STRIKING A BALANCE MANAGING EL NIÑO AND LA NIÑA IN CAMBODIA’S AGRICULTURE William R. Sutton, Jitendra P. Srivastava, Jawoo Koo, Ioannis Vasileiou, and Angga Pradesha Striking a Balance Managing El Niño and La Niña in Cambodia’s Agriculture William R. Sutton, Jitendra P. Srivastava, Jawoo Koo, Ioannis Vasileiou, and Angga Pradesh © 2019 International Bank for Reconstruction and Development/The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the govern- ments they represent. 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. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distrib- ute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: William R. Sutton, Jitendra P. Srivastava, Jawoo Koo, Ioannis Vasileiou, and Angga Pradesh. 2019. Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture. RepNo 132065. World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This trans- lation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover photo: © Chor Sokunthea/World Bank. Used with permission. Further permission required for reuse. Cover design: Shepherd, Inc. Photo credits: All photos used with permission. Further permission required for reuse. page xxii, 14, 52, 68, © Chor Sokunthea/ World Bank; page 44, © Masaru Goto/World Bank. Contents Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Roadmap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. ENSO Affects Cambodia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3. ENSO Affects Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 4. ENSO Contributes to Economic Impacts . . . . . . . . . . . . . . . . . . . . . . . . 23 5. ENSO Contributes to Social Impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 6. Cambodia Has Taken Actions Related to ENSO Preparedness . . . . . . 33 ENSO responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 State-run institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Policies and strategic plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Interventions by development partners . . . . . . . . . . . . . . . . . . . . . . 38 7. There Are Still Areas to Improve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 8. Policy Interventions Help Negate ENSO-Related Losses . . . . . . . . . . . 45 9. The Government Can Take Action to Improve ENSO Preparedness and Resilience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Preparedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Resilience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Annexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Annex 1: Methodological specifics . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Annex 2: National plans and agency roles related to climate . . . . . 73   iii iv Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Annex 3: Government initiatives for agriculture and farming households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Annex 4: Summary of El Niño–Southern Oscillation (ENSO) forecasts made available by technical agencies . . . . . . . . . . . . . . . . 81 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Figures Figure A: Regional Rainfall Differences between ENSO and non-ENSO Years in Each Quarter from 1980–2015. . . . . . . . . . . . . . xiv Figure 1: Integrated Analytical Framework. . . . . . . . . . . . . . . . . . . . . . . . . 4 Figure 2: Monthly Average of Rainfall Amount in ENSO Phases During 1980–2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Figure 3: Regional Rainfall Differences between El Niño and Neutral Phase in Each Quarter from 1980–2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Figure 4: Seawater Temperature Changes Indicating the Occurrence of El Niño and La Niña. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 5: Average Rainfall Differences between El Niño and Neutral Years. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Figure 6: Rainfall Comparison between 2015–2016 and the 31-year Annual Mean in the Plains (lowland area). . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 7: Agricultural Production as a Share of Total Value (2014–16) in Constant Dollars (2004–2006). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Figure 8: Agricultural Exports as a Share of Total Value (2014–16), in Constant Dollars (2004–06). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Figure 9: Annual Average Rice Production Changes from the Average of Previous Three Years, between 1995 and 2014. . . . . . . . . . . 18 Figure 10: Simulated Changes in Beans, Maize, Rice, and Vegetable Yields During El Niño and La Niña Years Compared to Average Years. . 18 Figure 11: Rice Cropping Calendar and Rice Agroecosystem in Cambodia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Figure 12: Rice Yields in Cambodia, 1961–2016 (kg/ha). . . . . . . . . . . . . . 20 Figure 13: Growth and Variability of Agricultural GDP in Cambodia. . . 24 Figure 14: GDP Losses During Typical El Niño Events ($ millions or percentage reductions). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 15: GDP Gains During La Niña Events ($ millions or percentage increases). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 16: Real Food and Agricultural Price Changes During ENSO Events (percentage). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Contents v Figure 17: Chronology of Notable Actions Taken in Response to the 2014–16 El Niño. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Figure 18: Household Consumption Losses During El Niño Events and Intervention Scenarios (gains if negative). . . . . . . . . . . . . . . . . . . . . 47 Figure 19: Household Consumption Losses by Expenditure Quintile with and without All Interventions Combined (gains if negative). Q1 Is the Poorest Quintile, Q5 Is the Wealthiest Quintile. . . . . . . . . . . . 48 Figure 20: Changes in National Poverty Headcount Rate and the Number of Poor People During Strong El Niño Events and Intervention Scenarios (percentage points and millions of people). . . . 49 Figure 21: Changes in Poverty Headcount Rate During Strong El Niño Events and Intervention Scenarios (percentage points). . . . . . . 50 Figure 22: Change in Poverty Headcount Rates During Strong El Niño Events and Intervention Scenarios by Location and Gender of the Household Head (percentage points). . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Figure 23: Timeline of Cambodian Legal Framework for Disaster Risk Reduction and Climate Change Adaptation. . . . . . . . . . . . . . . . . . . 76 Figure 24: MOWRAM’s Weather Forecast Products for the Fishery and Transportation Sectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Tables Table A:  Summary of Recommendations and Proposed Actions. . . . . . . xx Table 1: El Niño Events and Droughts in Cambodia, 1982–2016. . . . . . . 12 Table 2: National Economic Structure, 2014. . . . . . . . . . . . . . . . . . . . . . . 23 Table 3: Agriculture Food System GDP and Employment, 2014. . . . . . . . 25 Table 4: GDP Changes During Typical ENSO Events. . . . . . . . . . . . . . . . . 26 Table 5: Household Income and Consumption, 2014. . . . . . . . . . . . . . . . 30 Table 6: HRF Sector Coordinates Agency Activities in Priority Sectors. . . 39 Table 7: GDP Changes During El Niño Events and Intervention Scenarios. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 8: Rural and Urban Household Consumption Changes During El Niño Events and Intervention Scenarios (percentages). . . . . . . . . . . . 49 Table 9: Recommendations and Proposed Actions to Build pre-ENSO Preparedness and Resilience. . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Maps Map A: Cropland Areas of Cambodia. . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv Map 1: Cambodia’s Four Major Regions. . . . . . . . . . . . . . . . . . . . . . . . . . . 8 vi Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Map 2: Annual Rainfall (millimeters). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Map 3: Most Drought Vulnerable Provinces in Cambodia. . . . . . . . . . . . . 9 Map 4: Crop Land Use Patterns in Cambodia, 2015. . . . . . . . . . . . . . . . . 17 Map 5: Forest Fires in Cambodia During the 2016 Drought. . . . . . . . . . . 22 Boxes Box 1: Methodological Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Box 2: Distinguishing ENSO from Climate Change. . . . . . . . . . . . . . . . . . . 5 Box 3: Experience from the 2014–2016 El Niño. . . . . . . . . . . . . . . . . . . . 11 Box 4: Early Warning System 1294. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Box 5: Roadmap to Address the Impacts of El Niño (RAIN) in the Philippines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Box 6: Rainwater Harvesting in Brazil. . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Box 7: Turning Forecasting into Action: South Korea’s Anticipatory Response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Box 8: The Drought Tolerant Maize for Africa (DTMA) Project. . . . . . . . 57 Box 9: China’s Grain Reserve System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Box 10: Ethiopia’s Productive Safety Net Program Integrated with Early Warning Systems and Disaster Risk Management. . . . . . . . . 60 Box 11: Integrating Disaster Risk Management and Climate Change Adaptation into the Social Economic Development Plans for the Tra Vinh Province in Vietnam. . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Box 12: Challenges Associated with Index-based Agricultural Insurance in India. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Box 13: Alternative Wetting and Drying (AWD) in Vietnam and Bangladesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Abbreviations $ United States dollar AFS Agriculture Food System ASEAN The Association of Southeast Asian Nations CEDAC Cambodian Centre for Study and Development in Agriculture CHF Cambodia Humanitarian Forum CBDRM Community-based disaster risk management CGE Computable general equilibrium CRP CCAFS CGIAR Research Program on Climate Change, Agriculture and Food Security ECMWF European Centre for Medium-Range Weather Forecasts ENSO El Niño–Southern Oscillation ESCAP  United Nations Economic and Social Commission for Asia and the Pacific EWS Early warning system GDP Gross domestic product ha Hectare HRF Humanitarian Response Forum IFPRI International Food Policy Research Institute kg Kilograms km Kilometers m3 Cubic meter MAFF Ministry of Agriculture, Forestry and Fisheries MoE Ministry of Environment MOWRAM Ministry of Water Resources and Meteorology MRC Mekong River Commission NAP-DRR National Action Plan for Disaster Risk Reduction NAPA National Adaptation Program of Action to Climate Change NCDM National Committee for Disaster Management   vii viii Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture NIDM National Institute of Disaster Management NSDP National Strategic Development Plans NGO Nongovernmental organization ODA Overseas Development Assistance ONI Oceanic Niño Index SAM Social Accounting Matrix SEA-DRIF South East Asia flood monitoring and risk assessment SNAP Strategic National Action Plan on Disaster Risk Reduction THI Temperature Humidity Index UNDP United Nations Development Program UN-ESCAP United Nations Economic and Social Commission for Asia and Pacific UN-FAO United Nations Food and Agriculture Organization USAID United States Agency for International Development WMO World Meteorological Organization Foreword In a world of climate change and headline grabbing cyclones, El Niño is one of the most unspoken climate risks in East Asia and the Pacific. It is a cyclical event that consis- tently ravages the region’s economies and agricultural sectors with droughts and water scarcity. In turn, La Niña, the cool phase which typically occurs the year after an El Niño event, often brings extensive damage from floods and heavy rainfall. El Niño has occurred eight times since 1980, with the most recent event, from 2014 to 2016, being the most severe, causing billions of dollars in damage to the region. Experts forecast another El Niño event, predicted to affect East Asia and the Pacific in the winter of 2018–2019. Given the cyclical nature of the El Niño–Southern Oscillation (ENSO) events, it is crit- ical that governments have plans in place to face the threat. The research presented here is the first to carry out in-depth economic modeling to calculate changes in agricultural production, gross domestic product, household welfare, and poverty levels from both El Niño and La Niña in East Asia and the Pacific. It also estimates how certain policy inter- ventions could mitigate these impacts. As such, the Striking a Balance reports could be important tools for policy makers in Cambodia, Lao PDR, Myanmar, the Philippines, and ­Vietnam—the five countries examined in the series. The reports’ findings are concerning: the authors estimate El Niño produces GDP, con- sumption, and income losses for all households, in all countries, regardless of income level, urban-rural location, or gender. El Niño threatens to raise food prices, with women and poor people set to suffer the most because they spend more of their income on food. Because of this, El Niño could threaten the region’s poverty reduction and food security advances from the past decade. Fortunately, the reports also find there may be opportunities to harness the heavier rainfall which occurs during La Niña to achieve some GDP, consumption, and poverty reduction gains.   ix x Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Regional governments have made inroads in preparing for climate events like floods and other natural disasters, but more could be done to prepare for ENSO specifically. This includes building resilience and preparedness by investing in early warning sys- tems, developing national action plans, and cooperating with other East Asia–Pacific countries on ENSO-related challenges, which are regional in nature. Striking a balance among these various policy options, and between El Niño and La Niña management, demands concerted effort. It is our hope that this report will catalyze collective action and help governments and other national and subnational stakeholders achieve that balance. Juergen Voegele Senior Director Food and Agriculture Global Practice The World Bank Acknowledgments This report was prepared by a team led by William R. Sutton, Lead Agricultural Econ- omist in the World Bank’s Agriculture Global Practice, East Asia and Pacific Region, together with Jitendra P. Srivastava, Ioannis Vasileiou, and Maximillian Ashwill, and in collaboration with a team from the International Food Policy Research Institute (IFPRI) led by Mark Rosegrant and James Thurlow. We are grateful to Nathan M. Belete, Practice Manager (GFA02), for his valuable support, guidance, and oversight throughout the course of this study. We are also grateful to Ellen Goldstein, the Coun- try Director for Myanmar, Cambodia, and Lao PDR, and Inguna Dobraja, the Country Manager for Cambodia, for their support. The World Bank team also comprised Mudita Chamroeun, Tam Thi Do, Asa Giertz, Vanna Pol, Narya Ou, Huyi Zhang, Manuela Hernandez, and Xinyu Weng. The IFPRI team also comprised Jawoo Koo, Angga Pradesha, Ricky Robertson, Rowena Valmon- te-Santos, and Leocadio Sebastian (International Rice Research Institute). The report was edited by Maximillian Ashwill. We would like to thank the peer reviewers for this study, Abigail Baca, Erick Fernandes, and David Treguer, for their valuable advice. Helpful guidance was also provided by Sudhir Shetty, Zoubida Allaoua, Marc Sadler, Steven Jaffee, Frauke Jungbluth, Paavo Eliste, Pauline Zwaans, Francesca de Nicola, and Lilanie Magdamo. From the Government of Cambodia, we are grateful for the guidance and support provided by the Ministry of Agriculture, Forestry and Fisheries, the Ministry of Envi- ronment, and the Ministry of Water Resources and Meteorology. The study greatly benefitted from the important contributions made through valuable inputs, comments, advice, and support provided by academia, research institutions, civil society, NGOs, farmers, and development partners.   xi xii Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture We gratefully acknowledge the generous financial support provided by the contributors to the Multi-Donor Global Food Price Crisis Response Trust Fund: the governments of Australia, Canada, the Republic of Korea, and Spain. The team also acknowledges the support provided by the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS). Overview The purpose of this report is to help Cambodia’s policy makers and stakehold- ers prepare for future El Niño–Southern Oscillation (ENSO) events. It does this by providing information on ENSO’s poverty, economic, and agricultural impacts in Cambodia and outlining ways forward. The report finds that ENSO’s impacts vary from region to region and harm Cambodia’s people, economy, and agricultural sector. The country has made inroads in preparing for climate events like floods and other natural disasters, but more could be done to prepare for ENSO specifically. Being proactive to prepare for ENSO in Cambodia is important because of the country’s high exposure to climate shocks, the prominence of the agricultural sector in the national economy, the large rural population and its climate vulnerability. Furthermore, it is likely that Cambodia will face another El Niño by the winter of 2018/2019. ENSO has important impacts on Cambodia’s climate, agriculture, economy, and society Cambodia is vulnerable to ENSO-related climate shocks. Between 1980 and 2015, there were eight El Niño events in Cambodia. Historical data show that ENSO’s two phases, the cool La Niña phase and the warm El Niño, see increased and decreased average rainfall, respectively (Figure A). This has led to major flooding and severe water shortages. During the 2014–2016 El Niño, Cambodia saw major droughts. By 2016, water levels in Cambodia’s rivers fell to 50 to 70 percent of the interannual aver- age.1 This made 2016 one of the driest years in the country’s history, with one of the worst droughts of the last 50 years. Droughts affected almost every province and 2.5 million people experienced water shortages.2 That said, it is flooding that causes the greatest climate-related economic damage in Cambodia.3 In 2015, ENSO was blamed 1 Open Development (2016). 2 ACAPS (2016); Chakrya and Sassoon (2016); HRF (2016); Save the Children (2016). 3 National Institute of Disaster Management (2014).   xiii xiv Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE A: Regional Rainfall Differences MAP A: Cropland Areas of Cambodia. between ENSO and non-ENSO Years in Each Quarter from 1980–2015.4 Q1 Q2 Q3 Q4 Plain El Niño Lake Coastal Mountain Plain La Niña Lake Coastal Mountain –50% 50% –50% 50% –50% 50% –50% 50% Rain diff. Rain diff. Rain diff. Rain diff. neutral (%) neutral (%) neutral (%) neutral (%) for torrential rains of 50 to 100 millimeters per hour in Phnom Penh, causing flash- floods and infrastructure damage.5 El Niño events have negative effects on agricultural production in Cambodia, while La Niña events have a slightly positive effect. Most of Cambodia’s agricul- tural production occurs in the Mekong River area in the Plains region (Map A). The lower Mekong is mainly cultivated with rice and the upper Mekong with agro-indus- trial crops, such as rubber and cassava. The Lake region focuses more on freshwater fisheries. The Coastal region includes coastal and marine fisheries and some crops. The Mountain region includes agro-industrial crops, like rubber and cassava, and for- estry. Combining all ENSO events together and de-trending historical data reveals that, on average, rice production declines by 10 percent during El Niño and increases by five percent during La Niña. From 1997 to 2002, droughts caused a 20 percent decline in rice production.6 2010’s El Niño–related droughts damaged 14,000 hect- ares of transplanted rice, 3,500 hectares of rice seedlings, and 5,500 hectares of other crops.7 Additionally, rice production may be adversely affected in flood-prone areas in La Niña years. ENSO events likely contribute to declines in Cambodia’s livestock and fishery sectors. La Niña contributes to more days with high humidity temperatures per year, which can cause heat stress for livestock. From 1951 to 2010, La Niña years have 16 more such days than neutral and El Niño years, which have about 100 such days 4 Authors’ reanalysis using UEA CRU-TS v4.0. 5 UN-ESCAP (2016). 6 NAPA (2006); Chhinh (2015).  