is the tricts. The nationwide analysis and �eld surveys national agency in charge of environment directly informed each other, strengthened the con�dence under the prime minister. In addition, a number in the PEN �ndings, and provided wider geo- of other government agencies participated in the graphic relevance. The �eld surveys were imple- implementation of the study and provided critical mented by a study association led by ERI/STEA, data for the study (Table 3.3 and Box 3.1). comprised of research groups from the national Dialogue with these partners—and �ndings ministries and their agencies. from the initial PEN analysis (e.g. an extensive On the recommendation of the Government PEN study background report and Table 3.2)— of Lao PDR, the �eld surveys took place in two led to the selection of �ve focus areas, three of provinces in the far north of the country and in which are related to natural resources and two in two provinces in the far south. The government the area of water supply and sanitation (Table 3.3). selected one district in each province for the rural Four of the focus areas relate to rural develop- �eld surveys and the two smallest provincial cap- ment and one to urban development, reflecting itals for the urban �eld surveys (Table 3.4). a rural population share of more than 75 per- All four districts where the rural focus stud- cent in Lao PDR. ies took place are located in the upland districts. TABLE 3.3 PEN Focus Studies and Implementing Agencies Poverty and environment focus Implementing agency Natural Resources NTFP management National Agro-Forestry Research and Livelihood Institute MAF) Road access and development Environment and Social Division of MTCPC UXO contamination Environmental Research Institute (STEA) Environmental Health Rural water supply| and sanitation Department of Hygiene (MoH) Urban water supply and sanitation Urban Research Institute (MTCPC) PEN STUDY IN THE LOWER MEKONG SUB-REGION 79 POVERTY AND ENVIRONMENT IN LAO PDR BOX 3.1 Methodologies and Data The methodology employed in each of the �ve focus studies entailed the combination of national mapping and correlation analysis of poverty and environment indicators with village and house- hold surveys. The research teams, especially those working on the rural �eld surveys, were con- fronted with substantial challenges in carrying out local research in remote locations, so the methodology was carefully adapted to overcome these challenges (see chapter 1 for more detail). In addition, the study team took great care to ensure consistency between the indicators collected in the national databases and those collected through the surveys. The study utilized precise sampling criteria for villages and households. Poverty levels in the surveys were assessed through a scoring system based on both productive and non-productive household assets (asset indicators, in turn, were collected at the household and village levels). Three of the PEN focus studies conducted comparative analysis of poorer and less-poor house- holds, while the two others completed a simple comparison at the village level only. For some of the variables, information could not be updated from the �rst phase of the PEN study conducted in 2003–04. For example, the 1993–97 forest data were used, even though pat- terns of deforestation may have changed signi�cantly in the last eight years. Selected variables at district level (ADB 2001) from the LECS II (1997/98) surveys were analyzed in correlation with vari- ables from the PEN database. While the LECS III data provided further insights into some of the same indicators at household and province level, reliable district level data have not been pro- duced from LECS III. Moreover, the new Population Census 2005 can be an important source of analysis of PEN linkages. Partner Data assistance National Statistical Center (NSC) LECS III, and economic and demographic data MCTPC Road network data at district level Lao UXO program of�ce UXO contamination data from the national UXO survey in 1997 NAFRI NTFP database for 39 districts (the PEN focus study) URI Urban water supply for several provinces Environmental Health Division (MoH) Water supply breakdowns in rural villages TABLE 3.4 PEN Field Survey Districts and Towns District and towns Location Poverty incidence (1998) Notou district Phongsaly province (North) 79% Phongsaly town Phongsaly province (North) na Namor Oudomxay province (North) 93% Kaleum district Sekong province (South) 63% Lamam town Sekong province (South) na Phouvong Attapeu province (South) 71% National average 39% Source: District poverty incidence is from ADB (2001) based on LECS II. Poverty incidence is not available (na) for the two towns. 80 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR They are among the poorest within the group of POVERTY AND NATURAL 46 priority poor districts identi�ed in NGPES. RESOURCE MANAGEMENT The proportion of ethnic minority people is Close to 40 percent of land area in Lao PDR was higher than 98 percent in three of the selected classi�ed as forest in 1997, according to satellite districts, and is 72 percent in the fourth. The imagery data. Agricultural land, including fallow Mon-Khmer ethno-linguistic group accounts land, occupied 13 percent of the total territory. for more than half of the district population in Shrub and grassland constituted 43 percent of three of the districts. total land area (differentiating between fallow All four districts border either China or Viet- and shrub and grassland, however, is subject to nam, and two of the borders are located at high debate). There are substantial differences in land elevations. The marginalization of these districts use and vegetation across the three regions in is decreasing, however, as international roads and Lao PDR. Forest land was 65 percent of total borders are opening up. area in the South, 48 percent in the Center, and As the PEN �eld surveys focused on the upland only 21 percent in the North. Evergreen forest districts listed as priority poor districts in NGPES, constituted 36 percent of total forest (Map 3.3). the surveys have a strong de facto linkage to the on- Deforestation in 1993–97 was highest among going national policies of rural development based the districts with the highest poverty incidence, on “focal areas� and of stabilization of shifting and, geographically, in the northern provinces, cultivation. These policies are discussed in the which have some of the highest poverty incidence context of individual PEN focus areas and at the and lowest forest cover. A total of 1,118 km2 of end of the chapter. forest was lost during these four years in the MAP 3.3 Forest Cover 1997 and Deforestation 1993–97 Sources: PEN STUDY IN THE LOWER MEKONG SUB-REGION 81 POVERTY AND ENVIRONMENT IN LAO PDR North, representing more than 5 percent of nam in terms of total forest cover, that is, dif- remaining regional forest cover in 1997. There ferences in overall forest cover between border are at least two important dimensions to this sit- districts and the rest of the country are all uation. First, these �ndings alone are not suf�- explained by the aforementioned factors. Dis- cient to conclude that poor people’s livelihoods tricts bordering Vietnam do however have more are signi�cantly affected by this deforestation, evergreen forest than can be explained from the and even less so that the poor are responsible for factors discussed above, and districts bordering some of this deforestation. Second, the �ndings Thailand have less evergreen forest. refer to 1993–97, and patterns of deforestation The higher household poverty incidence in may have changed signi�cantly in the last eight districts bordering Vietnam could be a particular years. For instance, losses and scarcity of ever- concern for PEN linkages. These districts still green forest are a particularly growing concern have higher rates of forest cover than the average arising from the PEN focus study of NTFPs. district in Lao PDR. A poverty reduction strategy Demographic, socioeconomic, and geographic that opens the Vietnam border districts with road factors seem to have influenced the long-term linkages to Vietnam may want to pay particular trends in forest cover in Lao PDR.10 Population attention to forest conservation objectives. density and agricultural land area are associated These findings confirm the existence of a with lower forest cover. Road access and road den- poverty and environment nexus between poverty, sity is not found to have a negative effect. Steep- forestry, and land in Lao PDR: there are strong ness of land is found to be negatively associated linkages between indicators, poverty groups, and with forest cover. This is particularly the case for regions, but causal relationships that would help land with steepness above 30 percent. No asso- explain these factors are not identi�ed through ciation between overall forest cover and flat land this analysis. is found beyond the influence of population Many factors potentially affecting poverty density and agricultural land. Conversely, flat- are directly or indirectly related to natural ness of land is more influential in explaining resources (Figure 3.1). The largest difference in lower evergreen forest cover than population poverty incidence (i.e. substantially lower inci- density and agriculture. dence) is associated with literacy, road access, Compared to the Central region, the North- non-agriculture as main income source, and ern region is found to have signi�cantly less and urban household location. Poverty is also lower the Southern region more forest cover than in communities with development projects and expected from steepness of land, population land and forest allocation programs, and those density, and agriculture. No “border effect� was communities that have less UXO contamination found for districts bordering Thailand or Viet- and more irrigated land. Proximity to district cen- TABLE 3.5 Poverty & Environment Nexus in Natural Resource Management PEN areas: Indicators: Poverty and geographic analysis: Poverty Poverty incidence 2002/03 Somewhat higher in the NORTH Highest in districts bordering Vietnam Forest Deforestation (% and km2) 1993–97 Highest in the NORTH and POOREST DISTRICTS Evergreen forest cover 1997 High districts bordering Vietnam Land Agricultural land per capita 1997 Lowest in the NORTH and POOREST GROUP Sloped land (% of total land) Highest in the NORTH and POOREST GROUP 82 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.1 Poverty Incidence in Relation to Select Indicators Development projectin village Read (>50%ofHH members) UXO problem (<50% of villages in district) Road access Main income source (non-agriculture) Irrigated land Landallocation program in village Urban 0% 10% 20% 30% 40% 50% 60% YES NO Source: Cross-tabulations from household and village data in LECS III 2002/03. ter and residing in the provincial capital district is incomes out of agriculture and poverty reduction. also associated with lower poverty incidence. In Poverty in relation to UXO contamination and contrast, poverty incidence in the Government’s road access is further discussed later in the report. 46 priority districts and in districts bordering to New information about villages with land Vietnam are higher than can explained by the and forest allocation programs and development factors discussed above.11 projects, and other factors was also revealed from As many of the factors are associated with the analysis: urban-rural household location, an analysis was undertaken for rural households only. The analy- • More than 40 percent of the villages in the sis con�rmed that all the factors are associated LECS III survey reported that land and forest with poverty incidence even within rural areas. allocation programs had been implemented The lower poverty incidence among house- there. Poverty incidence in these villages is holds that have irrigated land and that are not signi�cantly less than in villages without allo- dependent on agriculture as their main source of cation programs, even when comparing villages income deserves further analysis. The �rst indi- within rural areas. Conversely, the Participatory cates a relationship between more intensive agri- Poverty Assessment (ADB 2001) indicates that culture and poverty reduction, while the second there are signi�cant trade-offs between protec- indicates a relationship between diversi�cation of tion efforts and poverty reduction in the short PEN STUDY IN THE LOWER MEKONG SUB-REGION 83 POVERTY AND ENVIRONMENT IN LAO PDR to medium term in many districts. Many com- NTFPs as having substantial potential to improve munities depend heavily on the forest for their rural livelihoods in the poorest districts. There is a livelihood, and many villages reported that two-way bene�cial relationship between forest they no longer have suf�cient access to forest conservation and sustainable use of NTFPs. Mar- resources after the implementation of allocation keted NTFPs, in addition to self-consumed ones, programs. have historically been important export products • Nearly 50 percent of surveyed villages reported and remain so today. In contrast, the regulatory having an ongoing development project at the framework is vague, and only anecdotal quanti�ed time of the survey. Poverty incidence in these information on NTFPs is available. There is an villages was lower than in villages without a acknowledged risk of resource decline. The PPA project. However, a development project had (ADB 2001) reported that the quantity of NTFPs a somewhat stronger association with higher was declining in many village areas. Pilot rehabili- household consumption than lower poverty tation or cultivation projects are being initiated, incidence. This could suggest that develop- but a clear policy statement is missing. ment projects may be benefiting non-poor The government’s interest in NTFPs is closely households more than poor households, but related to the stabilization policy for shifting cul- further assessment of this issue is needed. tivation. NTFPs are often cited as an alternative • Last but not least, ethnicity, village migration, to shifting cultivation—for example, in villages and market access were not on their own asso- consolidated into “focal areas.� The relationship ciated with lower or higher poverty incidence. between sustainable NTFP management and swidden agriculture12 or village consolidation These findings confirm the complex linkages appears, however, to be a complex one, as there between poverty and natural resource manage- is contradictory evidence from �eld research. ment. Poverty and renewable natural resource Some researchers argue that most NTFP species linkages must be seen in the context of broader require some type of human intervention for economic development. These issues are highly maintenance. Swidden cultivation in the form complex and interdependent. They involve the of rotational agriculture would, therefore, have long-practiced traditional agricultural production a positive impact on several NTFP species. systems, forest protection and agricultural land Others, on the other hand, argue that swidden- allocation, availability and access to NTFPs, the ing in a context of population growth inevitably village consolidation program, and development impacts NTFP resources through a reduction in and participation in a market economy. Natural old-growth forest. NTFPs are, therefore, an entry resources are used not only by rural households, point to research poverty, environment, and but also by other stakeholders. No information shifting cultivation. can be derived on the latter from the datasets ana- The PEN study of NTFPs was designed to lyzed. These issues all deserve further quantitative answer two questions. First, the study looked analysis in combination with local research. at two potential NTFP development strategies: (1) sustainable harvest of natural NTFPs, and (2) NTFP cultivation. The objective was to seek Non-Timber Forest Product evidence about the viability of these strategies, Resources and Poverty especially for poor households. Second, the PEN Non-timber forest products (NTFPs) are impor- study explored the relationship between village tant to the vast majority of rural households in the migration and consolidation policies and the pres- uplands of Lao PDR. The Lao government views sure on NTFP resources in poor communities. 84 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR NTFPs in Lao PDR are not incorporated NTFPs of National and into a statistical collection system. A collection Regional Importance of data and information was therefore under- The districts reported altogether 37 different taken from the districts. A total of 39 districts marketed NTFP species, with three species on responded with detailed data and information on predominant types of NTFPs, trends in their average considered as very important (Map availability in district forests over time, trends in 3.5). Among these, eight are important mar- market demand, and other valuable informa- keted species in at least two among the three tion. Data is available for most districts in three regions (North, Center, and South) and five in provinces in the North, one in the Center, and one one region (Table 3.6). Cardamom was cited in the South. Only 15 of the respondent districts as important in almost 70 percent of the dis- are among the 46 priority districts (Map 3.4).13 tricts, and rattan in almost 60 percent. Only The large number of NTFP species used in benzoin, an important local species, was left Lao PDR is an important reason for the lack of out of this analysis, because the producing dis- clarity in the sector. The survey led to the iden- tricts in northeastern Lao PDR did not answer ti�cation of 13 major species, on which the the survey. analysis subsequently focused. Finally, NTFP While NTFP species are harvested from three incomes and the importance of swidden prac- clearly differentiated domains—old-growth tices are two topics that are dif�cult to address in natural forests, secondary forests, and young rural surveys, since households may be wary of fallows—classifying individual species into one taxation and the swidden control policy. Proxy category is not easy. Only three of the 13 main indicators were assembled in addition to these. species were reported as fully harvested from the The analysis was, however, able to use survey “natural forest� domain. On average, four main �ndings on NTFP incomes and upland rice NTFP species were reported as forest species and areas after data consistency was con�rmed. two as fallow species. MAP 3.4 NTFP Survey Districts in Relation to Ethnic Minorities and 46 Priority Districts PEN STUDY IN THE LOWER MEKONG SUB-REGION 85 POVERTY AND ENVIRONMENT IN LAO PDR MAP 3.5 NTFPs of National and Regional Importance Reported by Each District NTFPs on Fallow Land NTFPs in Forest Source: PEN study (NAFRI). All main species identi�ed, except for broom Northern districts reported an active market grass and paper mulberry, are mostly harvested trend, especially close to the China border. Price from natural forest. A marked contrast is observed increases were noted for forest species, with between forest species and fallow species. The declining populations and decreasing marketed districts reported a decline in resources of most quantities. Conversely, the two fallow species NTFP species, particularly forest species, and showed an increase in population, marketed only reported an increase for the two fallow quantities, and prices. species (Map 3.6). Declining quality was also Some domestication, predominantly in the reported, especially for species. District respon- Northern provinces, was reported for seven of the dents primarily attributed resource degradation to 13 main species and for �ve secondary species. overharvesting and shifting cultivation. However, only one species—eaglewood—was TABLE 3.6 Commercially Important NTFPs in 39 Districts 8 Species of national relevance 5 species of regional relevance Cardamom, rattan, orchid, bong bark, dammar, Malva nut, galangal, sugar palm fruit, eaglewood, bamboo shoots, broom grass meuak bark, paper mulberry 86 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR reported as partly cultivated by a larger majority MAP 3.6 Number of Increasing/Decreasing Main of the districts (about 90 percent). Paper mul- NTFP Species berry, which grows spontaneously in fallows, ranked second in cultivated species. Cardamom Increase Decrease and galangal, two species that are easy to culti- vate, are harvested from the wild in around 70 percent of districts. NTFPs in the Household Economy About 96 percent of households in the northern districts and 72 percent in the South reported cash income from NTFPs every year. Households reported average annual NTFP sales of US $73 in the Namo District in the North and US $22 in the southern district. This is respectively 16 percent and 5 percent of average cash expenditures in the provinces in the LECS III survey. However, the fact that households ranked NTFPs as their sec- Distance to NTFP collection sites increased ond highest income source—after food crops and markedly during the last 10 years, from less before animal sales—would indicate much higher than 5 km to almost 10 km on average in the incomes. Reported incomes, however, provide a northern district, and from less than 7 km to useful basis for comparison between households. 16 km in the southern district. This is an indi- Poverty was not a primary factor in NTFP cation of resource degradation in the North, collection. The lowest household income from and of village resettlement in the South. How- NTFPs was less than US $5, while the highest ever, walking distance is not an absolute obsta- was 1,000 times more. NTFP income was, how- cle to NTFP collection. Only 9 percent of ever, 40 to 50 percent higher in households in the households, all of them in the North, cultivate lowest wealth category than in the medium- and some NTFPs. high-wealth categories in the northern district.14 More than 90 percent of households in the There was a reverse, but not statistically signi�- North and slightly fewer in the South were aware cant, relationship in the southern district. of customary NTFP management. NTFPs for Other than poverty, variables that were signif- self-consumption are an open resource, while icantly correlated to NTFP incomes were the age NTFPs for sale are reserved for the local com- of the village, cultivation of upland rice, and mar- munity. No improved management system was keting channels. In the northern district, house- observed in any of the villages surveyed, the newest holds with more than one hectare of upland rice of which was two years old and the oldest thirty. area had 2.5 times more income from NTFPs The older villages have signi�cantly more income than households with less than 0.5 hectares. In the from NTFPs. The only new village with a sub- southern district, the households in the survey stantial NTFP income has the largest upland rice had resettled to a new paddy basin, upland rice area per household, and as much as 35 percent areas were small, and the difference in income of the surveyed households are members of a between households with less than and those with marketing group. Members of the NTFP market- more than 0.5 hectares of upland rice was mini- ing group have NTFP incomes more than mal (Figure 3.2). three times higher than non-members. There PEN STUDY IN THE LOWER MEKONG SUB-REGION 87 POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.2 NTFP Income in Relation to Upland Rice Farming Namo District Phouvong District 1200 1200 Kip (000) per year 1000 Kip (000) per year 1000 800 800 600 600 400 400 200 200 0 0 <0.5 0.5-1.0 >=1.0 <0.5 >=0.5 Upland Rice Farming (ha per household) Upland Rice Farming (ha per household) was an absence of extension of this successful The existence of regulations and the interest of organization to other villages. both communities and local governments in more sustainable resource management indicates, how- ever, that community-based NTFP management Discussion can be a viable option. Market-related commu- NTFPs are important for the rural population of nity groups appear to work well and to integrate upland areas in Lao PDR regardless of ethnicity, poorer households successfully. Such groups have gender, wealth classes, or distance to forests. more potential to limit harvest levels than quotas. Households are more likely to use NTFPs as an Local governments do set up quotas based on income source when they are engaging in shift- rough estimates of resource demand and supply, ing cultivation, because this gives them knowl- but these are dif�cult to manage. edge about resources, closer access, and time for The small number of marketable species of collection. These households may be poorer, and national relevance offers an opportunity for poli- use NTFP income to compensate food de�cits. cies and programs to address individual NTFP This is the case in some communities, but not species in a speci�c manner. Efforts to help culti- in others. vation take off could concentrate on those forest The existence of a declining NTFP resource species, for which domestication is technically fea- trend is obvious for species growing in mature sible and not on open-space species. Community- forests and calls for urgent action. The disappear- based resource management programs could ance of some of the NTFP resources would focus on other forest species that cannot be culti- substantially affect broad numbers of upland vated. Public access to open-space species does communities, where poverty incidence remains not threaten resources and is an appropriate high. Domestication has hardly started, so NTFPs option for the poorest. remain by and large a common property resource. Land use systems are evolving as a combined NTFP cultivation is, however, an option only for outcome of migration trends to roadsides and some NTFP species. Community-based resource areas with potential for paddy �eld development management will remain the only option for other and government policies. Stabilization of shift- species. Communities’ customary regulations are ing cultivation through land allocation gener- weak in the face of high market demand and in the ally shortens rotation cycles. One can expect an absence of a formal regulatory framework. increase in the availability of fallow NTFPs, but 88 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR these have provided limited incomes so far. the 1990s, after the national and provincial net- Encouragement of migration to new settlements works had been largely completed (MRCDP along roadsides with no option for paddy �elds 1996). The mid-2000s appear to be a turning will enhance this trend, and so will the merging point. All districts are now connected at least by of several small villages into a larger one under earth roads, projects have demonstrated the via- the village consolidation policy. bility of village feeder road construction, and Migration to roadsides, where paddy �elds are various options to fund and ensure village road available, may reduce households’ NTFP incomes maintenance are in place or being piloted. in the short term and impact the livelihoods of In the next stage, the government may either poorer households, but longer-term trends are continue to follow a step-by-step approach by unclear. In the absence of such policies, however, ensuring all-weather links to district towns, or the more valuable NTFP resources will continue opt for a feeder road network serving a broad to decline together with old-growth forest. number of villages. The first option is stated as NTFPs are valuable observation lenses to a priority in NGPES for poor districts. Policy remind policy makers of the complexities and choices need to take into account the environ- time dimensions of policies related to land use ment: the opening of roads is widely regarded as change in upland areas, where swidden agricul- an important factor behind deforestation (IUCN ture remains an important livelihood source. The 1997). Policy choices are also related to the gov- prevailing sentiment that there is a simple causal ernment’s overall policy of supporting migration relationship between shifting cultivation and the to roadsides in upland areas. “Focal areas� are decline of NTFP resources may overlook other stated in NGPES as a priority for road develop- critical factors behind the ongoing decline of ment next to district towns. NTFP resources. A central question—whether to build a rural What is needed is neither a rapid shift of road network with fewer roads of higher quality, land use systems out of swidden cultivation nor or to expand the network to a broad number of the status quo. Instead, communities need gov- villages—is an important area of policy decision ernment support in managing their forest re- making. The latter option might be viable, pro- sources, including NTFPs, particularly in the vided that lower, cheaper standards are selected. context of an active market. Communities liv- The PEN study seeks to compare opportunities ing in remote locations and away from roads and constraints of these two options in terms of are equally in need of such support. The stabil- poverty reduction and preservation of environ- ity of villages and slow land use change may be mental resources. important factors for success. The PEN study encountered signi�cant dif�- culties both in national analysis of district-level Roads, Environment, and Poverty indicators and in the village survey. The percent- The Government of Lao PDR faces the challenge age of villages within 6 or 11 km from a main or of building a sustainable road network in a coun- secondary road is a useful road access indicator,15 try where both poverty and natural resources are but no update was available. Actual village access concentrated in upland districts. The NGPES may be more valid in a country where village acknowledges a high correlation between the lack location evolves rapidly. NGPS compiled data of road access and poverty. The policy frame- for the selection of priority poor districts and work of the transport sector in Lao PDR follows NSC prepared the population census, but the a step-by-step approach. The government and two datasets are largely inconsistent. This may the donor-funded projects started to direct their be related to the dif�culties of tracking numbers efforts toward rural roads in the second half of of villages. De�ning road access also is dif�cult.16 PEN STUDY IN THE LOWER MEKONG SUB-REGION 89 POVERTY AND ENVIRONMENT IN LAO PDR In addition, local governments may sometimes cover, but is associated with lower district ever- have limited incentives to report accurate infor- green forest cover.17 This indicates that opening mation on village roads. Road density, used in roads may lead to some deforestation, and that conjunction with density of population and vil- this potential impact needs to be managed. This lages, would conversely be a convenient indica- is particularly important in districts bordering to tor. Unfortunately, rural roads are not recorded Vietnam where most of the remaining evergreen separately from paths. forest in Lao PDR is located. These dif�culties have limited the scope of Overall, 72 percent of villages in Lao PDR the national data analysis but have shed light on already have at least seasonal road access. Villages the obstacles that the government is facing in may be served by these roads, by districts roads, monitoring and planning its local road network. or by rural roads linking villages only. Around The �eld surveys provided a valuable compre- 36 percent of the 30,000 kilometers of roads are hensive picture about the actual status of local rural roads; an unknown proportion of these are road and transportation systems in two marginal paths that are not suitable for motorized access. districts. Interview �ndings were mostly assem- Village access to roads is uneven among districts. bled at the community level. Average road density in the country is slightly The survey team found it dif�cult to avoid above 12 km per 100 km2, but it ranges from 1 to selecting villages recommended by local author- 35 in the rural districts. In 22 provinces, more ities. The villages surveyed mostly turned out to than 50 percent of the villages have no road access. be recent or upcoming bene�ciaries of road con- In short, the country has already achieved struction. This has limited the scope for using some coverage of village road access despite low the PEN study’s methodology of comparative road densities. A simple theoretical assessment of analysis between poorer and less-poor villages village densities provides a rationale for this. Rural and households, but it has provided valuable population density was below 17 people/km2 in insights into the processes used by local govern- half of the districts in 1995, and below 10 in half ments when selecting bene�ciary villages in their of the NGPES priority districts.18 Villages in rural road programs. districts19 have an average population of not less than 380 people nationwide and 325 people in priority poor districts. Only 23 districts have an Nationwide Perspective average village size of less than 250 people. This LECS II (1997–98) demonstrated that poor dis- means that a district with 10 people per km2 has tricts had substantially lower access to roads, as on average less than 3 villages per 100 km2. Road measured by percentage of villages with main or investment cost per village served would increase secondary roads within 6 kilometers. LECS III rapidly in correlation with lower village densi- con�rmed this correlation by showing that ties, if villages were equally distributed over the poverty incidence is almost twice as high among national territory. households in villages with no access to a road This, however, is not the case. In the four dis- compared to households with access. However, tricts surveyed, only the southern half of the correlation between road access and poverty Namo District in the North has a traditional set- incidence is not a simple causal relationship. For tlement of villages scattered over hilly terrain. example, the road network is being built at a Elsewhere, and in many districts, judging from markedly slower pace in upland districts, where the available detailed topography maps, most various poverty factors are combined. villages are distributed in strings along paths or Road access and road density were not found rivers, which are the traditional transport modes. to have a negative effect on overall district forest In many such districts, 10 kilometers of rural 90 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR roads per 100 km2 may be suf�cient to offer full an increasing number of villages are accessible by road access. This is equivalent to 10 meters of trucks during the dry season. The government road per inhabitant, or $150 per capita based on has not mentioned specific plans for building a rough estimate of $15,000 per km of earth village access roads. roads, a �gure that can be smaller if local com- The road network maps available at the munities provide part of the labor. national level are misleading. They mention sev- There is also a direct link between low com- eral rural roads with their registration number. munity density and limited environmental In Nhot Ou, these are the traditional paths that impact. Should road opening affect a corridor of continue to serve for travel by foot or with horse 50 meters on each side of the road, the result- animals. A substantial proportion of villages are ing impact would still be relatively limited as a distributed along or close to these paths. In percentage of total land. Environmental risk Kaleum, many of the paths are not used much, analysis, therefore, relates more to areas of spe- since most villages in Kaleum are located along cific environmental value than to overall forest rivers, and boats are the traditional travel mode. cover rate. Enterprises are intensively involved in road construction in both districts. A Chinese enter- prise has built the new village road in Nhot Ou. A District and Village Perspective The main road was also reportedly built through Kaleum District in the South and Nhot Ou Dis- a Chinese enterprise, although no con�rmation trict in the North have both very low population could be obtained. In Kaleum, the enterprise density and very limited village road access (Table controlling the metal scrap market from UXOs 3.7). Both districts cover large territories of more in the district has built a dry-season road to access than 300 km2, which is 70 percent above the villages in the eastern part, and a State Forest national average. Nhot Ou is crossed by an all- Enterprise (SFE) from Vietnam reportedly built season road from the South to an international another road.20 Unfortunately, enterprises do not border with China. Only one village other than build bridges, so the rainy season access problems villages along this main road has a motorable either remain unsolved or increase with road road, and plans for future road construction are degradation from the enterprise trucks. Informa- limited to one other village. In Kaleum, the road tion regarding road construction by enterprises, to the district town itself is hardly passable outside even location, was dif�cult to access in the dis- the dry season. This road serves only two rural vil- tricts and is not available at the national level. lages. All other villages are cut off from the district Village discussions provided detailed accounts town during the rainy season floods. However, of how a direct road link makes a critical difference TABLE 3.7 Rural Road Network Indicators in the Two Districts Surveyed Road Villages with % of Villages density no road villages with no dry Rural population Km/ access/total with no season cart Village density density 100 Km villages road access per 100 Km2 people/Km2 Kaleum 3 53 / 60 85% 23 / 60 1.9 4 Nhot Ou 4 62 / 86 72% 57 / 86 2.8 8 Source: national database and in-district checks PEN STUDY IN THE LOWER MEKONG SUB-REGION 91 POVERTY AND ENVIRONMENT IN LAO PDR to villages today, now that active market outlets and expedite the process of completion. In Nhot are within reach.21 Frequent market trips with Ou, in the first village community to obtain a buses become feasible. They increase both income road, villagers started the work by themselves and opportunities and social links. Farm product sales brought small tractors in spare parts before the and job opportunities increase substantially in the road was built. The second village awaiting a road rainy season, which remains a food de�cit season appears to be using its opium producer status and for many. Increased incomes allow purchase of position right on the border to be recognized as a transportation means and access to media, which poor village and a focal area, therefore qualifying in turn reinforce all other advantages. Informa- for road construction. tion exchange and support from the local govern- It is, therefore, not surprising that the vil- ment increase as well. lages interviewed made strong statements that Consensus on the difference that road access they were confident in their ability to manage makes for a village has become so strong that all their forest resources after a road was opened parties, from government to villagers, tend to and to restrict access from outsiders. Data does equate poverty with absence of road access. Dif- show some increase in NTFP collection for sale ferences between poorer and less-poor villages in villages along roads, but this increase was not among the ten villages studied indicate, albeit on significant. Spontaneous answers during inter- a non-representative village sample, that roads views also tend to validate the NTFP study are one of the key factors that influence the inter- findings regarding the potential of community action. Ethnicity, not roads, stands out as the participation in natural resource management �rst causal factor for poverty in the survey. systems. Villages that are gaining road access are making Preferred transportation tools have regional substantial efforts to receive government approval