7 Leng (2014). Overview xv per year.8 Poultry are sensitive to heat stress, with strong declines in feed intake, egg size, and egg quality.9 Swine and cattle are also sensitive to heat stress, with negative outcomes when temperatures are higher.10 Overall, cattle are the most heat tolerant, followed by poultry, then pigs, which are the least heat tolerant.11 ENSO contributions to livestock feed declines and flooding can also affect livestock health and production. Fisheries, which are central to the livelihoods of millions of Cambodians and account for about 80 percent of animal protein in household diets,12 are also impacted by ENSO events. The 2016 El Niño–related drought caused Cambodia’s fish production to decrease by 17 percent, leading to higher fish prices.13 Dry conditions and extreme heatwaves in the center of the country also killed about 65 tons of fish in Lake Tonle Chhmar.14 ENSO’s impacts on Cambodia’s agricultural sector has economy-wide ramifica- tions. Agriculture in Cambodia contributes to 30 percent of gross domestic product (GDP) and accounts for 45 percent of employment.15 This rises to 36 percent of GDP and over 50 percent of employment when one considers the entire Agricultural Food System (AFS), which includes downstream economic activities like food processing and the trading and transporting of agriculture-related products. This makes agricul- ture the country’s largest source of employment, especially for the poor. As such, any shocks from ENSO to agriculture has major impacts on welfare, food security, and national poverty levels. Cambodia experiences notable economic losses during El Niño events, but these are nearly offset by gains during La Niña events. Total economic losses from three El Niño–related historical droughts (1994, 2002, and 2005) were estimated at $138 million.16 According to simulations conducted for this report, national GDP falls by 0.4 percent during a typical El Niño event relative to a neutral non-ENSO year. Even small percentage reductions in national GDP can imply substantial monetary losses. For example, a 0.4 percent drop in national GDP is equal to $61 million in lost val- ue-added.17 Losses are most pronounced in the agriculture sector, where GDP falls by 8 Kalnay  and others (1996). 9  Lara and Rostangno (2013). 10  Wegner and others (2016). 11 Xin and Harmon (1998); Myer and Bucklin (2001).  12  Baran, Schwartz, and Kura (2009). 13 HRF (2016). 14Gaulter (2016). 15  CGE model and the 2014 SAM. 16 Miyan (2015). 17 Measured in 2014 prices. xvi Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture 1.4 percent, and in the mountainous region of the country, where crop and livestock18 GDP falls by 2.5 percent. By contrast, during a typical La Niña year, the economy expands. Simulations indicate national GDP gains of $55 million during La Niña events, which almost offset GDP losses during El Niño events. Poor rural households consume less and suffer more from ENSO events than nonpoor households. Cambodia’s rural population is 79 percent of the total popu- lation,19 high by both regional and global standards. Further, the rural poverty head- count ratio of 21 percent exceeds the urban poverty headcount ratio of only 6 percent.20 Given the large rural population, this means that poverty is rampant in Cambodia: 14 percent are poor21 while another third are “near-poor.”22 Household surveys find that rural consumption is only 38 percent of urban consumption levels, while the rural poor’s consumption is only 23 percent of urban consumption levels. Food insecurity is a major concern that ENSO threatens to make worse.23 Simulations show that both urban and rural household consumption declines during El Niño. Rural households are more likely to be farmers who benefit from higher food prices. This is not true for poor rural households who have less access to land than wealthier rural households and, therefore, earn fewer returns from farming. Poor households also tend to spend more of their incomes on food, meaning that when ENSO affects cereal production and increases prices, it has larger implications for poorer consumers. Women suffer disproportionately from El Niño. In Cambodia, just as many women work in agriculture as men.24 Women play an important role in water resource man- agement, both in the home and on the farm.25 Simulations show slightly more female- headed households than male-headed households fall into poverty during strong El Niño events. This is true in both rural areas and nationally. Also, reports show that female households tend to own smaller plots of land, are more indebted, and are more likely to be poor, compared to their male counterparts.26 Female-headed households are less likely to have access to agricultural lands than male-headed households. Moreover, 18 Crops and livestock are subsets of agriculture that exclude forestry and fisheries. 19 2017 World Bank estimates based on the United Nations Population Division's World Urbanization Prospects: 2014 Revision. Available at: https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS. 20 2012 figures: http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators. 21 2014 data; World Bank (2017). 22 2017 data; World Bank (2017). 23 2015 data from WB-WDI (2018). 24 Employment in agriculture, female (% of female employment) (modeled ILO estimate) (International Labour Organization, ILOSTAT database. Data retrieved in November 2017). 25 Chum, Tep, and Nevill (2016). 26 http://www.fao.org/3/a-i5489e.pdf. Overview xvii for every five male-headed agricultural households that receive loans, only one female- headed agricultural household receives the same.27 Cambodia prepared for ENSO, but there is room for improvement Cambodia’s government is committed to responding to weather-related shocks, but could do more to prepare for ENSO. Cambodia has contingency plans for floods and droughts, and receives overseas development assistance (ODA) to respond to natu- ral disasters. But, ENSO awareness and its effects on drought and floods remains lim- ited. Consultations with representatives of relevant ministries and other stakeholders28 reveal that government resources for climate change and natural disasters are mostly for ex post risk interventions, such as relief operations, but there are fewer resources for early warning and ex ante mitigation and preparation. This was evident during the El Niño that began in 2014. It was only in 2016 that emergency response measures where implemented, but these only helped affected regions cope and recover. They did little to prepare the country for future events. That said, in recent years the country has prioritized climate resilience and gender equity through various strategy documents and action plans. The 2014–2016 El Niño response exposed areas to strengthen in Cambodia’s ENSO preparedness efforts. These include the following: Government areas to strengthen • Many government plans and strategy documents associated with ENSO or slow-­ onset droughts have not been fully implemented, including plans to improve agri- cultural infrastructure, human capacity, and early warning systems. • There could be better cooperation on ENSO-related issues among government agencies at all levels and nongovernmental partners. There are many local and national level agencies that carry out work in climate resilience, but only one min- istry focuses specifically on ENSO. 27 http://www.fao.org/fileadmin/templates/ess/documents/meetings_and_workshops/APCAS23/ documents_OCT10/APCAS- 10-26_-Gender.pdf. 28 During the Consultative Stakeholder Workshop October 9–10, 2017. xviii Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture General data areas to strengthen • Cambodia would benefit from better data and comprehensive vulnerability maps. Although the government has collected data on drought and ENSO, there are still no ENSO vulnerability maps.29 • There is a lack of ENSO-related information in Cambodia. Most of the climate-­ related data and information that exists in Cambodia is on natural disasters and not specifically related to ENSO. This appears to have impeded the response to past ENSOs because policy makers were not adequately informed of potential ENSO events, despite information existing on when ENSOs would likely occur. General capacity areas to strengthen • Cambodia lacks sufficient human and institutional capacity to properly prepare for ENSO events. More specifically, there is not enough technical expertise to manage ENSO—or disaster-related databases and other systems that could produce timely information on impending disasters. This is especially true at the local or village level where El Niño and La Niña remain unfamiliar concepts. • It has been a challenge to scale up good practices and make them sustainable. There are many good nongovernmental organization (NGO) initiatives and internationally funded projects that could be expanded, but there are no dedicated institutions for learning from these efforts and no funding sources to scale up the best practices. • Efforts have been made by the government to integrate gender into disaster preven- tion, but more could be done, especially in terms of ENSO. This partly reflects the general lack of a codified focus on ENSO, and the general underrepresentation of women in climate and natural resource management interventions. Policy interventions help negate ENSO-related losses Introducing drought-tolerant varieties is an effective policy intervention to mit- igate GDP losses from a strong El Niño. In depth modeling carried out under this study of six policy interventions—including introducing drought-tolerant crop vari- eties, expanding irrigation, restricting rice exports, storing and distributing grains, expanding social protection coverage (social transfers), and applying all of these policy interventions simultaneously—show that on-farm investments, like drought-­ tolerant seeds and irrigation infrastructure, are the most effective interventions at 29 Based on interviews with officials in the Department of Climate Change on October 10, 2017. Overview xix limiting El Niño–related GDP losses in Cambodia. Introducing drought-tolerant vari- eties reduces national GDP losses from $61 million to $34 million. Expanding irriga- tion use is not as effective, but also significantly offsets GDP losses by making crop production more resilient to climate shocks. Overall, when all simulated interventions are combined and implemented concurrently, there are still national GDP losses equal to $31 million and food system GDP losses of $46 million during El Niño events. In contrast, distributing stored grains, restricting rice exports, and providing social transfers to poor households are less effective options for mitigating El Niño–related GDP losses. These three interventions improve consumer access to food by curbing price increases or income losses, rather than limiting declines in GDP or agricultural production. Providing social transfers and introducing drought-tolerant seed varieties are among the most effective policies to mitigate poverty and consumption losses during El Niño. Without interventions to mitigate impacts, a typical El Niño event causes the national poverty rate to increase by 1.4 percentage points, which is equiv- alent to an additional 210,000 people living below the poverty line. Simulations show that introducing drought-tolerant varieties prevents consumption losses for all house- holds and reduces the number of households that fall into poverty during El Niño. Social ­transfers—represented by cash transfers in the modeling—by contrast, do not reduce consumption declines during El Niño for all households, but drastically reduce consumption declines for poor households. Cash transfers are also effective at mitigat- ing poverty increases during El Niño. This is because cash transfers target poor house- holds but are financed by more affluent households through higher tax payments. When all policy scenarios considered for this report are implemented concurrently, poverty increases and total consumption losses are almost eliminated during El Niño. The impacts from policy interventions on poverty outcomes during El Niño vary across regions and genders. Simulations show El Niño causes poverty increases in all regions, but the most in the Lake and Mountain regions. That said, poor households in the Lake and Mountain regions benefit the most when all policies are applied. Intro- ducing drought-tolerant crop varieties benefits households in all regions, but this is especially effective in the Lake region. Irrigation, export restrictions, and stored grain distribution have little effect on poverty in the Mountain region, but these policies mit- igate poverty increases substantially in the Lake region. That said, storing and distrib- uting grains and restricting exports create market distortions that can harm producers and trading partners. Simulations also show both male- and female-headed households benefit from policy interventions, but when all policies are applied, female-headed households benefit the most. Again, introducing drought-tolerant seed varieties is the xx Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture most effective in mitigating poverty losses during El Niño. Social transfers, which directly target the poorest, who in this case are female-headed households, are also very effective. Distributing stored grains is slightly more beneficial for male-headed households. The government can take action to improve ENSO preparedness in Cambodia There are many opportunities to improve ENSO preparedness and resilience. In Table A, recommendations are divided into two groups: preparedness and resilience. Preparedness are measures specifically geared toward ENSO and should, ideally, be in place before the next ENSO event occurs. These actions will significantly empower people to cope, respond, and recover from damaging ENSO events. Resilience, by con- trast, are measures that are not specifically tailored to ENSO, but that will build individ- uals’ and organizations’ ability to adapt to multiple forms of risks and shocks without compromising long-term development. Recommendations in green are a high priority, recommendations in orange are a moderate priority. The last two columns denote which actions are short term (S), or should be completed within a year, and which actions are medium-to-long term (M/L), or would not be achievable in less than a year. TABLE A:  Summary of Recommendations and Proposed Actions. Recommendation Actions S M/L Prepare response • Establish a high-level ministerial committee tasked X measures for when with ENSO response and preparedness. ENSO-related • Develop local ENSO contingency plans and budgets. X climate events occur • Ensure emergency response inputs and supplies are available and connectivity infrastructure is X adequate. Develop a • Assign an ENSO task force to develop a X comprehensive comprehensive ENSO strategy. Preparedness ENSO framework • Carry out current action plans on disaster response X and climate resilience. • Empower communities through effective and X timely provision of services and benefits. Harness La Niña’s • Increase water-use efficiency. X X rebound • Reduce flood risk in flood-prone areas. X X Invest in forecasting • Identify forecasting needs and areas to orient the X capacity systems toward ENSO. • Develop accurate ENSO risk maps. X • Make this information accessible and X X understandable. Overview xxi Recommendation Actions S M/L Promote improved • Expand research mandates for research institutes to X seed varieties develop tolerant crop varieties. and agricultural • Improve extension services, technologies, and practices X management methods. Preparedness Improve grain • Assess the current national grain storage capacity. X storage capacity • Stock grains before the next ENSO. X X • Safeguard producers from market distortions. X X Expand and adjust • Assess constraints to expanding SSNs. X social safety nets • Implement current government guidelines on SSNs. X X (SSNs) • Target vulnerable areas and populations during X X ENSO. Improve • Streamline ENSO-related government and identify X government government agency roles. capacity and • Develop a mechanism for state-level stakeholders coordination X to work with provincial-level and local-level actors. • Exchange information through forums or other X dialogues. • Improve community participation and public X X awareness of ENSO. Integrate a regional • Begin dialogues with potential regional and X approach to ENSO international partners in ENSO. • Identify areas of cross-border vulnerability and X Resilience potential infrastructural cost sharing. Develop • Evaluate the pilot programs to potentially scale up. X risk transfer • Activate appropriate risk transfer mechanisms once mechanisms X a disaster begins. • Look into reallocating credit to provide emergency X X assistance. Invest in rural • Assess the prevalence and condition of rural X infrastructure infrastructure. • Where feasible, develop plans to improve roads X and expand irrigation and water-use efficiency. Target women • Develop a strategy targeting women who are X vulnerable to ENSO. • Empower women in leadership positions in local- X or national-level efforts to combat ENSO impacts. CHAPTER 1 Introduction The purpose of this report is to help Cambodia’s policy makers and stakehold- ers prepare for future El Niño–Southern Oscillation (ENSO) events. It does this by providing information on ENSO’s poverty, economic, and agricultural impacts in Cambodia and outlining ways forward. The report finds that ENSO’s impacts vary from region to region and harm Cambodia’s people, economy, and agricultural sector. The country has made inroads in preparing for climate events like floods and other natural disasters, but more could be done to prepare for ENSO: most efforts to respond to the 2014–2016 El Niño, which caused massive water shortages, were reactive. Being proactive to prepare for ENSO in Cambodia is important because of the coun- try’s high exposure to climate shocks, the prominence of the agricultural sector in the national economy, the large rural population and its climate vulnerability, and the lack of research on ENSO in Cambodia. Cambodia is vulnerable to ENSO-related shocks. Historical data for Cambodia show that ENSO’s two phases, the warm El Niño phase and the cool La Niña phase, see decreased and increased average rainfall, respectively. This has led to major water shortages and floods. Moreover, Cambodia is ranked the ninth most disaster-prone country in the world.30 During the 2014–2016 El Niño, Cambodia saw major droughts related to El Niño. By 2016, water levels in Cambodia’s rivers fell to 50 to 70 percent of the interannual average.31 This made 2016 one of the driest years in the country’s history, with one of the worst droughts of the last 50 years. Droughts affected almost every province and 2.5 million people experienced water shortages.32 30 UNU (2016). 31 OpenDevelopment (2016). 32 ACAPS (2016); Chakrya and Sassoon (2016); HRF (2016); Save the Children (2016).   1 2 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture ENSO’s impacts on Cambodia’s agricultural sector has economy-wide ramifica- tions. Agriculture in Cambodia contributes to 30 percent of gross domestic product (GDP) and accounts for 45 percent of employment.33 This is the country’s largest source of employment, especially for the poor, who live in rural areas and work in agri- culture. As such, any shocks from ENSO to agriculture has major impacts on incomes and people’s livelihoods. Agriculture is even more important when considering its linkages with downstream sectors and consumers in domestic and foreign markets. As such, any shocks to agriculture lead to reverberations across the entire economy, with serious implications on welfare, food security, and national poverty levels. Cambodia has a relatively large rural population, which is vulnerable to ENSO-related climate events. Cambodia’s rural population is 79 percent of the total population,34 high by both regional and global standards. Further, the rural poverty headcount ratio of 21 percent exceeds the urban poverty headcount ratio of only 6 per- cent.35 Given the large rural population, this means that poverty is rampant in Cambo- dia. Food insecurity is also a challenge that ENSO threatens to make worse: 15 percent of the population is food insecure36 and 28 percent of the population under five-years- old is undernourished.37 Since such a large portion of Cambodia is poor and works in agriculture, they are vulnerable to ENSO. This is because natural resource-based livelihoods are affected by ENSO events, such as droughts and floods, and poor farm- ers have few resources to respond. Further, most farmlands are not resilient to climate shocks, with over 90 percent of agricultural lands not irrigated. This report is timely given the lack of research on ENSO in Cambodia and the few efforts made by the country to prepare for ENSO. There is only a limited amount of available empirical evidence on ENSO’s impacts in Cambodia. There is more evidence on disaster and climate change impacts on agriculture and the wider economy, but these are also limited. Furthermore, no studies until this one analyzed how certain policy interventions could mitigate ENSO-related losses to GDP or con- sumption. Because ENSO’s effect on weather patterns can be forecast, albeit with varying degrees of accuracy, it is possible to design policies and response strategies in advance of extreme shocks. But to date, Cambodia has done very little to prepare the country for ENSO events. This lack of preparedness was evident during the country’s late and reactive response to the last El Niño in 2016. This report is especially timely 33 CGE model and the 2014 SAM. 34 2017 World Bank estimates based on the United Nations Population Division's World Urbanization Prospects: 2014 Revision. Available at: https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS 35 2012 figures: http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators 36 2015 data from WB-WDI (2018). 37 2010 data, http://www.foodsecurityportal.org/cambodia Introduction 3 because, at the time of writing, forecasts predict at least a 70 percent chance of another El Niño event occurring by December 2018.38 Roadmap This report evaluates El Niño and La Niña’s impacts on agricultural production— particularly for crops, livestock, and fisheries—and how these have implications for the national economy and household poverty levels. It then looks at some of the actions the Cambodian government and international partners have taken to mitigate the losses associated with past ENSO events. These actions include 2016 El Niño preparedness and response measures. The report examines some of the shortcomings of these actions, and then simulates how well certain policy options mitigate ENSO-related GDP and consump- tion losses. It concludes by recommending actions to build Cambodia’s preparedness for future ENSO events. Box 1 provides a brief description of the research’s methodology. BOX 1: Methodological Overview. The report synthesizes available evidence and presents new analyses on ENSO’s impacts on Cambodia’s weather, agricultural production, poverty levels, and the broader economy. The evidence comes from a secondary literature review, an analysis of historical data, and a series of simulations. Simulations were carried out using two models. The first is the Decision Support System for Agrotechnology Transfer (DSSAT) model, which is widely used by agri- cultural researchers to understand crop production system dynamics and simulate different farm management and environmental changes, including climate variability associated with ENSO. The second is a Computable General Equilibrium (CGE) model. The models simu- late: (i) potential agricultural productivity under various conditions, including water scarcity, different planting months, and low or high fertilizer use; (ii) the indicative impacts of ENSO on livestock using the Temperature Humidity Index (THI); (iii) economic outcomes associated with these productivity losses (usually during El Niño) and gains (usually during La Niña); (iv) economic outcomes, including spillover effects, associated with certain policy changes, such as providing drought-resistant crop varieties, increasing irrigated land, restricting rice exports, increasing grain storage and distribution, expanding social protection coverage (social transfers), and a combination of these; and (v) poverty impacts on rural households with male and female household heads. Put simply, the model simulates the impact of historical ENSO events if they were to reoccur today and affect the current economy. This is valuable given the lack of systematic historical analyses of ENSO’s impacts on poverty and broader economic growth. The models do not simulate water supply constraints because of a lack of useable data. For a more detailed description of the methodology, refer to Annex 1. (continued) 38 NOAA (2018); http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/ensodisc .shtml 4 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture BOX 1: Continued. Figure 1 depicts this report’s analytical framework. FIGURE 1: Integrated Analytical Framework. Weather and climate Rainfall; temperature Temperature humidity index (max, min) Crop yield impacts Spatial crop models Crop management (DSSAT) Seed varieties; chemical fertilizer; irrigation; crop Biophysical outcomes calendar Yields by crop and region Livestock & fisheries Animal deaths; heat stress; ocean capture stocks Economy-wide impacts Infrastructure Dynamic spatial CGE Economy outcomes Roads and ports; and microsimulation GDP, poverty, etc. agricultural capital model Policies Trade policies; price policies (subsidies, taxes); social safety nets CHAPTER 2 ENSO Affects Cambodia ENSO, or the El Niño–Southern Oscillation, describes naturally occurring ocean and atmospheric temperature fluctuations across the east-central equator of the Pacific Ocean. These temperature fluctuations are considered ENSO when they are greater than 1 degree Celsius. They can have large-scale impacts on ocean processes and global weather patterns. ENSO consists of two opposite phases: El Niño and La Niña. El Niño is ENSO’s warm phase, while La Niña is ENSO’s cold phase. El Niño and La Niña episodes are unpredictable, cyclical, and occur every four years and typi- cally last nine to twelve months, but can persist longer. Refer to Box 2 for a distinction between ENSO and climate change. BOX 2: Distinguishing ENSO from Climate Change. ENSO and climate change are two separate phenomena that share many attributes, and some key differences. Both are slow onset climate events that may not be immediately perceived by farmers or the wider society. This slow evolution of impacts creates policy challenges for both, since fast and large-scale disasters, such as hurricanes and earthquakes, are more likely to compel action, but slow moving disasters, like drought, may not compel change. As such, ENSO preparedness and climate change preparedness share many of the same actions and priorities. However, there are also key differences. While ENSO is cyclical with impacts that will come and go, climate change impacts are more permanent and are predicted to keep getting worse. Because of this, ENSO interventions in agriculture can be implemented in the short- and medium-term during ENSO years, and discontinued during non-ENSO years when climate conditions return to normal. Whereas climate change interventions in agricul- ture are long-term and permanent. In other words, with climate change there is no returning to normal. Climate change can also alter the impacts of ENSO. For example, climate change may make areas previously suitable for agriculture, unsuitable. It may also make areas drier and more water scarce, and in the process, make those areas more vulnerable to El Niño.   5 6 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 2: Monthly Average of Rainfall Amount in ENSO Phases During 1980–2015. Average rainfall (mm/month) (1980–2015) 400 300 200 100 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec El Niño Neutral La Niña Source: Authors’ reanalysis using the University of East Anglia Climatic Research Unit-Time Series (UEA CRU-TS) v4.0. ENSO’s main climate impact in Cambodia is on average rainfall, which declines during El Niño and increases during La Niña. On average, the warm El Niño phase is drier than the cool La Niña phase across Cambodia—especially from October to May (Figure 2). A statistical analysis on Cambodia’s rainfall history from 1980 to 2015 shows average annual rainfall is about 22 percent lower between October and June during El Niño compared to neutral, non-ENSO years. Alternatively, average annual rainfall is about 10 percent higher between October and December and 19 per- cent higher between April and June during La Niña compared to neutral years. Cambodia is vulnerable to extreme weather events. Besides ENSO, Cambodia’s climate is also influenced by the Tibet Plateau’s climate, which influences surrounding areas, and the Indian Ocean Dipole, an irregular sea-surface temperature oscillation in which the western Indian Ocean becomes alternately warmer, then colder than the eastern Indian ocean.39 Floods and droughts are associated with these teleconnections and are the most severe perennial weather events in Cambodia.40 Tropical cyclones in Cambodia are often not strong, since they weaken after crossing Vietnam. How- ever, these cyclones still cause heavy rainfall, worsening the country’s flood situation; 2009’s Typhoon Ketsana is one example. Human and economic costs are associated 39 Chhinh (2015). 40 Vathana and others (2013); Leng (2014); Chhinh (2015). ENSO Affects Cambodia 7 with these weather events: natural disasters from 1996 to 2016 led to 1,000 deaths and $242 million in economic losses.41 ENSO contributes to damaging flooding. Flooding is the natural disaster that causes the greatest economic damage in Cambodia.42 Much of this flooding occurs in the Mekong River areas, where water levels fluctuate significantly between the wet and dry seasons. Moreover, southwest monsoon rains produce recurrent flooding in Kan- dal, Kratie, Prey Veng, Stung Treng, Svay Rieng, Takeo, and Kampong Cham prov- inces. Flash floods commonly occur in tributaries of Lake Tonle Sap.43 In 2015, ENSO was blamed for torrential rains of 50 to 100 millimeters per hour in Phnom Penh, causing flashfloods and infrastructure damage.44 ENSO contributes to drought. Cambodia’s dry season is traditionally marked by drought and water scarcity, but El Niño contributes to this by causing an early end to the wet season or an early start to the dry season. These changes have major implications on wet season cropping and household livelihoods.45 ENSO also leads to heatwaves and higher temperatures, which cause water shortages.46 The last major drought, in 2016, cause widespread damage and required humanitarian relief operations (see Box 3). Unfortunately, still little is known about the impacts and areas affected by drought. Unlike floods, droughts have not been comprehensively assessed in Cambodia.47 ENSO has different impacts on different subnational regions in Cambodia. For the pur- poses of this report, Cambodia is divided into four subregions: the southwest coastal area, the southeast plains, the northeast mountains, and the northwest lake (Map 1). Map 2 shows the Coastal region tends to have the greatest rainfall, followed by the Mountain region. The Lake and Plains regions are the driest. Map 3 shows Cambodia’s drought vulnerability. Among the eleven most drought-prone provinces, drought is most severe in Prey Veng, Battambang, Kampong Thom, and Kampong Speu located in the Lake and Plains regions. Figure 3 shows the regional rainfall differences during El Niño and La Niña compared to non-ENSO years. On average, the year’s first quarter from January to March shows the greatest El Niño impacts: the Lake region saw 50 percent less rainfall from January to March, while the Coastal region saw 32 percent less rainfall from January to March. During La Niña, increased rainfall was only notable from April to June, particularly in the Lake and Mountain regions. 41 Kreftand others (2017). 42 National Institute of Disaster Management (2014). 43 National Institute of Disaster Management (2014). 44 UN-ESCAP (2016). 45 HRF (2015). 46 Leng (2014). 47 Leng (2014); Chhinh (2015). MAP 1: Cambodia's Four Major Regions.48 MAP 2: Annual Rainfall (millimeters).49 800–1,400 1,400–2,000 2,000–2,600 2,600–3,200 48 49 3,200–3,800 >3,800 48 Source: https://en.wikipedia.org/wiki/Provinces_of_Cambodia. 49 DANIDA (2015). ENSO Affects Cambodia 9 MAP 3: Most Drought Vulnerable Provinces in Cambodia.50 FIGURE 3: Regional Rainfall Differences between El Niño and Neutral Phase in Each Quarter from 1980–2015.51 Q1 Q2 Q3 Q4 Plain El Niño Lake Coastal Mountain Plain La Niña Lake Coastal Mountain –50% 50% –50% 50% –50% 50% –50% 50% Rain diff. Rain diff. Rain diff. Rain diff. neutral (%) neutral (%) neutral (%) neutral (%) Cambodia has experienced many ENSO events of differing severities over the past several decades. Between 1980 and 2015, there were eight El Niño events, when sea surface temperature anomalies, which are used to assess the development of El Niño or La Niña, exceeded 1 degree Celsius. During this timeframe, Cambodia suffered from nine severe droughts—in 1982–83, 1986–87, 1994, 1997–98, 2002, 2005, 2010, 2012, and 2016. Except for 2012, these droughts coincided with El Niño events (Figure 4).52 Figure 5 shows each ENSO event had a different level of severity. The most consequential 50 HRF (2015). 51 Authors’ reanalysis using UEA CRU-TS v4.0. 52 Miyan (2015); Leng (2014). 10 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 4: Seawater Temperature Changes Indicating the Occurrence of El Niño (above red line) and La Niña (below blue line).53 3 Dec-15, 2.3 2.5 Nov-97, 2.3 2 Dec-82, 2.1 Sep-87, 1.6 Dec-06, 0.9 1.5 Jan-92, 1.6 Jan-05, 0.7 Jan-10, 1.3 1 0.5 0 –0.5 –1 –1.5 –2 –2.5 Jan-82 Jan-83 Jan-84 Jan-85 Jan-86 Jan-87 Jan-88 Jan-89 Jan-90 Jan-91 Jan-92 Jan-93 Jan-94 Jan-95 Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16 Jan-17 El Niño event occurred in 1997–1998, when the region’s rainfall fell by 27 percent rela- tive to historical levels. Box 3 shows some of the weather-related impacts from the most recent El Niño, from 2014 to 2016, and Table 1 shows the relative impacts from the eight El Niño events that occurred since 1980, plus the 2012 non-ENSO-related drought. ENSO events, and their severity, are difficult to predict with accuracy, especially over the long term. Like other countries, Cambodia must rely on international ENSO forecasts as the basis for their own forecasts. Future ENSO events are predicted using Coupled Ocean-Atmosphere General Circulation Models (CGCMs). These mod- els simulate the complex interactions between oceans and climate systems, based on historical Sea Surface Temperature (SST) trends. There are still many technical and scientific challenges to modeling these complex oceanic and atmospheric processes. Moreover, ENSO forecasting is an actively researched area, and understanding is not complete. Internationally, the most reliable forecasts are made by combining and calibrating the output from various independent forecasting systems. As such, it was multi-model ensembles that successfully predicted the timing of ENSO’s onset, peak, and decline.54 However, it is not known why some ENSO-related weather events are more extreme than others. In fact, ENSO severity predictions are largely unreliable. The United States’ National Oceanic and Atmospheric Administration (NOAA) assessed its ENSO forecast accuracy with different lead times.55 They concluded that ENSO sever- ity forecasting was highly accurate one month in advance, but much less accurate four months or longer in advance. 53 Rayner and others (2003). 54 Stockdale, Balmaseda, and Ferranti (2017). 55 Barnston (2014). ENSO Affects Cambodia 11 FIGURE 5: Average Rainfall Differences between El Niño and Neutral Years.56 1982–1983 –12% 1986–1987 –12% 1987–1988 –17% 1991–1992 –10% 1997–1998 –27% 2002–2003 4% 2009–2010 –14% –30% –25% –20% –15% –10% –5% 0% 5% Rainfall difference (%) (El Niño–Neutral) BOX 3: Experience from the 2014–2016 El Niño. The most recent El Niño significantly altered Cambodia’s weather patterns. This El Niño began in late-2014 and lasted until mid-2016 causing higher temperatures, declined rainfall, and, eventually, drought. Initial signs of drought were seen in June 2015, and conditions worsened dramatically in March 2016, continuing until July 2016.57 Temperatures in June 2015 were warmer than the maximum recorded temperatures since 1990.58 The maximum temperature in 2015 in the Coastal and Mountain regions was 0.5 and 1.5 degrees Celsius higher, respectively, than the maximum temperature over the previous 24 years.59 August rainfall in many regions was below the long-term average, especially in Kampot, Battambang, Kampong Speu, and Banteay Meanchey, and there was almost no rain from October 2015 to June 2016.60 Water levels in Cambodia’s rivers fell to 50 to 70 percent of the interannual aver- age.61 This made 2016 one of the driest years on record, with one of the worst droughts of the last 50 years. Droughts affected almost every province and 2.5 million people experienced water shortages.62 (continued) 56 using UEA CRU-TS v4.0. Authors’ reanalysis 57 UN-ESCAP (2016). 58 UN-ESCAP (2016). 59 UN-ESCAP (2016). 60 Cambodia Humanitarian Forum (2016). 61 Open Development (2016). 62 ACAPS (2016); Chakrya and Sassoon (2016); HRF (2016); Save the Children (2016). 12 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture BOX 3: Continued. Figure 6 shows how monthly rainfall deviated from average historical rainfall patterns in 2015 and 2016. FIGURE 6: Rainfall Comparison between 2015–2016 and the 31-year Annual Mean in the Plains (lowland area). 300 250 200 150 100 50 0 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Mean of rainfall 1985–2015 Average rainfall 2016 Average rainfall 2015 Source: United Nations Economic and Social Commission for Asia and the Pacific (UN-ESCAP) 2016. TABLE 1: El Niño Events and Droughts in Cambodia, 1982–2016. Peak Oceanic El Niño Drought El Niño El Niño Niño Index duration event event strength (ONI) (months) Impacts 1982–1983 1982–1983 Strong 2.1 15 120,000 ha of crop area damaged 1986–1987 1986–1987 Moderate 1.6 18 N/A 1994 1994–1995 Moderate 1 6 5 million people affected by food and water shortages 1997–1998 1997–1998 Strong 2.3 13 N/A 2002 2002–2003 Moderate 1.2 9 2 million people affected by food and water shortages; $22 million in economic losses ENSO Affects Cambodia 13 Peak Oceanic El Niño Drought El Niño El Niño Niño Index duration event event strength (ONI) (months) Impacts 2005 2004–2005 Weak 0.7 10 2 million people affected; 62,700 ha of crop damage; $21 million in economic losses 2010 2009–2010 Moderate 1.3 10 12 provinces affected by drought; about 23,000 ha of crop area damaged 2012 No El Niño N/A N/A N/A 11 provinces affected by drought; over 14,000 ha of rice fields damaged of which 3,151 ha were totally damaged 2016 2014–2016 Strong 2.3 18 18 provinces affected by drought; 2.5 million people affected by food and water shortages Source: Summarized from Vathana et al, 2013; Leng, 2014; Chhinh, 2015; Miyan, 2015; El Niño information from http://ggweather.com/enso/oni.htm CHAPTER 3 ENSO Affects Agriculture Rice production dominates agricultural land use in Cambodia. In 2016, one-third of the country’s total land area of 18 million hectares was devoted to agriculture.63 Rice represents about 58 percent of total agricultural production value. Cassava is the country’s next most important crop, followed by livestock, including pigs and cattle (Figure 7). Maize production, by contrast, is much lower than other southeast Asian countries. The livestock sector in Cambodia also produces significant amounts of poultry and buffalos. Rice also accounts for nearly have of all agriculture exports FIGURE 7: Agricultural Production as a Share of Total Value (2014–16) in Constant Dollars (2004–2006). Dry beans Other Maize Milk 1% 9% 1% 1% Fresh vegetables 2% Pig meat 3% Cattle meat 4% Paddy rice Cassava 58% 21% Source: FAOSTAT, July 2018. 63 WB-WDI (2018).   15 16 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 8: Agricultural Exports as a Share of Total Value (2014–16), in Constant Dollars (2004–06). Raw Cassava starch sugar Other Palm oil 2% 1% 6% 2% Cigarettes 3% Pet food 3% Refined sugar 4% Dried cassava Rice 4% 48% Dry rubber 27% Source: FAOSTAT, July 2018. (Figure 8), while rubber accounts for over a quarter. The top three export markets for fresh foods and raw agro-food products are, in order of importance, Vietnam, China, and France, while the biggest markets for processed food are Vietnam, the United States, and Malaysia (2016).64 Most of Cambodia’s agricultural production occurs in the Mekong River area in the Plains. The Plains region consists of floodplains. These are mainly cultivated with rice in the lower Mekong and with agro-industrial crops, such as rubber and cassava, in the upper Mekong (Map 4). The Tonle Sap, or Lake region, focuses more on fresh- water fisheries. The Coastal region includes coastal and marine fisheries and some crops. The Mountain region includes agro-industrial crops, like rubber and cassava, and forestry. 