speci�cities. In Kaluem, mostly transportation trucks are available, which require larger roads and high maintenance costs. Motorbikes and TABLE 3.8 Selected Village Indicators in small tractors, available through the Chinese mar- Relation to Poverty ket, would allow the construction of narrower and cheaper roads. Tractors have a dual advan- Category of village Poorer Less Poor tage; they can be used for both plowing and transportation. % Lao Loum households 0 55 % Households with means 5 34 of Transportation Discussion % Villages located by 17 50 All-Season Road The case of Kaleum illustrates the validity of the % Households with 22 55 Media Equipment priority given in NGPES to completion of the all- % Villages with Seasonal 50 75 weather road network to district centers. How- Road ever, both Kaleum and Nhot Ou demonstrate a Hours Walk to All-Season 2,8 3,5 need for further road outreach to the rural areas. Main road % households with NTFP 74 75 Roads are one element among multiple poverty Income factors. Direct road access to existing villages % Villages with Land Use 83 75 makes a critical difference to communities in Planning Hours Walk to Old Growth 2,3 0,8 upland areas because it gives them access to an Forest active market. Communities in remote locations are willing to invest tremendous energy in secur- Source: From PEN survey in 10 villages (MCTPC). ing road access. Market access is not only a means 92 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR to improve cash incomes, but also an important enterprises. The role of private enterprises other social link in isolated areas. than construction companies in developing the The district surveys shed some light on the bar- road network is already important in poor dis- gaining process between local governments and tricts located on the borders. The role of govern- village communities in applications for a new vil- ment remains critical in planning and monitoring lage road. This is a very competitive process, given as well as in ensuring additional investments, the modest level of on-going road construction for example in bridges. In contrast, there is sur- projects. In this process, the successful communi- prisingly little information available on these ties are generally those that have developed spe- enterprises. ci�c strategies. These villages may be better off, Linkages between road construction and stronger communities, or communities with spe- increased natural resource extraction are somewhat cial features that make them qualify as focal areas. ambiguous. Communities appear highly moti- This trend raises the question of how to orga- vated to manage their resources well, provided nize and rank applications. Donor-funded road they are given the responsibility to do so. Con- sector projects have piloted prioritization through versely, the involvement of large enterprises in a scoring system based on community participa- building roads against logging contracts is likely to tion and cost in relation to population served. The generate impact. What is needed may not so much survey highlights this need for prioritization in dis- be to question current practice that allows them to tricts that are located on borders that have large log a limited area on both sides of the road as it is investment needs and/or have a multi-ethnic pop- to ensure sound implementation and monitoring. ulation with contrasting socioeconomic status. Nhot Ou has all these features. UXOs, Environment and Poverty Full village coverage by the road network is a distant prospect in marginal districts. This may Unexploded ordinance (UXOs) dating from partly explain the absence of a strategy in future the mid-1960s and mid-1970s affect 15 of Lao rural road network development at the national or PDR’s 18 provinces. Linkages between UXO local level. No comprehensive �nancial frame- contamination and poverty are obvious, but the work is in place. Such a strategy is needed, how- issue has not achieved high visibility. Deconta- ever, and could be designed to address poverty mination started in 1996/97 under the national and environment at the same time. Building Lao UXO program, but it is a very expensive cheaper tracks instead of roads does not appear process and is proceeding at a slow pace. to meet current local expectations. This might Little is known about linkages between UXO change rapidly, however, as smaller transportation contamination and the environment. Some argue means have become more available through bor- that farmers are coping with UXOs by farming on der trade. The widespread view that building a slopes, while others claim that UXOs could limit road network in a sparsely populated territory is a encroachment on forests and exploitation of tim- challenge further limits active planning of a rural ber and biodiversity degradation. In the latter road network. Figures show instead that a well- case, UXO contamination would have an unfor- distributed network may serve strings of villages tunate effect on communities’ health and liveli- as soon as the country’s GDP makes investments hood but positive linkage to the environment, of less than $200/capita economically viable. while in the former case the negative linkage to The information flow between national and the environment would be an additional reason to local government levels regarding the status of support UXO-affected communities. the road network is incomplete and not up-to- The Government of Lao PDR has created the date. This is especially the case for roads built by national Lao UXO program—which provides a PEN STUDY IN THE LOWER MEKONG SUB-REGION 93 POVERTY AND ENVIRONMENT IN LAO PDR channel for funding decontamination and public The national Lao UXO program and the awareness initiatives—into the NGPES. The pro- local government provided logistical support gram is described as relating strongly to poverty and translation from the local Mon-Khmer reduction and to the shifting cultivation stabiliza- languages. Information was mostly collected at tion policy. Decontamination is to be planned in community level through focus group discus- areas identi�ed as priority under these policies. sions. The national research team mostly used The PEN study sought to analyze poverty and household interviews to confirm information environment linkages of the UXO issue by ascer- derived from group discussions. taining the impacts of UXO presence on house- The limited possibilities of directly observing holds’ agriculture and overall land use systems. detailed land uses could not be fully overcome. And it sought to analyze poverty and environ- However, quantitative indicators were assembled ment linkages of the current prioritization process for the six villages surveyed, and discussions with in activities related to the UXO decontamination, elder people, who could remember the pre-war and to identify potential environmental linkages situation, indicated qualitative trends. A realistic specific to forested areas that might arise from set of �ndings was derived after supplementing decontamination. this information with the national database analy- The study benefited from the national Lao sis and with district interviews. UXO program database. The database includes findings of the comprehensive 1996/97 survey UXO Contamination of UXO impact (Handicap International 1997) as well as the updated status of village operations. The national survey of UXO contamination These data were analyzed by linking district level revealed in 1997 that at least 22 percent of all vil- UXO contamination and LECS II–III poverty lages in Lao PDR had a problem with UXOs, data. The methodology for field research took affecting close to 25 percent of the total popula- into account substantial obstacles, which were tion.22 In the Central region, nearly 50 percent expected. The research only took place in the of all villages are affected (Map 3.7). Kaleum district. UXO contamination is affecting the poor dis- proportionately. An estimated 28 percent of the poor and 20 percent of the non-poor lived in vil- lages with a UXO problem in 1997/98.23 More MAP 3.7 Percent of Villages in District Having than 70 percent of villages are affected in the UXO Problems poorest districts of the southern region. The 1997 survey ascertained that more than 450 UXO- UXO problem High UXO problem related accidents had occurred annually between 1974 and 1996. Casualties are only decreasing slowly: the Lao UXO program has recorded 110 accidents annually from 1999 to 2004, but these �gures only cover 59 districts. Both LECS II and LECS III show that UXO contamination in the South and Central regions is associated with higher poverty, less productive assets, and lower food security. In addition, LECS III shows that living in districts of the South and Central regions, where more than 50 percent of villages are affected by UXOs, is associated with 94 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.3 Left: UXO Contamination in Lao PDR (1997). Right: Percent of Villages Affected by UXOs in Relation to Poverty in the Southern Region (1997/98) Southern Region 60% 50% 80% 70% 40% 60% 30% 50% 20% 40% 30% 10% 20% 0% 10% North Center South 0% 12% 32% 56% Least Poor Less Poor Percentage of Districts Districts Poor Districts Poorest Districts Villages Affected %of Villages Affected 52% 50% 52% 73% Sources: Left: calculated from the national UXO survey 1997; right: Data are from LECS II 1997/98 and UXO national survey 1997. lower improved water supply and sanitation cov- TABLE 3.9 UXO Contamination and Key Indicators erage rates, more household time spent on fuel 2002/03 in the Southern Region wood and water collection, and lower education levels and health status. This difference between districts highly affected by UXOs and other dis- UXO > 50% UXO < 50% tricts is especially high for access to safe water in Poverty Incidence 2002/03 41% 22% the South (Table 3.9). Irrigated land (percent 4% 4% Villages in highly contaminated areas are also of households) less likely to have ongoing development projects, No Rice (months per year) 3.6 2.6 Fuel wood collection 0.9 0.6 and some villages are experiencing a higher rate (hours per day) of out-migration. As seen from the NTFP study, Water collection 1.2 0.8 there is also an indication that districts with high (hours per day) No toilet facility 72% 64% UXO contamination have less income from No water supply (using 20% 7% NTFPs, a source of income that could poten- open water source) tially compensate for the lower availability of Emigration rate 1.6% 1.1% (last 12 months) paddy rice and irrigation. Literacy (> 50% of house- 63% 83% Two positive �ndings were observed: (1) road hold members can read) access is equally available in districts with low Death rate (last 12 months) 0.6% 0.5% Development project 48% 62% and high UXO contamination, and (2) forest in village resources are relatively abundant in the South and Central regions. UXO > 50 (or < 50): households living in districts where more (or less) than The Government’s 46 priority districts 50 percent of villages are affected by UXOs. Source: The indicators are gen- erated from LECS III 2002/03. UXO data are from the national UXO survey account for 49 percent of highly affected vil- 1997. lages and 46 percent of recorded deaths since 1999. They only account for 30 percent of the PEN STUDY IN THE LOWER MEKONG SUB-REGION 95 POVERTY AND ENVIRONMENT IN LAO PDR agricultural land decontaminated between 1999 of the 22 contaminated eastern villages and has and 2004. been quite successful with support from an inter- The geographical distribution of UXO con- national project. The number of villages that had tamination in the six southern provinces is not moved out is unclear. unrelated to the distribution of Mon Khmer eth- Village consolidation in Kaleum is also active, nic groups. Mon Khmer people form the vast with or without migration.25 The 1997 UXO majority of villages in the Indochinese Cordillera. survey mapped 54 villages affected by UXOs out UXO-impacted villages are widely distributed in of a reported total of 115 villages. The research the six provinces except in the Mekong corridor, team recorded 60 villages. One village surveyed so there is no correlation between the proportion was formed by merging three nearby villages. of Mon Khmer people and proportion or num- The new location was severely contaminated, ber of villages affected by UXOs. It is, however, and reportedly no decontamination support was likely that several of the small Mon Khmer made available. There is also out-migration to groups mostly or fully live in districts with a high other districts in the province, both spontaneous proportion of UXO-affected villages. The map of and government-sponsored. District population Agent Orange missions shows close concentra- reportedly has declined by 7 percent in the 1990s tion of impacts on the Cordillera (Lao UXO pro- (Chaignon 2000). gram 2004 & Chaignon 2000). Bombings within villages precisely targeted the No association was found between forest cover Ho Chi Minh trails. This has clearly resulted in and percent of villages with UXO contamination higher UXO contamination along rivers and val- in a district. This �nding is perhaps not surpris- ley floors and in village centers. People are exposed ing, since affected areas range from the densely to UXOs in daily activities such as cooking or �sh- forested Cordillera to the Xieng Khouang plateau, ing. Paddy cultivation is especially dangerous, where forest cover is low. However, in districts both when reclaiming new paddy �elds and when with high UXO problems, forest cover is lower plowing the land. UXO decontamination is a pre- than can be explained by region, population den- requisite in construction works for roads and com- sity, agriculture, or topography. This finding munity buildings like schools. Since 2000, areas of deserves further analysis. paddy reclamation range from zero to only three hectares in all villages except for one, which has bene�ted from a donor-supported program. This Kaleum District old village is easy accessible from the district The distribution of villages affected by UXOs in town and had been selected for decontamina- Kaleum derives from two factors: (1) the initial tion for paddy �eld opening in the coming year. bombing impacts, and (2) active migration trends Altogether the Lao UXO program has cleared since the 1990s. All quantitative data in Kaleum 19 hectares of paddy �elds and 30 hectares for should be regarded as tentative.24 Most villages construction works between 1999 and 2004. with UXO impact are located today in the west- Swidden agriculture is comparatively one of ern third of the district closer to the district the least dangerous activities, because UXOs town. Several of the seven severely affected villages explode spontaneously when the fallow is burned, are right next to the district town. Villages from and crops are sown without plowing. Swidden the eastern part of the district are strongly encour- agriculture nowadays takes place close to the vil- aged to migrate to the west. The bombings fol- lages with very short rotations as a result. It is not lowed the Ho Chi Minh trails that extended possible to determine to what extent this is a cop- throughout much of the district. One village ing strategy due to UXOs, or is the result of the surveyed was created from the migration of two shifting cultivation stabilization strategy. The fact 96 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR that hunting nowadays also takes place close to mal assets may be an indicator of dif�culties in the villages—and that only one out of six villages adjusting. Half of the villages had lost large ani- reports substantial NTFP collection—points to mals since the war, and another 50 percent (not the former explanation. The fact that all villages necessarily the same ones) have decreasing or sta- have a land use plan points to the influence of ble animal numbers today. government policy, and so does the fact that local communities have been coping with high UXO Discussion presence on their own from 1973 until 1999. For example, they have become accustomed to col- The PEN study confirms that UXOs do not lecting UXOs either to store them away or more protect forests from encroachment and that recently to sell metal scrap. The latter has become decontamination has no potential for negative a major income source in the district. Villagers consequences on forest cover. First, decontam- stated that they saw several thousand—up to ination is proceeding at a very low pace. About 13,000—UXOs in their own village within one 5,500 hectares have been decontaminated be- year and that many more are still unidenti�ed. tween 1999 and 2004, an average 16 hectares Households appear to be adjusting their liveli- per district per year. Second, UXO contamina- hoods in several ways. Interestingly, only one out tion is a powerful factor limiting reclamation of of six villages is developing “classical� improved paddy fields and decrease of swidden agricul- agriculture based on paddy and livestock. All ture. The industrial forestry sector is important other �ve villages are reporting alternatives and in Kaleum and elsewhere in the Indochinese plans that range from migration to innovative Cordillera26 but, except for NTFPs, it is mostly income sources (Table 3.10). Villagers seem to be unrelated to households and was not covered by moving around without limitation from UXOs. the study. UXOs are certainly an issue for this sec- Inhabitants in all villages collect NTFPs and go tor, too, and its linkage to environment deserves to the market regularly by foot. Decreasing ani- analysis. TABLE 3.10 Past & Current Trends in Six Villages Surveyed Village 1 2 3 4 5 6 New or old village Migrated Old Migrated Migrated Migrated Old village from east, village from from recently consolidated nearby nearby from village village village nearby, consolidated village UXO impact Low/Moderate Moderate Moderate High Severe Moderate/ High Animal assets Since war Decrease Increase Decrease Decrease Increase Increase Since 2000 Increase Decrease Stable Increase Decrease Stable Paddy development Large area Small area None Small area Small area Small area since 2000 95% HH 10% HH Metal scrap 95% HH 20% HH 10% HH Livelihoods Today Paddy & live- Seasonal Tourism NTFPs Out- Metal scrap Gold stock dvt migration migration panning Tomorrow Sawmill Fruit trees Paddy dvt Source: PEN village survey (ERI). HH = household. PEN STUDY IN THE LOWER MEKONG SUB-REGION 97 POVERTY AND ENVIRONMENT IN LAO PDR UXOs may be contributing to rotational agri- UXO program was planning to work only in dis- culture with much shortened cycles together with tricts listed as poor in NGPES, including two- the government’s policy of shifting agriculture thirds of very poor districts (Lao UXO program stabilization. The combination of both issues may 2004). With widely spread UXO contamina- create unique dif�culties for the local population. tion, it is understandable that the more densely Making the new farming systems sustainable is populated lowland and urban areas may have challenging. Few locations in UXO-affected areas attracted support in the past. Yet, within districts, are appropriate for opening suf�cient areas of UXO-related activities—with the important paddy �elds. There is no indication of UXOs lim- exception of awareness raising, which is already iting animal husbandry, but the dif�culties of covering broad numbers of villages—do need to adjusting livelihoods appear to result in several be targeted for expansion. communities’ selling out their animal assets. Targeting villages with easy road or river A clear national policy statement is needed to access may be the only solution, but doing so in link UXO decontamination to shifting cultiva- the framework of the village resettlement and tion stabilization, but actual results appear to be consolidation policy is particularly questionable. limited in the district surveyed. Very few villages Kaleum is not the only district where UXOs are are prioritized for the opening of paddy �elds. concentrated in areas targeted for resettlement.28 Access is obviously one of the conditions of village Local governments in marginal districts do not selection. Within a village, some communities opt have the capacity to identify safe locations for for sharing the new paddy �elds, while in others new settlements and to plan timely decontami- paddy �elds are operated by a few families only. nation. Their energies would be better employed UXOs are not so much a natural resource in planning infrastructure for existing villages management issue as a strongly limiting factor and supporting communities’ initiatives for liveli- to overall rural development. The PEN study hood diversi�cation. demonstrates that even villages with no UXOs are indirectly affected by them in districts with POVERTY AND heavy UXO contamination. This is especially ENVIRONMENTAL HEALTH the case in districts where UXO impact is con- centrated closer to the district town. Water Supply & Sanitation The livelihood strategies observed in the vil- lage survey also point to the limitations of agricul- Achieving equitable access to sustainable water ture in most villages and to the need for support of supply and sanitation services is essential for pro- a more diversi�ed rural economy. However, pri- tecting the health of the poor and improving their oritization of decontamination efforts may be quality of life. Poor households tend to have more more critical for construction works than for young children and elderly than non-poor house- agriculture. A high proportion of the districts holds. These age groups are most vulnerable to dis- use decontamination mostly for purposes other ease. Poor households have less access to quality than agriculture.27 This is perhaps why in 1998, health services, and can least afford the cost of roads—and not UXOs—were cited as the main medical treatment and income losses from disease. problem by village leaders (Chaignon 2000). Increasing household access to improved water The private sector is also playing its share in supply and basic sanitation is viewed by the decontamination. Government of Lao PDR as an important part Prioritization of UXO-related support is, of poverty alleviation and socioeconomic devel- therefore, especially relevant at the district level opment. The Sixth National Socio-Economic and may relate to all sectors. In 2004, the Lao Development Plan 2006–10 aims at increasing 98 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR national water supply coverage rates by several MAP 3.8 Population with No Water Supply percentage points in the next �ve years. and Sanitation Substantial progress has been achieved in pro- viding the population with safe and accessible No water supply* No sanitation facilities water supply in the last decade. Improved water supply coverage increased from less than 20 per- cent in 1995 to 55 percent of the population in 2003.29 The population share with protected groundwater tripled during the period and the share with unprotected groundwater declined by nearly 10 percentage points. The most striking achievement was the reduction in the popula- tion share relying on open, unprotected water sources, such as rivers, ponds and streams, which declined from nearly 50 percent to 22 percent. Other types of water supply, including gravity- fed systems, increased from practically nonexis- Source: LECS III 2002/03. *Population with drinking water directly from tent to 18 percent of the population. rivers and other open, unprotected water sources. The increase in sanitation provision has been substantially slower than the increase in improved disparity. More than a third of the poorest rural water supply, but is still signi�cant. The popula- households depend on surface water compared tion share with no sanitation facilities (“no toilet/ to 23 percent of the richest rural households.32 latrine�) declined from 70 to 50 percent from Almost 80 percent of the poorest rural house- 1995 to 2003, and pour-flush toilets now serve holds do not have basic sanitation, compared to 37 percent of the population. 43 percent of the richest (Figure 3.4). The poor have not benefited enough from The reliance on open, unprotected water these improvements. Of particular concern for sources (surface water) and the lack of sanitation environmental health is the population continu- varies greatly across provinces and for non-poor ing to rely on open, unprotected water sources, and poor households. More than 50 percent of such as surface water. In 2003, about 30 percent the poor in Luangphrabang and Houaphanh of the poor relied on surface water, compared to provinces use surface water, and more than 20 percent of the non-poor (LECS III). While this 80 percent of the poor have no toilet/latrine represents a substantial decline from 1995, there in Phongsaly, Luangphrabang, Khammouane, was no relative gain for the poor.30 Savannakhet, and Saravane. Access also varies dur- The situation is more pronounced for basic ing the rainy and dry seasons. In the South, there sanitation. Nearly 70 percent of the poor did not are signi�cant differences across the two seasons. have toilet facilities in 2003, compared to slightly In 1995, district population water supply and more than 40 percent of the non-poor The non- sanitation coverage rates were strongly associated poor population with no toilet facilities declined with urban population share, villages’ access to by 21 percentage points, while for the poor, the roads and their education level, and to some decline was only 13 percentage points from 1995 extent UXO contamination.33 This continues to to 2003. The gap between the poor and the non- be the case today. About 28 percent of the rural poor therefore seems to have increased.31 population relies on surface water, compared to The inequity in access to improved water only 4 percent in urban areas. Population with supply and sanitation is not only an urban-rural toilet facilities exceeds 85 percent in urban areas, PEN STUDY IN THE LOWER MEKONG SUB-REGION 99 POVERTY AND ENVIRONMENT IN LAO PDR FIGURES 3.4 Reliance on Surface Water and Lack of Sanitation in Rural Areas in 2003 40 % 35 % 30 % 25 % 20 % 15 % 10 % 5% 0% 1 2 3 4 5 Surface Water 35% 33% 28 % 27% 23 % 80% 70% 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5 No Sanitation 79.0% 72.4% 60.8% 55.2% 42.5% Source: Calculated from LECS III, grouping households from poorest (=1) to richest (=5) but is only 40 percent in rural areas. Of the pop- An analysis was undertaken to assess the role ulation with access to roads, less than 20 percent or association of each of these factors with use surface water and nearly 60 percent have toi- household use of surface water and lack of a toi- let facilities. Of the population without access to let facility.35 The analysis enabled a comparison road, more than 40 percent use surface water of households that are generally similar except and only 20 percent have toilet facilities. Having with respect to one factor of the assessment. The no water and sanitation services is also strongly analysis resulted in the following conclusions: associated with household literacy level. Only 18 percent of households rely on surface water in • Rural households are 7.7 times more likely which more than half of the members can read, than urban households to rely on surface while nearly 35 percent of households use sur- water. face water in which less than half of members • Households in villages with no access to roads can read.34 Rural areas lag far behind the urban are 1.9 times more likely to rely on surface with respect to literacy levels. water than households with road access. 100 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR • Households in which more than half of the ity of communities and their perceptions of pri- members cannot read are 20 percent more likely orities to obtain and sustain water and sanitation to rely on surface water than households in services. Remote communities often have less which more than half of the people can read. contact with and less access to water and sani- • The presence of UXO problems in a village is tation decision makers. Moreover, less edu- associated with a 34 percent higher likelihood cated communities may be less skilled in of using surface water. communicating needs. They may also have less • Controlling for the above factors, being poor appreciation and/or knowledge of health risks increases the likelihood of relying on surface associated with lack of water supply and sani- water by 15 percent.36 tation. Poverty also influences demand for water and sanitation, as poor communities have There is an even stronger association between fewer resources to contribute to construction the same factors and lack of sanitation; that is, and operation/maintenance in rural areas and no toilet/latrine: to water connection fees and water tariffs in urban areas. • Rural households are 8 times more likely than urban households to have no sanitation facilities. Urban Water Supply and Sanitation • Households in villages with no access to roads The Water Supply Authority Lao PDR (WASA), are 3.4 times more likely to lack sanitation facil- responsible for urban water supply, emphasizes the ities than households with road access. importance of social fairness, quality of services, • Households in which more than half of the consumer satisfaction, and �nancial sustainability members cannot read are 4.3 times more likely in the water sector (Annual Water Sector Perfor- to lack sanitation than households in which mance Report 2003). more than half of the members can read. There are more than 140 urban centers in • Having UXO problems in the village is asso- Lao PDR, corresponding to district capitals. ciated with a 7.8 times higher likelihood of Half of these centers have a population less than not having sanitation. 4,000–5,000. In addition, there are many rural • In addition, controlling for the above factors, villages with urban characteristics. Nearly 60 per- being poor increases the likelihood of not hav- cent of the urban population resides in the ing sanitation by 60 percent. 18 provincial centers, of which more than three- fourths live in the largest �ve.37 These factors have supply and demand aspects. Nearly 75 percent of urban households have Household or community location and road improved water supply, while 25 percent con- access are likely to have influenced supply deci- tinue to rely on unprotected well water and open sions in the water and sanitation sector, with pri- water sources. However, according to a 2002 ority given to urban areas, easily accessible rural report entitled Small Towns Water Supply and areas, and areas with low risk of UXO accidents Sanitation Initiative in Lao PDR, more than 100 and/or low cost of UXO clearance. These supply of the district centers do not have a piped water decisions, among other issues, may be driven by supply with house connections. cost considerations as well as concerns about the By household living standard, 60 percent of sustainability of water and sanitation services in the poorest urban households and more than remote rural areas. 80 percent of the richest urban households have On the demand side, road access, education improved water supply (Figure 3.5). The dispar- and literacy, and poverty may influence the capac- ity is, however, larger for piped water supply PEN STUDY IN THE LOWER MEKONG SUB-REGION 101 POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.5 Urban Household Water Supply by Living Standard Quintile 10 0% 80 % 60 % 40 % 20 % 0% 1 2 3 4 5 Piped Water Protected groundwater "Other" water Unprotected groundwater Open watersources Source: Estimated from LECS III (2002/03). Note: Improved water supply is here de�ned as piped water supply, pro- tected groundwater, “other� water (mainly gravity-fed systems), and rainwater (1 to 2 percent of urban households; not included in the �gures above). Note: Living standard quintiles are consumption-based estimates from LECS III, with quintile 1 being the poorest households. with a house connection. Nearly 30 percent of A Look at Two Small Towns the poorest households have house connection, while more than 45 percent of the richest have A PEN study was carried out in Phongsaly and connection. Lamam towns to gain a better understanding of Over 90 percent of urban households have the needs and priorities in the water and sanita- basic sanitation facilities. Nearly 80 percent of tion sector for the poor in small towns in Lao the poorest households and more than 95 per- PDR (Box 3.2). The towns are provincial capi- cent of the richest households have basic sanita- tals of two of the poorest provinces in the coun- tion (Figure 3.6). Quality and hygienic safety of try, and have populations less than 10,000. Both these facilities vary substantially within and towns have a piped water supply network serv- between urban areas and across levels of living ing a majority of the households, but no sewage standard. network. FIGURE 3.6 Percent of Urban Households with Toilet Facility by Living Standard Quintile 10 0% 80 % 60 % 40 % 20 % 0% 1 2 3 4 Toilet Facility 7 8 .2 % 8 5.4% 8 8 .8 % 9 0.7% Source: Estimated from LECS III (2002/03). Note: Living standard quintiles are consumption-based from LECS III, with quintile 1 being the poorest households. 102 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR BOX 3.2 The Urban Research Institute (URI) in Vientiane implemented a study on poverty and urban water supply and sanitation in Phongsaly and Lamam towns.* The study assessed types of household water supply and sanitation facilities, piped water connection coverage, user satisfaction, and factors influ- encing demand for piped water supply and improved sanitation in poor and non-poor households. A speci�c aim of the study was to identify policies and interventions to promote equitable access to water and sanitation services that provide health protection and improved living conditions for poor and vulnerable households. The study included a survey of about 550 households, questionnaire interviews with the district water utilities (Nam Papa), and interviews with district authorities and community leaders. Six out of nine urban villages were surveyed in Phongsaly town and six out of six in Lamam town. Asset scores and self-reported living standard levels were used to identify poor households. *URI. 2006. Water Supply and Sanitation Situation in Relation to Poverty in Two Urban Districts, Lao PDR. PEN report prepared by URI. Vientiane. Piped water connection. In Phongsaly, the piped location of the house relative to the main road also water connection rate is less than 40 percent affects the water connection rate (Figure 3.7).38 for the poorest households but more than The lowest connection rate is therefore among 80 percent for the richest. Affordability alone poor households with low education level who does not explain this disparity. Household do not own their property and who reside along education level, land and house ownership, and the walking paths. FIGURE 3.7 Percent of Households with Piped Water Supply in Phongsaly Household Living Standard (1=low; 5=high) Household Owns House and Land 100% 10 0% 80% 80 % 60% 60 % 40% 40 % 20% 20 % 0% 0% 1 2 3 4 5 No Yes Education Level of Household Respondent Location of House 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% No schooling Primary Secondary Upper 0% school school secondary Walking path Small road Secondary Main road school road PEN STUDY IN THE LOWER MEKONG SUB-REGION 103 POVERTY AND ENVIRONMENT IN LAO PDR In Lamam, piped water connections are also In Lamam, more than 85 percent of house- associated with living standards and household holds are satis�ed with the piped water supply, ownership of house and land. The influence of although satisfaction reaches only 60 percent in education levels is weaker than in Phongsaly, one of the villages. Households that reported time and the connection rate is not affected by house- savings as an important motivation for connect- hold location relative to roads. ing to the water supply were more satis�ed than households for which time savings was of minor Piped water user satisfaction. Nearly 80 percent importance. This is contrary to the �ndings in of households in Phongsaly reported satisfaction Phongsaly, but is consistent with the substantially with the piped water supply. The highest rates higher volume of water supplied by the system in of satisfaction were in locations closest to the Lamam. The rate of satisfaction is about the same water distribution point. Only 60 percent of for poor and better-off households, and satisfac- households were satis�ed in the villages most far tion is marginally higher among households with away from the distribution point. Satisfaction is higher education level. also lower among households reporting that It is not clear why satisfaction with a piped time savings were of high importance in their water supply is associated with household edu- decision to connect to the piped water supply, cation level. One reason could be that more edu- and among households reporting that the cated households have a higher appreciation for piped water supply is insufficient to provide the potential hygiene and health bene�ts of hav- their most important water needs. These ing convenient access to increased quantity and finding are understandable in light of the low quality of water. quantity of water that the piped water network is capable of providing in Phongsaly. Most Water tariffs. Water tariffs in Phongsaly are households therefore need to supplement their among the highest in the country. Nevertheless, water demand from other sources, which in 85 percent of the poorest households and most cases is a time-consuming endeavor. 70 percent of other households reported that Education level also seems to have a subtantial tariffs are reasonable. In Lamam, nearly 50 per- influence on satisfaction with the piped water cent of households—poor and rich—reported supply in Phongsaly (Figure 3.8).39 that tariffs are reasonable. FIGURE 3.8 Household Satisfaction and Education Level in Phongsaly 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% No schooling Primary school Secondary school Upper secondary school 104 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR TABLE 3.11 Household Water Bill in Phongsaly and Lamam Small user Large user Phongsaly Lamam Phongsaly Lamam Per capita water consumption per day (liters) 20 60 100 250 Household water bill (000 Kip per year) 60 130 575 770 Water bill % of household total expenditure 1.1% 2.4% 2.0% 2.7% Water bill % of household non-food expenditure 5.7% 12.4% 4.0% 5.3% Note: Calculated based on block tariffs and annual household water consumption data from DCTPC, and urban house- hold expenditure quintiles from LECS III. The lowest quintile (poorest households) is applied in the calculation of water bill as a percent of consumption for “small user� and the highest quintile is applied for “large user.