64 http://www.intracen.org/layouts/CountryTemplate.aspx?pageid=47244645034&id=47244652178 ENSO Affects Agriculture 17 MAP 4: Crop Land Use Patterns in Cambodia, 2015. Cropland Rice paddy Shrub Tree Urban Source: Authors’ analysis using ESA (2018). Crops El Niño has a significantly negative effect on rice production, while La Niña has a slightly positive effect. Combining all ENSO events together and de-trending his- torical data reveals that, on average, rice production declines by 10 percent during El Niño and increases by 5 percent during La Niña (Figure 9). Moreover, model- ing ENSO’s impacts on different crops at the subnational level shows consistent but slightly different results. Figure 10, which shows CGE modeling results, indicates that El Niño generally has detrimental effects on all regions, especially for maize and vegetables, or its proxy crop of tomatoes. Interestingly, Figure 10 also shows maize and vegetables (tomatoes) benefit the most during La Niña events. In the models, rice sees yield changes during both ENSO phases of less than 3 percent, but this low rate is attributable to the fact that rice also occupies the largest land area, so on average, gains and losses will be diluted compared to other crops. Beans are a minor crop and appear 18 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 9: Annual Average FIGURE 10: Simulated Changes in Beans, Maize, Rice, Rice Production Changes from and Vegetable Yields During El Niño and La Niña the Average of Previous Three Years Compared to Average Years.66 Years, between 1995 and BEANS MAIZE RICE VEGETABLES 2014.65 10% 9% 7% from average Yield change Mountainous Average rice production difference 3% 5% 1% 0% [wrt. avg. 3 previous years] 5% 0% –1% –6% –10% –8% 0% 10% from average Yield change 6% 4% –5% 2% 0% 1% Plain 0% 0% –5% –5% –10% –10% –10% 10% El Niño La Niña 10% from average Yield change 1% 0% 0% 0% 0% Lake 0% –2% –6% –10% 10% 8% from average Yield change 2% Coastal 1% 2% 0% 0% –1% 0% –4% –10% El Niño La Niña El Niño La Niña El Niño La Niña El Niño La Niña largely unaffected by ENSO. During La Niña, there may also be flooding risks that are not captured by the crop model. Droughts, ENSO-related and non-ENSO-related, cause rice production losses. From 1997 to 2002, 20 percent of total rice losses were caused by drought.67 In 2010, El Niño contributed to severe droughts in 12 provinces, damaging 14,000 hectares of transplanted rice, 3,500 hectares of rice seedlings, and 5,500 hectares of other crops.68 From June to August 2015, drought delayed rice planting in certain areas, causing a 2 percent decline in national rice production from the previous year.69 In August 2015, the Ministry of Agriculture, Forestry, and Fisheries (MAFF) reported 185,000 hectares of “maturing phase” crops in 13 provinces were adversely affected by low rainfall and water deficits.70 Other research showed droughts from 1996 to 2013 accounted for 31 percent of total rice production losses—a figure projected to increase in the future.71 65 Authors’ reanalysis from FAOSTAT. 66 Authors. 67 NAPA (2006); Chhinh (2015). 68 Leng (2014). 69 ACAPS (2016). 70 David and Harfenist (2015). 71 NCDM and UNDP (2013). ENSO Affects Agriculture 19 Crop production impacts depend on how El Niño affects the rainy season. The rainy season in Cambodia runs from May to November. Figure 11 shows the rice cropping calendar in lowland rice areas (rainfed and irrigated), which comprise about 95 percent of the rice area planted. El Niño events that delay the onset of the rainy sea- son severely affect short-duration wet season crops, which represent about 25 percent of the rice area planted each year. By contrast, in some cases, a shorter rainy season may have minimal or even positive effects, by reducing flooding and increasing sunny days, on medium- and long-duration wet season crops. FIGURE 11: Rice Cropping Calendar and Rice Agroecosystem in Cambodia. CAMBODIA: Seasonal Crop Calendar Rice Short wet Medium wet Long wet Dry season Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Planning Mid-season Harvest Rainfed upland Rainfed lowland Deepwater/ rice rice floating rice Early Medium Late Sources: United States Department of Agriculture, 2013; Men Sarom, 2001. It is difficult to disentangle the effects of ENSO events in historical production data from other historical events like war or other shocks. Figure 12 shows national rice yields did not decline compared to the trend line during El Niño events from 1980 to 1994, while yields were generally 3 percent above the trend line during La Niña years. Since 1995, El Niño years still show little yield deviations from the trend, while La Niña years again show a small positive effect on yields. Perhaps more importantly, all possible associations between ENSO and rice yields are within typical year-to- year variability, making it extremely difficult to statistically isolate ENSO-specific impacts. Maize broadly follows a similar pattern to rice, but with even greater vari- ability. Part of the reason for this is the low quality of time series data in Cambodia and the significant turmoil the country has experienced over the years. For example, rice production plummeted during the civil war in the early-to-mid 1970s and reached an all-time low when the Khmer Rouge were in power in the late 1970s. It only began 20 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 12: Rice Yields in Cambodia, 1961–2016 (kg/ha). 3,500 3,000 2,500 Yield (kg/ha) 2,000 1,500 1,000 500 1960 1970 1980 1990 2000 2010 2020 Yearly values El Niño warm phase La Niña cold phase Trend line to accelerate rapidly after 2000. Most other major crops—including cassava, maize, soybeans, and dry beans—show similar historical growth patterns, often mirroring political developments. Livestock There are several mechanisms through which ENSO can impact livestock, but evidence has not established a causal relationship between ENSO and livestock production. Reports during flood events show livestock often die during flooding, and a lack of livestock vaccinations cause livestock to die from disease, such as swine cholera and blue-ear swine disease. It can also be assumed that low feed production during droughts detrimentally affects livestock health and production. Cambodia’s livestock sector is already subject to significant heat stress, which ENSO may make worse. Poultry are sensitive to heat, with strong declines in feed intake, egg size, and egg quality occurring at 80 degrees Fahrenheit.72 Swine are also sensitive to heat, with strong negative outcomes evident at temperatures above 68 degrees Fahrenheit. The reproductive performance of sows can also be negatively affected by high tempera- tures.73 Overall, cattle are the most heat tolerant, followed by poultry, then pigs, which are the least heat tolerant.74 72 Lara and Rostangno (2013). 73 Wegner and others (2016). 74 Xin and Harmon (1998); Myer and Bucklin (2001). ENSO Affects Agriculture 21 La Niña can cause heat stress on livestock. Once the daily Temperature Humidity Index (THI) exceeds 75, cattle begin to experience the ill effects of heat stress, such as slower growth or declining milk productivity.75 THI measures heat stress based on temperature and humidity; as such, La Niña’s high humidity can cause high THI. From 1951 to 2010, La Niña years have 16 more days with a maximum THI of 80 to 90, which produce moderate to severe stress, than neutral and El Niño years, which have about 100 such days per year.76 NASA’s POWER dataset77 shows La Niña causes up to 224 high THI days. While El Niño does not seem to contribute to more high THI days, evidence suggests El Niño, without considering humidity, is generally associated with higher temperatures. Again, it is unclear how much higher THI or more high THI days impact livestock in Cambodia. Fisheries ENSO-related drought can negatively affect fisheries. Fisheries are central to the livelihoods of millions of Cambodians, and account for about 80 percent of animal protein in household diets.78 The 2009 El Niño caused low water levels in the Mekong River, leading to salt water intrusion,79 affecting fish stocks. Reduced water levels in 2009 also lowered fish productivity in Lake Tonle Sap and in inland fisheries.80 This was also observed during the 2015 drought, which led to lower water levels and reduced fish production. Fishing households resorted to backyard aquaculture to aug- ment incomes and food consumption, but this required new investments that put many households in debt.81 The 2016 drought also caused widespread fish deaths as Cam- bodia’s fish production was 17 percent lower, leading to higher fish prices.82 Dry con- ditions and extreme heatwaves in the center of the country also killed about 65 tons of fish in Lake Tonle Chhmar.83 In Cambodia, there is also a link between ENSO, forest fires, and fisheries. The ENSO-related drought of 2016 increased forest fire risk, with water shortages mak- ing it difficult to combat these fires.84 According to the Cambodian Ministry of Agri- culture’s Inland Fisheries Institute, forest fires reduce fish production in Cambodia.85 75 West (2003). 76 Kalnay and others (1996). 77 See http://power.larc.nasa.gov/ 78 Baran, Schwartz, and Kura (2009). 79 Adamson and Bird (2010). 80 Adamson and Bird (2010). 81 ACF (2015). 82 HRF (2016). 83 Gaulter (2016). 84 ACAPS (2016). 85 Vidal (2016). 22 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture MAP 5: Forest Fires in Cambodia During the 2016 Drought (in red dots). Source: NASA (2016). Map 5 shows hundreds of fires occurred throughout Cambodia in 2016, particularly in the northeastern mountain region and the northwestern Lake Tonle Sap region. An estimated 250,000 hectares of the Lake region were destroyed by fire during the 2016 drought.86 This had major implications for fisheries. The lake is a critical aquatic eco- system for breeding many migratory fish species that eventually populate the Mekong River. Of course, forest fires also have implications for cropping.87 86 Vidal (2016). 87 ACAPS (2016); Cambodian Humanitarian Forum (2016). CHAPTER 4 ENSO Contributes to Economic Impacts Agriculture remains an important economic sector in Cambodia. Table 2 shows that according to CGE modeling, the agricultural sector contributes to 30 percent of national GDP. The industry sector nearly matches that contribution at 27 percent, but the service sector contributes the most at 43 percent. That said, agriculture is the larg- est employer in Cambodia by a significant margin, with an 11 percent higher share than services, the next largest employer. Table 2 suggests most agriculture is used for domestic purposes, with agriculture produce only accounting for two percent of national exports. This also suggests that Cambodia is largely food self-sufficient with TABLE 2: National Economic Structure, 2014. Share of total (%) GDP Employment Exports Imports All sectors 100 100 100 100 Agriculture 30.1 44.8 2.0 0.5  North 7.2 11.6 1.5 0.5  Center 8.2 13.0 1.5 0.5  South 2.3 3.6 1.5 0.5 3.7 6.0 1.5 0.5 Industry 27.1 22.4 68.4 84.3  Mining 0.2 0.0 0.0 0.1  Manufacturing 26.9 22.3 68.4 84.2   Agro-processing 2.2 2.2 1.3 8.8   Other manufactures 16.4 15.2 67.1 74.1   Other industry 8.2 4.9 0.1 1.3 Services 42.8 32.9 29.6 15.3 Source: Cambodia CGE model and 2014 SAM. Note: GDP is gross domestic product; employment is workers in primary jobs.   23 24 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 13: Growth and Variability of Agricultural GDP in Cambodia. GDP per capita (constant 2010 dollars) 18 300 16 250 14 Annual growth rate (%) 12 200 10 150 8 6 100 4 50 2 0 0 –2 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Annual agricultural GDP growth Agricultural GDP per capita Source: Own calculations using World Development Indicators (World Bank 2018). agriculture accounting for only a half percent of national imports. Industry makes up the majority of national imports and exports by far. The agricultural sector in Cambodia is growing but has become less important relative to the national economy. Figure 13 shows agricultural GDP per capita con- sistently rose from 2000 to 2013. In 2013, population growth outpaced agricultural growth, causing per capita agricultural GDP to fall slightly. Figure 13 also shows that annual agricultural growth rates vary significantly year-to-year, though generally remain positive, which reflects the sector’s exposure to agricultural shocks and low resilience. However, as Cambodia’s economy grew, agriculture’s relative contribution to the economy shrunk. From 2000 to 2016, agriculture’s contribution to national GDP declined from 38 percent to 27 percent according to the World Bank.88 Employment in agriculture also fell from 74 percent in 2000 to 42 percent in 2017.89 That said, 42 percent is still a very large portion of Cambodian society that derives its income from agriculture; this is especially true for the poor. However, World Bank estimates are slightly lower than CGE model estimates. Cambodia’s overall agricultural food system (AFS) represents a larger share of the national economy. AFS includes agriculture’s downstream economic activity, such as input production, agricultural processing, and agriculture-related trading and transporting. Table 3 shows that Cambodia’s AFS accounts for 36 percent of national 88 WB-WDI (2018). 89 WB-WDI (2017). ENSO Contributes to Economic Impacts 25 TABLE 3: Agriculture Food System GDP and Employment, 2014. Share of national total (%) GDP Employment National economy 100 100 Agriculture food system 36.4 50.4   Direct production 32.4 47.0   Agriculture 30.1 44.8   Agro-processing 2.2 2.2   Input production 0.7 1.1   Agriculture 0.4 0.8   Agro-processing 0.3 0.3 Trade and transport 3.3 2.3   Agriculture 1.5 1.1   Agro-processing 1.8 1.3 Source: Cambodia CGE model and 2014 SAM. Note: GDP is gross domestic product; employment is workers in primary jobs. GDP: 6 percent higher than agriculture’s direct contribution. Employment in the AFS also includes just over half of the employed population. Because of the importance of the AFS to the national economy, any agricultural shocks, such as those related to ENSO, can have economy-wide implications and affect sectors and people that are only indirectly linked to agriculture. El Niño events cause notable economic losses, but these are nearly offset by pos- sible gains during La Niña events. Total economic losses from three El Niño–related historical droughts (1994, 2002, and 2005) are estimated at $138 million.90 According to the CGE model (Table 4), national GDP falls by 0.4 percent during an El Niño event relative to a neutral climate year. Losses are larger in the agriculture sector, where GDP falls by 1.4 percent. The mountain region suffers the greatest losses, where crop and livestock91 GDP falls by 2.5 percent. Even small percentage reductions in national GDP can imply substantial monetary losses. For example, a 0.4 percent drop in national GDP is equal to $61 million in lost value-added, or national income (mea- sured in 2014 prices). Changes in agricultural productivity during a strong La Niña year expand the economy. The CGE analysis suggests gains ($55 million) during La Niña events almost offset losses ($61 million) during El Niño events. 90 Miyan (2015). 91 Crops and livestock are subsets of agriculture that exclude forestry and fisheries. 26 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture TABLE 4: GDP Changes During Typical ENSO Events. El Niño events La Niña events Percentage change in GDP (%) National –0.39 0.35 AFS –1.24 1.12 Agriculture –1.39 1.26 Crops and livestock –1.91 1.75   Plain region –1.94 1.62   Lake region –1.67 1.71   Coastal region –1.71 0.78   Mountainous region –2.53 2.67 Absolute change in GDP (US$ million) National –60.88 54.73 AFS –70.59 63.82 Agriculture –65.54 59.76 Crops and livestock –64.19 58.61   Plain region –22.01 18.41   Lake region –21.36 21.86   Coastal region –6.07 2.78   Mountainous region –14.75 15.57 Source: Simulation results from the Cambodia CGE model. GDP losses in Cambodia’s AFS are larger than losses in primary agriculture during El Niño. Figure 14 shows the AFS suffers $71 million in losses during a typi- cal El Niño event, while primary agriculture suffers $66 million in losses. Percentage GDP losses are largest in crop and livestock production. Percentage losses become smaller when considering the entire agriculture sector, and smaller still when consider- ing the broader AFS and national economy. However, absolute value losses generally increase as the focus broadens from crops and livestock to the AFS. This is because of the larger overall GDP contribution of AFS and negative spillovers between sectors. For example, lower agricultural production constrains the supply of raw materials to agriculture-related trading and processing. Figure 15 shows that GDP losses are nearly offset by La Niña gains across agricultural-related subsectors of the economy. Food and agricultural product prices increase only slightly during El Niño events. Price changes are a key mechanism through which agricultural production losses are transmitted to other sectors and households. CGE modeling shows that real agricul- tural prices increase by 1.4 percent and food prices increase by 0.8 percent during ENSO Contributes to Economic Impacts 27 FIGURE 14: GDP Losses During Typical El Niño Events ($ millions or percentage reductions). $70.6 2.5% $65.5 $64.2 $60.9 1.9% 1.9% 1.7% 1.7% 1.4% 1.2% $22.0 $21.4 $14.8 0.4% $6.1 National National National National Plain Lake Coastal Mountain economy AFS agriculture Crops and Livestock Source: Simulation results from the Cambodia CGE model. FIGURE 15: GDP Gains During La Niña Events ($ millions or percentage increases). $64.4 $60.3 $59.1 $55.4 2.7% 1.8% 1.7% 1.7% 1.3% 1.1% $21.9 $18.9 0.8% $15.6 0.4% $2.8 National National National National Plain Lake Coastal Mountain economy AFS agriculture Crops and Livestock Source: Simulation results from the Cambodia CGE model. ENSO events in Cambodia (Figure 16).92 Food products include commodities pro- duced by downstream food industries that are not directly affected by ENSO but that use agricultural products whose production is directly affected. Together with income changes, price changes explain why real consumption falls and poverty rises, albeit only slightly, for both rural farmers and urban consumers (see Table 8). 92 These are “real” price changes, which means they are the net general change in the consumer price index. 28 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 16: Real Food and Agricultural Price Changes During ENSO Events (percentage). 1.37% 0.82% –0.73% –1.27% El-Niño events La-Niña events Agricultural products Food (including processed) Source: Simulation results from the Cambodia CGE model. When agricultural production faces an ENSO-related shock, the economy adapts by reallocating labor. When ENSO causes agricultural production declines, work- ers who would normally be engaged in agriculture, migrate to work in other sectors outside of the AFS. In Cambodia, the reallocation of labor outside of the AFS out- weighs the lower demand for non-AFS products, leading to slightly smaller absolute GDP losses in the national economy ($61 million) compared to the AFS ($66 million). Overall, the CGE model estimates about a tenth of El Niño’s economic damages to the AFS occur outside of agriculture itself. Economic linkages also explain the sectoral distribution of GDP gains during strong La Niña events (Figure 15). However, during La Niña, higher crop and livestock productivity attracts workers from other parts of the AFS and the broader economy, and explains why national GDP has smaller gains than agricultural GDP during La Niña. CHAPTER 5 ENSO Contributes to Social Impacts Poverty in Cambodia has declined rapidly over the past two decades but still remains prevalent, especially in rural areas. Cambodia is classified as a least devel- oped country.93 Despite this, poverty has declined markedly from 50 percent in 2003 to 13.5 percent in 2014.94 That said, another third of the country’s population was still categorized as near-poor in 2017.95 Most of this poverty occurs in rural areas: the rural poverty headcount ratio (20.8 percent) is three times higher than the urban poverty headcount ratio (6.4 percent).96 As of 2016, four out of every five Cambodians lived in rural areas.97 Poor households consume less and suffer more from ENSO events than more affluent households. Table 5 shows Cambodia’s population consumed, on average, $841 of goods and services per person in 2014.98 However, rural consumption levels were only 38 percent of urban consumption levels. The rural poor consume even less: only 23 percent of urban consumption levels. Poor households also spend more of their incomes on food consumption, meaning when ENSO affects rice and maize produc- tion (cereals), for example, it will have larger implications for poorer consumers. This is because when production declines, prices increase. Table 8 shows both urban and rural households are affected: urban households because they are net consumers of 93 The U.N. definition of a least developed country is a nation that does not have a population greater than 75 million and meets all three criteria: (1) a GNI per capita of $1,025 or less, (2) a human assets index criterion—that includes health and education indices—of 60 or less and (3) an economic vulnerability index of 36 or less that measures vulnerability to economic and environmental shocks with eight indicators. 94 World Bank (2017). 95 World Bank (2017). 96 2012 figures: http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators 97 WB-WDI (2018). 98 At market exchange rates unadjusted for purchasing power parity.   