� Urban Vientiane Municipality is excluded to better reflect urban household expenditure and living standards in smaller towns. Even though water tariffs per cubic meter in total expenditure, but less than a small user as a Lamam are less than half the tariffs in Phongsaly, percent of non-food expenditure (Table 3.11). household water expenditure is higher in Lamam. This may explain why fewer households in Water connection fee. The household water Lamam report that water tariffs are reasonable. connection fee in Phongsaly is Kip 500 thousand Household water consumption is several times (nearly $50), and Kip 900 thousand (nearly higher in Lamam than in Phongsaly, mainly $90) in Lamam. In Phongsaly, this is nearly because water is less expensive and supply is more 10 percent of annual household expenditure on abundant in Lamam. all goods and services for the poorest households, Household expenditure on piped water ranges and 2 percent of total expenditure of the richest from 1 to 3 percent of total household expenditure households. As a share of household non-food on all goods and services, and from 4 to 12 percent expenditure, the connection fee is 47 percent for of non-food expenditure. A large user spends more the poorest households and 3 percent of the than a small user as a percent of the household’s richest households (Figure 3.9). In Lamam, the FIGURE 3.9 Connection Fee as Percent of Annual Household Consumption Phongsaly Lamam 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 1 2 3 4 5 1 2 3 4 5 % of total % of non-food % of total % of non-food Note: 1 is lowest living standard quintile and 5 is highest. PEN STUDY IN THE LOWER MEKONG SUB-REGION 105 POVERTY AND ENVIRONMENT IN LAO PDR water connection fee is as much as 87 percent of quent in poorer households. Education levels also annual non-food expenditure for the poorest have some influence on the type of toilet. House- households and 6 percent for the richest house- holds with higher education had flush toilets holds. This may explain why Lamam has a lower more frequently than households with low edu- connection rate than Phongsaly. cation.40 Pour-flush toilets were much more fre- In Phongsaly, 75 percent of the households quently used in households with elderly than in without piped water, but with network available households without elderly household members. in their section of the town, stated that the high In contrast to factors influencing piped water connection fee was the most important reason connection, the location of a house in relation to for not connecting to the piped water supply. In the type of road and ownership of house and land Lamam, all households without piped water were not found to influence households’ choice supply reported that the high connection fee was of toilet facility in Phongsaly. the main reason for not obtaining a connection. In Lamam, four factors were identi�ed that influenced whether households had a toilet facil- Basic sanitation. More than 95 percent of the ity and the choice of facility. Education level surveyed households had toilet facilities in was found to have a larger influence in Lamam Phongsaly, ranging from 90 percent among the than in Phongsaly. House location also mat- poorest households to 100 percent among the ters. Households living away from main roads richest households. In Lamam, nearly 80 percent and along walking paths are less likely to have of surveyed households had toilet facilities. Less a pour-flush toilet with septic tank or to have a than 45 percent of the poorest households had toilet at all. Households not owning a house toilet facilities, compared to more than 95 percent and land had a pour-flush toilet with septic tank of the richest households (Figure 3.10). less frequently. Living standards, education levels, and the presence of elderly family members in the house- Vulnerable croups. Incidence of diarrheal illness, hold explain much of the difference in the type and mortality from diarrheal illness, is generally of toilet facilities in households in Phongsaly. highest among young children and elderly Pour-flush toilets were much more frequent in individuals. Piped water supply and sanitation richer households, and dry pit toilets more fre- facilities, which makes it easier to practice good FIGURE 3.10 Percent of Households with Toilet Facility Phongsaly Lamam 100% 100% 80 % 80% 60 % 60% 40 % 40% 20 % 20% 0% 0% 1 2 3 4 5 1 2 3 4 5 Note: Lowest living standard=1 and highest living standard=5 106 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR hygiene, are therefore particularly important for A sustainable water supply and sanitation sys- households with young children and elderly. tem is de�ned as the system providing an accept- In both Phongsaly and Lamam, the piped able level of services throughout its design life. water connection rate is about the same for This includes acceptable water quality, quantity, households with children or elderly as in house- and continuity, and the physical reliability of the holds without children or elderly. With respect system to deliver services. To this de�nition to basic sanitation in Phongsaly, households should be added healthy hygienic conditions of with children have the same type of facilities as the system, and a community and household households without children. Households with hygiene that helps provide the maximum health elderly individuals, however, more frequently bene�ts of the system. have pour-flush toilets than households with- The PEN study was carried out in 32 villages out elderly individuals. In Lamam, households in four provinces in the North and South of Lao with children and elderly are less likely to have PDR under the leadership of the Department of a pour-flush toilet and more likely not to have Hygiene at the Ministry of Health (Box 3.3). a toilet facility. The water supply consisted of gravity-fed sys- tems (GFS) in 23 villages, and of bore holes in nine villages (Table 3.12). More than 60 percent Sustainability of Rural Water of the water supply systems had experienced Supply and Sanitation breakdowns in the prior 12 months, and 30 per- Over the last decade, there has been signi�cant cent of the systems had more than �ve break- progress in providing improved water supply and downs. In nearly 40 percent of the cases, it took sanitation systems to the rural population in Lao more than five days to repair the systems. The PDR. Water and sanitation village committees perception of households and village leaders were promoted to improve the sustainability of and committees was that poor management was these systems and community and household an important factor in the breakdowns. About hygiene. The water and sanitation systems have, 75 percent of village leaders and committees however, not been free of problems. The Ministry stated that poor management was one of the of Health (MoH), which is responsible for rural causes, and 35 percent of households stated that water and sanitation, was therefore interested in poor management was the main cause of break- a study on factors influencing the non-sustain- downs (Figure 3.11). ability of rural water supply and sanitation. Such The sanitary inspections of the gravity-fed a study was included as a PEN study, with a par- water supply systems found that proper mainte- ticular focus on poverty. nance and precautionary measures to prevent the BOX 3.3 The Department of Hygiene and Prevention at the Ministry of Health implemented a study on sustainability of rural water supply and sanitation in four poor districts in Lao PDR.* The study set out to identify factors influencing non-sustainability, looking at the quality of community management, community participation, inspections and maintenance, and community and household hygiene. Survey questionnaires were developed and administered in 32 villages to over 600 households, 32 village committees or village leaders, and several district of�ces. Sanitary inspections were conducted for all the water systems, 100 latrines, and more than 250 households. *Ministry of Health. 2005. Factors influencing the non-sustainability of improved water supply and sanitation programs in rural areas in poor provinces in Lao PDR. PEN report prepared by Environmental Health Division, Department of Hygiene and Prevention, Ministry of Health, Vientiane, Lao PDR. PEN STUDY IN THE LOWER MEKONG SUB-REGION 107 POVERTY AND ENVIRONMENT IN LAO PDR TABLE 3.12 Study of Rural Water Supply and Sanitation Sustainability Province Phongsaly Oudomxay Sekong Attapeu District Notou Namor Kaleum Phouvong Number of Villages 8 8 8 8 Number of Households per District 156 160 147 160 Type of Water Supply System in villages GFS 7 GFS GFS bore holes 1 bore hole Number of Latrine Inspections 45 40 0 15 Household Hygiene (number of 64 64 64 64 households per district) Notes: GFS=gravity fed system risk of water pollution or to keep good hygiene Sanitary inspections of 21 bore holes in the conditions were inadequate in more than 30 per- nine villages with bore holes also found conditions cent of cases. This included accumulation of in need of improvement. Inadequate fencing stagnant water around the stand post, risk of pol- around hand pumps to keep out animals, stagnant lution through faulty pipes, unscreened and water accumulations, and other nearby sources of unfenced water intake installations, and damages pollution were observed in more than 40 percent to stand post platforms. Risk of pollution from of cases. Latrines within 10 meters of the hand upstream human settlements, farm animals, and pump were also found in a few cases (Figure 3.13). crop production was also identi�ed, but much Soap was available in only 6 out of 100 latrines, less frequently (Figure 3.12). and water tanks were absent in 60 percent of the FIGURE 3.11 Water Supply (WS) Break-Downs WS break-down in the last 1 year Number of days to repair is > 5 days Break-downs > 5 times per year Poor management is one of the causes of break-down Poor management is main cause of break-down 0% 10% 20% 30% 40% 50% 60% 70% 80% 108 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.12 Sanitary Inspections of the Gravity Fed Water Supply Systems Stagnant water around stand post Risk of pollution through water pipe Unscreened intake installation Unfenced water intake installation Damages to stand post platform Upstream farm animals polluting the water source Damages to water tank Upstream crop production polluting the water source Upstream human settlements polluting the water source 0% 10% 20% 30% 40% 50% cases. Cleanliness was unsatisfactory in 40 of the village. Villages with ethnic minorities seemed latrines, and 20 to 30 of the latrines suffered also to face more dif�culties in avoiding break- from poor physical conditions (Figure 3.14). downs, and so did poorer villages (Table 3.13). Several factors were associated with water Demand for the water supply systems origi- supply breakdowns and frequency of break- nated from the villagers in nearly 70 percent of the downs in the 32 villages. Some of the most cases, or the village in nearly 95 percent of the important factors included low demand respon- cases. More than 90 percent of households pro- siveness; no �nancial fund for proper operation vided labor and materials. Post-project monitor- and maintenance and repairs; no regular inspec- ing took place, however, in only 55 to 60 percent tion of systems; and low education levels in the of the cases, and regular sanitary inspections of the FIGURE 3.13 Sanitary Inspections of the Bore Holes Inadequate fencing around hand pump Stagnant water within 2 m of hand pump (faulty drainage) Other sources of pollution within 10 m of hand pump No concrete floor around the hand pump Latrine within 10 m of hand pump 0% 10% 20% 30% 40% 50% PEN STUDY IN THE LOWER MEKONG SUB-REGION 109 POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.14 Household Latrine Inspections No soap in the latrine house No water tank in the latrine house Bad smells Bad condition of latrine house Bad condition of latrine floor 0% 20% 40% 60% 80% 100% systems were performed in less than 40 percent of many poor communities: rural location, no vil- the villages. The poorest performance was with lage access to a road, illiteracy or low education, respect to having a well-functioning water and or UXO contamination. The poor have lower sanitation village committee, training of village coverage rates of improved water supply and committees, and household collections for an basic sanitation in urban areas too. Low educa- operations and maintenance fund (Figure 3.15). tion and unaffordability of services explain some of this disparity. The study in Phongsaly and Lamam con�rms that these factors are important Discussion in small towns too, pointing to the high water Over the last decade, there was significant connection fees and low education levels as the progress in providing improved water supply and main reasons for low connection rates among basic sanitation. The poor, however, lag far poor households. Other factors include house- behind, and no relative gains were achieved. Four holds not owning their property or a location factors explain much of the low coverage rates in away from main roads. TABLE 3.13 Factors Associated with WS System Breakdowns Low demand responsiveness (e.g. village participation) No �nancial fund for proper O&M and repairs No post-project monitoring by government and international agencies No regular inspection of WS system Low education level Ethnic minorities Poverty Note: Factors were statistically signi�cant at 95 percent level (Chi-square). 110 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR FIGURE 3.15 Community Participation and Water Supply System Management Contribution of labor from village/villagers Demand for WS originated from village Contribution of material from village/villagers Demand for WS originated from villagers Government monitoring of WS after completion International agency monitoring of WS after completion Financial contribution from village/villagers Village Committee sanitary inspections > 8 times/year Village has WSS Village Committee Village Committee participated in training Money collected for WSS fund 0% 20% 40% 60% 80% 100% Provision of rural water supply and sanitation hygiene. These factors all point to the need for will need to be further invigorated if equity in increased emphasis on community capacity build- access to services is to be achieved. Particular ing with proper management training, commu- attention is needed for communities with no nity participation, information and awareness access to a road and in areas with high UXO con- campaigns, and instituting procedures to establish tamination, as these communities are among the village funds for maintenance and timely repairs. poorest in the country. Sustainability of services In urban areas, the household connection fee for these communities is an important considera- is of critical importance. Its high level is one of the tion. They tend to have higher illiteracy or lower main reasons for poor households not obtaining a education levels, pointing to a strong need for water connection. The fee is a substantial share of developing capacity for sustainable management poor households’ annual expenditure on all goods of water supply and sanitation services. and services and non-food expenditure. As �nan- Poor rural communities seem to be more cial sustainability of water agencies is an impor- inflicted with water supply system breakdowns tant objective of the government, any policy to than better-off communities. Inadequate manage- adjust the connection fee would need to strive to ment is most frequently reported as the main cause be revenue-neutral. or important cause of these breakdowns. Poor Household education level is an important communities are also less likely to establish village determinant of demand for improved water funds for maintenance and repairs. Hygiene supply and basic sanitation. Information and improvements are also much needed to fully ben- public awareness programs about the bene�ts of e�t from improved water supply and sanitation. improved services—such as opportunities for This includes regular inspection and proper improved household hygiene and potentially hygiene maintenance of water supply systems, better water quality—can therefore be an impor- appropriate community behavior in relation to tant instrument for encouraging household con- water use, and improved household and personal nections. This is particularly important among PEN STUDY IN THE LOWER MEKONG SUB-REGION 111 POVERTY AND ENVIRONMENT IN LAO PDR poor households, because they generally have tle is known about the level of exposure to fuel less education. Public awareness programs should wood smoke and associated health effects in Lao also target households with young children and PDR. Initial research by Laotian institutions41 elderly household members. These groups are seems to con�rm concerns raised by WHO that more at risk of disease. eye and lung irritation from fuel wood smoke Satisfaction with the piped water supply sys- appears to be a substantial issue, especially among tem was found to be higher among households women and young children. with higher education levels. More educated Poor people do not have suf�cient funds to buy households may appreciate the value of having simple energy saving wood stoves. About 95 per- piped water, just as their demand for piped water cent of the non-poor population and as much as is higher than households with lower education 98 percent of the poor population are using wood levels. Public awareness programs communicat- for cooking. Similarly, 98.7 percent of all house- ing the values of having improved water supply holds are using wood for cooking in the group of may therefore also help increase satisfaction in poorest districts compared to 80.9 percent in the addition to increasing connection rates. group of least poor districts. The poor are, there- Poor households in urban areas are also affected fore, likely to be disproportionately affected by by network availability. Piped water is sometimes indoor air pollution. However, the number of not available in the area of the town where the people affected depends on factors such as indoor- poor live. This is the case in some areas of both outdoor cooking practices, type of housing and Phongsaly and Lamam towns, and especially for ventilation, and type of fuel wood consumed. households living along walking paths and away from main roads. As these households tend to be Malaria poorer than other households, it is important that According to WHO data, the total number of planning of network expansions should consider con�rmed malaria cases declined from 80,000 in equity. 1996 to 27,000 in 2001. Based on regression For new water and sanitation projects in analysis of province-level LECS III data, an esti- small towns in Lao PDR—and for network mated 1 to 2 percent of the poor and 0.2–0.4 per- expansions and service improvement in towns cent of the non-poor were affected by malaria. with existing piped water supply—an increased This, however, only relates to con�rmed cases of emphasis on a participatory approach, and malaria and is likely to underestimate the number information and awareness campaigns in project of actual cases. preparation can help enhance user satisfaction, A substantially higher number of people may ensure appropriate service levels, allow a tariff be also affected by the threat of malaria in the sense structure and connection fees that promote that they may restrict their activities during the equity, and help ensure that household expecta- peak malaria season to avoid contracting the dis- tions are realistic and enable households to make ease. The PPA (ADB 2001) states that 64 percent informed decisions with their scarce �nancial of participating villages reported malaria to be a resources, particularly in poor communities. main health issue. While six years of data are insuf- �cient to draw �rm conclusions about trends, it appears that the non-poor have bene�ted more OTHER ENVIRONMENTAL ISSUES from the reduction in annual malaria cases dur- ing the 1996–2001 period: 72 percent of malaria Indoor Air Pollution cases are estimated to have occurred among the More than 95 percent of the population in Lao poor in 1996, a �gure that increased to 80 per- PDR uses fuel wood for cooking (LECS III). Lit- cent in 2001. 112 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR Natural Disasters TABLE 3.14 Estimated Number of Poor and Non-Poor The impact of floods and droughts on the poor Affected by Natural Disasters versus the non-poor was estimated through a Non-Poor Poor regression analysis using 1995, 1996, 1997, and 2002 data from the Natural Disaster Manage- Affected by flood (number of people) 149 000 16 000 ment Of�ce in Lao PDR.42 The results obtained Affected by flood (percent of people) 5.1% 1.0% point to a strong indication that floods largely Affected by drought (number of people) 28 000 5 000 affect the non-poor, while drought and other Affected by drought (percent of people) 1.7% 0.2% disasters largely affect the poor. It also seems likely that the number of poor affected by drought Note: Figures are based on 1995 population and do not take into account population growth between 1995 and 2002. is higher than the number of poor affected by floods, even though the area affected by flooding is substantially higher than the area affected by drought. which the Lao PDR government seeks to sup- port natural resource conservation in its upland areas. The government has set impressive targets: POVERTY-ENVIRONMENT IN THE stabilization of shifting cultivation and eradica- NATIONAL CONTEXT tion of opium production. The land use planning The �ndings from the PEN focus studies in the and land allocation policy is a main instrument North and the South, along the nationwide in the stabilization of shifting cultivation. The analysis, are broadly relevant to upland districts, “focal area� policy, whereby government efforts remote border districts along the Vietnamese focus on a small number of villages closer to the border and the mountainous sections of the road network, was �rst designed as the rural Chinese and Cambodian borders, districts with development policy of Lao PDR. It is presented a high share of ethnic minority people, and dis- as the policy for community-driven rural devel- tricts with high UXO contamination. A large opment in poor districts in NGPES. percent of districts in Lao PDR and all 46 prior- Village consolidation is proceeding at a steady ity districts share at least one of these criteria pace. In 2003, the number of villages was down (Table 3.15). to less than 11,600 villages or 74 percent of The natural resource management PEN stud- the number of villages in 1997.43 Although the ies, although located in marginal districts, shed datasets clearly indicate that the policy has been light as a whole on the set of policies through implemented more rapidly in lowland areas, it is TABLE 3.15 Number of Districts Sharing PEN Focus Study District Criteria Villages with Remote Ethnic Minority Mon Khmer High UXO Border People Above People Contamination Districts Upland 1/ Districts 50 % Above 43% Above 15% Lao PDR 74 70 78 39 37 NGPES priority 33 24 42 25 18 poor districts 1/ De�ned as districts with maximum elevation above 500 m. PEN STUDY IN THE LOWER MEKONG SUB-REGION 113 POVERTY AND ENVIRONMENT IN LAO PDR also taking place in upland areas, often in combi- PEN study provides further evidence of their nation with the other rural development policies. importance. The PEN studies provide evidence from three dif- ferent perspectives and lead to a single conclusion: Rural water supply & sanitation. Five key while these policies may well reflect a sound vision recommendations arise from the PEN study for the longer-term future of the uplands, their to achieve equity in access, sustainability of rapid pace of implementation runs counter to services, and health protection of the poor in their stated objective of poverty reduction, and rural communities: has achieved limited gains in terms of resource conservation. • Rural communities without access to roads and The Mon-Khmer ethnic groups have a special in areas with high UXO contamination are dis- position in the analysis of poverty and environ- proportionately lacking improved water supply ment linkages. They have long been recognized as and sanitation. These are among the poorest indigenous groups with a strong tradition of sus- communities in Lao PDR. A strategy to reach tainable rotational agriculture. It is increasingly these communities needs to be developed. The understood that they face speci�c poverty issues. strategy will need to take into account village The PEN study �ndings appear to have high rel- consolidation plans and carefully evaluate cost evance for the Khamu people in the northern and maintenance implications of water supply region and the various small groups along the options for these communities. Vietnamese border in the South.44 • Sustainability of water supply and sanitation The poverty and environment challenges services is essential for achieving equitable facing the Mon Khmer people in the South are access and health protection of the poor. Estab- reflected in the PEN focus studies. They often lishment of well-functioning village commit- live in villages with high UXO contamination. tees therefore needs to be further emphasized, They live in sparsely populated, highly forested and periodic assistance might be needed to areas, in which the national policy framework ensure their continuity and ef�ciency. This is is conducive to proactive resettlement. Yet the particularly important in poorer communities, PEN focus studies tell vivid stories of commu- which seem to have more problems with water nities that are adjusting, diversifying their liveli- supply system breakdowns. hoods, entering the market economy, and looking • Communities with high illiteracy or low edu- ahead at a more community-based management cation are more likely to not have improved of local natural resources. Innovative solutions water supply and basic sanitation. Particular are needed to address poverty and environment attention is required in developing manage- jointly in these circumstances, and the PEN ment capacity in these communities. This studies only provide some initial insights about includes proper training in water supply sys- such solutions. tem management, developing mechanisms for community participation, establishing regular inspections and maintenance procedures, and Policy Implications and strengthening of monitoring and evaluation of Recommendations water and sanitation projects. The following section provides some policy • Village funds are essential for proper system implications and recommendations that emerge maintenance and timely repairs. Building com- from the PEN study. While many of these rec- mitment and procedures to institute a fund, ommendations are not necessarily new, the even if gradually, therefore needs to be further 114 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR emphasized, even in the poorest communities. tation is therefore an important ingredient Further consideration should also be given to in achieving equity. In planning new piped evaluating system options in light of mainte- water supply systems in other towns, under- nance and management requirements. standing perceptions of these bene�ts is there- • Good community, household and water sup- fore a prerequisite. ply system and latrine hygiene is essential in order to fully benefit from improved water Transport. The PEN road study identi�ed four supply and sanitation. Hygiene promotion recommendations for poverty reduction and programs therefore need to be developed and sound environmental management in Lao PDR: implemented. Improved hand-washing and point-of-use drinking water disinfection are • The NGPES’s focus on ensuring all-weather particularly effective in protecting health access roads to all district towns is relevant, (Fewtrell and Colford 2004). but policy and technical decisions to help improve the local road network beyond that Urban water supply & sanitation. Avail- stage need to be taken now. A village road ability of improved water supply and sanitation network will have substantial poverty reduc- is much more widespread in urban than rural tion impact, provided an appropriate prior- communities. However, much is still to be itization of villages to be served first is achieved in small towns, especially for the poor. ensured. This will require long-term plan- Three priority recommendations have been ning and management. Experience from identi�ed: donor-supported pilot projects deserves to be summarized and mainstreamed. Risks of • Equitable access to piped water supply in urban negative environmental impact may be lim- areas requires a review of water connection fees. ited overall, but should remain a concern in The fees are unaffordable for many poor house- and around areas of environmental value. holds. A revenue neutral policy change could be Confirming local responsibilities in natural considered whereby water tariffs are raised to resource management appears to be a viable compensate for a reduction in connection fees. and efficient option. This would prevent compromising the �nan- • A comprehensive and up-to-date monitoring cial viability of water agencies and quality of system of the local road network at the national services. level is a key tool to manage the poverty and • Network expansions should be equitable so environmental aspects of rural road develop- that the poor are not the last households to ment. There is room for substantial improve- receive water supply. This implies that social ment. A combination of improved database considerations receive an equal footing with tools and of more transparent information �nancial and technical considerations in net- flows between national and local levels of gov- work expansion. Particular attention is required ernment is needed. for so-called “temporary� housing and settle- • There is an urgent need to recognize the risks ments further away from main roads. and opportunities of the involvement of private • Affordability is not the only reason the poor enterprises from various sectors in road con- do not have water connections or basic sani- struction in Lao PDR’s many border districts. tation. Education level is equally important. The capacity of enterprises to invest in roads is Information and public awareness of the an opportunity in terms of poverty reduction. benefits of improved water supply and sani- The government has an important role in mon- PEN STUDY IN THE LOWER MEKONG SUB-REGION 115 POVERTY AND ENVIRONMENT IN LAO PDR itoring environmental impacts at the national shifting cultivation, development of larger level. Enterprises need to be encouraged to communities (preferably along roads), and select sustainable road construction technology, focus of government efforts on these so that the roads create a long-term poverty communities, may be an appropriate vision in reduction effect. This objective may require the the very long term. However, all the PEN creation of �nancial incentives. studies provide evidence that the current high • Narrow tracks are more cost-efficient and pace of implementation of this policy is environmentally friendly than roads. Build- counterproductive, both in terms of sustainable ing tracks instead of roads may attract more poverty reduction and environmental pro- interest, as small tractors for farmers and tection. A slow, gradual transition in the land motorbikes for government staff visits to vil- use systems of upland communities appears lages become more widely available. It also to be a much more appropriate option. The remains an option worth investigating in the Government of Lao PDR should use this near future. increasing evidence provided by its researchers and donor-funded projects to turn to more Agriculture and forestry. The PEN study provides people-centered, service-oriented ways of the basis for two recommendations related to supporting communities that wish to migrate NTFPs: and/or intensify their land use practices. Providing support to communities off the road is a necessity, • Systematic monitoring of NTFPs in terms of if this is to be achieved. This is not an easy task, resources, products, cultivation, and markets is and innovative programs will be needed. a necessity in order to better manage risks of The UXO study demonstrates that active resource decline and market collapse. Both donor contribution to UXO decontamination would disproportionately affect the poor. Mak- should be encouraged. UXO decontamination ing cultivation of NTFPs a fully viable and rec- is a win-win solution in terms of poverty reduc- ognized economic activity is also needed in that tion, improved environmental management, regard. This objective will require not only con- and providing safe water supply and sanitation. tinued applied research but also improvements Decontamination efforts will, however, require in the legal and policy framework. a very long period of time, and in the mean- • The land use planning and land allocation pol- time, all coping options for local communi- icy, as stated in NGPES, needs to be made ties should be supported. In southern Laos, more appropriate to the poverty reduction the incompatibility of meaningful support for needs of upland villages. The high motivation UXO-affected communities with a proactive of local communities to better manage local resettlement policy must be recognized and resources is an opportunity. Mainstreaming addressed. Diversified community initiatives, successful local pilots in community participa- with or without migration and inside or out- tion can help take advantage of this opportu- side of agriculture, deserve careful attention nity. NTFP harvesting and marketing are an and support. ideal entry point for overall community forest Any joint solution in terms of poverty reduc- management in areas where NTFP collection tion and environmental protection merits careful for sale remains an important income source. geographical targeting and improved coordination between stakeholders, especially more intensive Cross-sectoral interventions. The comprehensive community participation. Table 3.16 provides a policy of the Government of Lao PDR for its summary of PEN study �ndings related to each upland regions, which includes stabilization of recommendation. 116 PEN STUDY IN THE LOWER MEKONG SUB-REGION TABLE 3.16 Lao PDR Summary of Priorities for Programs with Joint Positive Poverty Reduction-Environment Impact Coordination needs between Development areas Program priority Geographical targeting sectors and agencies All sectors Integrated livelihood investment program in All villages within districts, All sectors, district government, participation priority poor upland districts both along roads and of village communities, NGOs away from roads Support to diversi�ed development strategies Districts along Vietnam Lao-UXO program, district government, par- in UXO-affected districts borders, with a focus on ticipation of village communities, NGOs • Diversi�ed income generation including Mon-Khmer villages in livestock and off-farm the South • Slow transition out of swidden agriculture Agriculture Community participation in natural resource Uplands districts Agriculture & forestry with transport & forestry management: Northern region for NTFPs District government, participation of village • More participatory land use planning communities allowing for slow transition out of swidden agriculture • NTFP harvesting and marketing used as an entry point in community resource management Sustainable NTFP production: National government • Comprehensive monitoring system for District governments NTFP production and markets Agriculture & forestry department • Policy statement, legal framework and Research organizations applied research promoting cultivation (continued ) PEN STUDY IN THE LOWER MEKONG SUB-REGION 117 POVERTY AND ENVIRONMENT IN LAO PDR 118 TABLE 3.16 Lao PDR Summary of Priorities for Programs with Joint Positive Poverty Reduction-Environment Impact (Continued ) Coordination needs between Development areas Program priority Geographical targeting sectors and agencies Health Rural clean water & sanitation services: Poor districts in the North Namsaat, Ministry of Health and support POVERTY AND ENVIRONMENT IN LAO PDR • Improving the sustainability of services in and South and along the from national and provincial government poor communities that already have services border region to Vietnam • Community and household hygiene promo- tion programs • Targeting of service provision to poor, marginalized and disadvantaged commu- nities (the ones that are least likely to PEN STUDY IN THE LOWER MEKONG SUB-REGION attract services on their own) Urban clean water & sanitation services: Small provincial & district WASA, Nam Papa, urban village leaders and • Review of water network connection fee in towns community organizations small towns • Public awareness campaigns to increase demand for improved water supply and sanitation in poor households • Due consideration to equity in network expansion planning Transport Mainstream basic road access prioritization Marginal districts National, provincial, and district levels with and monitoring approach into national Especially border districts village community participation road construction program Agriculture and forestry for accompanying community-based natural resource management Regulatory framework and monitoring system Border upland districts National, provincial, and district levels with for enterprise-based road construction enterprises POVERTY AND ENVIRONMENT IN LAO PDR Costs of Implementation scale independent providers in a few towns in Lao PDR is in the range of US $5–15. A range of US Implementation of the recommendations pro- $10–40 per capita suggests a total investment posed above involves various levels of costs. Some requirement of US $6–23 million to provide all costs can be expected to be relatively modest, remaining small towns, i.e., about 85 towns, with such as monitoring, awareness and information improved water supply by 2020, reflecting a programs, and hygiene improvement programs, small-town population growth rate of 3.5 percent but do nevertheless require sustained commit- per year.45 ment and human and logistic capacity in order to While signi�cant resources are required to be effective and sustainable. Other interventions implement many of the proposed recommenda- such as water supply and sanitation and rural tions from this PEN study, the effectiveness roads involve capital investments. The level of and sustainability of implementation clearly also these investments depends, among other things, depend on provincial and local government com- on service level. While it is dif�cult to provide mitment, capacity and contributions; programs exact estimates of investment requirements, some to raise awareness and capacity, and for making �gures are available to provide a plausible cost informed choices in local communities; and range for provision of rural and urban water sup- demand by these communities and local authori- ply and sanitation. ties for programs that addresses poverty reduction Available capital cost estimates in Lao PDR and natural resource management and provides for rural water supply are on the order of US environmental services such as water supply and $10 per capita, which include construction mate- sanitation. Factoring in these conditions for effec- rials and construction of facilities. In addition, tiveness and sustainability is needed in regional communities generally contribute about 40 per- poverty reduction targeting, such as for the Gov- cent of local cost in-kind, labor, local materials ernment’s 46 priority districts and other districts and cash. External support cost of construction in need of poverty-environment interventions materials for basic sanitation facilities is estimated and programs. In this context, as part of the PEN as low as US $2 per capita, with communities study, a research paper on the economics of contributing about 85 percent of local costs. regional poverty-environment programs with These cost estimates suggests an overall invest- application to Lao PDR was prepared to shed ment requirement of US $22 million to reach an some light on central government or donor allo- 85 percent rural coverage rate of improved water cation of resources to provinces and local com- supply and basic sanitation in Lao PDR by 2020 munities. The paper takes into account trade-offs as targeted by the NGPES. This cost is in addi- between multiple objectives, and differences in tion to local community contributions of labor investment and administrative costs across regions and local materials. within a country.46 There is more uncertainty as to the cost of pro- viding urban water supply and sanitation. A pri- ority is provision of improved water supply to Endnotes small towns which lag behind main urban centers. 1. Lao Expenditure and Consumption Survey 2002/03 The historic per capita cost of urban water supply (LECS III) and 1992/93 (LECS I). in Lao PDR is about US $135. This however is 2. One of the districts, in Vientiane Municipality, has likely to be substantially higher than what is “graduated� and there remain now 46 priority districts. 3. The district poverty data are from ADB (2001), based required for small towns. The cost in small towns on poverty incidence in 1998 from LECS II. in Vietnam and Cambodia are estimated at US 4. Forest data are from satellite imagery for the period $23–40 per capita, while the capital cost of small- 1993–97. These data were the most recent consistent PEN STUDY IN THE LOWER MEKONG SUB-REGION 119 POVERTY AND ENVIRONMENT IN LAO PDR data available at the time of the PEN analysis presented urban population data for the nine districts created in the this report. after the 1995 census. 5. This does not include land in fallow/re-growth, but 20. Kaleum borders A Luoi district in Thua Tien Hue only cultivated land. province in Vietnam. The timber sector there is active 6. Linkages between reliance on open unprotected surface in conjunction with the Da nang port. The newly ren- water and health effects are discussed in the water and ovated Ho Chi Minh road passes through A Luoi and sanitation section. this might increase attraction toward villages in Kaleum. 7. Access to roads is measured as the percentage of villages 21. The border in Nhot Ou became an of�cial international with main or secondary roads within 6 kilometers. border in recent years. It crosses to the fast-growing 8. Poverty incidence is from LECS II as reported in ADB Simao prefecture, Yunnan province. People in Kaleum (2001). go to Saravanh, the next provincial capital. 9. It should be noted that the data on malaria incidence 22. The survey took place in 72 percent of villages in Lao reflect con�rmed cases. From the surveys in ADB PDR; 31 percent of surveyed villages were found to be (2001) households do report however that malaria is a impacted by UXOs. major threat to their health. 23. This was estimated with a district level regression equa- 10. The results of a regression analysis of the variations in tion by combining district poverty incidence data from district forest cover in 1997 are provided in the PEN LECS II 1997/98 (ADB 2001) and the UXO national background report for Lao PDR. survey data. 11. All the factors were assessed in a regression analysis 24. Even the �gure for total land area could not be con- using LECS III household data (combined with UXO �rmed. contamination data). They were all found to be statis- 25. Village communities have only 32 households on aver- tically signi�cant. This means that each factor is statis- age (20 in the six villages surveyed) and half this num- tically associated with poverty, after controlling for the ber of houses due to extensive occurrence of extended families. other factors. The regression results are available in the 26. Phouvong, the other PEN study district in the South, PEN background report. is a good example. 12. Swiddening is used here as a neutral term when refer- 27. Non-agricultural land accounts for more than one- ring to the agricultural practice itself, and the term third of total land area decontaminated by the Lao shifting cultivation is used when referring to the policy. UXO program in 29 districts out of 57 where activities 13. This is a reminder of the limitations of analyzing have taken place. poverty at a single level, the district. It might also indi- 28. Phouvong, the other district studied in the South, is cate a lower importance of marketed NTFPs in the pri- a national focal area (JICA 2002) where the newly ority districts, or a lower interest in this non-of�cial reclaimed lowlands are heavily affected by UXOs. survey. 29. Population Census 1995 and Lao Expenditure and 14. A wealth indicator was created for each household sur- Consumption Survey 2002/03 (LECS III). Improved veyed based on household durable goods and produc- water supply is here de�ned as piped water, protected tive assets. groundwater, rainwater, and “other water supply.� 15. Households living in villages within a 6 km radius from “Other water supply� mainly refers to gravity-fed sys- a road can generally make a round trip to a road within tems, but it is unclear from the LECS III data if this the same day. category also includes any types of unimproved water 16. NGPES has de�ned low access as presence of at least supply. trails accessible by cart during the dry season. Donor- 30. In 1995, about 60 percent of the poor and 40 percent of funded projects on basic road access had proposed a the non-poor used surface water. The change from 1995 higher but easier to observe standard of motorable to 2003 was therefore 50 percent for both the poor and access 10 months per year (WSP civils 2004). non-poor, with no relative gain for the poor. The �gures 17. The results are from a regression analysis of district level for 1995 are estimated from the Census 1995 and LECS data, controlling for agricultural land, population den- II. The �gures for 2003 are from LECS III. sity, border location, and regional and topographic dis- 31. Estimated from the Census 1995 and LECS II for 1995 trict characteristics. and LECS III for 2003. 18. Population growth is high and assessed at 2.5 to 32. Twenty percent poorest and richest as de�ned by con- 2.8 percent, but strong migration trends from more sumption quintiles from LECS III. marginal to less marginal areas result in a stable or even 33. From a district-level regression analysis of data com- decreasing population in a fair proportion of districts. bined from the 1995 census, poverty incidence from The 2005 census will provide more accurate popula- LECS II, and UXO data from the national survey in tion trends. 1997. 19. 15 districts have an urban population of more than 34. The preceding analysis is from LECS III 2002/03. 30 percent and may be de�ned as urban; there is no 35. Logistic regression analysis of LECS III household data. 120 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN LAO PDR 36. Recall that poor households are 50 percent more likely References (To be completed) to rely on surface water than non-poor households. However, the results are after controlling for the other ADB 2001. Lao PDR Participatory Poverty Assessment. factors. Those factors are associated/correlated with Bouahom, Bouthong and others, eds. 2005. “Poverty reduc- poverty. The result (i.e., 15 percent) is therefore lower tion and Shifting Stabilization in the uplands of Lao than the “uncontrolled� result (i.e., 50 percent). PDR: Technologies, Approaches and Methods for 37. WASA, DHUP, and URI. 2002. “Small Towns Water Improving Rural Livelihoods.� Proceedings of a Work- Supply and Sanitation Initiative in Lao PDR.� Pre- shop Held in Luang Prabang, Lao PDR, January 27–30, 2004. Vientiane: National Agriculture and pared with support from WSP-EAP. Forestry Research Institute. 38. These results are from a regression analysis of the Buys et al (2004). The Economics of Regional Poverty- household data from the URI survey. Environment Programs: An Application For Lao PDR. 39. This result is from a regression analysis of the house- Prepared for EASEN, World Bank. hold data, controlling for household location and Chaignon, Jacquelyn. 2000. “UNDP Sekong Indigenous importance of time savings. People’s Development Program: Inception Report & 40. After controlling for living standard. Extended Program Strategy.� City: publisher. 41. Lao Women’s Union Gender Resource Information Goudineau, Yves, ed. 1997. “Resettlement and Social Char- and Development Center, 1999. acteristics of New Villages: Basic Needs for Resettled 42. The following equation was estimated, one for floods, communities in the Lao PDR.� An ORSTOM Survey. one for drought: Vols. 1 and 2. Vientiane: UNDP. N = a + bp POPp + bnp POPnp + e Government of Laos (Lao PDR?). 2003. The National Poverty Eradication Program (NPEP): a Comprehensive where N is number of people affected, and POPp and Approach to Growth and Development. Eighth Round POPnp are number of poor and non-poor respectively Table Meeting, City, September 2003. in each province affected by flood or drought. The Handicap International. 1997. Living with UXO: National constant (a) was restricted to zero in the regression to Survey on the Socio-Economic Impact of UXO in Lao reflect that N will be zero if there are no poor or non- PDR. Final report. Prepared for the Ministry of Labour poor people. As the data contained no information on & Social Welfare, Lao National UXO Programme. the number of people affected, a conversion from City: publisher. hectares damaged to affected people was undertaken IUCN. 1997. Manual of EIA Procedures for Road Projects in based on agricultural land per household and average the Lao PDR. Prepared for MTRPC with Assistance household size. This conversion provides an order of from SIDA. City: publisher. magnitude. Two regression equations were also esti- JICA. 2002. Government’s Efforts to Rural Development in mated with slope dummy variables for provinces that Consideration with Ethnic Minorities in Lao PDR. City: were not affected by flood or drought, little affected Lao Consulting Group. Kakwani et al 2002. title. City: publisher. and medium affected. The estimated coefficients from Lao National UXO Program. 2004. Turning Point: UXO these equations allow for an estimate of the number Lao Work Plan 2004. city: publisher. of poor and non-poor affected by flood or drought Lao UXO Program. 2002. “UXO Lao Community Aware- across the country. ness In-depth Study, Khammouane.� City: publisher. 43. This was the newly reported number of villages for Lao Women’s Union Gender Resource Information and the next population census. An even lower number of Development Center, 1999. title. City: publisher. villages, 10,700, was reported in 2003. The reason MTCPC. 1996. “Notes from the Proceedings of the Lao behind this difference between 2003111111 and 2005 Road Sector Donor Coordination meeting.� Paper pre- is unclear. sented at the Lao Road Sector Donor Coordination 44. Similar or closely related groups live on the Vietnamese Meeting, February 6 and 7, 1996, city. side of the border. Several of them are of�cially recog- SIDA/MTCPC. 2004. “Community Road Management nized as ethnic groups with special dif�culties in Vietnam Implementation Guide.� City: publisher. (chapter 5). World Bank. 2004. “Upland Livelihoods Improvement Proj- 45. The cost estimates of rural and urban water supply and ect Concept Note.� Washington, DC: World Bank. sanitation in Lao were provided by Thomas Meadley, World Bank. 2005. “World Developement Indicators.� World Bank, Vientiane Lao PDR. Washington, DC: World Bank. 46. See Buys et al (2004). The Economics of Regional WSP Civils 2004. Technical Assistance report to the Basic Poverty-Environment Programs: An Application For Access Component of the Lao Swedish Road Sector Lao PDR. Prepared for EASEN, World Bank. Project 3. PEN STUDY IN THE LOWER MEKONG SUB-REGION 121 4 Poverty and Environment in Vietnam INTRODUCTION The methodology of the PEN research, just as in the two other countries, starts with observing In 2004, some 15.6 million people in Vietnam— correlations between poverty and environment 19 percent of the population—were still living indicators through mapping and correlation analy- below the poverty line despite the impressive sis. Local quantitative and qualitative research on poverty reduction gains made in recent years. This speci�c issues is carried out to shed light on some is equivalent to 80 percent of the total population of the causes behind these correlations. Evidence of Cambodia and Lao PDR. The Government collected at various levels was then combined to of Vietnam faces the challenge of continuing its answer the speci�c research questions de�ned in poverty reduction efforts while ensuring preser- each problem area. The national research groups vation or restoration of environmental resources. have provided significant help in assembling Are there win-win policy options and interventions national datasets. They have carried out all local that can jointly address poverty and environment? research activities. How can a donor community that supports the Vietnam is a highly diverse country, and the government’s growth and poverty reduction strat- PEN research covers as much of this diversity as egy best contribute to these win-win solutions? possible. Poverty and environment linkages in These are the questions that the PEN research Vietnam can be structured into two broad cate- seeks to answer. gories, environmental health and natural resource The PEN research is a collaborative effort use. Diverse poverty issues and environmental between a World Bank team and three ministries, issues are at stake in each of these two categories. two of which are cross-sectoral agencies (Box 4.1). The scope of the PEN research encompasses the This enables the research to encompass policies rural, urban, and industrial sectors. In the rural and programs with a multi-sectoral scope. Land sector, the research targets “poverty pockets,� administration and sustainable development geographical areas trapped in very high poverty are two of these horizontal areas. As part of the rates and deep poverty. It also covers the various strengthening legal framework in Vietnam, the policy tools of the national policy framework land law is in a process of �nalization in the form (Box 4.2). of a “code.� The Cau River basin is the �rst area A set of �ve detailed research areas remains selected for a pilot sustainable development pro- insuf�cient to cover the diverse dimensions of gram under the national Agenda 21 program. poverty and environment in Vietnam. The spe- PEN STUDY IN THE LOWER MEKONG SUB-REGION 123 POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.1 PEN Research Partners in Vietnam – The Ministry of Natural Resources and Environment is in charge of both the environment and the management of land and other natural resources. MoNRE was the PEN research partner for detailed research on industrial pollution and on land administration. – The Ministry of Health was the PEN research partner for detailed research on water supply & sanitation. – The Ministry of Planning and Investment is in charge of comprehensive advice on socioeconomic development strategies, programs, and plans. MPI’s National Of�ce for Sustainable Development was the PEN research partner for detailed poverty and environment analysis in the Cau River basin. The following groups have undertaken research activities: the expert group at Hanoi Medical Col- lege for water supply & sanitation; the Research Center for Energy & Environment for industrial pollution; the Center of Occupational and Environmental Health of the Vietnam Association of Occupational Health and the University of Economics, Ho Chi Minh City, for pesticides; TECOS under MoNRE for land administration; and the Socio-Economic Development Center for the Cau River. This chapter includes an analysis of the databases assembled by the research partners, while the discussion on searching for causes behind poverty and environment correlations is largely based on the national reports by the research groups. ci�c issues of protected areas are largely left aside, The PEN research uses a large recent data and so is the detailed analysis of natural resource set, the National Health Survey, and has itself use and poverty in the lowlands, which is a lower produced a new database by pooling together priority compared to the uplands. Field observa- statistical data from different administrative tions mostly also leave aside border areas and the lines. Some important datasets have, however, Central Highlands. The national datasets, how- remained outside of the scope of the study. Dis- ever, allow drawing conclusions on a broader trict-level data was assembled whenever possi- scale after careful analysis of the relevance of the ble. Since provinces normally assemble district case studies in the national context. data but only provide province-level data to the BOX 4.2 Detailed PEN Research Areas and Policy Framework Environmental health and poverty Water supply, sanitation, and poverty: a national analysis of the recent Vietnam National Health Survey. Policy framework: sector policies and area-based programs in health, agriculture, and rural development. Industrial pollution and poverty: a set of case studies combined with analysis of the National Health Survey. Policy framework: 1/ national remediation program for large polluting industries; 2/ regional remediation program for household-based polluting industries in the Red River Delta; and 3/ strengthening of the legal framework in the urban sector. Pesticides, health, and poverty: a case study on pesticide use in the Mekong River Delta. Policy framework: national program for the reduction of agricultural inputs. Natural resource use and poverty Land administration, environment and poverty: a case study in three provinces on the impact of the new land law in rural upland regions combined with the analysis of a national land use database. Policy framework: strengthening of the legal framework in the rural sector. Integrated poverty and analysis research Poverty and environment in the Cau River basin: a case study in two provinces on natural resource use in the upstream section of the river basin and environmental health in the downstream section. Policy framework: National Agenda 21 program. 124 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.3 National Data Sources for PEN in Vietnam The Vietnam Living Standard Surveys (VNLS) are large-sample household surveys designed to mea- sure trends in poverty and associated factors. The most recent was released in 2004. Previous sur- veys were carried out in 1997–98 and 1992–93. The surveys are representative at the province level. Small-area estimation techniques use correlations between poverty and statistical data available at more local levels in order to assess poverty down to the commune level. Unless otherwise indicated, all poverty data in the report are from these sources. The Vietnam National Health Survey (VNHS) carried out in 2001–02 provides a unique opportu- nity to assess the impact of poverty and various other factors on environmental health in Vietnam. The survey covers 36,000 households and 150,000 individuals in all provinces. It includes information on drinking water supply, sanitation, household behavior, water pollution sources, prevalence rates, and indicators of severity of diseases in all age groups, as well as living standards, education levels, and housing characteristics. Unless otherwise noted, this survey is the source for all national analysis in water & sanitation and health. The PEN II land use database was assembled by TECOS from datasets provided by CEMMA, GSO, MoNRE’s 2000 and 2003 land census, land use and detailed land allocation statistics, the Ministry of Transportation, and the Center for Hydro-Meteorology and Environmental Network. Unless otherwise noted, all variables in the natural resource analysis use this database. Additional GSO datasets. Unless otherwise noted, all industrial output data is directly compiled from GSO data. GSO also contributed critical datasets to complement other sources. Reliability of statistical data. The research teams have carefully crosschecked data for quality. Some variables, often in relation to more remote provinces or districts, remain unreliable. While absolute quantitative values for speci�c locations may be questionable, the datasets provide solid information regarding overall trends in time and in geographical distribution. national level, this has occasionally required association of uplands with high poverty manual transcription of district information. incidence. At the local level, a complicating factor in rural – Trends in forest resources are a combined surveys was the dual administrative levels of outcome of how local governments, for- commune and village. Statistical data is avail- est enterprises, migrants and households, able only at the commune level, while poverty especially ethnic minority groups, inter- issues and environment issues tend to differ act locally. among villages within a commune. We used – Legal framework improvements, espe- the village level when defining samples. cially in land administration, are one of the leverage points that have potential to improve sustainability of natural resource 4.1. NATURAL RESOURCE USE use and reduce poverty, but the poorest AND POVERTY communities are largely by-passed by these improvements. – Three questions are equally relevant to Key �ndings identify win-win poverty and environment – Issues of sustainable natural resource use are policy options: where and how are the poor linked to poverty in the uplands, not in the disproportionately affected by resource lowlands. decrease or degradation? where and how are – There is a mechanical correlation be- they at risk of natural disasters? and where tween poverty, fragile soils and location or and how are is their lack of access to the loss of forest resources that is largely due to local resource base an issue? PEN STUDY IN THE LOWER MEKONG SUB-REGION 125 POVERTY AND ENVIRONMENT IN VIETNAM 4.1.1. Geographical Dimensions Correlation of the spatial distribution of poverty is visible at all available levels of analysis: region, The distribution of poverty in the country, when province, district, and commune (Maps 4.1). defined as the percentage of population living Historical processes, through which the admin- below the poverty line, is strikingly similar to a istrative units in Vietnam have been delineated, map of elevations (Figure 4 in chapter 1). Poverty have generally followed an upland/lowland limit. reduction in Vietnam has been very rapid over Recent changes in administrative units are often the last decade. Poverty incidence had already reinforcing this trend. Thai Nguyen Province in declined to 36.5 percent in 1998–99 and further the Cau River basin in northeastern Vietnam, declined to 19 percent in 2003–04. According for example, was divided into two provinces: an to the latest poverty update of the World Bank, upland province, Bac Kan, and a midland pro- poverty incidence in Vietnam in 2004 is only a vince, Thai Nguyen. The same is often observed third of the poverty incidence recorded in 1993. when larger districts are divided into two. Variations in poverty incidence remain across The large majority of the poor in Vietnam provinces, ranging from 0 percent poverty in Ho live below 500 meters of elevation (Figure 4, chap- Chi Minh City to a maximum of 74 percent in ter 1). This is an effect of the extreme imbalance Lai Chau Province, the northwest province that of population densities between the lowlands and borders Laos. Provincial average per capita expen- uplands (Map 2, introduction). Almost half of diture also varies substantially. Vietnam’s population was concentrated in 1999 MAPS 4.1 Poverty Incidence at Various Levels Left: overall poverty, regional level (2004); center: overall poverty, province level (2004); right: rural poverty, district level (1999). 126 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM in the 24 small, flat lowland provinces around TABLE 4.1 Poverty and Environment Indicators in Hanoi and HCM City, constituting only 16 per- Lowland, Upland, and Other Districts cent of the total land area of Vietnam, with a 28 percent average poverty incidence. Correlation Hilly & of poverty incidence with elevation is however not District category 1/ Lowland Mixed Mountainous visible when looking at mean elevations within administrative divisions. Elevation alone is not Number of districts 282 195 134 Mean elevation (m) 45 441 925 signi�cantly correlated to poverty at district level Number of poor 12.4 10.5 5.0 (IFPRI & IDS 2003) or province level. There are (M people, 1999) lower elevation areas or more fertile areas within % poor 28 43 62 Rural density (inhab/km2) 418 185 45 most upland or hilly provinces and districts. High- % ethnic minority people 4 15 61 elevation plateau regions in the Northeast and in % agricultural land 65 24 12 the Central Highlands regions also have a mix of Paddy per rural capita (m2) 958 471 549 high and medium overall lower poverty incidence. % of area with slope < 8° 98 59 32 % of area with slope > 30° 0 3 8 Association of high poverty incidence with Forest land area (M ha) 1.6 4.4 5.6 elevation implies mechanical linkages between % forest land in land area 22 38 37 poverty and forest resources and between poverty and fragile lands. Correlation does not imply that 1/ CEMMA de�nes six categories of districts. In addition to lowland and causal relationships between these factors are at mountainous districts, other categories include partly hilly districts (80 dis- tricts), hilly districts (73), hilly and mountainous districts (39), and partly play. There is nevertheless a striking association mountainous (3). between poverty and these indicators. At the provincial level, provinces around Hanoi and HCM City have about 60 percent of their land 47 percent of Vietnam’s 10.5 million ethnic cultivated. Only about 6 percent of land is forest minority people.1 land in these two regions, compared to the national The severity of poverty (or poverty depth) average of 34 percent. Hilly and mountainous is substantial in districts with highest elevation provinces have 90 percent of Vietnam’s forest (IFPRI & IDS 2003). Poverty rates have decreased land. A comparison of district-level poverty and signi�cantly between 1999 and 2004, but more environment indicators between categories of slowly than elsewhere. The six provinces with districts de�ning of�cially lowland and upland more than 50 percent of the poor population in districts con�rms this contrast (Table 4.1 and 2004 are the three provinces of the Northwest Map 4.2). It also shows the importance of hilly region, the three mountainous border provinces and mixed districts alongside higher elevation in the Northeast region, and Bac Kan Province areas in poverty and in forest resources. in the Cau River basin. The three provinces with Many of Vietnam’s high elevation areas can 40 to 50 percent of the poor population in 2004 be described both as environmentally fragile areas are the Central Highland provinces other than and as poverty traps. Forest resources are spread Lam Dong (the more developed province recently across mountainous and hilly areas. The Cordillera, incorporated into the region). with its ridge forming most of the border between Communes in upland areas often have scat- Vietnam and Laos, is a rich area of biodiversity tered villages with contrasting issues poverty and (Map 3, introduction). Almost all protected natural resource use. The commune is the lowest areas are located in the high elevation regions. administrative level, but each commune is com- Mountainous districts have 55 percent of “com- prised of several villages. Remote villages in these mercial forests� and 51 percent “environmen- areas are typically home to more disadvantaged tal forests� (Box 4.4). They are also home to ethnic groups. Forest resources also tend to be con- PEN STUDY IN THE LOWER MEKONG SUB-REGION 127 POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.4 Forestry Indicators and Databases Forest cover rate is a widely used indicator in Vietnam, and one of the Millennium Development Goal indicators. The Government has a strategic goal to recover a 43 forest cover rate (Poverty task force 2003). There are three ways of de�ning forest cover rate. – Under the land administration in MONRE, the land census defines forest land and its sub-classifications. This classification is for zoning purposes only. – MARD regularly carries out forest inventories of forested areas. – FAO data provides actual forest cover from satellite images. Coupling the MoNRE and MARD databases could provide data on the percentage of land zoned as forest land that is actually under forest cover. In practice the datasets are dif�cult to reconcile. Trends indicated by the different data sources are equally dif�cult to reconcile. FAO indicates that deforestation in Vietnam may have been as high as 2.3 percent per year from 1980 to 1990, while forest cover increased marginally from the mid–1990s. The 1994–99 forest land data indicate that deforestation in Vietnam continued at about 1.3 to 1.4 percent per year. The datasets are best used to assess geographical patterns, not trends over time. BOX 4.4 (continued) Land Census De�nitions Special use forest Protection forest Production forest Plantation forest Plantation special use Plantation protection Plantation production Natural forest Natural special use Natural protection Natural production Equivalent forest management regime Plantation forest Environmental forests: Environmental forests: Commercial timber plantation Natural forest Protected areas Watershed protection Natural forest with on-going commercial exploitation centrated in some of the villages—often a legacy protected wetland area in Vietnam. Coastal and of the state forest-farm zoning system. Moun- marine degradation was assessed as a medium tainous provinces have more villages in each com- level poverty-environment priority. The linkage mune. The Northwest, Northeast, North-Central between �sheries and poverty, however, is not coast, and Central Highland provinces have 10 to very strong. A regression analysis of provincial 11 villages per commune, whereas other regions �shery employment in relation to poverty inci- have 4 to 7 villages per commune. This �gure dence carried out at the beginning of the PEN reaches 14 in Sonla and Tuyen Quang provinces. research has found that around 2.6 percent of the The number of villages has increased between non-poor and only 0.5 percent of the poor are 1995 and 2000 by 17 percent in the Northwest employed in this sector. Typhoons and floods region and by 41 percent in the Central High- are an exception, since they affect a number of lands, which indicates active in-migration. poor coastal provinces. The lowlands have their own environmental threats, but most natural resource management 4.1.2. Forest Resources and Poverty issues do not appear to impact the poor dispro- portionately. Aquaculture and fishing in the Deforestation mostly takes place in poorer pro- lowlands have developed at a rapid pace, and vinces (Table 4.2). This is partly a simple conse- increased protection of the marine resources quence of the geographical superposition of the has not followed. In 2003, there was only one forestry sector and forestry. 128 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM Several indicators point to an important con- TABLE 4.2 Disparities in Provinces with Decreasing trast in linkages between poverty and forestry Forest Land in 1994–99 resources between the northern regions and the central regions. During 1994–99, half of the Provinces with Provinces with provinces experienced a net increase in forest land Poverty incidence 0–30% poverty 50–80% poverty in provinces (1999) incidence incidence area, with a total increase of 1.14 million hectares, or 11 percent of the 1994 forest land area. The Number of provinces 11 14 largest losses of forest land were in regions with Loss of forest land (hectares 13 145 high poverty incidence. Loss of forest land in the per 1,000 people) central third of the country—starting from Quang Loss of forest land 4% 9% (% of total land area) Binh Province—continued throughout the 1990s Loss of forest land (% of 25% 20% and totaled about 1.25 million hectares from 1994 natural forest area) to 1999. Poverty rates in all these provinces except Gross forestry output 15 106 (m3 per 1,000 people) one are well above the Vietnam average. Con- versely, poverty incidence has been associated with an increase in forest land area in the northern part of the country during the same period (Map 4.4). forest they manage under the protection of for- This fact points to an absence of proof that the est management boards are more prevalent in poor are one of the causes of deforestation or the northern regions, while those that have that they are poor because they are signi�cantly become fully commercial SFEs are more pres- affected by deforestation. It is however clear that ent in the central regions. This trend is contin- in the poor provinces, the poor do not bene�t uing in subsequent changes in forest land between enough from forestry resources, whether they are 2000 and 2003. logged or protected, to raise their incomes above While less than 20 percent of forest land is the poverty line. allocated to enterprises in most districts in the Forest types also differ between the northern North, this proportion is higher than 40 percent and central regions, largely in association to the in around 106 districts, most of them located structure of the State forestry sector. “Environ- in the Central regions (Map 4.2). More than mental forests� are largely dominant in northern 40 percent of total land area is allocated to upland areas, while natural production forest— forestry enterprises in 21 districts in these regions. an indicator approaching land zoned for com- The land census records allocation of forest mercial forestry—is nearly absent in the North land to various categories of stakeholders. In but remains important in the central regions 2000, 3.5 million ha of forest land, or 31 percent (Maps 4.2). There is concentration of natural of the total land zoned for forestry, were allo- production forest in the central third of the cated to enterprises, just twice the percentage of country. In 2000, 682 SFEs were managing forest land allocated to households. By 2003, 5.5 million ha, or 17 percent of the total land the proportion of forest land allocated to house- area in the whole country, and undertook a mix holds had increased to 22 percent, while forest of watershed management and commercial log- land allocated to enterprises was unchanged. ging activities (MoNRE 2000). Starting in the Allocation of forest land to enterprises has early 1990s, the SFEs that were in charge of decreased in 20 provinces, but has continued to management, exploitation, processing, and dis- increase in 18 provinces. tribution of the country’s forest resources have Finally, dominant types of ethnic minority undertaken a reform process. Enterprises dis- groups differ in the northern and in the central continuing commercial operations to put the regions. CEMMA de�nes 54 minority national- PEN STUDY IN THE LOWER MEKONG SUB-REGION 129 POVERTY AND ENVIRONMENT IN VIETNAM MAPS 4.2 Regional Contrast in the Forestry Sector Left: change in forest land area (1995–2000). Center: Percent of protection and special use forest in forest land (2000). Right: Percent of natural production forest in forest land (2000). ities. There is no widely accepted grouping of the 4.1.3. Soil Resources and Poverty 54 groups into broader categories. We make a tentative grouping can based on ethno-linguistic In almost all upland areas, there is a trend among groups (Khong Dien 2002). If a “small southern subsistence farming systems toward intensi�ca- groups� category is de�ned by including 24 Mon- tion with the opening of new paddy �elds and an Khmer and Malayo-Polynesian groups, a “main increase in multiple cropping in the paddy �elds. northern groups� category (including the four From 2000 to 2003, paddy area in Vietnam main Thai family groups and the Muong) and a decreased overall by close to 246,000 hectares, or “small northern groups� category (including small 6 percent of the total paddy land area, but it Tibeto-burman group), a picture with largely increased in 21 of the country’s 62 provinces, all disjointed regions of residence appears (Map 4.3). located in the upland northern and central upland There is an apparent paradox between the fact that regions. All provinces in the Northwest and Cen- the small southern groups, which have tradition- tral Highland regions increased their paddy area. ally been described as having more sustainable The increase reached 28 percent in Lai Chau systems of rotational agriculture, live in defor- Province, which has the highest rate of rotational estation areas, while the northern groups live in agriculture in the country. Of the 36 provinces reforestation areas. that had an increase in the cropping index in 130 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM MAPS 4.3 Forest Land Allocation and Ethnic Minorities Left: forest land allocated to enterprises (percentage of total land area, to update to percentage of forest land) (2000). Center: forest land allocated to households. Right: districts with high share of “main Northern� and “small Southern� ethnic groups. paddy �elds, 16 are in the three upland regions— of 3 million farm close to 550,000 ha of rubber the Northwest, Northeast, and Central Highlands. and coffee, an 80 percent increase from 1995. The This indicates that household land use systems impact of subsistence farming on poverty and in the North and in several Central provinces are environment in the region is limited compared mostly on a virtuous path of intensi�cation; that to this large-scale development of perennial crops. is, away from rotational agriculture toward more Subsistence farming is however changing fast. sustainable systems centered on paddy �elds. How The increase of paddy area in Kontum Province much the poorest communities and households reached 37 percent between 2000 and 2003. benefit from these improvements is however Linkages between degradation of fragile soils limited by the fact that poorer communities and from farming activities and poverty have been households generally have fewer or even no discussed over many years in Northwest Vietnam, paddy �elds. the region with highest area of sloped farmland. Migration makes linkages between agriculture They remain dif�cult to quantify. All provinces indicators and poverty and environment much in the three upland regions, as well as in the less visible. In the Central Highlands, the region North-central coast region, have fewer paddy with highest in-migration, the rural population �elds than sloped farmland (Map 4.9c). This PEN STUDY IN THE LOWER MEKONG SUB-REGION 131 POVERTY AND ENVIRONMENT IN VIETNAM proportion reaches �ve in Lai Chau and eight in floods. There were 23 coastal districts with poverty Sonla. Flood events are geographically associated incidence of 50 percent in 1999. with cultivation of sloped land. The Black River basin was the main area in Vietnam where three or more flood events were recorded in 2002 and 4.1.4. Land Administration, Natural 2003. The Black River has been at the center of Resource Use and Poverty forest rehabilitation policies in the 1990s. It is The geographical linkage between land allocation also a region where maize is being developed on to poor households and other stakeholders indi- a broad scale for the feed industry. The three cates how important land administration is in Northwest provinces alone, with only 3 percent understanding the poverty and environment nexus of rural population in Vietnam, grow 15 percent in Vietnam. The national legal framework for land of the national area of maize. The area of maize administration has been modernized extensively in these three regions grew again by 22 percent in the past decade. A land law was created in 1987 between 2000 and 2003. While commercial and was partly or fully updated three times. The monoculture of maize on steep slopes has been latest version of the land law—implemented in documented to generate erosion, there is also early November 2004—is a key version, since �eld evidence that maize introduced in rotational the following one planned for 2008 would be a agriculture beside upland rice can improve over- long-term land code. all sustainability of the rotation. The 2004 land law, although primarily Whether related to human activity or not, the designed to resolve urgent land administration natural disasters in Vietnam are often located in issues in urban and lowland rural areas, addresses, areas with a high proportion of poor. In addition two broad fields of special relevance in poor to floods in the Black River basin and other basins upland areas, land titling and land use planning in the North, the poor in the Central coastal (Table 4.3). National stakeholders expect the law region are impacted by typhoons and subsequent to be an instrument to speed up the allocation of TABLE 4.3 Transition Stages in Land Administration in Vietnam 1. Early