29 30 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture TABLE 5: Household Income and Consumption, 2014. National Rural Rural poor Urban Population (millions) 15.3 12.0 2.8 3.3 Consumption per capita (US$) 841 625 369 1,625 Share of food consumption (%) 100 100 100 100   Cereals and roots 20.5 23.3 27.4 15.9  Vegetables 9.6 10.1 10.4 8.8   Meat, fish, and eggs 35.3 37.4 40.3 31.7   Milk and dairy 0.7 0.6 0.4 0.8   Pulses and oilseeds 2.9 3.4 3.9 2.2   Other foods 31.0 25.2 17.6 40.7 Food share of consumption (%) 41.5 44.5 57.2 37.3 Processed foods in total foods (%) 39.4 32.7 25.6 50.6 Total household income (%) 100 100 100 100   Agricultural returns 12.5 21.4 14.1 0.4   Labor remuneration 42.3 39.4 59.0 46.2   Capital profits 43.0 35.2 26.6 53.7   Other sources 2.2 4.1 0.4 –0.3 Source: Cambodia CGE model and 2014 SAM. Note: Food consumption excludes meals prepared outside the household. Processed foods exclude products processed and consumed within the household. Other income sources include social and foreign remittances. food products, and rural households because they tend to spend more of their income on food consumption. However, rural households are also more likely to be farmers who benefit from higher food prices. That said, poor rural households earn less of their income from agricultural land than higher income rural households, mainly because poorer households have less access to land and, therefore, earn fewer returns from farming. Poorer households rely more on incomes from low-skilled labor. Cambodians generally do not have diverse diets, resulting in malnutrition. Cereals remain the most important source of food energy in Cambodia, accounting for 68 percent, with rice comprising over 63 percent of cereal intake. The lack of fat in Cambodian diets contributes to low absorption rates of Vitamin A and other fat-soluble micronutrients. Because families spend so much of their income on food already—70 percent by some measures99—there is little room for families to further diversify and improve their diets. As a result, undernourishment remains high, with 15 percent of the population undernourished. This is an improvement from 2000 when 99 UN Food and Nutrition Security Profiles (2014). ENSO Contributes to Social Impacts 31 29 percent of the population was undernourished.100 Unfortunately, child wasting—or children under-five who are underweight for their height—has not decreased, actually increasing from 8 percent in 2005 to 9 percent in 2014.101 Additionally, stunting, or the impaired growth of children, is still very high at 40 percent for children under five years of age.102 A lack of proper water, sanitation, and hygiene facilities can also con- tribute to health challenges. A study of eight Cambodian provinces during the 2015 ENSO showed only 12 percent of schools had latrines, while slightly over half of those did not function because of water shortages.103 Strong El Niño events have a greater impact on household consumption than on the national GDP. Households experience a 1.8 percent decline in consumption, or welfare, during a strong El Niño event, with similar losses for poor and non-poor house- holds. This decline is greater than that of GDP losses, partly because GDP changes are absorbed across households, investors, and the government, but also because damages to agriculture and food prices is of greater importance for household spending than for investment or government spending. Poor households’ ability to smooth consumption, by selling assets, during El Niño shocks may be more limited than higher income households’ ability, but this is not considered in the CGE modeling of consumption losses (reported in Figure 18). Women play an important role in agriculture and water management in Cambo- dia, making them vulnerable to ENSO’s impacts on these sectors. In Cambodia, women work in agriculture just as much as men: in 2014, 46 percent of employed women worked in agriculture compared to 44 percent of employed men.104 Women tend to be more vulnerable to extreme climate events because of limited access to timely and reliable information on changing climate conditions.105 Women also play a role in managing forests and biodiversity and often augment the family’s food and income needs.106 Women’s roles in water resource management is particularly strong, as they must manage water needs for both the home and the farm. Women in Koh Kong province, for example, reported several of their challenges related to water scarcity. These include: higher water prices; finding water for livestock; difficulties 100 WB-WDI (2018). 101 WB-WDI (2018). 102 FAO (2010). Cambodia—Food and Nutrition Security Profiles 2010. Available at: http://www.fao. org/3/a-at706e.pdf 103 Save the Children (2016). 104 NIS (2015). 105 Hing (2016). 106 Hing (2016). 32 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture maintaining crops and growing seedlings; less water for personal consumption such as drinking and hygiene; and migration to Thailand or other Cambodian provinces.107 Female-headed households are more affected by strong El Niño events than male- headed households. This is true in both rural areas and nationally. Figure 22 shows poverty impacts are only slightly stronger for women-headed households than male- headed households. One reason for this parity is that ENSO impacts are evenly felt among rural and urban economies and because the CGE model does not differen- tiate between male and female impacts, only between the relative expenditure lev- els of male- and female-headed households. Other studies do, however, differentiate between male and female impacts. A 2016 FAO report, for example, shows nearly a fifth of all Cambodian households are headed by women. These households tend to own smaller plots of land, are more indebted, and are more likely to be poor, reducing their resilience to ENSO-related impacts.108 Also, female-headed agricultural house- holds are 3 percent less likely to have access to agricultural lands than male-headed agricultural households. Moreover, for every five male-headed agricultural households that receive loans, only one female-headed agricultural household receives the same.109 The recent shift toward agriculture mechanization creates additional challenges for female-headed households, who receive fewer trainings and extension services.110 107 Chum, Tep, and Nevill (2016). 108 http://www.fao.org/3/a-i5489e.pdf 109 http://www.fao.org/fileadmin/templates/ess/documents/meetings_and_workshops/APCAS23/ documents_OCT10/APCAS-10-26_-Gender.pdf 110 http://documents.worldbank.org/curated/en/805091467993504209/pdf/96308-ESW-KH-White-cover- P145838-PUBLIC-Cambodian-Agriculture-in-Transition.pdf CHAPTER 6 Cambodia Has Taken Actions Related to ENSO Preparedness ENSO responses Cambodia’s government is committed to responding to weather-related shocks, but so far has done little to prepare for ENSO. Cambodia has contin- gency plans for floods and droughts, and receives funding from many overseas development assistance (ODA) projects to respond to natural disasters. But, ENSO awareness and its effects on drought and floods remains limited. Consultations with representatives of relevant ministries and other stakeholders111 reveal that government resources for climate change and natural disasters are mostly for ex post risk interventions, such as relief operations, but there are very few resources for early warning and ex ante mitigation and preparation. Refer to Annex 2 for more details on national plans and agencies related to climate-­ related responses. Cambodia’s efforts to mitigate impacts from the 2014–2016 El Niño were mainly in emergency response and little in preparation. Although the El Niño event began in 2014, it was not until April 2016 that the Cambodian Prime Minister, Hun Sen, announced a plan to assist those affected by the drought. This included sending trucks of drinking water to the most impacted areas.112 The Prime Minis- ter also called on the armed forces, civil servants, the Cambodian Red Cross, and all political parties to pitch in and help get water to those in need.113 After just one month, on May 25, the Prime Minister declared an end to the relief campaign.114 In total, the government allocated $125,000 for the campaign.115 Hoping to mitigate 111 During the Consultative Stakeholder Workshop, October 9–10, 2017. 112 Thul (2016). 113 Press (2016). 114 Gaulter (2016). 115 Samean and Sassoon (2016).   33 34 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 17: Chronology of Notable Actions Taken in Response to the 2014–16 El Niño. Armed forces, civil Peak of El Niño. servants, and MOWRAM started Cambodian Red PM Hunsen a “dry and hot” Cross were declared the end climate for the mobilized for of government’s Rainfall El Niño started next 6 months humanitarian aids relief campaigns started Nov Aug Nov Apr May End of Jun 2014 2015 2015 2016 2016 May 2016 2016 Some provinces PM Hunsen called Caritas Cambodia started to feel the for help from all launched an affects of drought elements in emergency appeal Cambodian society on drought, seeking funds to support 7 provinces future calamities, the Ministry of Water Resources and Meteorology (MOWRAM) provided some instructions to the communities to address future drought issues. These included promoting water conservation until the arrival of the rainy season; deterring dry season rice planting; and promoting rice storage for consumption.116 It is not known if any of these awareness raising measures had an impact. Fig- ure 17 shows the slow response to the 2014–2016 El Niño. State-run institutions Cambodia’s Ministry of Water Resources and Meteorology (MOWRAM) is the national-level focal point for ENSO issues. Within this role, MOWRAM is imple- menting a four-year (2015–2019) project to improve the natural disaster early warning system in Cambodia. The project is supported by the United Nations Development Program (UNDP) and aims to provide natural disaster warnings. Through this project, MOWRAM is expected to generate and simultaneously disseminate plans and early warning advisories to the media and local authorities. The warnings focus on improv- ing agricultural preparedness and emergency response.117 Within the M ­ OWRAM, the Department of Meteorology and the Department of Hydrology and River Works, which operate the hydrometeorological station network, are responsible for communi- cating and disseminating ENSO-related information.118 116 CHF (2016). 117 http://www.kh.undp.org/content/cambodia/en/home/operations/projects/environment_and_energy/early- warning-systems.html 118 http://www.mowram.gov.kh/index.php/en Cambodia Has Taken Actions Related to ENSO Preparedness 35 The Ministry of Agriculture, Forestry and Fisheries (MAFF) is in charge of agricultural-related issues. More specifically, the MAFF oversees forestry, fisher- ies, crop production, livestock production, and farmers’ livelihood development. In 2014, MAFF developed a Plan of Action for Disaster Risk Reduction in Agriculture for 2014–2018. This plan marked a departure from previous disaster strategies in that it gave equal interest and resources for prevention, preparedness, emergency response, and recovery. Traditionally, the Cambodian government’s disaster strategies were largely reactive. The Ministry of Environment (MoE) is responsible for climate change, envi- ronmental protection, and natural resources conservation. MoE is responsible for managing and protecting the country’s forests as part of its natural disaster pre- paredness. MoE also negotiates and implements international treaties and agreements related to climate change. The National Climate Change Committee, chaired by MoE, coordinates and implements climate change policies. The National Committee for Disaster Management (NCDM) leads Cambodia’s disaster risk management. The NCDM is chaired by the Prime Minister and brings together members from 22 different ministries, the Cambodian Armed Forces, Civil Aviation Authority, and the Cambodian Red Cross.119 The NCDM’s functions include: coordinating concerned ministries, NGOs, UN agencies, local authorities, interna- tional partners, national associations, and local donors in raising funds for emergency response and disaster relief; conducting research on natural disasters, preparedness, and emergency response plans; implementing capacity building activities for offi- cials in charge of disaster risk management at all levels; and other functions. The NCDM manages Cambodia’s Early Warning System 1294 (EWS 1294).120 The sys- tem, which delivers urgent text message alerts to subscribers and is described in Box 4, was first tested in Pursat province in 2013 and has been replicated in five other prov- inces, including Kampot, Banteay, Meachey, Siem Reap, Battambang, and Kampong Chhnang. Cambodian cellular providers provide the EWS service for free, and the system now has over 50,000 registered users. Notably, the system is not designed for slow onset disasters like droughts, but rather quick onset events like, “storms, floods, fires, or disease outbreaks.” For example, in July 2018, EWS 1294 sent two early warning messages to communes in Kampot affected by flood. Later, the messaging system advised locals to evacuate.121 119 Asian Development Bank (2014). 120 Initiallydeveloped for the NCDM by the Czech NGO, People in Need. 121 https://www.humanitarianresponse.info/sites/www.humanitarianresponse.info/files/documents/files/ cambodia_hrf_sitrep_flood_2018-_23072018_.docx 36 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture BOX 4: Early Warning System 1294. Source: http://ews1294.info/about. The Ministry of Women's Affairs leads the government’s gender equality efforts. One of these efforts, related to ENSO, is the Gender and Climate Change Strategic Plan122 (GCCSP), which was launched in February 2013. Its main objective is to inte- grate gender into climate change responses and policy making. The GCCSP’s recom- mendations translated into the Gender and Climate Change Action Plan123 (GCCAP). The GCCAP prescribes action in four areas, including: building institutional capacity and cross-sectoral coordination; raising awareness on women’s roles in climate change adaptation and mitigation; reducing climate vulnerability for women; and reducing greenhouse gas emissions by introducing climate-friendly, low carbon economic activities for women.124 122 http://www.camclimate.org.kh/en/documents-and-media/library/category/117-sectoral-ccsp.html? download=527:sectoral-ccsp-for-gender-and-climate-change 123 http://www.camclimate.org.kh/en/documents-and-media/library/category/128-climate-change-action- plan.html?download=644:ccap-for-gender-2014-2018-en 124 Robinson (2015). Cambodia Has Taken Actions Related to ENSO Preparedness 37 Policies and strategic plans The Cambodian government abides by several current and former international agreements, strategic documents, and planning frameworks related to climate. Some of the former agreements and documents, which are now expired, include: • The Strategic National Action Plan on Disaster Risk Reduction 2008–2013 (SNAP), which was shaped by the Hyogo Framework for Action 2005–2015, directed govern- ment disaster planning until 2013. • The Strategy for Agriculture and Water 2010–2013 was carried out by MAFF and MOWRAM and served as a policy guide. • The Climate Change Strategic Plan for Water Resources and Meteorology ­2013–2017 was launched by MOWRAM to mitigate natural disaster impacts. • The Nationally Determined Contributions document described the government’s priorities to alleviate the impacts of extreme events.125 Among the active strategies, the Law on Disaster Management126 regulates Cam- bodia’s emergency disaster response and preparedness. It defines standard operat- ing procedures for disaster relief operations. The law prioritizes greater early warning capacity and contingency and emergency response plans as the main pre-disaster preparedness activities. In the short term, the law directs local authorities in disas- ter-affected areas to mobilize people and resources for disaster relief and emergency response operations. Over the long term, the law emphasizes the need for prevention during the pre-disaster period. It calls for strengthening public awareness and develop- ing risk prevention programs, including climate change adaptation measures. It does not directly address ENSO. The National Strategic Development Plan (NSDP) 2014–2018127 outlines Cambo- dia’s development policy priorities. This document describes strategies to achieve these priorities and potential obstacles to the process. In this plan, natural disasters are one of the most pressing concerns, and it assigns various tasks for different gov- ernmental agencies to reduce these risks. This includes building early warning capac- ity, strengthening legal and institutional frameworks, raising public awareness, and building government capacity. Based on this NSDP (and the previous NSDP), the NCDM developed two five-year national action plans for disaster reduction. The first, the Strategic National Action Plan for Disaster Risk Reduction (SNAP-DRR), covered 125 MoE and NCSD (2016). 126 Royal Decree NS/RKM/0715/007, July 10, 2015; http://www.ifrc.org/Global/Publications/IDRL/ DM%20acts/Cambodia%20DM%20Law_English.pdf 127 http://cdc-crdb.gov.kh/cdc/documents/NSDP_2014-2018.pdf 38 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture the period of 2008 to 2013,128 and the second, the National Action Plan for Disaster Risk Reduction (NAP-DRR) covered the current period of 2014 to 2018.129 These two action plans promoted community resilience, capacity building for disaster manage- ment institutions, risk assessment and improved early warning systems, and emer- gency response and recovery capabilities.130 The government has made some efforts to integrate women into climate and disas- ter programing. This includes integrating gender into the 2014–2018 NSDP and the 2014–2023 Cambodia Climate Change Strategic Plan (CCCSP). In the NSDP, “Gen- der Equity” is a strategic objective with action plans for women’s health, education, legal protection, and economic empowerment. The NSDP also promotes women’s roles in climate change and green growth. In the CCCSP, reducing gender vulnerabil- ity and health risks to climate change are strategic objectives.131 Additionally, MAFF developed the 2016–2020 Gender Mainstreaming Policy and Strategic Framework, as a key document for gender and food security. This framework outlines the difficulties of rural women in the agricultural space, particularly the rapid mechanization of agri- culture, and the risks of development and modernization leaving women behind. The framework acknowledges women’s underrepresentation in community-based organi- zations and rural services, and women’s limited access to resources necessary for eco- nomic empowerment.132 The National Adaptation Program of Action to Climate Change (NAPA) was put into effect in 2006. The NAPA framework is carried out by MAFF and MOWRAM and directs climate change adaptation initiatives. Many of these are relevant for ENSO events, including: climate risk mapping, climate coping mechanisms, and legal and institutional arrangements for climate change adaptation.133 The NAPA also includes solutions to ensure water availability for irrigation, strengthen community disaster pre- paredness and response capacity, and diversify livelihoods in vulnerable areas. Interventions by development partners Many local and international organizations work on ENSO-related issues, such as natural disaster risk prevention, in Cambodia. These organizations include: 128 General Directorate of Agriculture (2013). 129 Hou (2016). 130 IFRC (2017). 131 UNDP Cambodia (2014). 132 http://documents.worldbank.org/curated/en/805091467993504209/pdf/96308-ESW-KH-White-cover- P145838-PUBLIC-Cambodian-Agriculture-in-Transition.pdf 133 General Directorate of Agriculture (2013). Cambodia Has Taken Actions Related to ENSO Preparedness 39 ActionAid, Caritas, the World Bank, Save the Children, the World Food Program, CARE International, the Asian Development Bank, the International Federation of the Red Cross and Red Crescent, European Civil Protection and Humanitarian Aid Operations, the United Nations Office for the Coordination of Humanitarian Affairs, the United States Agency for International Development (USAID), and many others. Support from these organizations comes in various forms: in-kind, such as food and water aid; cash, such as transfer funds and financial support for infrastructure develop- ment and rehabilitation; and technical assistance in areas such as planning, mapping, and capacity building. These organizations operate at the national and subnational levels. At the subnational level, they coordinate with provincial, district, and commu- nity-level committees, local disaster management teams, and the Cambodia Humani- tarian Forum (CHF).134 The Humanitarian Response Forum (HRF) coordinates international support for disaster risk prevention and reduction. The HRF was established in 2011 to jointly coordinate disaster aid efforts among the United Nations agencies, international NGOs, and other international partners. The HRF collaborates closely with govern- ment agencies and ministries, particularly the NCDM, to ensure international human- itarian support satisfies the government’s needs. The HRF is co-chaired by ActionAid and the World Food Program. HRF coordinates specific agencies in certain priority areas (Table 6). The HRF’s Cambodia Contingency Plan supports the government’s preparation for droughts and floods. The plan calls for immediate responsive activities; fuel for irrigation; capacity building; security and protection; drought detection and preparedness; food, water, sanitation and health care provisions; basic income sup- port and income generating activities; and water resource and irrigation infrastructure maintenance. TABLE 6: HRF Sector Coordinates Agency Activities in Priority Sectors. Priority area Agency leads Food security and nutrition (FSN) WFP, Oxfam Water, sanitation, and hygiene (WASH) UNICEF, World Vision Shelter IOM, PIN Health WHO, World Vision Education UNICEF, Save the Children Protection UNICEF, Save the Children Source: HRF, 2015. 