transition 2. Mid transition 3. Late transition 1. Land tenure Allocation of land use 1.2. Issuance of Land 1.3. Fully transferable rights Tenure Certi�cates land titles Forestry land: not Agricultural land: just Urban land: new Law completed completed Forestry land: starting 2. Land use planning 2.1. Decreasing weight 2.2. Shift from control 2.3.1. Local plans man- of plan targets, of productive use aged by community increasing mapping categories to control authorities to regu- and participation of use transfer late land market categories Upland areas: ongoing New Law: controls Urban areas: starting transfers out of paddy 2.3.2. “Unused land� and environmental category cancelled forest Not started yet In italics: stages for different areas and types of land at the start of the 2004 land law. Source: summary of stakeholder interviews 132 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.5 A Land Administration Survey in Six Upland Districts Are land administration instruments an appropriate and viable answer to natural resource use linkages to poverty and environment? How does conversion of land uses, especially paddy �elds and forests, affect poor households versus non-poor households and what is its impact on the environment? Can the environmental and social impacts of the new law be monitored, and how? A survey carried out in three upland provinces where implementation of the new law started earlier—Tuyen Quang in the Northeast region, Nghe An in the North-central coast, and Binh Dinh in the South-central coast—was undertaken to answer these questions. One poor commune was surveyed in each of six contrasted districts in terms of land use trends and structure of the forestry sector. Implementation of the land law was reviewed at district and commune level through statis- tics and local interviews. A 174-household survey allowed a poor-non poor comparative analysis using poverty scores based on household assets. Aerial photographs allowed the research team to compare actual land use with reported data. A survey fully or partly based in the lowlands, the main area targeted under the new law, was initially considered but rejected in favor of comparative analysis of various upland areas for two reasons. First, global research indicates that land use transfers from agriculture to urban or indus- trial uses have no net impact on poverty, since farmers move from agricultural to manufacturing jobs. Second, ongoing studies are focusing on lowland issues with poverty and environment implications: land consolidation, and loss of paddy areas to aquaculture. The full report of the national research team is available on a separate CD Rom (TECOS 2005). formal titles on forest land, the land category where level, but did not match commune-level �gures. this process is lagging behind. The new law also Two out of the six communes could not report introduces for the �rst time a controlled transfer numbers of households with paddy land allocated, of paddy areas and of “environmental� forest land while the others simply reported 100 percent, to other uses through land use planning. a figure that is not consistent with the survey Limited reliability of quantitative data on land sample. None of the local of�cials could recall the use and land administration is an important fac- 2000 situation for land use rights. Capacity of tor to take into account in the analysis of land local staff was assessed on a scale of 1 to 5 depend- administration reform. Discrepancies between the ing on their education level and other indicators, land census and actual land use are well-known with 1 indicating highest capacity. Capacity scores in Vietnam. They reach a dramatic level in the were 2.7 and 3 in less-poor communes and vil- six communes surveyed in percentage terms for lages respectively, and 3.3 and 4 in poorer com- paddy �elds, and in absolute terms for forest land munes and villages. (Table 4.4). Interestingly, there is more forest MoNRE is facing difficulties in receiving observed in the photographs than forest reported information from the provinces. Land titling data by local of�cials. could only be assembled at province level during Limitations in data availability also reflect lim- itations in access to information on land admin- istration and handling capacity among local cadres TABLE 4.4 Actual Land Use/Statistical Land Use in marginal areas. Each commune has at least one staff member dedicated to land administration, Average 6 communes Minimum Maximum but these staff members had not received infor- mation about the new land law. Land census sta- Total land area +0.4% −0.4% +1.2% tistics are compiled at the district level, but were Forest land area +38% −0.4% +104% not available for the commune level. Data on Paddy land area +140% +23% +296% forest land allocation was available at the village PEN STUDY IN THE LOWER MEKONG SUB-REGION 133 POVERTY AND ENVIRONMENT IN VIETNAM the PEN research, not district level, and only 21 forest land, a small proportion being transferred out of 61 provinces provided full information. into agricultural land. Unused land decreased by The information gap is highest for forest land titles 60 percent in one of the northern districts between issued to enterprises. Even data on land titling 1995 and 2000. The decrease is slowing down, progress for paddy fields are inconsistent. The but was still 10 percent in 2000–2003 in one datasets remain useful to monitor time and of the central districts. The forest land domain, regional trends. Paddy land titling is well advanced already covering at least 40 percent in the com- in the Northeast and Northwest regions and munes, and up to 90 percent in one north-central lagging behind in all Central Highland provinces. commune, continues to increase. Forest land titling is less advanced for households Forest rezoning creates contrasting local situ- than titling for enterprises. Twenty-eight provinces ations. In the northern province, although no do not report any household Land Tenure Cer- rezoning took place in the two communes studied, ti�cates (LTCs) on forest land, whereas only four some other communes have had large land trans- provinces do not report any enterprise LTC on fers from special use forest to protection forest, forest land. bringing income opportunities from forest pro- Conversion from paddy to other uses, the tection contracts but raising questions of bio- �rst control point for land use transfers in the diversity protection. Other communes have had new land law, does occur in upland areas, but so a substantial decrease in production forest and far linkages to poverty are indirect only. Paddy an increase in protection forest, reducing oppor- land area decreased by 300–400 ha in of each of tunities for sloped land development. In Nghe the three surveyed provinces between 2000 and An, land rezoning as production forest is substan- 2003, but in only two of the six surveyed districts tial in some communes and negligible in others. with higher population density. Paddy conver- In Binh Dinh, forest land is being rezoned into sion takes place in less-poor communes. We also agricultural land in some communes in the Kinh collected oral testimony regarding the value of district, while in the ethnic minority district com- paddy �elds to both the poor and less-poor upland munes rezone land into production and protec- farmers. Paddy area per capita, from 240 m2 in one tion forest. commune in Nghe An to 750 m2 in one commune Actual use of sloped land zoned as production in Tuyen Quang, is not smaller than the provin- forest appears to be unrelated to land allocation cial average, but paddy remains an important food or land titling (Table 4.5). Land allocation may security element. Paddy conversion—totaling have contributed to more decision making in only only 2 ha—was reported in two communes and one surveyed village in Tuyen Quang. In this mostly relates to �sh pond development by a few village, 58 percent of households have forest land better-off households in one village. Things might allocated, and they are developing orange pro- change, however. Of the households surveyed, duction, unlike the other village with a state forest 3 percent indicated their interest in turning paddy enterprise and only 4 percent of households with �elds into perennial crop plantations. forest land allocated. It is local rapport—among Changes in forest land zoning are actively stakeholders, state forest enterprises, national park taking place, albeit at a lower pace than in the staff, and community/households—and the pres- 1995–2000 period. Transfers from environmen- ence or absence of market opportunities that tal forests to other uses, which are due to be con- appears to explain why households use or do not trolled through land use plans, are limited. Most use the forest land they have been allocated. The change relates to gradual rezoning of land zoned ethnic dimension is a visible factor in this stake- as unused, generally rotational land and related holder rapport. Provincial and district policies fallow. This land is mostly rezoned production also play an important role, for example, with 134 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM TABLE 4.5 Speci�c Uses of Sloped Land in the Six Communes Surveyed Commune Uses of sloped land Important stakeholders % Poor Ethnic groups TQ 1 Orange in village 1 SFE in village 2, 66% 84% MNG with conflict TQ 2 RA forbidden by local Protected area 87% 53% MNG government but practiced NA 1 Tourism: 1 village resettled Protected area 69% 94% MNG RA restricted by local government NA 2 RA organized by district District 67% 95% MNG & SSG with land allocation BD 1 Cassava, cashew SFE, without conflict 53% No BD 2 RA, cashew SFE, with conflict 80% 97% SSG RA = rotational agriculture; MNG = “Main Northern ethnic groups�; SSG = “Small Southern ethnic groups� (see definition in national analysis). regard to rotational agriculture. One district in land allocated between 4 and 60 percent in the Nghe An promotes regulated rotational agricul- villages studied. In Binh Dinh, no forest land ture on allocated forestland, while the neighbor- allocation is reported in the villages surveyed. ing one restricts it. Economic organizations, mostly state forest Among the three land instruments analyzed— enterprises, are important stakeholders in the land use planning, land allocation, and land tenure three provinces, albeit at a variable degree. The certi�cates—new procedures for land use plans area of forest LTCs issued to economic organi- were found to be largely not taken into account zations is three times that issued to households yet. Land use plans have largely remained eco- and six times in Binh Dinh province. nomic planning instruments used by technical bureaus at the provincial level with limited par- ticipation from more local levels. Land use plans FIGURE 4.1 Allocation of Forest Land in the Six Districts Surveyed are in operation in only four of the six districts surveyed. Only Nghe An, an upland province that is generally fairly advanced in land admin- Other forest land 160 000 istration, had land use plans in both districts sur- veyed. None of the six communes had a land use 120 000 Forest land plan in place. allocated to HH Progress in forest land allocation is highly 80 000 dependent on the different provincial policies. Other forest Allocation of land titles on forest land has started 40 000 allocated to EO in the three provinces, although not in the com- 0 munes surveyed. All households in the Nghe An TQ TQ NA NA BD BD Natural production Province communes have land use rights on 1 2 1 2 1 2 forest allocated to forest land or are about to receive land titles, EO though this land only accounts for 11 to 19 per- cent of total forest land (Figure 4.1). In Tuyen Source: PEN Case Study Database. Quang, proportions of households with forest TQ = Tuyen Quang, NA = Nghe An, BD = Binh Dinh PEN STUDY IN THE LOWER MEKONG SUB-REGION 135 POVERTY AND ENVIRONMENT IN VIETNAM No correlation between more secure land can be spontaneously developed with environ- tenure and poverty is visible in the survey, either mental impacts. at the household or village level.2 Poverty scores of households with and without forest land allo- Policy implications cated are equal. While correlation of large areas In the absence of innovative approaches to land of forest land to wealth status is a frequent occur- administration in marginal rural areas, a double- rence in Vietnam (see Cau River section), land tracked system in the land administration legal allocation itself is not associated with less poverty. framework is being created. Urban areas are For the most part, the households surveyed are accessing a more comprehensive legal framework aware of the possibility of applying for LTCs in with the new land law, while poorer, marginal the near future. Only 4 percent mentioned areas may continue to operate largely outside the administrative dif�culties among the reasons land administration framework and to lag behind why they did not use their sloped land. Land in land administration improvements. allocation even with formal titles is not perceived Capacity-building among local cadres is crit- as secure land tenure. Only forty-�ve percent of ical to avoiding this risk of a modernized land households indicated they were planning some administration bypassing marginal rural areas, change of land use on their forest land. Only and the new law offers an opportunity for a sub- 19 percent stated they were planning to develop stantial information and training program. This perennial crops. 55 percent said that the issue capacity issue is compounded by the fact that vil- was insuf�cient knowledge to do so, and 82 per- lage leaders are legally in charge of handling land cent that they were not planning to use their use matters. Community participation is recog- land titles to access credit. nized in Vietnam as a means to overcome the The prevalence of large land managers on absence of reliable statistics and to improve local sloped land, and of agriculture and forestry decision making, but local capacity to use these administration in decision making, suggests that approaches needs to be enhanced as part of an forest land allocation has only created some information and training program. tenure security when provincial and district gov- A transition from top-down land use plans to ernments encourage decision making by the participatory, problem-solving oriented land use households. When this support is absent, defor- plans has the potential to enhance local decision ested sloped land remains largely under-used, or making in poor communities vis-a-vis other stake- holders. Land use plans are also needed to avoid environmental impacts from sloped land develop- ment. A balance needs to be found between FIGURE 4.2 Households with and without Forest Land Allocated by Poverty Level identifying critical land-use control points, pos- sibly other than those in the land law, and avoid- ing the detailed directions of land use that still 31-35 prevail and are in themselves a powerful deter- rent to local initiatives. Improved land use Poverty score 29-30 With (82 HH) planning will require disentangling the dis- 28 Without (92 HH) crepancy issue between statistics and actual 26-27 land use, and modern tools such as aerial pho- tographs and GPS may be an opportunity to 21-25 do so. 0 20 40 60 80 100 % Formal land use titles are a necessary condi- tion for land tenure security, but not a suf�cient 136 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM one. The perception of security is as important as Rural and peri-urban craft industries legal land titles in ensuring that title holders can have rapidly raised household income in use their sloped land to secure or improve their craft villages, but with serious pollution livelihoods. Land use plans can support this and health effects. Non-household labor process, and local access of the poor to credit, agri- hired in craft industries is disproportion- cultural inputs, and markets are a pre-condition— ately from low-income households, and still often missing in marginal rural areas. are exposed to serious air pollution and Monitoring poverty and environment impacts toxic substances. Dug wells, mostly used of the Land law can be done through better use of by low-income households, are at risk of existing instruments and local experience rather contamination from untreated wastewater than by creating a new monitoring system. The generated by craft industries. land census, with its detailed information requests on land use types and stakeholders receiving land Drinking water and sanitation use rights, is a perfect tool. What is needed is a bet- There is a signi�cant level of inequity in ter dialogue between central level and provinces to access to safe water and sanitation through- ensure that reliable data are assembled and used. out Vietnam. The wealth of experience being assembled by Signi�cant numbers of lowland poor provinces and districts through pilot land using dug wells or surface water are affected administration schemes deserves an evaluation by lack of access or pollution of drinking process and the exchange and dissemination water. of lessons learned. Monitoring and evaluation The role of ethnicity in higher incidence of the land law can only be successful if it of diseases related to water and sanitation is incorporates the impact of forestry policy and visible but not fully explained. reform through improved linkages between the two agencies. Pesticide use Pesticide use has become prevalent 4.2. ENVIRONMENTAL HEALTH throughout Vietnam including in poor AND POVERTY provinces. The poor are more exposed to toxic Key �ndings pesticides than the non-poor. Industrial pollution The poor as aware as the non-poor of There has been a trade-off between poverty the risk related to pesticide use but have reduction and environmental quality in the less access to information that could help last decade in particular from the non-state them avoid these risks. sector The urban poor are disproportionately employed in industry with high occupa- 4.2.1. Lack of Safe Water and tional health risk and exposure to pollution, Sanitation and Poverty particularly in small-scale industry. Young children and elderly—the age The Vietnam National Health Survey groups most susceptible to health effects How do the poor differ from the non-poor from pollution—are disproportionately in access safe water and sanitation? What concentrated in poor households in the is the level of health risk, particularly for urban industrialized areas. diarrheal illness, of various forms of drink- PEN STUDY IN THE LOWER MEKONG SUB-REGION 137 POVERTY AND ENVIRONMENT IN VIETNAM poorest segments of the population, which have ing water supply and sanitation in Viet- much lower coverage rates than the rest of the nam? How many people are at risk of ill- population. More than 17 percent of the poorest ness, especially in relation to the poorer 20 percent of the population use surface water as segments of the population? What are the drinking water, while only 3 percent of the richest resulting priorities in water and sanitation 20 percent use surface water. improvements, and should they have a Beyond this direct link between poverty and geographic or socioeconomic focus? These questions can be answered by the National access to clean water and sanitation, no obvious Health Survey database. All data, unless geographical linkage between poverty and indi- otherwise mentioned, were obtained from cators of environmental health in relation to water a multivariable logistic regression analysis supply. The overall difference between rural and using detailed information for the more urban areas in coverage of clean water supply is than 150,000 individuals in the survey. limited. The disparity is within rural and urban areas. About 8 percent of people in urban areas Vietnam has achieved impressive gains in water and 22 percent in rural areas are without safe supply and basic sanitation coverage rates. About drinking water; that is, they use surface water or 80 percent of the population has access to clean dug wells with a nearby pollution source. drinking water. The two main unsafe water The disparity between rural and urban areas sources—surface water and wells, both of which in access of the poor and the non-poor to toilet have high risk of pollution—declined respectively facilities is substantial. In urban areas, almost all from 20 percent to 10 percent and from 50 per- of the richest households have toilet facilities, while cent to less than 35 percent between 1992–93 this proportion is 3 percent in the rural areas. In and 2001–02. At the same time, the share of the the urban areas, 24 percent of the poorest do not population with drilled-well water increased from have toilet facilities, while in rural areas this pro- 5 percent to over 20 percent. While access to water portion raises to 38 percent. supply and sanitation facilities has improved sig- Unsafe water supply combined with low ni�cantly, it still remains a major issue for the sanitation and poor hygiene behavior is highly FIGURE 4.3 Drinking Water Supply in Vietnam % 60 50 40 1992 - 93 30 1997 - 98 20 2001 - 02 10 0 Tap water Drilled well Rainwater Dug well Surface Other water source Drinking water source Ministry of Health de�nes clean drinking water as piped city water, drilled well water, rain water, protected moun- tain spring water, and dug well water without a pollution source within a 7-meter radius. A pollution source is de�ned as, for instance, a toilet, sewage, or a cattle shelter. 138 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM SP10 correlated to diarrheal illness in many countries FIGURE 4.7 Population without Safe Drinking Water (left) (Curtis and Cairncross 2003; Fewtrell and Colford or Toilets (right) 2004; Esrey and others 1991). Hand washing alone is often found to reduce diarrheal illness by 35% as much as 45 percent (Curtis and Cairncross 31% 30% 2003). The poor are generally disproportionately 25% 25% 20% 24% 20% affected by disease, because they tend to have less 15% 17% 17% access to medical services (Figure 4.4), less �nan- 10% 11% 9% cial resources to pay for quality services, more 6% 5% 3% income losses from illness, and are more often 0% Poor Near Poor Average Better-off Rich self-employed or work in the informal sector. Rural Population Urban Population They often have fewer resources to undertake averting behavior (actions taken by individuals and households to reduce the risk of exposure 40% 38% to health risks). Averting behavior observed 35% in Vietnam includes purchase of bottled drink- 30% 25% 24% ing water or point-of-use drinking water treat- 20% 17% ment; use of pesticide protection equipment; 15% 10% 13% 12% residence in less polluted neighborhoods; and 5% 7% 7% 3% hygienic behavior. 2% 0.3% 0% Poor Near Poor Average Better-off Rich Unsafe drinking water sources and lack of toilet facilities are causing substantially higher Rural Population Urban Population rates of diarrheal illness. While this �nding applies to many countries, a speci�c feature in Vietnam is that two very different types of unsafe drinking water, surface water and polluted dug wells, con- tinue to be an important source of drinking water (Figure 4.5). FIGURE 4.4 Percent of Households Living in Commune with Doctor 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Poor Near Poor Average Better-off Rich PEN STUDY IN THE LOWER MEKONG SUB-REGION 139 POVERTY AND ENVIRONMENT IN VIETNAM wells are the main water source in various low- FIGURE 4.5 Population Drinking Water Supply 2001/02 land areas, except in the Mekong River Delta. The use of dug wells exceeds 75 percent in some Other water Bought water provinces, but is below 5 percent in the Mekong Piped mountain spring water River Delta. There are six provinces with more Dug well (no pollution source) Dug well (with pollution source) than 25 percent of the population with a pollution Surface water source near their dug well. This �gure reaches more than 40 percent in some of the provinces Rain water with the highest reliance on dug wells. The prevalence rate of diarrheal illness varies considerably across provinces. An index of the Drilled well water Tap water burden of diarrheal illness, created from the sur- vey data and reflecting both prevalence rate and duration of illness, shows that disease burden is ADI = acute diarrheal illness highest in the Northwest and Central Vietnam, as well as in a few of the provinces in the Northeast Use of surface water is common, not only in and the Mekong River Delta. The correlation marginal areas but also in more developed areas. between the disease burden and poverty incidence Surface water is the main water source for more is pronounced (Figure 4.6). than 50 percent of the pollution in six Northwest While pollution of water sources may arise provinces and in the Mekong River Delta. Polluted from human or other sources, such as sewage and MAP 4.4 Population Using Unsafe Water Sources Percent of population using Percent of Population Percent of dug wells with surface using dug wells nearby pollution source water 140 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM FIGURE 4.6 Prevalence of Diarrheal Illness by Age and Living Standard ADI prevalence by Age Group and Quintile 10% 8.2% 8% 7.1% 6.6% 6% 5.4% 4.2% 4% 2.2% 2.3% 2.1% 2.2% 1.9% 2% 0% Poor Near poor Average Better-off Rich Under 5 5 and older ADI = acute diarrheal illness animal waste, the lack of toilet facilities is asso- five years than in older children in the poorest ciated with both low living standards and higher households, but “only� twice as high in the richest incidence and longer duration of diarrheal illness. households. For children under �ve, the average In six provinces in the northern and central parts duration is 3.3 days in the poorest households of the country, more than 50 percent of the pop- and 2.8 days in the richest households. ulation lack toilet facilities. Polluted dug wells are a critical element of the Poverty is associated with signi�cantly higher poverty and environmental health linkage. Nearly diarrheal illness in children, but not in adults. 35 percent of the population relies on dug wells. Young children in poor households are at higher Around 20 percent of all dug wells, which corre- risk not only of illness, but also of diarrheal mor- spond to nearly 8 percent of the total population, tality. In the overall population, the use of open, have a pollution source within a 7-meter radius. unprotected water sources for drinking water is Of the population that use dug wells, 27 percent found to increase the risk of diarrheal illness by of the poorest people have a pollution source near 51 percent in children under �ve years of age the well, while only 13 percent of the richest have and by 33 percent in the population over �ve a pollution source near the well. The risk of diar- (compared to the average for all other sources of rheal illness in all age groups is 15 percent higher drinking water). Children in households having for individuals using drinking water from dug no toilet facility have a 34 percent higher risk of wells with a nearby pollution source. The risk of illness than children in households with a toilet diarrheal illness is 36 percent higher for those facility. There is a 23 percent increase in risk for without a toilet facility compared to those with the population over �ve. Children under �ve are a toilet facility. substantially more affected by diarrheal illness. Once households have access to simple facil- At the national level, the diarrheal prevalence rate ities for drinking water and toilets, there is no in children under the age of �ve is nearly twice advantage—in terms of environmental health— as high in “poor� compared to “rich� households. of moving toward improved facilities. Risk of Comparing across age groups, the prevalence rate diarrheal illness between tap water, clean dug is almost four times higher in children under wells, drilled wells, rain water, or piped spring PEN STUDY IN THE LOWER MEKONG SUB-REGION 141 POVERTY AND ENVIRONMENT IN VIETNAM MAP 4.5 Index of Disease Burden and Pollution Sources Index of disease burden Population with Pollution Population without toilets Source near drinking water FIGURE 4.7 Population Using Dug Well with Nearby Pollution Source 30% 27.1% 26.1% 25% 23.2% 20% 20.3% 15% 13.1% 10% 5% 0% Poor Near poor Average Better-off Rich 142 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM water is similar. The same applies to types of the Mekong River Delta. 63 percent of the poor- toilet facilities, such as simple pit latrines, pour- est households always boil their drinking water. flush, and modern flush toilets except, again, This is lower compared to 87 percent of the when drinking water with a nearby pollution richest households but still a remarkably high source is used. This �nding is worth noting, since proportion. households have high demand to shift to more Rainwater is increasingly used as an alterna- modern toilets. The use of flush toilets increased tive to polluted water sources. More than 50 per- from less than 10 percent to over 20 percent, cent of households use rainwater in several of while the category “other toilet,� which includes the provinces in the Red River Delta, and in the toilets without safe drainage, declined to less than Mekong River Delta rainwater use ranges from 10 10 percent over the last 10 years. This demand to 30 percent. The poor are less likely to have this continues as the use of simple toilets remains option. Only 6 percent of the poorest households widespread (at over 30 percent of the popula- are using rainwater compared to 14 to 16 percent tion) and sharing a toilet is common (a almost of the “middle-income� groups. No elevated risk 15 percent). of illness was found in this study from rainwater Three options are available to households compared to other improved or protected sources with a polluted local water source to increase of drinking water. More than 20 percent of the security of their drinking water. These options, by households report that they apply chemical or decreasing frequency, are boiling drinking water, mechanical treatment to their drinking water. the use of rainwater, and chemical or mechanical Of the poorest households, 13 percent report they treatment of water. The poor are less likely to use such drinking water treatment. About 27 per- access any of these three options. An overwhelm- cent of the richest households apply chemical or ing majority of the population boils drinking mechanical treatment to their drinking water. water. Nationally, 78 percent of households report Behavior, and particularly behavior among eth- that they always boil their drinking water, ranging nic minority people, appears to be a critical factor from well over 90 percent of households in most explaining diarrheal illness. Two factors likely to of the provinces in the Red River Delta and some be associated with behavior are the mother’s of the Northeast provinces, to less than 50 percent education and ethnicity. Children whose mothers of the households in many of the provinces in have not completed primary education have a FIGURE 4.8 Boiling of Drinking Water in Relation to Poverty Population always boiling drinking water (quintile) 100% 87.3% 83.3% 75.1% 79.1% 80% 62.9% 60% 40% 20% 0% Poor Near poor Average Better-off Rich PEN STUDY IN THE LOWER MEKONG SUB-REGION 143 POVERTY AND ENVIRONMENT IN VIETNAM 27 percent higher risk of illness than children holds that take the risk of not boiling water, with mothers who have completed either a pri- although small, is clearly located in the provinces mary or a higher level of education. Similarly, in with a high share of ethnic minority population the population above �ve years, lack of primary and in the southernmost provinces of Vietnam. education is associated with a 24 percent higher The relative share of behavior and lack of risk of illness. “Poor� households are more likely access to improved water cannot be fully assessed. than “rich� households to rely on surface water This derives from the fact that, although such for drinking water, to have no toilet facility, and a high share of the population boils its drink- to have a lower level of education. No such influ- ing water, the use of open, unprotected water ence is found for the over-5 population. sources (such as surface water) is still associated While mothers in ethnic minority groups are with an elevated risk of diarrheal illness. Insuf- likely to have received less education (if any in some ficient water quantity for hygiene purposes, groups), ethnicity in itself is found to explain no such as hand washing and domestic cleanliness, less than 44 percent of higher risk of illness among is likely to result in increased risk of illness, the over-5 population. This �nding controls for even if drinking water quality after boiling is of all factors commonly associated with ethnicity, sufficient quality. including education. One possible explanation for the fact that poverty quintile, not ethnicity, Policy implications is revealed in the analysis for children under 5 is Safe water and sanitation programs face a chal- that poverty reflects the influence of ethnicity of lenging issue of targeting. Sound policies could parents, as ethnic minority people tend be poorer simply give priority to providing safe water than the majority of the population in Vietnam. supply and basic sanitation to all those without There is, of course, no simple linkage between these services. In reality, competition to access ethnicity and behavior. The proportion of house- improved services arises between poor and non- FIGURE 4.9 Relative Risk of Diarrheal Illness Children Under 5 Years Population Over 5 Years 60% 51% 50% 50% 40% 40% 44% 34% 30% 27% 30% 33% 20% 10% 20% 24% 23% 0% 10% -10% -14% -20% 0% Using No toilet Mother not Quintile Using No toilet Unfinised Ethnic surface finished surface water primary minority water primary 144 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM MAPS 4.6 Percent of Households Boiling Drinking Water Percent of Households Always Percent of Household Rarely or Never Boiling Drinking Water Boiling Drinking Water poor households or communities in a context 4.2.2 Industrial Pollution, of limited budgets. The 135 program allows Poverty and Health targeting at the commune level. Appropriate tar- Industrial pollution and poverty geting of poorer villages within these communes is critical. Industrial development in Vietnam has been a Pro-poor solutions to the dug-well pollution driving force in the impressive reduction in poverty issue will require affordable technical innovations. and the growth in living standards in the last Pollution from local domestic and human sources decade. The higher the industrial output per can be addressed through improved waste man- capita, the lower the poverty incidence and the agement. It is not so, however, for diffuse pollu- higher the per capita expenditure (Figure 4.10). tion of agricultural and, increasingly, industrial On the one hand, industry, and especially indus- origin. Making sure that the poor have access to trial growth, contributes to poverty reduction and affordable alternative solutions to dug wells will an increased standard of living. On the other hand, be critical. local communities, which may or may not directly Improving safe water and sanitation in areas bene�t from the industrial growth in their area, with substantially lower coverage requires close are affected by pollution. Industrial pollution coordination between the agencies in charge of presents a serious health threat to workers and setting up the infrastructure, health services, the the public, particularly for low-income groups, organizations in charge of information and edu- which can least afford ill-health. cation, and in some cases the organizations in In line with their high industrial concentra- charge of ethnic minority affairs. tion, the highest pollution loads are found in two PEN STUDY IN THE LOWER MEKONG SUB-REGION 145 FIGURE 4.10 Living Standards and Industrial Output by Province 80% 12000 70% Per Capita Expenditure 2002/03 10000 Poverty Incidence 2004 60% 50% 8000 40% 6000 30% 4000 20% 10% 2000 0% 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Industrial Output per Capita Industrial Output Per Capita 2004 Poverty incidence is from VLSS 2004, per capita expenditure from VNHS 2002/03. BOX 4.6 Methodology of the Industrial Pollution and Poverty Case Study The PEN study involved a “macro� and “micro� approach and incorporated existing national, provin- cial, district, commune, village and household data on industry, poverty, water supply, and health. At “macro� level, the Vietnam National Health Survey (VNHS 2002) was used to analyze health conditions, employment and demographic characteristics, and household water supply. The national research team collected district level data from of�ces in 14 provinces, assembled a national data set of seriously polluting establishments (SPEs) at the commune level, and categorized “craft� industry villages at the commune level in the Red River Delta from a study by MoST/INEST. At “micro� level, the national team, with initial support from the Korean Institute for Industrial Economics and Trade (KIET), conducted �eld observations and interviews in several cities and “craft� villages. Linkages taken into account reflect the complexity of determinants of impacts of poverty on the poor. The following relations between industry, pollution, human exposure, natural resource damage, and impacts on the poor were all taken into account in the analysis: Pollution Determinants: Industrial Industrial growth Structure Heavy industry Craft Industry Process technology Pollution abatement technology Technology Operation and maintenance Impact Determinants: Industrial location Population concentration Household microenvironments Occupation Water supply and agriculture Exposure Averting behavior Vulnerable groups Health status Poor Non-Poor Socio-economic status Households Households Medical care Income protection 146 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM regions, the Red River Delta and the Southeast generate both more rapid poverty reduction and region around Ho Chi Minh City. Inter-regional more environmental problems. Vietnam’s indus- differences in industrial output are not pronounced trial production has increased about 15 percent when presented as output per capita. The South- per year between 1995 and 2004 (Maps 4.7). east and Red River Delta have a slightly higher Regional industrial output growth during that per capita output. However, when looking at total period was 17 percent in the Red River Delta. industrial output, a more likely indicator of poten- Eight of the ten provinces with the fastest tial pollution and health issues, about 70 percent industrial growth are in the RRD, Northeast, of national industrial output is concentrated in and Northwest regions. The lowest growth was the Red River Delta and the Southeast. Nearly in the Central Highlands and the Mekong 50 percent of this output comes from the South- River Delta. east. Little of the industrial output is produced Linkages between poverty and geographical in the Northwest and the Central Highlands distribution of industrial development and pol- (Figure 4.11). lution strongly relate to the type of industrial The poor are more likely to be industrial work- sector—State versus non-State. Non-state indus- ers with higher occupational health risk. In the trial output is strongly associated with higher Red River Delta and the Southeast, 30 to 35 per- living standards and lower poverty. Provincial cent of urban low-income workers are employed poverty incidence is 0.2 percent lower for every in industry, compared to 15 to 25 percent from 1 percent higher industrial output per capita from high-income households. The difference is even the state sector, and 0.4 percent lower for every more pronounced in the six provinces with the 1 percent from non-state industry. Similarly, the highest industrial output in Vietnam (Figure 4.12). provincial per capita expenditure level is 0.1 per- In Hai Phong, 55 percent of low-income work- cent higher for every 1 percent higher industrial ers are employed in industry, compared to only output from the State sector, and 0.15 percent 13 percent of the high-income group. higher for every 1 percent from the non-State Within the two most industrialized regions of sector. Compared to the state sector, the larger Vietnam, the Red River Delta stands out for influence of the non-State industrial sector on having rapid industrial growth, which is likely to provincial living standards might be due to the FIGURE 4.11 Regional Share of National Industrial Output in 2004 30 25 20 15 10 5 0 Red River North East North West North South Central South East Mekong Delta Central Central Highlands River Delta Coast Coast Output per Capita Output per Urban Capita Figures are Million Dong (1994 constant prices). PEN STUDY IN THE LOWER MEKONG SUB-REGION 147 POVERTY AND ENVIRONMENT IN VIETNAM non-State sector’s generally smaller scale and FIGURE 4.12 Industrial Employment in Urban Vietnam higher labor intensity.3 70% The non-State sector is at the same time the 60% sector with the highest correlation to poverty red- 50% uction, and the sector in which more polluting— 40% or more pollution-intensive manufacturing 30% sectors—are mostly found.4 High-pollution- 20% intensive sectors represent almost 30 percent of 10% total manufacturing output. While non-State 0% manufacturing output was 27 percent of total RRD SE region HANOI HCMC Binh Ba Ria Hai Dong Nai region Duong Vung Phong manufacturing output in 2004, 37 percent of Tau output from high-pollution-intensive sectors came Low income High income from non-State manufacturing. State- and for- eign-invested manufacturing growth is relatively Source: Derived from VNHS 2002. “pollution neutral� since State manufacturing is particularly high in the chemical and basic metals sectors and foreign-invested manufacturing is highest in the chemical and tanning and leather MAPS 4.7 Provincial Industrial Output Industrial Output Industrial Output Provincial Output Growth Per Capita 2004 Per Urban Capita 2004 1995-2004 148 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM FIGURE 4.13 High-Pollution-Intensive Manufacturing Output, 2004 20000 16000 12000 8000 4000 0 Chemicals; Tanning; Fabricated Rubber; Plastic Basic metals Paper; Paper Coke; Refined Recycling Chemical Leather metal products products products petroleum products products products State Noe-Sate Foreign invested Billion Dong in 1994 constant prices product sectors. Non-State manufacturing dom- Among the non-state sector industries, the inates in the fabricated metal products and rubber craft industry has witnessed phenomenal growth, and plastic sectors (Figure 4.13). These high- especially in the Red River Delta region, where pollution-intensive non-State sectors grew in the 60 percent of the craft industry is located. Accord- range of 25 to 30 percent per year, nearly twice ing to MoST/INEST (2005), there are more than as fast as low-pollution-intensive non-state sectors 840 craft villages in the RRD region. The concen- (Figure 4.14). tration of the craft industry is particularly high FIGURE 4.14 Manufacturing Output Growth, 1995–2004 25% 20% 15% 10% 5% 0% Total State Non-state Foreign invested Low pollution intensive sectors High pollution intensive sectors PEN STUDY IN THE LOWER MEKONG SUB-REGION 149 POVERTY AND ENVIRONMENT IN VIETNAM in Ha Tay and Thai Binh provinces. A quarter environment in Vietnam. However, pollution of the rural communes in the RRD have at least from industrial zones is now increasingly impact- one craft village, with some communes having ing the nearby environment, including water more than �ve. resources, paddy �elds, and aquaculture farms. Rural industries were in the past limited to A major focus of Vietnam’s expanding indus- mines. Mines are located in a mix of poorer and trial development and its possible pollution impacts less-poor provinces. They are likely to draw poorer (also on different socio-economic groups) has workers, and their concentration creates high been the expansion of craft village (CV) industries, potential environmental impact on those living now tentatively accounting for almost 20 percent nearby. There are 16,000 hectares formally allo- of Vietnam’s Industrial Output Value (KEI 2005). cated to mining in Vietnam. The area is con- More than 55 percent of Vietnam’s CVs are centrated in 11 provinces. In these provinces, located in the Red River Delta in North Vietnam, 0.1 percent of total land area, often concentrated with a particular high concentration in Ha Tay in a few districts, is allocated to mining. Many of and other provinces like Thai Bihn and Nam these provinces are in the Cau River basin. Dihn (�gures 4.15a and 4.15b). Initial mapping Industrial pollution in Vietnam is shifting from shows that the same pattern appears for the most an urban-only issue to a mixed urban-rural issue. polluted CV industrial enterprises with highest The government has made signi�cant progress in concentrations in areas located North West and relocating industrial enterprises away from many South & South East of Hanoi (�gures 4.15b). A of the major city areas to industrial zones. This critical pattern is that CV industries are creating has alleviated some of the pollution in and near pollution patterns in rural communities that were residential areas. Some enterprises in the cities spared earlier from industrial pollution. have also installed pollution control technology. There was no difference in poverty incidence These improvements are not complete; numerous in rural communes with and without craft villages small- and medium-scale enterprises and remain- in the RRD in 1999. Average incidence was ing large enterprises continue to pollute the urban remarkably close to the national average. More FIGURE 4.15 Craft Villages in the Red River Delta 300 250 200 150 100 50 0 Ha Tay Thai Nam Hai Bac Hanoi Hung Vinh Ha Ninh Hai Binh Dinh Duong Ninh Yen Phuc Nam Binh Phong Source: From data provided by INEST, Hanoi Polytechnic University, Vietnam. 150 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM recently, poverty incidence may have declined linked to poverty. Air pollution causes cardio- faster in communes with craft villages, but no pulmonary mortality in adults, acute respiratory systematic data is available to con�rm this sup- mortality in children, and respiratory illness in position. The craft industry typically consists of children and adults, resulting in hospitalizations, household enterprises. On average, about 60 per- lost work days, and lost school attendance). It is cent of households in the craft villages are engaged the elderly and young children that are most at in the industry in the RRD, and in a quarter of risk of health effects from air pollution.5 In urban the villages more than 80 percent of the house- children under 5 years old, nearly 35 percent of all holds are engaged. Households engaged in the cases of acute respiratory illness (ARI) occur in the craft industry usually have a substantially higher 20 percent lowest income households. In contrast, income than the agricultural households in their only 12 percent of the cases are in the highest villages and in the RRD in general. As the house- income households (Figure 4.18). In the popu- hold enterprises grow, they hire workers who are lation over �ve years of age, the prevalence of ARI generally poor people. This is most pronounced is two times higher in low-income than in high- in Ha Tay and Thai Binh provinces, the provinces income households.6 People from low-income with the highest number of craft industries. Thirty urban households also die younger. The average percent of low-income workers are engaged in age at death was 60 years in the three lowest industry in these provinces, while only 13 to income groups, compared to 66 years in the two 17 percent in the high-income group are engaged highest income groups (Figure 4.18). Numbers in these activities (Figure 4.16). of children are disproportionately high in lower- Communes with lower poverty incidence in income households. About 30 percent of children 1999 tend to have craft industry with higher pol- in urban areas in Vietnam live in the 20 percent lution intensity (Figure 4.17). Craft villages engage lowest income households. in a variety of production activities that range from Visual observation of areas of residence of the non-polluting to highly polluting, some with long poor revealed the degree of higher exposure of and important cultural traditions and some with the poor to air pollution. The poor in urban areas a recent commercial and industrial nature. Activ- tend to live in cheaper but more polluted and ities include highly skilled embroidery, weaving, environmentally unsafe areas. The urban poor pottery work, and ceramics to food processing, also are disproportionately affected by occupa- construction materials, and waste recycling, such tional health risks. They are over-represented in as plastic and metals. While the causes of this industries with high occupational health risk and relationship are not clear, it could suggest a trade- exposure to pollution, particularly in small-scale off between income generation and environ- industries with no pollution control technology mental quality in relation to the development of and worker protection. In a lead recycling craft craft industry. village, for example, workers operate open fur- Analysis of the Vietnam Health Survey reveals naces to melt batteries. In a tanning enterprise, substantial inequality in health conditions in workers are in direct contact with wastewater. In urban areas. While it is not possible to establish addition to these pollution and health issues in a relationship between health effects and air pol- urban areas, evidence is mounting of serious health lution in urban areas from this survey, this �nding effects from craft industry pollution in rural and does suggest that controlling air pollution from peri-urban areas (Box 4.8). industry and other sources could provide substan- Craft industries generate both air and water tial health bene�ts for low-income households. pollution. Craft industry wastewater pollution The burden of ill health in urban areas, which constitutes a health risk to poor households in mostly affects children to the elderly, is clearly particular. Untreated wastewater from craft indus- PEN STUDY IN THE LOWER MEKONG SUB-REGION 151 152 MAPS 4.8 Craft Village Concentrations and Pollution Hotspots in the Red River Delta POVERTY AND ENVIRONMENT IN VIETNAM PEN STUDY IN THE LOWER MEKONG SUB-REGION Source: Maps generated by combining INEST CV database and MONRE’s classi�cation of polluted enterprises. POVERTY AND ENVIRONMENT IN VIETNAM tries increases the risk of contamination of dug FIGURE 4.16 Industrial Employment in Rural Vietnam wells. The poor are less likely to own craft enter- prises. They are often migrants including migrants 35% from minority ethnic groups. The owners are 30% often also exposed to pollution but not to direct 25% occupational health risks, and they have more 20% 15% access to alternative drinking water sources and 10% to health services. 5% 0% Ha Tay Thai Binh Nam Dinh Hai Duong Bac Ninh RRD region Policy implications Low income High income Environmental regulatory improvements can achieve a “win-win� poverty and environment Source: Derived from VNHS 2002. impact if efforts are directed toward reducing the impacts of industrial pollution on the health of low-income groups. Environmental health in Korea have faced in the development of SMEs relation to industrial pollution relates more to (Box 4.9). policies for low-income groups than to poverty Impact of craft industries in and around the reduction. There is a broad need for reform in Red River Delta on the health of the poor requires the regulatory framework and for parallel enforce- urgent attention. Uncontrolled development of ment of this framework in the workplace and in small and medium industrial enterprises would residential areas. result in a drastic increase in pollution. Preventing At present, the challenges that Vietnam is fac- this uncontrolled development will require a mix ing in the context of craft industry development of measures. First, a national regulatory frame- are similar to the challenges that China and work for craft industries needs to be in place and FIGURE 4.17 Craft Village Pollution Intensity and Poverty Incidence 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1 low poverty medium poverty high poverty incidence incidence incidence Source: Derived from data prepared by Tai, Thang et al (2006). Villages are classi�ed low=1, medium=2, and high=3 pollution intensity based on type of craft production. Industrial Pollution and Health PEN STUDY IN THE LOWER MEKONG SUB-REGION 153 POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.7 Craft Village Studies in Vietnam Three comprehensive studies of the craft industry have been carried out in recent years. MARD/JICA (2005) identi�ed over 2,000 craft industry villages, de�ned as having at least 20 percent of house- holds produce the same craft or at least 20 percent of village income from craft products. MoST/HPU (INEST) (2005) identi�ed 1,450 craft villages, including 840 in the Red River Delta, de�ned as villages with craft production constituting at least 50 percent of village income or having 30 percent of households engaged in craft production. MONRE/VEPA (2003) focused on 366 craft villages in six provinces in Nhue-Day River basin. The studies by MoST and MONRE concentrated on the envi- ronmental aspects and workers’ occupational health risks in craft villages. KEI, Korea (2005) also reviewed environmental conditions in craft villages and explored potential environmental management solutions to the problems. FIGURE 4.18 Burden of Ill Health in Urban Areas Percent of ARI Cases (Children under 5 years) Average Age at Death (last 3 years) 40% 67 66 35% 65 30% 64 25% 63 62 20% 61 15% 60 10% 59 58 5% 57 0% 56 1 2 3 4 5 1-3 4-5 Living Standard Quintiles Living Standard Quintiles Source: Calculated from VNHS 2002. BOX 4.8 Main Findings by the Vietnam Background Study – Industrial pollution in many cases impacts all people, poor and non-poor alike but the poor have fewer mechanisms to cope with the impacts from pollution. In Chi Dao commune, in a lead battery recycling craft village, both the poor, non-craft house- holds, and wealthier craft households were impacted signi�cantly by the different types of pol- lution. The craft households (there were only 25 in the 500-household village) live and work nearby heavily polluted water. The wastewater, which has been measured to hold 15 times the maximum lead-content standard, ends up in ponds or in the ground near the craft households. The poor are usually hired by the craft households to do the melting and collecting of the recy- cled lead at the furnaces, and thus are directly exposed to the toxic fumes. The lead content in the air in Chi Dao village is estimated to be about 4,600 times higher than the standard. In Du Dai, a food processing craft village in Thai Binh Province, water pollution caused by livestock production and noodle steaming facilities affect the poor and non-poor indiscrimi- nately. The poor however cannot afford health care treatment or deeper wells. Many of the workers are ethnic minority people. – In the major cities, the poor live in locations with higher exposure to industrial pollution and higher risk to health damage and disaster In District 6 in Ho Chi Minh City, an inner-city district with the highest poverty incidence (9.19 per- cent) of the districts studied, an open drainage canal that flows through the district is the cause of serious environmental pollution throughout the district. The poorest people, who live in slum areas along the canal, are most severely affected from the exposure to this pollution source. Source: Tai, Nguyen Van and others (2006). 154 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.9 Relevant Experiences of Environmental Pollution of SMEs in China and Korea Since the late 1980s, the main engine of China’s industrial development has been industrial small- and medium-enterprises (SMEs), often referred to as Township and Village Industrial Enterprises (TVIEs). In China, TVIE’s share of gross industrial output value increased from about 25 percent in 1990 to 58 percent in 1998, while industrial output value itself increased 20-fold between 1985 and 2004. From 1989 to 1995, TVIEs accounted for almost the entire increase in industrial-based water pollution loads. TVIE-generated particulate emissions (industrial dust) tripled and TVIE-generated solid waste volumes more than doubled. While industrial state-owned enterprises (SOEs) were sub- ject to a set of government pollution control policies from the 1980s, the TVIEs were not subject to such regulations. When the pollution loads and their environmental impacts were realized through the nationwide survey of more than 1 million TVIEs in 1995 and 1996, the government decided in 1997 to close down about 70,000 small enterprises. As a result, pollution loads were driven down in 1998 and 1999. However, China is still struggling with the aftermath of the earlier “laissez-faire� TVIE policies. Some of the critical lessons are: – It was particularly in the period when the TVIE contribution expanded that major pollution increases took place. – The regulation to control pollution from TVIEs should have been in place at the outset of this expansion, not at its end. – Technical, economic, and organizational means to control pollution from TVIEs should have been developed during the �rst phases of its expansion, not after. Korea followed a similar path in industrial development and environmental pollution of SMEs. SMEs grew rapidly in the late 1970s and the early 1980s as government support was redirected away from the large conglomerates. The majority of SMEs were concentrated on high polluting industries such as textiles, dying, leather, foundries, metal plating, paper manufacturing, and cement manufacturing. SME-generated pollution loads were estimated to account for more than 80 percent of the entire increase in industrial pollution loads from the mid-1970s to the early 1980s. The 1963 Pollution Prevention Law failed to contain industrial pollution due to a social atmosphere that placed higher priority on economic growth over environmental protection. Fol- lowing increasing pollution due to the proliferation of SMEs, the1977 Environment Preservation Act introduced comprehensive environmental standards and total pollution load management, mainly targeting SMEs. Government’s �nancial support to SME investments in pollution preven- tion facilities proved to be the most ef�cient policy. From 1983, the Korean Government began to provide long-term low interest loans and tax incentives for environmental investments and the downward spiral of industrial pollution was gradually reversed. needs to be enforced. Second, mechanisms to mental health. Pesticides relate to the agricul- modernize production facilities and use improved tural sector and may cause diffuse pollution but technology are required. Attention to the impact their linkage to poverty is the potential environ- on low-income groups should be incorporated at mental health risks that they generate. all stages. While wastewater pollution control from Pesticide use in Vietnam started as early as the craft industry is a priority, immediate attention is 1950s. Pesticide consumption grew signi�cantly also needed in neighboring communes and areas during the 1990s, increasing from 20,000 tons with high reliance on shallow dug wells. Water in 1991 to more than 40,000 tons in ___ (FAO quality monitoring of wells should be expanded, 2004). These increases followed crop intensi- and programs implemented to provide local peo- fication and increased cultivation of crops that ple, especially the poor, with safe water supply. required higher pesticide applications, such as dry-season paddy or fruits and vegetables. In addi- tion to the intensi�cation of pesticide use, there is Pesticide Use and Poverty evidence of improper pesticide use and its impact Pesticides are the third and last speci�c research on health. Hospital admission records in Vietnam, area in linkages between poverty and environ- for example, trace nearly 11 percent of all poi- PEN STUDY IN THE LOWER MEKONG SUB-REGION 155 POVERTY AND ENVIRONMENT IN VIETNAM sonings to pesticide misuse, with approximately pesticide users, have access to appropriate agri- 840 pesticide poisonings in 53 provinces in 1999 cultural inputs and techniques. (Poison Control Center 2000). Available studies, Eighty-six percent of the surveyed house- however, do not provide any analysis of the bur- holds report spraying pesticides every year. Pes- den of misuse shared by the poor. ticide use is highest—92 percent—among the At the national level, the National Health Sur- poorer 20 percent of households, and lowest— vey reveals a clear pattern of higher pesticide use 75 percent—among the better-off 20 percent of among the rural poor than among the non-poor households. The Vietnam Health Survey data for (Table 4.6). There is also a pronounced pattern the Mekong River Delta is overall lower, with in the application techniques. Of those farmers 58 percent of rural households using pesticides, who use pesticides, the poor use protective equip- but the linkage between income group and pes- ment during pesticide spraying much less fre- ticide use is similar.7 quently than the non-poor. Poor farmers in the survey use on average less The lowlands do not show higher pesticide use than half of the total reported annual pesticide than the uplands. First, the use of pesticides is use among the non-poor households. This is not prevalent in the whole country (Map 4.2). Second, due to a more limited range of products used. the upland regions of the Northeast and the Cen- There is no difference in the use of insecticides, tral Highlands, which specialize in high-value fungicides, and herbicides among the poor and perennial crops, have several poor provinces with the non-poor. Crop intensity is similar. Eighty- the highest prevalence of pesticide use in the nine percent of poor farmers in the sample and country. 60 percent of the non-poor cultivate less than The government has started an extension pro- 1.5 ha of land. Rice is cropped three times a year. gram of good agricultural practices that include Application times over one year are 4.5 times for optimizing the use of pesticides. Just as in other the poor, which is slightly less than the 5-time countries in the region, this shift is necessary to average for the non-poor. Average quantities per adjust to international market quality standards, application, therefore, make the difference. and is also welcome in terms of improved envi- There is a striking difference in toxicity between ronmental management. If this shift is to be suc- pesticides used by poor households in the survey cessful, however, policies have to ensure that the and pesticides used by the non-poor. When poor and non-poor farmers, who are important quantities were weighted based on their toxicity, it appears that the poor use a higher amount of “risk-weighted� pesticides (Figure 4.19). Slightly more than half the total quantity of pesticides TABLE 4.6 Share of Rural Households Using Pesticides used is, however, only moderately toxic based on the WHO classi�cation (category II). None of HH using the pesticides that the poor or the non-poor report protective pquipment using are extremely hazardous (category Ia) and HH using Quintile pesticides Always Sometimes Never only 0.5 percent of the total quantity is from the second-highest category of toxicity (category Ib). 1 64% 57% 18% 25% The poor in the survey use masks and other 2 73% 64% 17% 18% protective measures less often; this is statistically 3 68% 67% 16% 17% 4 52% 70% 14% 16% signi�cant at the 10 percent level. Around 42 per- 5 35% 72% 11% 17% cent of the poor do not use a mask and are, there- Total 61% 65% 16% 19% fore, likely to use no protection other than a shirt and trousers, whereas only 36 percent of the non- 156 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM MAPS 4.9 Paddy Area Distribution in Vietnam BOX 4.10 The Pesticide Use Survey in the Mekong River Delta Why is pesticide use more prevalent among poor farmers? Is the use of hazardous pesticides also more prevalent among them? Do these facts relate to different awareness levels and behavior patterns about risks related to pesticide use? And do the poor differ from the non-poor in access to information on risks related to pesticide use and to protective measures? These are the ques- tions that a survey in the Mekong River Delta undertook to answer. The study was also seeking to analyze whether health impairments attributable to pesticide use were disproportionately higher among poor farmers. The study surveyed 603 rice producers in 10 districts about their cropping systems, pesticide use and practices, application precautions, averting behavior, and health effects. The Mekong River Delta, the largest region of intensive rice production in Vietnam, was selected because it offered the advantage of a simple and intensive production system. This avoided discussing differences in pesticide use among a large range of crops. Households with per capita income below 1.2 Million VND/year (13 percent of the sample) are de�ned as poor and compared with other households de�ned as non-poor. The full report of the national team is available on CD-Rom (Dasgupta et al 2005, Nguyen 2005) PEN STUDY IN THE LOWER MEKONG SUB-REGION 157 POVERTY AND ENVIRONMENT IN VIETNAM were uncomfortable, compared to 41 percent of FIGURE 4.19 Application Amount, Risk-Weighted Amount the non-poor. (kg), and Number Poor farmers also use alternative plant pro- of Applications tection methods less frequently. Only 9 percent of the poor state that they used integrated pest 6.0 management (IPM) technologies, a number sig- 5.0 ni�cantly lower than the 16 percent of non-poor 4.0 farmers who cited the method. Similarly, 79 per- cent of the poor farmers—versus 68 percent of the 3.0 non-poor—stated that they did not use any pest 2.0 control method other than pesticides. 1.0 By contrast, the poor appear to be at least as aware of risks from pesticide use and to have good 0.0 Application amount *** Risk-weighted amount No. of applications ** access to information. Both the poor and the (kg) (per 100 kg) Poor (n=79) Non-poor (n=524) non-poor in the survey make clear statements about their perceived risk from pesticide use. Only 3 percent of the poor and 1 percent of the non- ***Statistically signi�cant at 1 percent level of signi�cance poor did not answer the question. 30 percent of **Statistically signi�cant at 5 percent level of signi�cance the poor stated that using pesticides generated a high risk, a proportion almost 50 percent higher poor do not use a mask. The difference is more than for the non-poor (Table 4.7). pronounced for protective equipment with less Poor farmers have a fair level of access to infor- frequent use, such as gloves and glasses. When mation. Only 5 percent of the poor, or 2 percent- asked why, close to one-third of poor farmers said that masks were not available, compared to TABLE 4.7 Perceived Risk of Pesticides only one-fourth of the non-poor. Only 34 per- (above) and Access to Information (below) cent of the poor stated that they believed masks Source Poor Non-poor FIGURE 4.20 Use of Protective Equipments Public media 38.8 40.9 Pesticide companies 17.8 18.5 100 Other 14.4 12.8 90 Agricultural extensionnists 17.7 11.6 Pesticide retailer 5.5 9.1 80 No information available 5.5 7.1 70 NGO 0.2 0.1 Total 100.0 100.0 Percentage 60 50 Perceived Risk Poor Non-poor 40 30 High risk 30 21 20 Medium risk 25 26 Small risk 28 31 10 No risk 14 20 0 Don’t know 3 1 Shirt Trousers Mask Hat Gloves Glasses Shoes Total 100 100 Poor (n=79) Non-poor (n=524) Sample size (n=) 79 524 158 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM age points less than the non-poor, state that no ers, regardless of their living standards, but while information is available on pesticides. All farmers 60 percent of farmers in the two upper quintiles identify the public media as their primary source have received IPM training, only 37 percent of of information on pesticides. Pesticide compa- those in the lowest quintiles have participated in nies and the staff of the local agricultural exten- such training. sion of�ces rank second as information sources. Suppliers also play an important role in pes- The level of education probably plays an impor- ticide use. Seventy-six percent of farmers in the tant role in accessing information. Only 1 percent �rst quintile purchase pesticides on credit, com- of the poor in the survey have no primary edu- pared to 54 percent in the highest quintile. The cation, and more than 45 percent have studied poor thus depend on credit providers for product beyond primary school. Availability of informa- selection. The non-poor use pesticide retailers as tion from agricultural extension agents is also information sources, but may be in a better posi- important. These are a main source of informa- tion to compare. Only 21 percent of farmers in the tion for 18 percent of the non-poor, a proportion lowest quintile regularly change the products they 50 percent higher than for the non-poor. More buy; 32 to 42 percent of those in the upper two than 60 percent of farmers in the survey are aware quintiles do. Poor farmers state more often that of pollution from pesticides through water, and the important factor in their choice of pesticides 27 percent state that wildlife could be impacted. is price, not technical effectiveness or safety. The usefulness of accessible information, there- The poor may also be unable to avoid pesticide fore, appears to be a critical factor that triggers toxicity when they are hired for pesticide spraying. more cautious behavior in relation to pesticide This situation is not uncommon, as indicated by use. The poor appear to have less access to train- the substantial proportion of rice growers in the ing and to information that allows real choices. highest quintile of the survey who do not report Only 29 percent of the poor compared to 34 per- spraying pesticides (25 percent). Many of them cent of the non-poor in the survey say that they probably hire other farmers to spray pesticides have participated in formal training in safe han- on their �elds; these are likely to be poorer indi- dling of pesticides. While this difference is not viduals or even landless people. Rural poverty is signi�cant, the difference in access to IPM train- often related to landlessness in the Mekong River ing is. IPM is an appropriate method for all farm- Delta.8 ers, regardless of their living standards, but while Potential differences in health impacts are dif- 60 percent of farmers in the two upper quintiles �cult to identify. The tests carried out in this study have received IPM training, only 37 percent of may only have an indicative value, but they indi- those in the lowest quintiles have participated in cate substantial exposure to pesticides. Around such training. 60 percent of farmers experienced skin irritations, The usefulness of accessible information, there- headaches, dizziness, eye irritations, shortness of fore, appears to be a critical factor that triggers breath, and other acute short-term health effects more cautious behavior in relation to pesticide after spraying pesticides (Figure 4.21). use. The poor appear to have less access to train- There is no clear distinction in how the poor ing and to information that allows real choices. and the non-poor are affected, but there is high Only 29 percent of the poor compared to 34 per- awareness of the issue among both the poor and cent of the non-poor in the survey say that they the non-poor. When asked whether they believed have participated in formal training in safe han- that these symptoms were related to pesticide use, dling of pesticides. While this difference is not 81 percent of poor and 82 percent of non-poor signi�cant, the difference in access to IPM train- farmers responded af�rmatively. Different levels ing is. IPM is an appropriate method for all farm- between the poor and non-poor are visible in some PEN STUDY IN THE LOWER MEKONG SUB-REGION 159 POVERTY AND ENVIRONMENT IN VIETNAM to potential health impacts in an environment FIGURE 4.21 Self-Reported Health Impairments After Using Pesticides (% of farmers) where paddy �elds are predominant. Waterborne diffuse pollution can become a substantial prob- 70 lem for all those who live in this environment. 60 The national policy to reduce the use of toxic 50 pesticides is a sound policy with an overall positive impact both in terms of poverty and environment. Percentage 40 This policy can, however, gain from careful incor- 30 poration of problems that are speci�c to the poor. 20 Poor farmers tend to participate less in agricultural 10 extension unless a speci�c effort is made to reach 0 out to them. The relatively higher levels of educa- Skin Headache Dizziness Eye Other Shortness of Fever Vomiting Convulsions Diarrhea irritation irritation problems breath tion and access to information create an opportu- Poor (n=70) Non-poor (n=412) nity for information programs on pesticide use. These messages can help them change attitudes and practices when faced with a pesticide issue. cases, but may not be signi�cant. Blood tests indicated signi�cant exposure to organophoshates 4.2.3. Poverty and Environment in or carbamates among 42 percent of the poor Regional Programs: The Cau versus 32 percent of the non-poor.9Thirty-eight River Basin percent of the poor versus 31 percent of the non- poor reacted positively to patch-skin tests for Key �ndings contact dermatitis, indicating existing exposure In contrast with classical environmental aware- to pesticides. In subsequent speci�c tests of reac- ness messages, poverty and environment link- tion to three commonly used pesticides, only 15 to age areas cover not only forestry but also lack 25 percent of the farmers tested positive, and access to safe water and sanitation and impact there was no clear distinction between poor and from the mining sector in the upstream. Land- non-poor. scape degradation is an issue in addition to industrial pollution in the downstream. Local- ized environmental impacts are at least as Policy implications important as impacts from one section of the There are three main reasons why the differences river basin on lower section one. observed among income groups in relation to Social differentiation within communities is pesticide use, although small in absolute terms, increasing rapidly and is accompanying indus- deserve attention in terms of policy making. First, trial pollution. This creates a need for public the large range of variables that show a correlation intervention to go beyond pollution remedia- with income groups provides a solid body of evi- tion. Social differentiation with the upstream dence of a poverty and environment linkage in and downstream section is extremely high and pesticide use. Second, many of the differences is generating a need for balance within the river observed are probably on an increasing trend. basin in the allocation of investment resources. As the non-poor become increasingly able to avoid The village level is a key level for targeting the use of more toxic pesticides, they may con- speci�c investments, while the province is a key level for environmental awareness and tinue to be directly exposed to these pesticides. stakeholder coordination. Third, the poor and the non-poor alike are subject 160 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM The beautiful landscapes and clean waters of the provincial development plans. Implementa- the Cau River in northeast Vietnam are described tion of the master plan, however, faces severe in a well-known traditional song in Vietnam. budget limitations. The Cau River basin was The river basin is today one of the three highest proposed in 2004 as a candidate to pilot an inte- national priorities in terms of environmental man- grated sustainable development program under agement. The Cau River originates in Bac Kan Vietnam’s national Agenda 21. Province. It then flows through Thai Nguyen City, a seriously polluted industrial zone in the Poverty and environment in the upstream country (Vietnam Environmental Monitor 2003). cau river basin Bac Ninh Province in the Red River Delta is one of the downstream provinces and is home to a Forestry in Bac Kan is an important part of land high proportion of the 200 craft industry villages use and is clearly household-based, but the sec- in the river basin. The Cau River basin is among tor provides limited income. Close to two thirds the most important areas in Vietnam in terms of of total land area within the river basin in the population concerned with water resource man- province and in the districts surveyed has been agement issues. An interim steering committee zoned as production forest land. Most forest is was set up initially to coordinate water resource natural forest zoned for production due to a sparse management in the river basin. A 13-year master forest cover. Satellite images show that active plan to protect the Cau River’s environment was deforestation took place in the 1990s within the released in 1998. Water resources and pollution river basin, and that the limited reforestation have been analyzed, and a substantial number of operations have mostly taken place in more acces- engineering solutions have been incorporated in sible areas (Castella and Dang 2002). Most of the BOX 4.11 Comprehensive Analysis of Poverty and Environment Linkages in the Cau River Basin How do various poverty and environment issues interact within the limited geographical domain of the river basin? Can these interactions serve as a basis to prioritize investments in the existing master plan and in the proposed sustainable development program? And what can be learned from these interactions to design environmental awareness activities within the framework of the national Agenda 21 program? A comparative analysis was carried out in two provinces in the Cau river basin to answer these questions. A 700-household survey in six communes was combined with a methodology of rural appraisal with focus group discussions, direct observation, and interviews in Bac Kan Province, the upstream province, and Bac Ninh Province, one of the downstream provinces with a high concentration of craft industries. Actual household incomes have been recorded, which allows for analysis by quintile. The survey was initially designed to analyze natural resource use in the upstream river basin survey, but the magnitude of mining and drinking water issues led to include these as well. The establishment of a multisectoral commune-level database was envisioned but turned out to raise substantial dif�culties10. Instead, we use district-level data from the PEN II land use database and the national health survey. The master plan is designed to bene�t the 54 districts of the 6 provinces that are members of the river basin committee11. The analysis takes into account this level as well as (1) the 30 river basin districts de�ned as affected by the Cau River in the master plan;12 (2) the impacted districts, or those districts plus the 12 districts of Hai Duong Province; and (4) the 16 riverside districts that directly create or receive pollution impacts along the main riverbed (Map 1). Area and population covered each vary from 1 to 3.7 between de�nitions (4) and (1). The full report of the national team is available on CD Rom (Socio-economic development center 2005) PEN STUDY IN THE LOWER MEKONG SUB-REGION 161 POVERTY AND ENVIRONMENT IN VIETNAM area is in a stage of natural regeneration. Ninety- available from local employment in mines. Average �ve percent of households have been allocated incomes recorded in that commune are 45 per- forest land. Land that remains under a state forest cent above the other two communes. However, enterprise provides fuelwood and bene�ts from social disparity also increases with mining and protection contracts to the households. other non-agricultural activities. The income of The four districts located within the physical the richest 20 percent of the surveyed households river basin are substantially less poor than the is more than 9.5 times the income of the poor- three more remote northern districts located est 20 percent in the river source commune with outside of the river basin. The Dao people, an the large mine. It is only 2.9 times greater in the ethnic group generally living in disadvantaged fully agricultural and forestry communes. Poor conditions compared to the main ethnic group households are often not in a position to avoid in the province, the Tay, account for a high pro- water pollution in dug wells. portion of the population in all districts but a Both large and small mines generate pollution higher one in the northern districts. Environ- issues. In one village within one of the three com- mental stakes differ as well. Forest in the northern munes surveyed, a boom in illegal gold mining in districts is mostly zoned for environmental pro- the early 1990s attracted several thousand infor- tection and includes the Bac Be National Park. mal workers. The activity has been banned since, Slopes are markedly steeper than within the river but households express high concern over water basin. The impact of rotational agriculture, a quality. They claim that they have lost domestic practice that the Dao have retained in all districts animals, an important household asset for house- to complement their small areas of paddy �elds, holds in poor communities, due to water toxicity. is likely to be higher than within the river basin. Social differentiation is equally visible in access Erosion within the river basin is generated by to forest resources. Poorer households have been causes broader than rotational agriculture and allocated 3.6 ha of forest land on average, signif- mining is an important cause. icantly lower than the average of 5 ha for the The mining sector is present in three of Bac entire sample of households. About 10 percent of Kan’s four river basin districts. Its environmental households have been allocated more than 10 ha impact on soil and water resources appears to be and up to 67 ha of forest land. Poorer households substantial. Cho Don district plays a crucial role are also allocated forest that is more distant.14 in the water balance of the whole river basin, Allocation of distant forest land means fewer because of its location at the source of the Cau opportunities for households to use this land to River. Cho Don is also the district with the largest grow timber or cash trees. The number of large proportion of land allocated to mining uses— domestic animals, another indicator of access to 2 percent of total land area13. forest resources since large animals in the area A combined issue of water quantity and water mostly feed in the forests, is also clearly linked to quality is stated by households as one of their income groups. main problems and affects the poor more than Differences in poverty levels between villages the non-poor who can afford alternative sources. are visible. The average income in higher-income Close to 85 percent of households mention water villages is 1.2 to 1.5 times the income of the shortages during the dry season. Twenty percent poorer villages within each of three communes of households in the water source commune have surveyed. These differences have a clear ethnic shifted from dug wells to gravity water pipes, dimension. Two of the villages surveyed are from because of declining water quality in the dug wells. the Dao ethnic group in a mixed Tay, Dao, and This desire to shift to a more modern type of Kinh commune, while the other villages belong water supply can be related to the higher incomes to fully Tay communes with only a few Kinh 162 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM MAPS 4.10 Cau River Basin Upper left: boundaries of river basin. Upper right: population change. Lower left: ethnic minority people in total population. Lower right: poverty incidence to replace Dao ethnic group. PEN STUDY IN THE LOWER MEKONG SUB-REGION 163 POVERTY AND ENVIRONMENT IN VIETNAM torical� stratum related to differences in status FIGURE 4.22 Forestry and Poverty Indicators in the Upstream Province between villages in each commune and a more recent stratum related to social differentiation within communities. In the �rst stratum, poorer N. districts 52 villages have smaller paddy �eld areas and con- Average elevation 103 tinue to farm sloped land, although they are grad- % poor ually reducing this practice through intensi�cation population 71 Bac Kan Province of the paddy �elds. They may access larger areas % Dao ethnic group 87 Bac Kan river basin of forest land but are less likely to derive income from it. Poorer villages have less access to collec- % forestry land 94 tive water networks and are at risk of signi�cant % environmental impact from point pollution. In the second stra- forests 57 tum, the poor are more exposed to dug well pol- 0 20 40 60 80 100 120 lution and less opportunities to derive income from their forest land. These strata are not visible in statistics alone. households. The two Dao villages rank second The income of the surveyed Dao households is and third poorest among the 11 villages surveyed. 