134 Leng (2014). 40 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture The Cambodia Humanitarian Forum (CHF) coordinates local disaster response efforts in Cambodia and was particularly effective during the 2014–2016 El Niño. Currently, the CHF has 121 local Cambodian NGO members. The CHF was very active during the 2014–2016 ENSO-related drought. In 2016, 11 CHF member orga- nizations135 conducted a rapid assessment of drought impacts and needs in different regions. It assessed water availability in lakes, canals, open wells, pump wells, and village ponds. Although the assessment was sound, only a few organizations provided follow-up assistance. These CHF member organizations helped communities dig wells and ponds and assisted in canal and pond rehabilitation in Prey Veng and Pursat provinces.136 Based on the CHF assessment, Caritas Cambodia sent out an emergency appeal for financial assistance for the drought-affected areas. Through this emer- gency appeal, Caritas Cambodia received almost $400,000 and was able to support 10,800 families in 24 districts and seven provinces by providing drought victims with drinking water, solar energy installations, tube wells for emergency relief, and water treatment and storage equipment. 135 1:Khmer Buddhist Association (KBA), 2: Vulnerability and Illiteracy Reduction (VIR), 3: Partnership for Development in Kampuchea (PADEK), 4: Alliance Association of Rural Restoration (AARR), 5: Akphivath Neary Khmer Organization (ANKO), 6: Environment Protection and Development, 7: Organization (EPDO), 8: Society for Action and Change for Rural Education and Development (SACRED), 9: Ponleu Ney Kdey Sangkhum (PNKS), 10: Phnom Neang Kangrei Association (PNKA), and 11: Development Association Resource Economic (DARE). 136 Cambodian Humanitarian Forum (2016). CHAPTER 7 There Are Still Areas to Improve The 2014–2016 El Niño–related drought exposed areas to improve in Cambodia’s ENSO preparedness efforts. While the severity of the drought could not have been predicted, international forecasts did predict there would be an El Niño during that time.137 Unfortunately, little was done to prepare for this possibility. As such, when dry conditions turned to drought, the government could have been better prepared. It was only in 2016 that emergency response measures were implemented, but these only helped affected regions cope and recover (Figure 17). They did little to prepare the country for future events. During this process several specific areas to strengthen ENSO preparedness were identified. These are described next. Cambodia needs better data and comprehensive vulnerability maps. Although the government has collected data on drought and ENSO and carried out ranking exer- cises, there are still no ENSO vulnerability maps.138 The government’s initial data col- prone areas and assessments lection efforts provided information on different disaster-­ on why certain areas recovered and others did not. The ranking exercises were con- ducted at the provincial level to determine which provinces are vulnerable to droughts. Despite this, more information is needed to help farmers prepare for El Niño. This includes ENSO forecasting for specific seasons, locations, and time periods. Related to this, the Department of Meteorology could develop their capacity to create practical and updated drought maps for users and the public.139 Various government plans and strategy documents associated with ENSO or slow onset droughts have not been fully implemented. For example, MOWRAM is still 137 to Annex 4 for potential sources for ENSO forecasts. Refer 138 Basedon interviews with officials in the Department of Climate Change on October 10, 2017. 139 UN-ESCAP (2016).   41 42 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture in the process of developing a nationwide early warning system for drought. Such a system is critical, particularly for rural communities that depend on rainfall for water needs and to farmers and fishers for their livelihoods.140 Infrastructure, especially small-scale irrigation facilities, must be maintained and upgraded as prioritized in the HRF’s Contingency Plan. Also, government and local response capacity could improve to properly prepare for ENSO-related climate events. This, despite being a priority of the NAPA, NSDP, NAP-DRR, SNAP-DRR, HRF Contingency Plan, and the Law on Disaster Management. Capacity building is also a stated aim of the NCDM, the Min- istry of Women’s Affairs, and many international development partners working in Cambodia. Cambodia lacks sufficient human and institutional capacity to properly prepare for ENSO events, especially at the community level. More specifically, there is not enough technical expertise to manage ENSO- or disaster-related databases and other systems that could produce timely information on impending disasters. There is insuf- ficient in-country capacity to take advantage of globally available information and new computer software and hardware for better forecasting. Moreover, at all levels of government, there could be greater awareness on the challenges and preparedness strategies for ENSO events. This is especially true at the local or village level where El Niño and La Niña are unfamiliar concepts, but concepts that have very real con- sequences for communities. Government extension workers are frontline responders to ENSO events, but most are not trained for ENSO-type emergencies. Most of their training has been in extreme weather events like cyclones and flooding, not slow-on- set drought, changing seasons, or persistent water scarcity. Moreover, the Cambodian hydrometeorological services could improve its ability to provide timely and accurate early warning information for farmers. There is a lack of ENSO-related information in Cambodia. Most of the climate-­ related data and information that exists in Cambodia are on natural disasters and not specifically related to ENSO. This appears to have impeded the response to past ENSOs because policy makers were not prepared for potential ENSO events, despite information existing on when ENSOs would likely occur. Considering that most of the rice and rice-based cropping systems in Cambodia are rainfed, the impact of El Niño on agriculture can be severe if it affects the length of the rainy season. Such infor- mation is needed, particularly at the local level, where farmers and fishers must be prepared for potential ENSO challenges 140 Chhinh (2015); ACAPS (2016). There Are Still Areas to Improve 43 There could be better cooperation on ENSO-related issues among government agencies at all levels and with nongovernmental partners. Cambodia’s approach to water resource management would benefit from being more integrated. For example, the MAFF and MOWRAM work together on integrated water resource management for the sustainable use of water resources, but this cooperation should include all rel- evant agencies, including the MoE, the NCDM, and local partners. Currently, only MOWRAM directly looks at ENSO, and there is no mechanism to cooperate with other institutions on ENSO-related issues. Such a mechanism would raise awareness on ENSO and prevent overlap on preparedness and response measures. Furthermore, there is a need for better coordination between the national level and provincial or community levels.141 This includes national-level participation in local forums and grassroots involvement in national dialogues around ENSO issues. Scaling up good practices and making them sustainable has been a challenge. There are many good NGO initiatives and internationally funded projects that could be expanded. These projects have been tested and piloted in many regions during many climate disasters, but most end when funding is exhausted. There are no dedicated institutions for learning from these efforts and no funding sources to scale up the best practices. Efforts have been made by the government to integrate gender into disaster pre- vention, but more could be done, especially in terms of ENSO. Research for this report uncovered little information on gender integration in ENSO issues.142 This partly reflects the general lack of a codified focus on ENSO, and the general under- representation of women in climate and natural resource management interventions, despite women’s large roles in both agriculture and natural resource management. That said, the government has made advances in this area and several local-level projects focus on women. For example, women’s groups in Kratie and Preah Vihea provinces received agricultural and water-use training in UNDP’s Promoting Climate-Resilient Water Management and Agricultural Practices project. Women participants showed how their gardening and water-saving techniques could be a benefit to the project.143 141 Hou (2016). 142 The CGE model is used to estimate poverty changes for male- and female-headed households. It should be noted that the CGE model itself is not disaggregated by gender, but it is linked to a survey- based microsimulation module, which includes more detailed information on all households. Changes in consumption spending for each household group in the CGE model are imposed on their corresponding individual households in the survey. Total consumption and poverty status for each household in the survey is then recalculated before and after the ENSO shock or policy intervention. 143 Hing (2016). CHAPTER 8 Policy Interventions Help Negate ENSO-Related Losses There are additional policy interventions that can help negate ENSO-related losses. This section simulates the effectiveness of six such policy approaches. These include: (i) Introducing drought-tolerant crop varieties to limit on-farm production losses. (ii) Expanding irrigation, which currently only covers 12 percent of rice lands in the Plains region and 10 percent of other regions’ rice lands. (iii) Restricting rice exports to curb food price increases in domestic markets, as the government did in 2001, 2008, and 2011. Rice export restrictions should be, at most, a short-term mea- sure that is removed once a crisis passes. This is because export bans carry longer term trade risks and can adversely affect food security in other countries. (iv) Storing more grains and distributing them through domestic markets. From the mid-1990s to the mid-2000s, Cambodia regularly added to and depleted national grain stocks, but since then this practice has lessened.144 This suggests the country may require new investments in storage systems and facilities, the costs of which are not considered here. As with export restrictions, intervening in grain markets benefits consumers but reduces the prices paid to farmers, therefore caution is warranted, even during emer- gencies. (v) Providing social transfers, which, in this simulation, refers to expanding social protection programs by providing conditional cash transfers to poor households. Currently, the government’s various social protection and pension schemes transfer, on average, about $1.2 per person per year. These schemes target poor households and only represent 1 percent of household income. Finally, (vi) Implementing all of the above policy interventions simultaneously.145 The model considers costs associated with implementing these interventions if they specifically mitigate ENSO’s impacts 144 This is according to FAO Food Balance sheets, which end in 2013. 145 simulations are indicative of broad categories of on-farm, market and social policies. For instance, cash transfers could be replaced by food aid or crop insurance payments to smallholders, and irrigation infrastructure could be replaced by improvements in crop water use efficiency.   45 46 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture but does not consider infrastructure investments, which lead to a stream of benefits that are not limited to ENSO years. See Annex 1 for methodological specifics for each of these simulated interventions. Introducing drought-tolerant seed varieties is an effective policy measure to miti- gate GDP losses from a strong El Niño. Table 7 reports GDP losses (in absolute and percentage terms) with and without the effects of the policy interventions. On-farm investments, like drought-tolerant seeds and irrigation infrastructure, are the most effective simulated interventions at limiting El Niño–related GDP losses in Cambodia. Introducing drought-tolerant varieties reduces national GDP losses from $61 million to $34 million. Expanding irrigation use is not as effective, but also significantly off- sets GDP losses by making crop production more resilient to climate shocks. Irriga- tion also makes agriculture more resilient by raising yields during normal years and protecting yields during drought years. Overall, when all six policy interventions are combined and implemented concurrently, there are still national GDP losses equal to $31 million, and food system GDP losses equal to $46 million during El Niño events. Distributing stored grains, restricting rice exports, and providing social trans- fers are less effective in mitigating El Niño–related GDP losses. The introduction of stored grains into the local market increases competition for farmers and lowers producer prices, hence it is ineffectual in reducing El Niño–related GDP losses, espe- cially in agriculture. Restricting exports worsens El Niño–related GDP losses because it reduces rice exporter earnings. Likewise, cash transfers are also ineffective because TABLE 7: GDP Changes During El Niño Events and Intervention Scenarios. With interventions Drought- Without tolerant Additional Export Grain Social All interventions varieties irrigation restrictions storage transfers combined Percentage change in GDP (%) National –0.39 –0.22 –0.34 –0.41 –0.39 –0.39 –0.20 AFS –1.24 –0.69 –1.08 –1.42 –1.28 –1.23 –0.81 Agriculture –1.39 –0.80 –1.22 –1.40 –1.54 –1.37 –0.80 Absolute change in GDP (US$ million) National –60.88 –33.85 –53.29 –64.16 –61.06 –60.62 –30.72 AFS –70.59 –39.55 –61.57 –81.19 –73.12 –70.19 –46.04 Agriculture –65.54 –37.67 –57.83 –66.20 –72.84 –64.86 –37.82 Source: Simulation results from the Cambodia CGE model. Policy Interventions Help Negate ENSO-Related Losses 47 FIGURE 18: Household Consumption Losses During El Niño Events and Intervention Scenarios (gains if negative). 1.75% 1.78% 1.59% 1.47% 1.77% 1.24% 1.08% 1.58% 1.55% 1.30% 1.06% 1.03% 0.14% –0.60% Without Drought-tolerant Additional Export Grain Social Combined interventions varieties irrigation restrictions storage transfers With interventions All households Poorest quintile Source: Simulation results from the Cambodia CGE model. they do not increase agricultural production and require offsetting tax increases for higher income households. However, as shown in Figures 18 and 20, there are consumption-side and poverty benefits from expanding from cash transfers. ­ Social transfers are effective at reaching poor consumers. Figure 18 shows con- sumption, or welfare, losses from El Niño events for all households compared to the poorest one-fifth of households. Cash transfers leave total consumption losses for all households largely unchanged: a 1.77 percent loss with no policy and 1.78 percent loss with cash transfers. This means that outcomes for higher income households worsen because of cash transfers as poorer households benefit. This is because additional social transfers are financed through higher taxes for more affluent households, who comprise most of the national tax base. In other words, the simulated cash transfer pro- gram is designed to be a transfer from higher to lower income households, which also explains why there was little change in national GDP losses under this scenario. That said, social trQansfers were the most effective policy for minimizing welfare losses for the poor. Introducing drought-tolerant seed varieties, was effective in minimizing welfare losses during El Niño for all households, including the poorest. When all the simulated policy scenarios are implemented, total consumption losses during El Niño are almost eliminated. Figure 18 shows that consumption losses from El Niño drop from 1.77 percent to 0.14 percent for all households when all 48 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 19: Household Consumption Losses by Expenditure Quintile with and without All Interventions Combined (gains if negative). Q1 Is the Poorest Quintile, Q5 Is the Wealthiest Quintile. 1.8% 1.8% 1.7% 1.8% 1.7% Severe El Niño event 0.5% 0.3% All interventions combined –0.2% –0.3% –0.6% Q1 Q2 Q3 Q4 Q5 National per capita consumption quintiles Source: Simulation results from the Cambodia CGE model. policies are applied. Figure 19 shows that the poorest households benefit the most, but that wealthier households also benefit greatly. In fact, on-farm investments, designed to make the agricultural sector more resilient, and social transfers, designed to cush- ion short-term consumption shocks, may be sufficient in negating most ENSO-related welfare losses. Introducing drought-tolerant seeds and expanding cash transfers are effective interventions at mitigating poverty increases from El Niño. The immediate impact of El Niño events on poor households can be measured by changes in the national pov- erty headcount rate, which shows the share of the population living below the national poverty line. Figure 20 shows how interventions reduce El Niño–related poverty rate increases and the number of poor people during an El Niño. Without interventions to mitigate impacts, a typical El Niño event causes the national poverty rate to increase by 1.4 percentage points. This is equivalent to 210,000 additional people living below the poverty line during that period. Implementing all interventions together would actually reduce the overall poverty rate by 30,000 people, in spite of the El Niño. Applying a rice export ban mitigates poverty losses the least among the simulated pol- icy options because they do not, directly or indirectly, target the poorest households. These bans can also adversely affect international rice markets. Policy Interventions Help Negate ENSO-Related Losses 49 FIGURE 20: Changes in National Poverty Headcount Rate and the Number of Poor People During Strong El Niño Events and Intervention Scenarios (percentage points and millions of people). 0.21 0.18 0.18 0.16 1.4% 0.14 0.13 1.2% 1.2% 1.1% 0.9% 0.9% –0.2% –0.03 Without Drought-tolerant Additional Export Grain Social Combined interventions varieties irrigation restrictions storage transfers With interventions Number of poor people Poverty rate Source: Simulation results from the Cambodia CGE model. TABLE 8: Rural and Urban Household Consumption Changes During El Niño Events and Intervention Scenarios (percentages). National Rural Urban Without interventions –1.75 –1.89 –1.54 Drought-tolerant varieties –1.08 –1.17 –0.95 Additional irrigation –1.59 –1.72 –1.41 Export restrictions –1.47 –1.64 –1.24 Grain storage –1.24 –1.56 –0.79 Social transfers –1.78 –1.75 –1.82 Combined –0.14 –0.23 –0.01 Source: Simulation results from the Cambodia CGE model. Both urban and rural households have consumption-side benefits from policy interventions, but generally El Niño affects rural consumption more. Table 8 shows the relative consumption losses under the various policy scenarios. Urban household welfare benefits more from policy interventions that reduce consumer prices, such as depleting grain stores and banning rice exports. However, flooding 50 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 21: Changes in Poverty Headcount Rate During Strong El Niño Events and Intervention Scenarios (percentage points). 