74 percent of the survey average, less than the Paddy accounts for 100 percent of agricultural income gap between households. The richest land in all Tay villages (except the two higher- commune is also the commune where the income income villages, where the land may have more of the poorer 20 percent of households is the pro�table uses). It accounts for only 15–20 per- lowest. Finally, differences between the poor and cent of agricultural land in the two Dao villages. the non-poor in land use are not visible when The development of tea and other perennial crops comparing average variables for income groups on deforested forest land is low −1.3 ha one across the district. average - in all villages due to limited forestry budgets available for the provision of seedlings. The Dao villages have planted the smallest areas. Poverty and environment in the Ethnicity sometimes relates to poverty and downstream cau river basin natural resource use in opposite directions. House- Bac Ninh Province, although downstream of Thai holds in the two Dao villages have been allocated Nguyen’s heavy industries, has important pollu- 6 and 11 ha of forest land on average, signi�cantly tion sources in the province itself. The develop- more than in the seven higher-income villages, ment rate of household-based craft industries in where forest land allocated to households is only Bac Ninh Province is staggering. The number of 4 ha on average. This is due to larger forested areas households running craft industries in Yen Phong, in the remote locations where the Dao villages the main district surveyed in Bac Ninh Province, are established. All Tay villages are increasing grew to more than 5,600 in 2004. In this district, their area of paddy �eld by opening new plots. One 20 percent of households run a craft industry. Dao village is engaged in the same trend, while High pollution impact is generated by craft indus- the other is not. Paddy intensi�cation through a tries, agricultural ventures and large enterprises shift from one to two crops a year is practiced in and is concentrated in a small number of villages the two Dao villages and in the higher-income within the communes. Paper recycling is one of Tay villages, but not in the other Tay villages. the heavily polluting craft industries. Phong Khe Poverty and environment linkages in the village, a well-known craft village part of the sur- upstream Cau River basin appear to have a “his- vey, turned to paper recycling in the 1990s after 164 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM �recracker production, its former production, last 12 months versus 6.4 days for the non-poor. was banned. The second surveyed commune has Both of these �gures are higher than the national a pig-raising and alcohol-brewing village. A local average of 5.8 days. The entrepreneurs see pol- coal power plant is the main pollution source for lution as an unavoidable cost of poverty reduction. the third survey commune, which has retained This view is supported by rapid reduction of agriculture as its main activity. State-owned indus- poverty incidence decreases as pollution increases. trial industries in Yen Phong district increased The proportion of households with incomes below from 5 to 24 over the same 10-year period, and $1 per day is 57 percent in the agricultural com- private enterprises grew from 23 to 154. mune, 33 percent in the alcohol brewing com- Social differentiation is growing rapidly and is mune, and 13 percent in the paper recycling substantially higher than in Bac Kan. The incomes commune. of the richest 20 percent of households in the The poor are at least as aware as the non-poor paper recycling village are 54 times the incomes about environmental risks. They use water sources of the poorest 20 percent in the third commune that are less safe but state concerns about unsafe (the commune that has not diversified out of water more often (Figure 4.23). They assess dug agriculture). The poor population in craft villages wells as an unsafe water source but are twice more is mostly comprised of workers coming from out- likely to rely on them than non-poor households. side the village to work in local household-based The poor even raise concerns for water quality industries. There are no less than 4,000 workers from drilled wells. in Phong Khe village alone. Occupational health District and commune leaders underline the and lack of healthcare system was mentioned as technical and �nancial limitations of the two pol- an issue by the respondents. lution remediation policies are in place, (1) hotspot Unclean drinking water, a contaminated gen- treatment for some of the large industrial estab- eral water environment, coal gas, and dust are lishments; and (2) the development of industrial the environmental issues mentioned as affecting zones to move craft industries out of residential daily lives. Commune health workers witness areas. The coal power plant has been covered higher incidence of several diseases in the craft under the pollution remediation program for large villages. The poor reported 9.8 sick days in the industrial sites but local residents believe that the FIGURE 4.23 Water Sources in Three Downstream Communes. Left: Percentage of Households Using Water Sources. Right: Percentage of Households Assessing Water Source as Unsafe 80 100 100 100 Poor households 70 Non-poor households 95 60 All households 90 90,6 90 87,1 Poor households 85,7 50 85 40 77,8 80 30 75 20 68,8 70 10 65 0 60 Drilled well Dugwell Rivers and River and Dugwell Rainwater Drilled well ponds ponds PEN STUDY IN THE LOWER MEKONG SUB-REGION 165 POVERTY AND ENVIRONMENT IN VIETNAM impact from the exhaust has remained signi�- absolute numbers of poor people in 1999, despite cant. An industrial zone was created in the paper the fact that 60 percent of its residents live below recycling village, but only 60 industrial estab- the poverty line. This �gure remained high in lishments, or around one-third of the total num- 2004, estimated at about 50 percent. ber, have had the means to assemble the required Bac Ninh had 445 poor people per km2 in investment to settle in that zone. The leaders stated 1999, the highest density in the country in rela- feeling as though they were left alone to address tion with very high rural population densities, local pollution issues with limited resources. 1,064 people/km2 in 2000 or more than 20 times They underline that the river basin master plan the population density in Bac Kan. With the assembled evidence of pollution charges many exception of Hanoi municipality, Bac Ninh is times above national standards as early as 1998, the province with the smallest territory and the and that yet limited changes have taken place highest population density in Vietnam. It is also since then. the third least forested province in Vietnam, with only 1 percent of land zoned as forest land. GDP Linkages between the upstream per capita in Bac Ninh is 85 percent higher than and downstream river basin GDP in Bac Kan, though the economy is now Bac Kan and Bac Ninh are two provinces only growing at the same rate of 2.3 to 2.4 percent per a short distance away from one another that year in the two provinces. Poverty incidence in reflect the extreme upland/lowland dimension Bac Ninh decreased from 38 percent in 1999 to of the poverty-environment nexus in Vietnam. 3 percent in 2004. Bac Kan Province was created when the former All household and community-level indicators Thai Nguyen Province was divided into two by of poverty are markedly higher in the upstream zoning out into the section of the province that districts reflecting both lower incomes and lower specialized in forestry. Bac Kan has the third low- access to services (Table 4.8). There are however est population density in the country (57 people/ two exceptions. Paddy �eld per capita is however km2.) Bac Kan, with 62 percent of its land area not lower in Bac Kan than in Bac Ninh. Area per under forest cover, is the third most forested capita is very small in Bac Ninh in relation to high province in the country. The province ranked population density, and progress in opening seventh in terms of poverty incidence in 2004, just new paddy �elds to increase the sustainability of as it did in 1999. It only ranked 55th in terms of farming has been made in Bac Kan. Children in TABLE 4.8 Poverty and Environment Indicators in the Upstream and Downstream Districts Cho Moi/ Yen Phong/ % Bac Ninh/ District/Province Bac Kan Bac Ninh Bac Kan Monthly income 1000 VND (VLSS) 191 326 171 % Kinh population 21 100 476 Agricultural land per capita 1347 545 40 % crop production in income 42 8 18 % agriculture and forestry in income 62 22 35 % households with temporary housing 26 1 5 % households without safe water 85 20 23 % households without sanitation 79 54 68 % communes with doctor 44 100 229 166 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM Bac Kan, although overwhelmingly from ethnic basin. Flood episodes followed by low water flows minority groups, have equal access to education. do put downstream provinces at risk and worsen In the future, job opportunities outside subsis- pollution impact but water shortages are also an tence farming and information in general might important issue in upstream communes. Local become equally accessible across the river basin. stakeholders mention a possible combination One of the most common environmental of factors that include disruption from mining awareness messages in Vietnam says that increas- operations at the river basin head and past degra- ing forest cover rates in upstream watersheds dation of forest resources, perhaps in combina- will preserve or improve water balance in the tion with climatic trends. This is consistent downstream watersheds and prevent soil erosion. with the fact that soil erosion in Bac Kan may Upland farmers in some of the poorest, heavily be a more serious issue in sections of the forested districts are receiving this message repeat- province that belong to river basins other than edly (Tessier 2002). This message is very far from the Cau River basin. reality in the Cau River basin. Forest resource degradation impacts the local economy in the Policy implications upstream section before generating potential The masterplan mostly proposes investments for impact in the downstream section. Most of the forestry development in the upstream section are forest land is now under natural regeneration, mostly for forestry development, and concern but this process will take a long time. for environmental issues is limited to the down- Assessing the relative responsibilities of sub- stream section of the river basin. By doing so, the sistence farming and of commercial forest exploita- masterplan omits the issue of access to safe water tion in forest degradation in the past is much less and control of spot pollution in the upstream of a priority than building alternatives for the section. It does recognizes the need for planting current period. Alternatives are in place for farm- trees and improving landscapes in the downstream ers who are starting to develop tea and other section. The continuing focus on forestry in the perennial crops but are more dif�cult to identify uplands is more dictated by the absence of bud- for the local government itself. The development gets available from national programs. of the mining sector, an alternative with a dis- The environmental improvements identi�ed proportionate environmental impact, is not an in the initial master plan signi�cantly exceed the appropriate solution. Forestry and related reha- �nancial resources available. The master plan bud- bilitation needs do not only relate to the upper get is $1 per person per year, if the six provinces river basin. Just as in the traditional song about are fully included. The master plan underlined the the Cau River, trees were an important element importance of more cost-ef�cient “soft invest- of landscapes all along the Cau River. ment� to overcome these important budget con- The downstream river basin suffers from an straints. Improving coordination between agencies uneven water balance, but there is no evidence of and exchange of information was identi�ed as an the impact of subsistence activities in the upstream important type of “soft investment� early on. The river basin. Much of the pollution increase in the focus of the Agenda 21 program on environmen- downstream section is reported as being generated tal awareness is a second type of “soft investment.� locally by a combination of heavy industries, Messages based on a comparison of representative unregulated craft village industries, agrochemicals, communes—one in the upstream section, one in and organic pollution. Heavy industries in Thai the downstream section—were initially planned. Nguyen not only generate pollutant emissions, The awareness messages would have described but also have high water uptakes that reduce water initial dif�culties and progress made through a availability for the downstream section of the river pilot program in these two communes. PEN STUDY IN THE LOWER MEKONG SUB-REGION 167 POVERTY AND ENVIRONMENT IN VIETNAM The Cau River basin offers an ideal setting for Community capacity building is another cost- innovative environmental awareness messages. ef�cient intervention with the potential to be a Messages that would tell the story of a need to “win-win� solution in the long term for poverty increase forest cover rates to restore water balance and the environment in the Cau River basin. in the downstream section can be replaced with Capacity building needs are high among local up-to-date, realistic messages. The strong con- cadres. With improved capacity access of poorer trast between less-favored forested uplands pop- communities to development opportunities or ulated by minority ethnic groups and heavily to environmental improvements can improve. polluted lowlands that also have an important Building community capacity through various poverty issue is an opportunity to develop mes- means, from training local cadres to formal en- sages based on two shared issues, the need to hancement in the regulatory framework, is an restore water quality and landscape degradation. indirect environmental intervention, but a direct Messages based on shared issues have higher poten- poverty reduction intervention that can signi�- tial to generate attitude change. cantly improve the environment. In a constrained budget environment hard investments are likely to remain focused on 4.3. POLICY OPTIONS: INTEGRATING selected locations within the physical river basin POVERTY AND ENVIRONMENT itself. The provincial level is however an appro- IN NATIONAL POLICIES priate level for cross-sectoral coordination and AND PROGRAMS environmental awareness and is recognized in the master plan. Some of the solutions to envi- 1. Geographical Targeting of Policies ronmental degradation within the physical river and Programs basin are located in province sections that are out of this river basin. Tourism development is an Prioritizing the upland sector or the lowland option worth exploring as an alternative to min- urban and industrial sector is not relevant. There ing in Bac Kan even though the Bac Be national is a dual need for interventions in the marginal park is outside of the physical river basin. Well- rural areas and in the urban or peri-urban areas functioning protected areas may be a powerful that are in a rapid industrialization process. There awareness-raising factor. They convey the mes- is, however, a striking contrast between existing sage that industrial pollution is not a necessary poverty reduction programs in upland rural areas evil in economic development. that are mostly conducive to environmental Interventions prioritized as “win-win� poverty improvements, despite room for improvement, and environment solutions are also cost-effective and environmental programs in urban and indus- solutions that can help alleviate budget constraints. trial areas that have so far largely missed a pro-poor Safe water access ranks first among these cost- dimension. The former deserve continuation in efficient solutions. Affordable options to give the long term, while the latter require innovative access to safe water to the poor in the downstream and pro-active approaches. The danger would be section are likely to be less investment-intensive to reduce “well-known� interventions in upland options than full coverage of the population with rural areas to concentrate on innovative inter- tap water systems. Access of speci�c villages to ventions in the lowlands. Some lowland rural areas safe in the uplands can be more cost-ef�cient than also have important poverty and environment larger-scale tree plantations. When safe water issues that are unrelated to industrialization. These access is secured. Households will continue to include the Southernmost areas of the Mekong shift from cropping sloped land to animal pro- for access to safe water and coastal areas with poor duction and perennial crops as tea. population subject to typhoons. 168 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM Upland rural areas with persistently high tantly capacity to prioritize village needs and poverty rates and pressing natural resource man- interventions needs to be enhanced among gov- agement issues will however deserve specific ernment and at commune level and speci�c tools intervention programs in the long term. The gov- are needed. ernment already has a long-standing policy of Finally, some poverty and environment issues speci�c support to these upland, ethnic minority would be best addressed at a nationwide level. areas and a speci�c agency, CEMMA. The initial Environmental health risks in relation to pesti- goal of this policy was centered on security con- cides, for example, are prevalent throughout the cerns in border areas, and economic development country. was mostly promoted through in-migration. This goal shifted to poverty reduction in the early 2. Priorities and Importance of 1990s (Communist Party of Vietnam 1989) and Multi-Sectoral Approaches several targeted programs are implemented, today with support from several international donors. The policy implications of each detailed PEN The 135 Program under the Ministry of Planning research area together create a framework of and Investment, initiated in 1999, is probably the priorities (Table 4.47). Other priorities that best-known program for upland areas along with have been covered under the PEN research include other programs under MOLISA. This program pro-poor disaster preparedness programs, envi- targets poor and remote communes primarily for ronmental health issues such as malaria and small infrastructure development. The number indoor air pollution, and issues specific to pro- of bene�ciary communes totaled 1,875 in 1999, tected areas. and increased to 2,335 in 2003. Potential “win-win� impact on poverty reduc- The Northern mountains region stands out tion and the environment lies more in incorpo- for its very high poverty incidence and poverty rating better attention to the environment in depth. Poverty and environment linkages are existing poverty reduction programs and taking equally important in the central regions. Regional a “pro-poor� approach in existing sector policies differences can best be taken into account through and environmental programs than in innovative programs with a regional dimension. And the interventions. speci�c need for an actively pro-poor continuation Most priorities identi�ed are multi-sectoral of forestry reform in the central regions needs to within broader rural or urban and industrial devel- be fully recognized by Government and donors opment areas. Interventions needed for balanced alike. and equitable industrial development for exam- The Red River Delta and adjoining provinces ple are multi-sectoral by nature. A poverty and has an outstanding issue of rapidly growing pol- environment approach needs to take into account lution and degraded river basins such as the Cau not only pollution control and prevention through river. These deserve geographically targeted inter- relocation of industries to industrial zones for ventions. Interventions covering both uplands example, but also how each solution may affect and lowlands and oriented towards soft invest- the poor differently from the non-poor. Urban ments such as environmental awareness can have development, including transportation, deter- high ef�ciency compared to pollution remediation mines the locations where the poor can afford to alone. work and live. Health services in urban areas can Targeting poor villages within communes is be pro-poor by focusing on children and the both a need and a challenge. Poverty and environ- elderly and on occupational health in high-risk ment issues are combined in some of the villages occupations. Social protection measures accessi- in some cases, separate in others. Most impor- ble to the poor are also a necessity. Although the PEN STUDY IN THE LOWER MEKONG SUB-REGION 169 POVERTY AND ENVIRONMENT IN VIETNAM TABLE 4.9 Summary of Program Priorities in Relation to Poverty and Environment Linkages Poverty & environment linkage area Program/policy priority Environmental Clean water & sanitation: improved services for rural households: health in rural a—with no improved water supply or sanitation areas b—using polluted dug wells. Information/education programs. Community leader information and training. Pesticides Strategic information campaign for users and retailers Environmental Industry regulatory framework and institutional capacity building health in urban Example: pro-poor national regulatory framework for craft industries and industrial Urban & peri-urban development areas Priority to access to safe water and sanitation Affordable improvements in working & living conditions Capacity building for land law implementation Natural resource Continued poor commune program with improvements in: use in upland Community capacity building areas Information Land administration Development of modern land use plans, with simpli�ed speci�cations in mar- ginal areas Community leader information and training on legal framework Linkage to pro-poor forestry reform in the transition from State to enterprise sector Natural resource Area-based integrated program use and Environmental awareness program combined with selected priority investments environmental health government is already implementing some multi- The specific difficulties of marginal upland sectoral interventions, the agencies that have taken areas that are home to ethnic minority people part in the PEN research or contributed data— deserve to be addressed by a separate poverty MoNRE, MPI, MARD, MOH, CEMMA—can and environment initiative for the Cordillera. improve the poverty and environment impact of Such an initiative could facilitate recognition these interventions through better coordination that, despite existing policies, inequitable access among themselves as well as among other stake- to resources or lack of capacity to compete with holders that include enterprises, local governments other economic operators creates powerful limi- or the media. tations in poverty reduction, sound environmen- tal management, or both. Building community capacity in areas such as land administration 3. Opportunities for Poverty and would also have potential in such an initiative. Environment Initiatives An initiative in the region could also focus There is an opportunity to give high visibility to on bringing pro-poor information that, un- broad poverty and environment issues by making like general mass media information, would them stand-alone initiatives. Four such initiatives be tailored to the practical needs of the local have been identi�ed (Box 4.12). population. 170 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM BOX 4.12 Potential Poverty and Environment Initiatives and Cross-cutting Limiting Factors There is potential for four poverty and environment initiatives: – A nationwide clean water and sanitation initiative – A poverty and environment initiative for ethnic minorities in the Cordillera (in relation to the existing partnership for poorest communes) – An area-based comprehensive initiative in degraded territories (already planned for the Cau River basin) to support innovations in pro-poor environmental health and pollution remediation. Access to clean water and sanitation is such a these innovative interventions. A regulatory frame- clear and broad poverty and environment priority work for craft industries and other major gaps in that it deserves a speci�c nationwide initiative. the legal or regulatory framework require urgent The contrasted issues of the marginal upland areas action. Improved capacity for multisectoral urban in the Cordillera and of the rural and peri-urban development policies is a critical skill. Improved areas that are increasingly affected by diffuse institutional capacity can bring in improved imple- pollution would justify a double initiative. A mentation and enforcement of the regulatory specific clean water access program—instead framework, enhanced coordination between stake- of the existing general infrastructure-building holders, and increased attention to innovation. 135 program—would serve to target those cur- The Ministry of Planning and Investment rently without access and would help balance (MPI) has expressed interest in Bank assistance investments between clean water and roads or to urgently address industrial pollution issues in other much-needed but expensive infrastructure. “hot spots� particularly in the Cau River Basin In the lowlands, affordable, broad-based solutions as a follow-up to this study (note that this proposal to address diffuse pollution need to be introduced. is being shaped at present in line with �ndings pre- Again, access to appropriate information for the sented in map 4.8). The planned pilot project poor would be an important element of such would provide technical assistance to support initiatives. (i) the government in strengthening monitoring An initiative in the urban and industrial sec- and enforcement of compliance with environ- tor would facilitate innovation, a critical element mental regulations, implementing environmental for success in these interventions. Innovative information disclosure programs, and introduc- approaches are needed for issues that are partly ing regulations for craft industries; (ii) industry rural, partly peri-urban, and partly urban. Inno- through the preparation of environmental man- vative technical management systems need to be agement audits, the introduction of self-monitor- set to address issues such as diffuse pollution, ing and environmental management systems such as well as in terms of institutions to deal with, as ISO 14,000, adoption of clean technology, and for example, pollution issues concentrated in vil- training programs for workers’ safety in high lages within communes. Disclosure of informa- health risk industries; (iii) the banking sector tion on pollution levels, wherever possible based through the preparation and adoption of envi- on improved pollution monitoring, can be a use- ronmental manuals. On the investment side, the ful measure to help the poor and the non-poor get project would pilot pollution management invest- relevant information. Public awareness campaigns ments including (i) the establishment of central- can also help promote risk-averting behaviors ized craft industry zones and centralized waste among low-income groups. The proposed Agenda water treatment systems for highly polluting craft 21 program in a degraded river basin is one of industries in areas with high pollution risk of PEN STUDY IN THE LOWER MEKONG SUB-REGION 171 POVERTY AND ENVIRONMENT IN VIETNAM potable water supplies and where the density and 12. The actual number of districts is changing. Three dis- size of craft industries make this economically jus- tricts have recently been merged or divided into two. In all districts,AQ4 at least part of the physical river basin ti�ed; and (ii) investments, in a cost sharing basis, are included. Six of Bac Giang’s 10 districts, which are in cleaner technologies and pollution control in the watershed of a lower tributary, are excluded. medium and large scale enterprises in high 13. Mostly under a joint venture between a state-owned growth-high pollution intensive sectors in enter- enterprise from Thai Nguyen province and two Asian enterprises. prises, which make commitments to introduce 14. Poorer households in other case studies were found to environmental management systems and to com- be allocated more distant forest land with poorer qual- ply with environmental regulations. MPI is con- ity land (Sunderlin and Huynh 2005). The survey only sidering the inclusion of an environmental health recorded distance to land with protection contracts, but a similar situation is expected for production forest. component into the proposal, reflecting health impacts polluting industries may have on certain populations, for example through the use of References unsafe drinking water sources (see �gure 4.23). PEN Case study reports in CD Rom Attachment Dasgupta, Susmita, Craig Meisner, David Wheeler and Jostein Nygard 2005. Poverty/Pesticide Use in the Endnotes Mekong Delta, Vietnam. Washington, DC: World Bank. 1. The latest World Bank poverty update for Vietnam Nguyen, Huu Dung. 2005. Poverty and Pesticide Use in shows that in 2004, 39 percent of the poor were ethnic Rice Production in the Mekong Delta, Vietnam. Ho minorities, a �gure that has risen from 20 percent in Chi Minh City: Environment Economic Unit, Univer- 1993. sity of Economics. 2. Since it was too early to examine land titles, the Socio-Economic Development Center. 2005. “Environ- analysis was done on allocation of land use rights on ment and Poverty Linkages in the Cau River Basin— forest land. Vietnam Poverty and Environment Nexus Study.� 3. Foreign-invested industrial output has no signi�cant Hanoi: background paper to the PEN study. association with provincial poverty or living standards. Tai, Nguyen Van, Nguyen Trung Thang, Nguyen Thi This might be influenced by the concentration of for- Thuy Duong, Nguyen Duc Minh, and Hoang Van eign-invested industry in a few provinces, and is not Hoa. 2006. “Poverty—Environment Nexus Program, necessarily inconsistent with the foreign-invested indus- Case study: Poverty and Industrial Pollution in Viet- try having important macroeconomic bene�ts, and thus nam.� Hanoi: background paper to the PEN study. indirectly contributing to higher living standards. TECOS (Consultancy Service and Technology Development 4. Pollution intensity de�ned as pollution load per unit of Company for Natural Resources and Environment). production output. 2005. Vietnam Poverty and Environment Nexus Study 5. World Health Report 2002. World Health Organi- Phase II. Land Resource—Poverty Study in Vietnam. zation. Hanoi: background paper to the PEN study. 6. Estimated from the VNHS 2002. 7. The health survey covers all rural households, while this Other references survey is focused on rice growers. Castella, Jean-Christophe and Dang Dinh Quang. 2002. 8. The study was focused on rice growers and therefore did “Doi Moi in the Mountains: Land Use Changes and not cover exposure of landless people to pesticide use. Farmers’ Livelihood Strategies in Bac Kan Province, 9. The blood test used a cholinesterase enzyme technique Vietnam.� Hanoi: The Agricultural Publishing House. measuring active cholinesterase enzymes in erythrocyte Communist Party of Vietnam. 1989. Politburo Decision on and plasma using the Test-Mate-Kit (EQMR-USA). Guidelines and Main Policies for Economic and Social 10. IRRI’s SAM project has established a detailed database Development in the Mountain Area. Document of the for Bac Kan Province. Communist Party No 22/TQ NW, November 27, 1989. 11. Hanoi municipality only has two river basin districts Curtis, V. and S. Cairncross. 2003. “Effect of Washing and is not a full-fledged member; Vin Phuc is a full Hands with Soap on Diarrhea Risk in the Community: member with six river basin districts, although the A Systematic Review.� Lancet Infectious Diseases 3. Cau River des not flow through the province; Hai Esrey, S. A., J. B. Potash, L. Roberts, and C. Shiff. 1991. Duong is a member because it receives downstream “Effects of Improved Water Supply and Sanitation impacts. on Ascariasis, Diarrhoea, Dracunculiasis, Hookworm 172 PEN STUDY IN THE LOWER MEKONG SUB-REGION POVERTY AND ENVIRONMENT IN VIETNAM infection, Schistosomiasis, and Trachoma.� WHO Bul- Swinkels, Rob 2004. “Draft Synthesis: Poverty and Envi- letin 69 (5):AQ2. ronment.� Hanoi: World Bank. Fewtrell, L. and J. Colford Jr. 2004. “Water, Sanitation and Tessier, Olivier. 2002. “World Bank Forestry Sector devel- Hygiene: Interventions and Diarrhea—A systematic opment Project. Initial Social Screening.� Montpellier, review and meta-analysis.� HNPAQ3 Discussion Paper. France: Tercia consultants. Washington, DC: World Bank. Vietnam Ministry of Planning and Investment. 2004. Socio- IFPRI, IDS, and Inter-Ministerial Poverty Mapping Task economic Development of the Poorest Communes: Force. 2003. “Poverty and inequality in Vietnam: Spa- Community Perspectives and Future Prospects. Partner- tial patterns and geographic determinants.� Washing- ship to Assist the Poorest Communes national confer- ton: IFPRI and Brighton: IDS. ence proceedings. November 26, 2004. Hanoi: MPI. Khong, Dien. 2002. “Population and Ethno-Demogra- Vietnam Poverty Task Force. 2002. “Ensuring Environ- phy in Vietnam.� Chiang Mai, Thailand: Silkworm mental Sustainability. Localization MDGs for Poverty Books. Reduction in Vietnam.� Hanoi: Government of Viet- Korea Environment Institute (KEI), 2005. “Vietnam– nam, Poverty Task Force. Regional Environmental Management for Traditional Vietnam Poverty Update. 2006. Power Point Presentation. Villages,� Final Report, Seoul, Korea. Washington, DC: World Bank. Sunderlin, William and Huynh Thu Ba. 2005. “Poverty World Bank. 2003. Vietnam Environment Monitor 2003. Alleviation and Forests in Vietnam.� Jakarta: CIFOR. Washington, DC: World Bank. PEN STUDY IN THE LOWER MEKONG SUB-REGION 173 III SUB-REGIONAL PERSPECTIVES 5 Subregional Findings INTRODUCTION potential to address poverty jointly with environ- ment in initiatives at that level. The three coun- The Lower Mekong River flows through Laos, tries are already part of cooperation bodies, most Cambodia, and Vietnam in its lower stretches. notably ASEAN and the Mekong River Com- Do the three countries form a “subregion� in mission.1 ASEAN is a driving force behind the terms of poverty and environment issues? current opening of borders and regional roads.2 The Lower Mekong River itself is not a com- According to MRC, the international body estab- mon thread when environment is analyzed from lished to promote and ensure the sustainable the point of view of its linkages to poverty. In management of water and related resources, the Cambodia, seasonal floods heavily affect large condition of the common resources “ha[s] impli- numbers of poor. At the same time, seasonal cations for how successful people are in raising flooding is important for many rural farmers. In their standard of living and bringing about related Laos, the Mekong floods affect the non-poor social improvements� (MRC 2003). groups that form the vast majority of Mekong Potential joint initiatives will be a focus topic plains inhabitants. In Vietnam, floods along the for discussion among national stakeholders at the northern central coast are related to local topo- subregional PEN conference. This draft chapter graphy and typhoons and not to the Mekong will be revised after the conference to fully incorpo- River, and only the southernmost section of Viet- rate the results of this upcoming event. nam, the Mekong River Delta, belongs to this river basin. This chapter highlights which areas of poverty Shared Features in Poverty and linkage have high magnitude in only one of the Environment Linkages three countries or are shared among two or three The PEN research has analyzed the magnitude countries. The practical solutions to these issues of several poverty and environment linkages in are often country-specific, but broader policy Vietnam, Laos, and Cambodia. When taking a options are largely similar. This creates opportu- horizontal view of poverty and environment, nities to jointly address poverty and environment linkages can be ranked as having high, medium, above the national level. or low magnitude or severity based on informa- Subregional environmental initiatives are tion on spatial distribution of the issue and of already in place, but there is a largely untapped poverty (Figure 5.1). PEN STUDY IN THE LOWER MEKONG SUB-REGION 177 SUBREGIONAL FINDINGS FIGURE 5.1 Magnitude of Poverty and Environment Linkages in the Three Countries VN = Vietnam, LA = Laos, KH = Cambodia Poverty and environment linkages of high relevance in more than one country include:3 – Environmental health issues with water supply and sanitation and urban pollution, – Natural resource use issues related to forestry issues (including NTFPs, other resources and the impact of roads), UXO contamination, and natural disasters. Environmental Health and Poverty • Low levels of community capacity and infor- The analysis of environmental health and pov- mation on environmental health (an issue erty linkages in each of the three countries has of special relevance in Vietnam and Laos), revealed three important shared features (Fig- and ure 5.1): • Special challenges in urban areas (Vietnam and Cambodia at the present stage). • The joint existence of a poverty and environ- ment nexus for clean water and sanitation and Access to safe water is spelled out in the national of important dif�culties in taking it fully into policy frameworks and is one of the Millennium account (in all three countries), Development Goals. Yet the gap between the • A combined issue of lack of access of the poor poor and the non-poor in accessing clean water to services and of widespread diffuse water and sanitation, and their lagging behind in the pollution (in all three countries), progress being made, remains a major environ- 178 PEN STUDY IN THE LOWER MEKONG SUB-REGION SUBREGIONAL FINDINGS mental health issue. Differences between coun- Natural Resource Use and Poverty tries are dwarfed by the shared importance of the Regarding linkages between natural resource use issue itself (Figure 5.1). and poverty, the analysis derived two important For the subregion as a whole, there is a risk of shared features in the three countries: diffuse water pollution of diverse origins. This has already materialized in Vietnam. Water pol- • Access of the poor to their resource base appears lution in upland areas originates not only from to often be a prominent issue compared to local organic sources, but also from pesticides. appropriate management of resources by indi- Water pollution in lowland areas can be a com- vidual households, and bination of organic, agricultural, and industrial • Communities’ capacity in marginal areas is pollution. The wide distribution of paddy weak; this stands out as a limiting factor for fields and water bodies in all regions, not only the implementation of potential poverty and in the lowlands, probably plays a significant environment policies. role in this problem. The poor and non-poor Access of the poor to their local natural resource share the issue of avoiding pollution when they base is restricted in some cases and totally open live in areas with polluted waters. The poor, in others. Issues of restricted access, generally to even in marginal rural areas, can also be con- forestry resources, are insuf�ciently recognized fronted with spot pollution. With high and in the subregion—in sharp contrast with other increasing inequity levels, the poor visibly find countries in Southeast Asia such as Indonesia it difficult to access the improved services being (Homes 2002). Stakeholders that compete for created. resource access include local and migrant house- The PEN research provides evidence that holds in migration regions, both of which are ethnic minority people have less access to clean possibly equally poor. They also include larger water and sanitation and demonstrates that com- operators, from small local entrepreneurs to plex demand problems are probably at play. The large international corporations. The weight of poor are at least as aware as the non-poor of the forestry sector in the three countries is a key pollution risks but, as pointed out in the case element in this context. Commercial timber is a studies, communities need to have the capacity major national commodity in Laos and Cam- and local institutions in place to access services bodian, and remains an important one in the and maintain them. Appropriate information southern two-thirds of Vietnam. This competi- for individuals is also often missing. The lack tion is openly discussed in the media in Cambo- of appropriate technology is also a problem, dia, but not in Laos and Vietnam. Dif�culties especially in Cambodia’s drought-affected pla- with open paddy fields in UXO-affected areas teaus and in the case of Vietnam’s polluted dug in Laos are another type of access restriction. wells. Examples of open access include most NTFPs. Inequitable access to safe working and liv- In Cambodia, most natural resources are under ing environments is high in the growing urban open access, unless a concession has been set up. areas in the subregion, as well as in Vietnam’s When the poor do not have suf�cient access to industrial areas. Communities do not have for- an environmental resource that is critical to their mal institutional capacity to manage local issues. livelihood, they generally turn to another, less- Information on workers’ rights and residents’ restricted resource that can generate negative rights is weak. Improved management systems, environmental impact while providing limited such as urban land use planning, are just start- livelihood improvements. ing to be put in place in Vietnam through the In situations of open access, the local tradi- land law. tional community management modes may have PEN STUDY IN THE LOWER MEKONG SUB-REGION 179 SUBREGIONAL FINDINGS allowed sustainable resource management in the located in the poor areas that link southern Laos past, but they are often disrupted. Population to southern Vietnam (Lao UXO program 2003). growth, migration, and the arrival of market The management modes of natural resources, demand are elements that combine to create dis- from the commercial logging sector to subsistence ruption. In situations of restricted access, weaker agriculture, are in a transition stage in the three communities may not be in a position to access the countries. This rapid change further weakens poor information they need to improve the situation. communities. In Vietnam, state interventions in Ethnic minority communities may have retained commercial logging are gradually replaced by cor- strong management systems in some instances, porate operators, where national and international but they are also more likely to have encountered corporations are taking over from state forest disruption. Communities’ capacity is also a criti- enterprises. In Cambodia, the concession system is cal element in the important and widely discussed being replaced with logging contracts. Farmers’ issue of balance between the need to build a rural subsistence land use systems are changing. The road network and forest conservation. The case most noticeable change relates to traditional rota- studies in Laos have documented communities’ tional agriculture (or shifting cultivation, Box 5.1). desire and con�dence in their ability to manage Rotational agriculture is gradually becoming more their own resources in a sustainable manner after intensive, a change that mostly generates benign or a road is opened, but dif�culties are likely to occur even positive environmental impacts. Paths of unless communities’ legal capacity in resource change are identical, but stages of progress are con- management is established and taken into account. trasted, Vietnam being the most advanced. Disruption of community capacity is partly a legacy of war. War legacy is one of the reasons Shared Geographical Features behind active rural-to-rural migration in Cam- that Shape Poverty and bodia, and is still visible after 30 years in the mar- Environment Linkages ginal areas of Laos and Cambodia. Land use Linkages between environmental health or nat- systems that may appear traditional have often ural resource use and poverty in the three coun- already been through several steps of adaptation. tries occur in a geographical context that displays UXO-contaminated areas are among the poorest three further and important shared features: rural areas in Laos and in Cambodia, while Agent Orange residues, which are believed to continue • A dichotomy between lowlands, where urban to generate serious heath impacts, are clearly centers and industrial growth are located, BOX 5.1 Rotational Agriculture, Poverty, and Environment Poverty and environment linkages have been repeatedly described as a downward spiral between environmental degradation and rural poverty. The transition of rotational agriculture in Laos and Vietnam is illustrative. As population pressure increases and the desire of more educated households for less strenuous farming tasks rises, fallow periods become shorter until they eventually disappear. The transition is technically dif�cult, especially in weed control. Declining yields may occur in the process. Interaction with government policies or with other stakeholders competing for the same land resources often makes things worse. When farmers are asked to rapidly reduce their rotational agriculture area, either fallow cycles are reduced too quickly to allow for adjustment, or subsistence farmers turn to opening other plots of land “behind the hill.� Soil erosion in upland areas remains an issue, but rotational cultivation is probably not the main culprit. River basin topography—especially for river basins flowing into Vietnam from the Himalayan plateau—as well as commercial agriculture and non-agricultural activities such as mining all play a role in soil erosion. 180 PEN STUDY IN THE LOWER MEKONG SUB-REGION SUBREGIONAL FINDINGS and the marginal rural areas that are lagging FIGURE 5.2 Upland and Lowland in Poverty behind; & Environment Issues in the • Legal communities that are composed of sev- Three Countries eral smaller village communities; and Qualitative assessment of magnitude from 1 = low to 4 = very high • The importance of rural-to-rural migration. The upland / lowland structure is present in each 4 of the three countries. Vietnam and Laos both dis- Poverty 3 play a strong contrast between their lowlands and incidence 2 their uplands, which by and large form the Anna- 1 mite Cordillera (Figure 5.2). The poor on both sides of the mountain range account for the largest 4 share of the rural population and live in a region Number of 3 rural poor 2 where national forest resources are located. Most 1 ethnic minority groups in the Cordillera are dis- tributed across the two countries. Yet an impor- tant difference between Vietnam and Laos is the Importance of 4 natural resources 3 national distribution of poverty. In Laos, which & poverty issues 2 has low population density throughout the coun- 1 try, the upland poor account for a substantial share of poverty in the country. The highest absolute Importance of 4 numbers of poor in Vietnam reside in the densely environmental 3 populated lowlands. Cambodia also has higher health 2 poverty in the plateau and mountain regions, & poverty issues 1 where there is a high concentration of ethnic Up- Low- Up- Low- Up- Low- minorities—but the overall land area and popula- land land land land land land tion within upland areas is much smaller, so that Laos Vietnam Cambodia the overwhelming majority of the poor reside in the lowlands, many of them in villages within communes on marginal plateau soils. Numbers of urban centers are not speci�c to the subregion but rural poor are also high in the floodable areas. are nevertheless an important feature. Migration Since 60 percent of Cambodia is under forest, no into rural areas, which are often as poor as the correlation is visible between rural poverty-trap regions of origin, is creating pressure and distur- areas and the location of forest resources. bance on natural resources. In spite of their differ- Communes in Cambodia and Vietnam are ent political systems, rural-to-rural migration is a composed of villages that tend to have unequal combination of government-sponsored programs endowments and therefore contrasting poverty and spontaneous migration in each of the three levels. (The commune is the legal institution, countries. Population migrates from resource- while the village is the human community.) Small poor areas to what are perceived to be resource- village communities in Laos are in the process of rich areas.4 Spontaneous migration is especially being merged into larger villages to obtain legal high in Cambodia, a combination of the post-war status. phase, where many still have not found a stable Migration is modifying these geographical pat- community to settle in, and of traditionally sea- terns in an important manner. The increasing sonal migration patterns. The government is in the environmental health issues speci�c to growing process of creating “social concessions.� Some of PEN STUDY IN THE LOWER MEKONG SUB-REGION 181 SUBREGIONAL FINDINGS the land will be provided to local residents, and the environmental problems they face. Although some may be provided to migrants. In Laos, the the need for increased community capacity has village consolidation and focal zone policies, achieved high visibility in decentralization and which are at the center of many debates among the poverty reduction policies and programs, its donor commodity and have been discussed in the importance for the environment is often less rec- country chapter, actively promote rural-to-rural ognized. There is also a need for better recogni- migration. Migration from the lowlands to the tion that the lack of appropriate and sustainable uplands in Vietnam, a sensitive issue that is often management systems, be it in terms of technol- not discussed openly, is also a mix of spontaneous ogy or in terms of institutional arrangements, trends and government programs. may continue to limit poverty reduction and Many of the correlations observed between environmental improvements. natural resource use and poverty in environmen- Four important policy implications derive tally fragile areas retain an unclear link to causal- from the combination of these limiting factors ity because of these two geographical factors. and are shared by the three countries: Does poverty play an important role in resource decline or degradation? Or are the poor affected • “Indirect� policies and programs focusing on by resource degradation? Are they at risk of nat- improvements in the four main PEN condi- ural disasters that can be largely unrelated to nat- tions are as needed as direct interventions on ural resource use? All these remain equally valid environmental issues; questions. • Poverty and environment interventions in the subregion need to fully take account of chal- lenges in relation to local institutions, espe- Shared Policy Options cially at community level; When poverty issues and environment issues are • Appropriate pacing of change is a key element observed to take place together and to have pos- in policy making; and sible causal links in Laos, Cambodia, or Vietnam, • International donors have a role to play to a combination of a small number of limiting fac- ensure that these are fully incorporated in tors is generally observed. These factors—access national policies and programs. to resources, information and education, manage- ment systems and community capacity—create Several donor agencies are in the process of sup- together a strong obstacle to sustainable devel- porting poverty reduction programs that fully opment. While they may be observed in many incorporate environment. While design proce- countries, what makes them so important in Laos, dures in international programs have generally Vietnam, and Cambodia is the contrast between led to screen environmentally friendly approaches their visibility and their low-key status in national to poverty reduction and to avoid negative social environmental policies and donor-supported pro- impacts in environmental programs, they now grams. Each of the ten PEN studies has yielded want to go one step further. However, direct strong evidence that two, three, or even four of investments in environmental improvements these factors were at play in poverty and envi- are unlikely to succeed unless improvements in ronment linkages, while at least one was not fully access, information, and education, as well as taken into account in the policy framework: capacity and management systems are facilitated Despite broad availability of information at the same time. Indirect interventions on these through the media and improvements in educa- limiting factors should also be considered. Per- tion levels, the poor continue to face more dif�- haps one important value of donor intervention culties in coping with and �nding solutions to is to facilitate cross-sector interventions. Just as 182 PEN STUDY IN THE LOWER MEKONG SUB-REGION SUBREGIONAL FINDINGS FIGURE 5.3 The Four Main Conditions for Effectiveness of PEN Interventions in Cambodia, Lao PDR and Vietnam sustainable development social growth dimensions equity poverty reduction environmental dimensions environ- mental appropriate health information access community management natural capacity systems resource use 4 conditions for effectiveness of PEN interventions limiting factors for poverty and environment are dia and Vietnam, and justifies targeting at that not within a speci�c sector, cross-sector solutions level. However, prioritizing poorer, often environ- are generally needed to generate win-win poverty mentally fragile villages is no easy task. The com- and environment solutions. Government inter- munes need to build capacity to identify priorities ventions are often sector interventions. Geo- and ensure they are taken into account. Migration graphically targeted polices in the uplands of can be so high that it may make geographical tar- Vietnam and Laos are designed to be cross-sector geting less relevant. Disruption from migration interventions, but they would gain from fully can be so substantial that migration will eventually incorporating issues of environmental health and fail to achieve poverty reduction outcomes, or to sustainable natural resource use. preserve the environment, or both. Community capacity is a legal and institutional Rapid change is taking place in the subregion issue compounded with social and cultural fea- in terms of economic growth, access to roads, com- tures. The two-tiered system of formal and infor- munication technologies and markets. Timing mal communities makes community capacity might be the key to successful policies and inter- building an even more complex issue. So does ventions. Progress in environmental health, espe- migration. Migrating communities have to rebuild cially in clean water and sanitation, is slow, while capacity and need time to do so. Targeting poorer much more proactive interventions could take villages is an option worth considering in Cambo- place. Conversely, national policy frameworks are PEN STUDY IN THE LOWER MEKONG SUB-REGION 183 SUBREGIONAL FINDINGS often designed to hasten change in natural management of water systems. Indoor air pol- resource use, while gradual, careful solutions are lution deserves inclusion in the messages. And needed to ensure success in terms of poverty community capacity-building obviously needs reduction. to include training of local cadres. Policy options in environmental Policy options in natural resource health and poverty use and poverty • Environmental health policies would be best Ensuring more equitable access of the poor to structured in the three countries as a dual set the local resource base or helping them break out of programs, one for marginal areas and one of traditional low-return unsustainable activities for urban areas. Marginal rural areas deserve are two policy options that have the potential to an intensi�ed initiative to provide clean water be successful poverty and environment strategies and sanitation access to the poor. Social dispar- but will require long-term efforts in the three ity in urban areas deserves attention in the three countries. Increasing access needs to go hand-in- larger urban areas in Vietnam and Cambodia. hand with the design of new community manage- Other urban areas will gain from addressing ment systems. Sustaining these systems when these problems early on. stakeholder relationships and the market environ- • Pro-poor improvements in environmental ment are changing quickly will require enhanced health in urban and industrial areas require capacity at the community level. Promoting legal framework improvements, together with non-traditional activities requires access to mar- information and awareness programs, real ini- kets and services in the marginal areas and is tiation of land use regulations, and institutional unlikely to be pro-poor, unless special attention reforms giving more capacity to the affected is paid to giving the poor access to these activi- communities. These programs are best struc- ties. And non-traditional activities may be more tured when they focus both on target residents harmful than subsistence agriculture to the envi- and their living neighborhoods and workers ronment unless new management modes includ- and their work environment. A number of ing appropriate technology are in place. pro-poor urban development programs are in Geographically targeted programs that incor- porate attention to these issues are already in place place and deserve expansion. in Vietnam and Laos, but there is a need for bet- • Programs should gain from being joint initia- ter recognition that these marginal areas deserve tives among the agencies building infrastruc- long-term support policies. It is this long-term ture, the agencies in charge of sanitation dimension, provided it is recognized, that allows development, of hygiene information and, policies and programs to focus on indirect inter- more generally, of information (including ventions to manage the environment better and the media), and—in Laos and Vietnam—the not on direct environmental interventions. agency in charge of ethnic issues. Such joint In this context of a need for sustained support efforts are required to trigger real change in to the marginal areas in the longer term, three the poverty trap areas. shorter-term policy options derive from the PEN • Information and education would greatly research: bene�t from careful tailoring to the speci�ci- ties of ethnic groups. Innovative information • Improvements in access and community capac- messages have potential to trigger local inter- ity can best be achieved if policies do not create est in accessing improved services, to facilitate additional disruption by promoting rapid attitude change toward clean water and sani- change in local land use systems. National pol- tation, and to raise interest in community icy frameworks, in sharp contrast with the 184 PEN STUDY IN THE LOWER MEKONG SUB-REGION SUBREGIONAL FINDINGS recent past, are starting to take this need for exchange programs tend to relate either to eco- slow adjustment into account and they deserve nomic development in general or to the environ- support in this process. In Lao PDR, the shift- ment alone. ing cultivation policy is formulated today as stabilization. In Vietnam, allocation of forest Poverty and environment issues may land for rotational agriculture has been legal- remain dif�cult to address through joint ized and allocation of forest use rights to com- subregional projects or policies munities has started. Conversely, migration There are two reasons for taking a cautious look policies are still viewed as a priority option in at subregional initiatives. First, there are remain- the subregion and have the potential to gener- ing security concerns in the marginal areas that ate poverty, environmental damage, or both. are also often border areas. These concerns can- • Addressing growing and emerging issues com- not be taken lightly. For example, the southern pared phasing-out issues will increase the effec- ethnic groups in Laos are neighbors to the Cen- tiveness of interventions. Rotational agriculture tral Highland groups in Vietnam. The Central is a primary example of a phasing-out issue in Highlands are regarded in Vietnam as a region the subregion. Improved management of mar- that has been recently facing unrest. keted NTFPs under strong market demand is Second, joint projects may not be feasible a growing issue. So is the need for increased when the legal and regulatory frameworks between transparency about stakeholders use of natural two countries differ markedly. Transboundary resources. protected areas, for example, are an effective and • Innovative awareness programs targeting poor necessary approach in terms of environment, but communities in environmentally fragile areas taking poverty reduction into account in these are, just as for environmental health, an impor- initiatives is far from obvious. Integrated protec- tant “forgotten� option. Existing programs have tion and development projects have proven to be long been sending simple messages relating to an appropriate approach on the Vietnamese side forest conservation and to reduction of unsus- of the Laos-Vietnam border (for example World tainable slope cultivation. These messages are Bank 1997, WWF 2005). In Laos, such projects obsolete, and national governments can bene- might still provide some encouragement for �t from introducing modern environmental those who would hope to bene�t from a devel- education, taking into account a broader array opment intervention to migrate into the pro- of issues. tected area. Similarly, in urban areas, Vietnam, Laos, or Cambodia are likely to have very differ- 5.2. Potential Subregional Initiatives ent answers in terms of urban land use planning, and lessons learned are more likely to be a valid Laos, Vietnam, and Cambodia: comparison �eld for exchange of experience than for joint and exchange of experience on poverty action. and environment Subregional exchanges should primarily Many of the poverty and environment issues are involve local governments and experts and shared between two countries, while the third should focus on the production of information faces a contrasting situation. Seven �elds stand and capacity building tools and on monitor- out as opportunities for subregional exchanges ing. Ethnicity should be taken into account (Table 5.1). Subregional exchanges focusing on when designing these tools. linkages between poverty and environment largely Country-to-country exchanges are also already remain an innovative approach, especially for taking place between local governments across rural areas and border areas. Country-to-country borders, but they tend to address immediate PEN STUDY IN THE LOWER MEKONG SUB-REGION 185 SUBREGIONAL FINDINGS TABLE 5.1 Opportunities for Subregional Exchanges Subregional poverty and Source of relevant environment linkage experience Scope of relevance Environmental Health Linkage 1. Water Access Infrastructure and Primary: Vietnam and and Poverty for Ethnic Minority formal community Lao PDR Communities capacity building in Secondary: Cambodia Vietnam Informal community management system in Lao PDR Linkage 2. Diffuse Pol- Diffuse Pollution in 3 countries lution in Wells Vietnam Technical solutions in 3 countries Linkage 3. Low-income Existing programs (Asia 3 countries groups in urban areas urbs and Citynet 2001) and projects Natural Resource Use Linkage 4. Natural dis- Pro-poor flood response Primary: Vietnam and and Poverty aster preparedness in Cambodia Cambodia Pro-poor typhoon Secondary: Lao PDR response in Vietnam Linkage 5. Livelihood Pilot projects in 3 countries improvement in UXO 3 countries affected communities Cross-sector Linkage 6. Targeted 135 program in Viet- Primary: Laos investment programs nam and related Secondary: Cambodia in upland areas donor-assisted projects Linkage 7. Environmen- Agenda 21 in Vietnam Primary: Vietnam and tal awareness pro- Miscellaneous proj- Laos grams in poor areas ect experience in Secondary: Cambodia 3 countries issues of illegal activities or to simply promote problem areas. There will be much added value in economic relations. There is potential to expand terms of environmental information and educa- the scope of these exchanges to cover priority tion if these tools take ethnicity into account. In issues among the seven subregional poverty and the past, communication between ethnic groups environment issues (Box 5.1). Researchers and living in different states has been—and often technical assistants often specialize in one of the remains today—a sensitive issue. Today, with three countries. Subregional exchanges would the development of IT, appropriate information allow them to learn from comparative analysis on environmental health and natural resource and to facilitate exchanges about lessons learned use can be best packaged for ethnic minorities at between the countries. the subregion level, while information delivery Researchers and technical assistants engaging could remain at the national level and below. in subregional exchanges would be in a position The PEN research has documented how to develop tools in the form of public information analysis is limited by the lack of consistent data. packages and media products to support national Several key poverty and environment indica- programs in priority poverty and environment tors, such as access to clean water and for forest 186 PEN STUDY IN THE LOWER MEKONG SUB-REGION SUBREGIONAL FINDINGS MAP 5.1 Poverty Yunnan province China Myanmar Transportation and trade GMS main transport routes Pilot open border Vientiane points (2005) under GMS transport Agreement Savanna- Other border crossingpoints (to khet be completed) Danang Examples of flow of Thailand timber, NTFPs, maize Major urban center (to add) Protected areas Bangkok Transboundary (existing) Transboundary (planned) Phnom Penh National PAs nearby main routes HCMC Sources : ADB 2005 ICEM 2004 PEN field observations PEN STUDY IN THE LOWER MEKONG SUB-REGION 187 SUBREGIONAL FINDINGS BOX 5.2 Border Features of Poverty and Environment Linkages Border areas have a special role to play in subregional initiatives for poverty and environment. The vast majority of border areas between Vietnam, Laos, and Cambodia are marginal, fragile areas with high poverty incidence. Forty-six priority poor districts in Laos—and many of Vietnam’s districts with a high share of communes listed as remote and poor—are upland border districts (Map 5.1). Border regions had long specialized—perhaps as many marginal areas in the world—in trade that ranged from informal to illegal, but this is changing as the opening of the border between the two countries is making rapid progress5—a remarkable change compared to recent decades. A shift from informal to more formal trade is taking place. This shift originates not only from the opening of borders but also from the use of information technology. Increasingly comprehensive regulatory frameworks in Thailand and China are also encouraging informal traders to turn to for- mal transactions. Opening borders in poor areas is a coin with two sides for the environment and for poverty. Legal trade can help reduce environmentally adverse illegal activities as local economic operators may �nd larger formal transactions more pro�table than smaller informal ones, provided interna- tional trade does not bypass the local operators of border trade. There is, however, a risk of larger-scale environmental threats from international trade with limited poverty reduction gains. From a biodiversity perspective, trade in plant and animal products has long been an area of con- cern for environmental groups. This focus rightly deserves to be continued, but legal trade requires as much attention as the persistent illegal activities. Examples of poverty and environment related to legal or informal trade: • Logging for export to China and Vietnam increase as stronger environmental regulations come into operation in these countries. • Active NTFP exports to China and Vietnam are depleting resources for some species, even when trade is legal. Quantities and prices often remain in the grey area of informal trade, and illegal exports of species with biodiversity value complement legal ones (PEN study on NTFPs in Laos). • The hill closure policy in China and the development of demand for animal feed in the subre- gion trigger rapid development of hybrid maize cultivation on sloped land in Laos. This often generates forest losses with limited income gains (Prodessa 2002). • Pesticides are often purchased by farmers directly on border markets in Northern Laos in a total absence of agricultural extension services (Ducourtieux 2004). • In Laos, roads built by enterprises outside of road network development plans can lead to for- est depletion with limited improvements in accessibility for local people (PEN study on roads). Local governments in marginal border areas are confronted with the issue of making improved decisions in terms of environment and poverty as trade is increasing. The transition stage between informal and formal exchanges creates speci�c challenges. Increased transactions are taking place while local government capacity, be it in controlling environmental impact or in ensuring poverty reduction gains, is limited. Two types of subregional initiatives can support this process: • Local-government-to-local-government exchanges with a focus on the subregional poverty and environment issues; • Design of information and training materials targeting local governments. cover, happen to be MDGs. With a joint effort ment. The subregion appears to be an ideal to use a shared methodology when assembling space for the collection and use of a consistent these indicators, the governments and the set of poverty and environment indicators. The donor agencies that support their poverty district level (except in Cambodia), the com- reduction efforts would have a solid basis for mune level, and the village level (for village making investment and policy decisions that samples) have been identified as important tar- fully incorporate both poverty and environ- geting levels for PEN analysis. A monitoring 188 PEN STUDY IN THE LOWER MEKONG SUB-REGION SUBREGIONAL FINDINGS program collecting MDGs at these levels on a Thailand, and the Socialist Republic of Viet Nam for the regular basis would be a cost-effective option. Facilitation of Cross-Border Transport of Goods and Peo- ple. February 2005. These priorities for exchanges and monitoring Asia-Urbs and Citynet. 2001. Cities, Poverty and the Environ- fully justify a third phase of the PEN study. Phase ment. Report and Proceedings from a Regional Workshop, I consisted of an extensive review of available geo- Hanoi, July 30–August 2, 2001. Citynet (Regional Net- graphically referenced data and an analysis of cor- work of Local Authorities for the Management of Human Settlements) and European Communities relations between poverty and environment. Phase (Europeaid Asia-Urbs Program). II turned to researching linkages and poverty Project for the Development of Sayaboury. (PRODESSA) and environment in terms of policy implications. 2002. “Improving the Quality and Marketing Channels A third phase could usefully focus on supporting of Maize and Other Crop Products in Southern Sayaboury Region.� Report for the French Committee exchanges as well as monitoring poverty and envi- for Co-operation with Laos. Paris: The French Devel- ronment issues and progress. opment Agency. Ducourtieux, Olivier. 2004. “Shifting Cultivation and Endnotes Poverty: a Complex Issue.� In: “Poverty Reduction and Shifting Cultivation in the Uplands of Lao PDR, Tech- 1. Vietnam, Laos, and Cambodia, which together formed nologies, Approaches and Methods for Improving French Indochina until 1954, also take part in the inter- Upland Livelihoods.� Proceedings of a Workshop held national cultural cooperation forum related to the French in Luang Prabang, January 27–30, 2004. language. Homes, Derek. 2002. “Indonesia: Where Have All the 2. Vietnam joined ASEAN in 1995, Lao PDR joined in Forests Gone?� Environment and Social Development 1997, and Cambodia in 1999. East Asia and Paci�c Region Discussion Paper. Wash- 3. Several poverty and environment linkages that have not ington: The World Bank. been analyzed in depth in one of the three countries are not International Center for Environmental Management included in the �gure but are however viewed as highly rel- (ICEM). 2004. “Protected Areas and Development in evant. These include indoor air pollution and malaria. the Lower Mekong River Region.� Queensland, Australia: 4. The subregion has a centuries-old history of migration ICEM. of ethnic groups from the North to the South. Lao UXO program. 2003. “Map of UXO Impact and Bomb- 5. Under the Greater Mekong Subregion Cross-border ing Data 1965–75.� In: “Claiming the future, the impact Transport Agreement that was initiated in 1999, major of UXO and Landmines in Lao PDR.� Herberton, Aus- roads will soon link major urban areas in the three coun- tralia: Australian Network of the International Campaign tries and in Thailand and China (Map 5.1). Full imple- to Ban Landmines. mentation of the agreement is expected for 2007–08. Five Mines Action Group (MAG). 2005. Laos Report 2005. pilot border points are fully open since 2005 (ADB 2005). London: MAG. Mekong River Commission (MRC). 2003 Social Atlas of the Lower Mekong Basin. Bangkok: MRC. References World Bank. 1997. “Vietnam Forest Protection and Rural Asian Development Bank (ADB). 2005. Agreement between development Project.� Staff Appraisal Report. Wash- and among the Governments of the Kingdom of Cambodia, ington: The World Bank. the People’s Republic of China, the Lao People’s Democra- World Wide Fund (WWF). 2005. Greater Truong Son Con- tic Republic, the Union of Myanmar, the Kingdom of servation Action Plan. City: publisher. PEN STUDY IN THE LOWER MEKONG SUB-REGION 189 Environmental and Social Development Unit East Asia and Paci�c Region The World Bank in Lao PDR THE WORLD BANK Patou Xay - Nehru Road 1818 H Street, N.W. P.O. Box 345 code 01004 Washington, D.C. 20433 Vientiane, Lao PDR Telephone: 1-202-473-1000 Telephone: 856-21-414-209 www.worldbank.org/eapenvironment www.worldbank.org/lao The World Bank in Cambodia The World Bank in Vietnam 113 Norodom Boulevard 63 Ly Thai To Street Phnom Penh, Cambodia Hanoi, Vietnam Telephone: 855-23-217-304 Telephone: 84-4-934-6600 www.worldbank.org/cambodia www.worldbank.org/vietnam