1.7 1.7 1.7 1.6 1.6 1.4 1.4 1.3 1.2 1.2 1.1 1.1 1.0 1.0 1.0 0.9 0.9 0.9 0.9 0.8 0.8 0.90.90.9 –0.1 –0.3 –0.2 –0.5 Without Drought-tolerant Additional Export Grain Social Combined interventions varieties irrigation restrictions storage transfers Plain Lake Coastal Mountainous Source: Simulation results from the Cambodia CGE model. the local market with cheap grain and banning exports can adversely affect producers and regional trading partners. Interventions that target production and cash transfers benefit urban households the least. Social transfers actually exacerbate welfare losses for urban households because urban households tend to be wealthier than rural house- holds, and therefore finance most of the transfer costs through higher taxes. However, when all interventions are applied together, welfare losses for urban households are eliminated completely, despite the El Niño. The impacts from policy interventions on poverty outcomes vary across regions. Figure 21 shows El Niño causes poverty increases in all regions, but the most in the Lake and Mountain regions. That said, poor households in the Lake and Mountain regions benefit the most when all policies are applied. Introducing drought tolerant crop varieties benefits households in all regions, but this is especially effective in the Lake region. Irrigation, rice export bans, and depleting grain stores have little effect on poverty in the Mountain region, but these policies mitigate poverty increases substan- tially in the Lake region. More female-headed than male-headed households enter poverty during El Niño, but female-headed households stand to benefit the most from policy interventions. Figure 22 shows how poverty headcount rates change during strong El Niño events. Policy Interventions Help Negate ENSO-Related Losses 51 FIGURE 22: Change in Poverty Headcount Rates During Strong El Niño Events and Intervention Scenarios by Location and Gender of the Household Head (percentage points). 1.7 1.7 1.5 1.4 1.4 1.4 1.4 1.4 1.4 1.3 1.3 1.3 1.2 1.2 1.2 1.2 1.2 1.1 1.0 1.0 1.1 1.1 0.9 0.8 –0.2 –0.2 –0.3 –0.3 Without Drought-tolerant Additional Export Grain Social Combined interventions varieties irrigation restrictions storage transfers National male National female Rural male Rural female Source: Simulation results from the Cambodia CGE and microsimulation model. Both male- and female-headed households benefit from policy interventions, but when all policies are applied, female-headed households benefit the most. Again, introduc- ing drought resistant seed varieties is the most effective policy option in mitigating poverty losses during El Niño. Social transfers, which directly target the poorest, who in this case are female-headed households, are also very effective. Distributing stored grains is slightly more beneficial for male-headed households. CHAPTER 9 The Government Can Take Action to Improve ENSO Preparedness and Resilience There are many opportunities to improve ENSO preparedness and resilience. In the section below these opportunities are divided into two groups: preparedness and resilience. While there is obvious overlap between these two concepts, for the pur- poses of this report they are defined as the following. Preparedness are measures spe- cifically geared toward ENSO and should, ideally, be in place before the next ENSO event occurs. These actions will significantly empower people to cope, respond, and recover from damaging ENSO events. Resilience, by contrast, are measures that are not specifically designed to respond to ENSO, but that will build individuals’ and organizations’ ability to adapt to multiple forms of risks and shocks without compro- mising long-term development. Included in this section are various best practices from around the world that can be emulated. Table 9 outlines these recommendations and prescribes steps that should be taken for each. Preparedness Prepare response measures for when ENSO-related climate events occur. Response measures take place during an emergency and include actions to save lives and prevent further property damage. A proper response system requires several plans and actions that should be in place prior to an El Niño or La Niña event. These include, first, establishing a high-level interministerial committee or commission tasked with ENSO response and preparedness. Second, developing local drought and flood contingency plans that can be activated when ENSO-related climate events occur. Third, making sure contingency budgets, or other ENSO-related budget lines, can be accessed to pro- vide humanitarian or relief support to affected areas and populations. Fourth, ensuring   53 54 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture that emergency response inputs and supplies are available. This includes relief sup- plies, such as water and food, and agricultural inputs, such as seeds and fertilizers. Fifth, ensuring that infrastructure needed for emergency responses—main roads, water ports, and other connectivity-related infrastructure—is sufficiently climate proofed. These will safeguard “last mile” delivery capacity. The government could develop a comprehensive framework for managing ENSO risk. The government has not developed a focused approach, nationally or subnation- ally, on El Niño or La Niña. Most of the country’s climate-related policies and pro- grams focus on naturally occurring natural hazards, such as floods, cyclones, and some drought. This lack of ENSO awareness contributed to a response to the 2014–2016 El Niño that could have been stronger. Moreover, Cambodia’s ENSO efforts were mostly ex post disaster relief. Less has been done to prepare for slow onset disasters, like drought, and this research uncovered no direct efforts to specifically prepare for ENSO events. To improve upon this, the government could appoint an ENSO task force, headed by MOWRAM or the Prime Minister’s office and with representatives from various government agencies, to develop a comprehensive strategy to enhance the country’s ENSO preparedness. This task force could emulate the Philippines’ El Niño Task Force, which developed a Roadmap to Address the Impacts of El Niño (RAIN). RAIN is described in Box 5. Harnessing the rebound from La Niña would significantly mitigate the damage from El Niño events. In other southeast Asia countries where CGE modeling was applied for ENSO, production increases during La Niña did not cancel out El Niño losses. However, in Cambodia they nearly do. Simulations show that La Niña events increase rainfall and positively impact agricultural production in Cambodia nearly as much as El Niño harms it. As such, there is an opportunity to take advantage of positive La Niña–related impacts, like increased rainfall and solar radiation. This could be done by expanding planting and improving water catchment during La Niña (see Box 6 for Brazil’s water catchment strategies). It would also be important to implement flood prevention policies during La Niña. The modeling does not account for damaging floods, but as discussed, these are a pervasive threat in Cambodia, especially during La Niña. Therefore, according to simulations, if policies are enacted to mitigate El Niño’s negative impacts and harness La Niña’s rebound while limiting flood damage, Cambodia could actually benefit from ENSO events. Invest in early, accurate forecasting to help farmers prepare for ENSO events. This forecasting should go beyond EWS 1294 (Box 4), which focuses on quick onset droughts and agriculture. ENSO forecasts are onset disasters, to focus on slow-­ The Government Can Take Action to Improve ENSO Preparedness and Resilience 55 BOX 5: Roadmap to Address the Impacts of El Niño (RAIN) in the Philippines. The Philippines developed a Roadmap to Address the Impact of El Niño (RAIN), which directly addresses ENSO. The Philippines’ National Economic and Development Authority (NEDA) leads the government’s El Niño Task Force, which developed RAIN. The president appointed146 NEDA to lead the interagency Task Force, which also has memberships from a variety of government agencies,147 during the 2015/2016 El Niño. NEDA’s appointment shows the government views ENSO as an economic development issue not just an agricul- ture issue. RAIN focuses on 67 provinces affected by El Niño and has several components, including cloud seeding, Cash-for-Work, food distribution, public information, water system improvements, and seeds and fertilizer distribution. NEDA claims148 RAIN stabilized food prices and food supplies by supporting crop production in unaffected or mildly affected regions at times when other areas were broadly impacted by drought. During the most recent El Niño, NEDA recommended allotting $165 million for 2015 and $257 million for 2016 to prepare for El Niño.149 Allocations included $83 million to manage water supplies, $20 mil- lion to provide food stamps, and another $40 million to support affected urban households. available from various global and regional technical agencies in both text and graphic formats, providing multi-model ENSO predictions with reasonable confidence levels and diagnostic interpretations.150 MOWRAM can use these data to interpret ENSO’s impacts on climate at national and subnational levels. MOWRAM’s Department of Meteorology already issues daily weather forecast information in sector-specific visu- alization products through the Cambodia Meteo website151 for fisheries, transportation, and tourism (see Figure 24 in Annex 3). MOWRAM could invest further in developing seasonal forecast products and expanding the sectoral coverage to include agriculture. To do so, MOWRAM may need to invest in both infrastructure and human capacity, particularly to improve dissemination and communication mechanisms. See Box 7 for how South Korea integrated drought and agricultural forecasting into actionable steps. 146 Philippine Star. (2016, April). Palace: Gov’t started preparing for El Niño in August 2015. Retrieved November 2017, from Philippine Star: http://www.philstar.com/headlines/2016/04/03/1569015/ palace-govt-started-preparing-el-nino-august-2015. 147 Other members of the El Niño Task Force include: Bureau of Animal Industry, BFAR, Bureau of Soils and Water Management (BSWM), National Disaster Risk Reduction and Management Council (NDRRMC), Mindanao Rural Development Program, Philippine Carabao Center, Philippine Center for Postharvest Development and Mechanization, Philippine Crop Insurance Corporation (PCIC), Philippine Rice Research Institute (PhilRice), and Sugar Regulatory Board. 148 NEDA (2016). 149 Philippine Star. (2015, October). P19.2-B budget recommended to fight El Niño. Retrieved October 2017, from Philippine Star: http://www.philstar.com/headlines/2015/10/15/1510871/ p19.2-b-budget-recommended-fight-el-nino. 150 https://reliefweb.int/sites/reliefweb.int/files/resources/El%20Nino%20report-%20finalized%20 ESCAP07082017.pdf 151 http://www.cambodiameteo.com 56 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture BOX 6: Rainwater Harvesting in Brazil. To avoid water shortages in the semiarid tropics of northeastern Brazil, efforts have been made to harvest rainwater. In this part of Brazil, annual rainfall varies widely across regions, from 200 to 1,000 mm, causing unreliable rural drinking water supplies, especially in rural areas. In response, a group of NGOs collaborated with the Brazilian government to construct one million rainwater tanks in the region. Most of these tanks were made of precast concrete plates or wire mesh concrete. The project was implemented over a five-year period and ben- efitted around five million people. Over the years, rainwater harvesting has been integrated Brazil—even into educational programs for sustainable living in the semiarid regions of ­ spreading to other urban areas. Over the past decade, many NGOs and grassroots organi- zations have focused their work on developing rainwater harvesting, and the irrigation of small-scale agriculture using subsurface impoundments. Source: UN Environmental Program, “Rainwater Harvesting and Utilization,” http://www.unep.or.jp/ ietc/Publications/Urban/UrbanEnv-2/9.asp BOX 7: Turning Forecasting into Action: South Korea’s Anticipatory Response. The Korean government fully integrated drought forecasting into a comprehensive Agricul- tural Drought Policy and Action Plan. Since 2010, South Korea has experienced an increase in droughts—the worst taking place in 2015. This plan, launched in 2015 and updated in 2017, includes (1) agricultural drought monitoring, (2) drought damage evaluation, and (3) an anticipatory action plan. The Ministry of Public Safety and Security oversees the forecasting system, known as the Agriculture Drought Forecasting and Warning System, which integrates meteorological and water-level data. The ministry coordinates closely with other government agencies. The forecasting system offers new information on drought preparedness and mitigation planning. This includes a long-term plan and water-saving strategies. It analyzes previous droughts using agricultural drought maps, predicts drought damages, and considers capac- ity constraints, such as those in irrigation facilities. The government uses these forecasting data to prioritize drought-prone regions and target assistance to the most vulnerable areas. The government also implements an irrigation plan to guide farmers on what crops are best and what are optimal planting times. The Action Plan emphasizes education campaigns for farmers on water management techniques. This campaign includes in-person workshops, animated videos, and printed promotional material. This new approach helps the Korean government better coordinate among agencies, determine drought responses, and improve the quality of decision making for effective drought management. Sources: Jong Hoon Park et al. 2016. “Agricultural Drought Forecasting and Early Warning System,” World Irrigation Forum http://www.icid.org/wif2_full_papers/wif2_w.2.2.23.pdf; Lee Kwangya. May 2018. “Agricultural Drought Policy and Action Plan: Republic of Korea” PowerPoint presentation. The Government Can Take Action to Improve ENSO Preparedness and Resilience 57 Enhance extension services to expand the use of drought-tolerant seeds. The CGE modeling showed that expanding the use of drought-tolerant seeds can generate large economic benefits during El Niño and other drought events. Using drought-­ tolerant seeds could also reduce production losses and reduce the number of people that fall into poverty during El Niño. The International Rice Research Institute (IRRI), working together with national research institutes, developed and disseminated drought-tolerant rice varieties in Cambodia, but the program is still limited. Scaling up the development and disbursement of these seeds would require improvements in both seed and farmer extension systems. It would also require additional research on which seeds could produce the greatest benefits. Box 8 shows how Sub-Saharan African nations successfully implemented such a large-scale drought-tolerant seed disbursement system. Efforts could be made to smooth agricultural price shocks, potentially by dis- tributing stored grains during El Niño. Managing food and agriculture price BOX 8: The Drought Tolerant Maize for Africa (DTMA) Project. The Drought Tolerant Maize for Africa (DTMA) project introduced drought-tolerant seeds to smallholder farmers in Africa. Taking place from 2007 to 2015, it developed and dis- seminated over 200 new drought-tolerant, high-yielding, disease-prone, and locally adapted maize varieties to 13 African countries. By the time the project concluded, 72 percent of all maize grown in Sub-Saharan Africa was derived from these seeds. In 2014 alone, the project produced nearly 54,000 tons of certified drought-tolerant maize seeds, benefitting an esti- mated 5.4 million households, or 43 million people, across DTMA countries. Compared to the eight years before the project, Malawi and Ethiopia doubled their maize yields. During the project years, Rwanda and Burkina Faso, which increased agricultural investment and established conducive policy environments during those years, were the only Sub-Saharan African countries that also showed increases in maize production. This underlies the effec- tiveness of the DTMA. There were some challenges and constraints with DTMA. For example, there were seed shortages, high seed prices, a scarcity of information, a lack of seed variety options, and an overall shortage of resources. Also, it is difficult to attribute the yield improvements solely to DTMA. Other factors may have also played roles, for example, conducive policy envi- ronments and favorable weather conditions. The project was funded by the Bill and Melinda Gates Foundation and jointly implemented through National Agricultural Research systems by the International Maize and Wheat Improvement Center (CIMMYT) and the Interna- tional Institute of Tropical Agriculture (IITA). Source: CIMMYT. A Quarterly Bulletin of the Drought Tolerant Maize for Africa Project, Vol. 4 No. 4. December 2015. 58 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture fluctuations should be a policy during ENSO events. The CGE analysis shows that even small changes in production levels can lead to large changes in food prices, which have severely adverse effects on household well-being. Improved farm technologies and management cannot completely mitigate production losses, so other mechanisms should be sought to smooth price fluctuations and offset food shortfalls. However, the CGE analysis shows that restricting rice exports has little effect on GDP; therefore, it may be more beneficial to increase grain storage capacity so the grains can be distrib- uted during production shocks. That said, caution should be taken when manipulating grain markets since this may adversely impact producers. An adequate rural transpor- tation system would be needed to quickly transport these grains to ENSO-­ affected regions. The CGE analysis shows utilizing grain stocks achieves only modest benefits, but this may be because of Cambodia’s limited grain stocking capacity. As mentioned, FAO Food Balance sheets show stock depletions in Cambodia have not exceeded 100,000 tons since the mid-2000s, a relatively small amount compared to what would be required to achieve greater benefits.152 As a result, assessments could be made on the cost-effectiveness of building additional grain storage capacity. Box 9 highlights China’s grain reserve system. Expanding social safety nets could mitigate poverty increases during El Niño. Results from the CGE analysis show that social transfers would reduce the num- ber of people who fall into poverty during ENSO events. However, Cambodia’s social safety net coverage is low. In 2011, the government issued policy guidelines to implement social protection policies, but the implementation and coverage of these guidelines were limited because of knowledge gaps, resource constraints, poor coordination, and institutional weaknesses. As such, any social safety net expansion would also require increased funding, skills development, and capacity building. Initial efforts to expand social safety nets should target rural areas where most of the country’s impoverished are located. This pro-poor targeting could come in the form of food aid, food-for-work programs, or expanded labor market programs, and not just cash transfers.153 Moreover, these programs would need to be adjusted and scaled up during ENSO to make them more responsive to droughts and other impacts. Box 10 shows how an Ethiopian social safety net program was adjusted to improve the country’s climate resilience. 152 FAO (2018). 153 Chey Tech (2012); Suy and others (2018). The Government Can Take Action to Improve ENSO Preparedness and Resilience 59 BOX 9: China’s Grain Reserve System. China’s national grain reserve system plays an important role in safeguarding the grain market from rising food prices. The system consists of two types of grain reserve pro- grams: the national temporary grain reserve program and the national strategic grain reserve program. The national temporary grain reserve program allows China to cope with grain yield reductions from natural disasters and stabilize the grain market during supply shocks. This allowed China to escape the steep grain price increases that hit other Asia-Pacific countries. China’s vertically managed grain reserve system has two levels of legal entities—the China Grain Reserves Corporation, or Sinograin, at the national level, and the provincial grain reserve corporations at the subnational level. The reserve corporations have a network of branch offices in major grain producing and consuming areas. In August 2003, China enacted regulations on the administration of central grain reserves. These regulations clar- ified legal responsibilities and grain use, storage, inspection, and supervision processes. In 2004, China liberalized its grain markets and, until recently, Sinograin and its provincial entities purchased grains from producers at floor prices, which were preset each year by the central government. As a result, China accumulated large grain reserves and, during major droughts in 2009 and 2010, the central government sent 1.4 million tons of reserve grains to drought-stricken southern provinces to alleviate local food shortages. This successfully stabilized local grain prices, protecting farmers’ incomes, mitigating the impact of steep price increases on poor consumers, and ensuring social and political stability. However, introducing such a large influx of cheap grain also created market distortions that can harm producers. Sources: Qin Zhongchun. 2009. Formation and Features of China’s New Grain Reserve System; FAO. 2009. “Country responses to the food security crisis: nature and preliminary implications of the policies pursued, Initiative on soaring food prices,” http://www.fao.org/fileadmin/user_upload/ISFP/ pdf_for_site_Country_Response_to_the_Food_Security.pdf Resilience Improve the coordination and capacity of government agencies and other insti- tutions to improve resilience for ENSO events. Government policies, programs, and departments dealing with ENSO and other climate shocks should be streamlined. A special committee, or national task force, could be devoted to addressing slow-onset disasters related to ENSO within MOWRAM. The committee could include members from different government agencies and civil society. Every effort should be made to provide “last mile” delivery of services and benefits to empower vulnerable commu- nities to prepare and take effective measures to reduce the adverse effects of ENSO. 60 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture BOX 10: Ethiopia’s Productive Safety Net Program Integrated with Early Warning Systems and Disaster Risk Management. Ethiopia’s Productive Safety Net Program (PSNP) is a large national social safety net (SSN) program that includes elements of climate resilience. The PSNP’s goal is to improve food security among Ethiopia’s poor, and mitigate the impacts from shorter term shocks, mainly droughts. The PNSP is implemented almost entirely through national government systems, which are decentralized through regional and local administrations. The unique aspect of this system is its incorporation of early warning and disaster risk management into its insti- tutional structure. The Ministry of Agriculture is responsible for program management, with the Disaster Risk Management and Food Security Sector tasked with overall program coordination. The Early Warning and Response Directorate provides early warning information on natural hazards and ensures the PSNP’s emergency responses are linked to relief and hazard response activ- ities. The Natural Resource Management Directorate oversees the public works and the Ministry of Finance and Economic Development oversees financial management.154 These federal implementation arrangements are replicated within the PSNP’s eight regions and 319 woredas (districts). Source: GFDRR 2013, “Ethiopia’s Productive Safety Net Program (PSNP) Integrating Disaster and Climate Risk Management,” http://www.wcdrr.org/wcdrr-data/uploads/482/SPL_DRM_TK_CS2_ Ethiopia%20PSNP.pdf There is a particular need for state-level stakeholders to work with provincial-level and local-level actors. Creating a better flow of information on ENSO-related challenges and solutions provides state-level actors with a greater understanding of local issues, and allows local-level actors to receive ENSO updates and gain capacity in national and international best practices. This information exchange could occur through forums or other dialogues. Given the newness of focusing on ENSO in Cambodia, these committees and forums would also create a better understanding of the roles and responsibilities of all stakeholders at all levels: local, national, and international. Box 11 highlights a coordination mechanism from Vietnam. The government should explore mechanisms for risk transfer and risk sharing. Risk transfer mechanisms are in their infancy in Cambodia. The successful155 pilot pro- gram for agricultural micro-insurance was launched in July 2015 by the Cambodian Centre for Study and Development in Agriculture (CEDAC) to insure farmers against 154 World Bank (2010). 155 An assessment of the micro-insurance pilot showed 153 farmers joined the scheme during its first season and 52 farmers received compensation for their drought-affected rice crops (Sokhorng, 2016). The Government Can Take Action to Improve ENSO Preparedness and Resilience 61 BOX 11: Integrating Disaster Risk Management and Climate Change Adaptation into the Social Economic Development Plans for the Tra Vinh Province in Vietnam. The Tra Vinh Provincial People’s Committee (PPC) mainstreamed Disaster Risk Manage- ment (DRM) and Climate Change Adaptation (CCA) into the National Target Program to Respond to Climate Change in Tra Vinh. Oxfam supported this process by supporting a local government-appointed Task Force, comprised of commune leaders and sector staff special- izing in CCA and DRM, and shared with them information on DRM and climate change vulnerability. As a result of these interactions, climate change, disaster vulnerabilities, and risk reduction issues that arose from the commune risk assessments were integrated into commune Social Economic Development Plans (SEDPs) and then district SEDPs. The Tra Vinh Department for Planning and Investment (DPI) drafted planning guidelines to integrate DRM, CCA, and gender into SEDPs. In 2014, with Oxfam support, DPI used this approach with 19 communes in two districts (Chau Thanh and Cau Ngang) to incorporate key DRM and CCA issues into their SEDPs. After several years, this participatory and integrated planning approach helped achieve national strategies and meet program targets. Similarly, people in the SEDP Planning Task Groups, technical support groups, and community-based groups benefitted from training, capacity building, and peer-to-peer support. The process integrated gender, climate change, and disaster risk into SEDP planning and raised awareness on these in local communities. Adapted from: Global Facility for Disaster Reduction and Recovery. 2017. “Toward Integrated Disaster Risk Management in Vietnam: Recommendations Based on the Drought and Saltwater Intrusion Crisis and the Case for Investing in Longer-Term Resilience.” World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/28871 License: CC BY 3.0 IGO Source: VUFO-NGO Resource Centre Vietnam 2015. rice crop losses from droughts, floods, or other climate events. With a total investment of $96,000, the insurance is only available in Kampot, Prey Veng, Takeo, Kampong Chnang, and Kampong Speu.156 However, during droughts or after destructive floods, financial resource delivery mechanisms, backed by a financial protection strategy, would ensure efficient and effective access to quick post-disaster resources for emergency response, especially for vulnerable and rural populations. Although these transfers would not be activated until a disaster strikes, the mechanisms would need to be in place beforehand. If the government decides to use an index-based agricultural insurance, they must dra- matically improve their data collection capability, including meteorological stations and damage assessment expertise. Box 12 highlights some of the shortcomings associated with India’s agricultural insurance system. 156 Raksmey and Kossov (2015). 62 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture BOX 12: Challenges Associated with Index-based Agricultural Insurance in India. Between 2012–2013, the National Agricultural Insurance Scheme, India’s national weather insurance program, covered 21 percent of farm holdings and 19 percent of crop acreage in the country. However, the insured sum covered only 5.5 percent of the value of total crop output. Because of this, insured farmers were not satisfied. However, to rectify this, data estimates indicate risk coverage would need to increase to 50 percent of crop output, and the program would be about ten times more expensive. Thus, the National Institution for Transforming India concluded that agriculture insurance was not working, and the country’s strategy should shift toward a comprehensive Agricultural Calamity Compensation Fund to offset partial crop losses faced by farmers. Under this new fund, the state would insure each crop field for yield or production losses, but premiums remain with the state rather than being paid to insurance. In this more cost-effective manner, instead of subsidizing the insurance premium for all farmers, the government is required to cover crop losses only for the farmers affected by the calamity.157 Invest in rural infrastructure, such as roads and irrigation systems. Modeling conducted for this report shows that increased investment in irrigation moderately mit- igates GDP losses during El Niño by making crop production more resilient to climate shocks. Irrigation also raises yields during normal years, further increasing longer term resilience. In Cambodia, most cropping systems are rainfed, with irrigation cov- ering only about 10 percent of rice paddies. In Vietnam, by contrast, over 80 percent of rice paddies are irrigated. This indicates Cambodia has room to expand significantly in irrigation infrastructure. Moreover, Alternative Wetting and Drying (AWD) techniques can be applied to irrigated fields to reduce water use without limiting yields. Box 13 described how AWD is being used successfully in Vietnam and Bangladesh. Although not explored in this report, there may also be opportunities to improve upon the rural transportation network by investing in road construction and maintenance. This would improve farmers’ access to input and output markets, and better link remote villages to services during ENSO events. Target women to lead agriculture projects or target them as beneficiaries. As discussed, much of Cambodia’s agricultural workforce is composed of women. They participate in every facet of agriculture and, as such, are well positioned to carry out agriculture-related ENSO preparation. This is because they are among the most knowledgeable of their crops, fish, and livestock and the climatic conditions of their 157 https://mail.google.com/mail/u/0/?tab=wm#inbox/164902e54cf622a5?projector=1&messagePartId=0.1 The Government Can Take Action to Improve ENSO Preparedness and Resilience 63 BOX 13: Alternative Wetting and Drying (AWD) in Vietnam and Bangladesh Vietnam and Bangladesh successfully use AWD to limit water use in irrigation. AWD is a paddy rice management practice that can enhance yields, improve climate hazard resilience, and reduce greenhouse gas emissions. In AWD, irrigation water is used a few days after soil moisture becomes low. Hence, the rice paddy becomes alternately flooded and dry. AWD reduces the number of times a field is irrigated, reducing water fees and allowing farmers to adapt to water scarcity challenges. For Bangladesh, AWD created water savings of 22 to 26 percent, saving between 2,580 and 3,590 cubic meters of water per hectare. For Vietnam, AWD created water savings of 40 to 50 percent.158 Source: Dinesh, D., Frid-Nielsen, S., Norman, J., Mutamba, M., Loboguerrero, A. M., and Campbell, B. M., 2015. “Is Climate-Smart Agriculture effective? A review of selected cases.” CGIAR. local environment. They are also technically proficient counterparts in agriculture at the local level. In addition, as we have seen, women lag behind men in many key socioeconomic indicators, such as income and access to credit. In this sense, targeting women creates the double benefit of establishing project counterparts and building the resilience of a vulnerable group. Cambodia should cooperate with other southeast Asian countries on ENSO-­ related challenges, which are regional in nature. For example, ENSO-related flooding, salinization, and water scarcity issues are persistent in all the countries of the Mekong Delta, which runs from Vietnam to China, and through Cambodia.159 Increased regional cooperation to develop preparedness measures and meteorological and hydrological services and forecasting would lower common infrastructure costs, and allow for more knowledge exchange on best practices. There are already a num- ber of regional and global institutions that could facilitate such cooperation. These include: • The World Meteorological Organization (WMO), which houses the Southeast Asia Flash Flood Guidance System, the European Centre for Medium-Range Weather Forecasts (ECMWF), and the Global Flood Awareness System; • The United Nations Economic and Social Commission for Asia and the Pacific (ESCAP), which coordinates a regional Typhoon Committee;160 158 Basak (2016). 159 World Bank GFDRR June 2018, “Strengthening the Regional Dimension of Hydromet Services in Southeast Asia: A Policy Note with a Focus on Cambodia, Lao PDR, and Vietnam.” 160 14 members: Cambodia; China; Democratic People’s Republic of Korea; Hong Kong SAR, China; Japan; Lao PDR; Macao SAR, China; Malaysia; the Philippines; Republic of Korea; Singapore; Thailand; Vietnam; and the United States. 64 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture • The Association of Southeast Asian Nations (ASEAN), which coordinates a Regional Climate Outlook Forum, runs a Subcommittee on Meteorology and Geo- physics, and implements the Southeast Asia Radar Network and Composite project; • The Mekong River Commission (MRC), which shares hydrological data among the Mekong basin countries, including Thailand, Vietnam, Cambodia, and Lao PDR;161 • The Asian Disaster Preparedness Center, which facilitates satellite imagery and geospatial technology exchanges among Mekong Basin countries;162 and • The World Bank, which is developing a South East Asia flood monitoring and risk assessment platform, or the SEA-DRIF Platform, to assess near-real-time flood impacts in Myanmar, Cambodia, and Lao PDR.163 TABLE 9: Recommendations and Proposed Actions to Build pre-ENSO Preparedness and Resilience. In Table 9, recommendations are divided into two groups: preparedness and resilience Recommendations in green are a high priority, recommendations in tan are a moderate priority. The last two columns denote which actions are short term (S), or should be completed within a year, and which actions are medium-to-long term (M/L), or would not be achievable in less than a year. Recommendation Actions S M/L Prepare response • Establish a high-level ministerial committee or measures for when commission tasked with ENSO response and X ENSO-related preparedness. climate events • Develop local drought and flood contingency plans occur that can be activated when ENSO-related climate X events occur. • Make sure contingency budgets, or other ENSO- Preparedness related budget lines, can be accessed to provide X humanitarian or relief support to affected areas and populations. • Ensure emergency response inputs and supplies are available. This includes relief supplies, such as water X and food, and agricultural inputs, such as seeds and fertilizers. • Ensure infrastructure needed for emergency responses—main roads, water ports, and other X connectivity-related infrastructure—is sufficiently climate proofed. 161 http://www.mrcmekong.org/about-mrc/ 162 World Bank GFDRR, June 2018, “Strengthening the Regional Dimension of Hydromet Services in Southeast Asia: A Policy Note with a Focus on Cambodia, Lao PDR, and Vietnam.” 163 Deltares, “South-East Asia platform for NRT flood impact assessment,” May 2018 (working paper for the World Bank). The Government Can Take Action to Improve ENSO Preparedness and Resilience 65 Recommendation Actions S M/L Develop a • Appoint an ENSO task force, headed by MOWRAM or comprehensive the Prime Minister’s Office and with representatives ENSO framework from various government agencies, to develop a X comprehensive strategy to enhance the country’s ENSO preparedness. • Bring in technical experts to assist in composing this X framework. • Develop a strategy document and action plan, X including standard operating procedures. • Create a system to mobilize local partners for disaster responses and mandate government facilities be X made available during disasters. • Carry out current action plans on disaster response and climate resilience, including implementation of X the Law on Disaster Management. • Empower communities through effective and timely X provision of services and benefits. Harness La Niña’s • Appoint a committee or task force to explore X rebound agricultural strategies for La Niña. Preparedness • Make plans to rehabilitate water storage facilities, X increase water catchment, and expand planting. • Take measures to reduce flood risk in flood-prone X areas. Invest in • Bring together administrators of NCDM’s EWS 1294 X forecasting and MOWRAM’s EWS with ENSO technical experts. capacity • Identify areas to orient the systems toward ENSO. X • Identify forecasting needs, including infrastructure (meteorological stations, communication systems) X and human capacity (technical proficiency). • Compile ENSO forecasts from global and regional technical agencies in both text and graphic formats X to provide multi-model ENSO predictions.164 • Develop accurate ENSO risk maps. X • Use this data to interpret ENSO’s impacts at national X and subnational levels. • Broadcast this information to farmers across the country through EWS 1294 or another communication system. To do so, MOWRAM may need to invest in X both infrastructure and human capacity, particularly to improve dissemination and communication mechanisms. • Make this information accessible and understandable. X X (continued) 164 https://reliefweb.int/sites/reliefweb.int/files/resources/El%20Nino%20report-%20finalized%20 ESCAP07082017.pdf 66 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture TABLE 9:  Continued. Recommendation Actions S M/L Promote improved • Expand research mandates for IRRI and national seed varieties research institutes to identify and develop early- X and agricultural maturing and drought- and flood-tolerant rice, or practices other crop, varieties. • Improve technologies and management methods. X • Scale up the dissemination and farmer awareness of X these seeds through improved extension services. Improve grain • Assess the current grain storage capacity in storage capacity Cambodia. Focus on constraints causing grain storage X to not exceed 100,000 tons since the mid-2000s. • Assess the cost effectiveness of building additional X grain storage capacity. Preparedness • Make plans to stock these spaces before the next X X ENSO in areas that are easily accessible. • Safeguard producers from potential market X X distortions. Expand and adjust • Assess constraints to expanding SSNs, which currently X social safety nets have low coverage. (SSNs) • Move toward increased funding, skills development, and capacity building in order to implement current X government guidelines on SSNs. • Target poor, rural areas with food-for-work programs, expanded labor market programs, and X X conditional cash transfers. • Adjust these programs to make them more responsive to droughts and other ENSO impacts X through better targeting or by expanding these systems in times of ENSO-related crises. Improve • Streamline government policies, programs, and government departments dealing with ENSO and other climate X capacity and shocks. coordination • Clearly identify the roles of each institution in X relation to these programs. • Assign a special committee, or national task force, devoted to addressing slow-onset disasters related X Resilience to ENSO within MOWRAM, including members from different government agencies and civil society. • Develop a mechanism for state-level stakeholders to work with provincial-level and local-level actors to X create a better flow of information on ENSO-related challenges and solutions. • Exchange information through forums or other X dialogues. • Improve community participation and public X X awareness of ENSO. The Government Can Take Action to Improve ENSO Preparedness and Resilience 67 Recommendation Actions S M/L Integrate a • Begin dialogues with potential partners in ENSO, regional approach including WMO, ESCAP, ASEAN, MRC, the Asian X to ENSO Disaster Preparedness Center, and the World Bank’s SEA-DRIF Platform. • Identify areas of cross-border vulnerability and X potential infrastructural cost sharing. Develop • Evaluate the pilot program for agricultural micro- risk transfer insurance launched in July 2015 by the Cambodian X mechanisms Centre for Study and Development in Agriculture (CEDAC) for areas to potentially scale up. • Assess the appropriateness of various mechanism X including index-based agricultural insurance. • Activate these mechanisms once a disaster begins. X • Look into reallocating credit to provide emergency X X assistance. Resilience Invest in rural • Assess the prevalence and condition of current infrastructure irrigation systems, water harvesting systems, and X rural transportation systems (roads). • Where feasible, develop plans to expand irrigation and water harvesting systems, especially in areas X prone to drought. • Develop plans to improve rural roads with a focus on X connecting agricultural production to markets. Target women • The Ministry of Women’s Affairs and MOWRAM should develop a strategy targeting women who X are vulnerable to ENSO (agriculture workers, rural inhabitants, natural resource managers). • Develop quotas or guidelines to empower women in leadership or coordination positions in local- or X national-level efforts to combat ENSO impacts. • Introduce resource-efficient, low-carbon practices for X women. • Target women in SSNs. X 68 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Annexes Annex 1: Methodological specifics This annex outlines the specifics of the methodology summarized in Box 1. General First, historic climate data is examined, including variability in rainfall and tempera- ture, followed by an assessment of the frequency of ENSO events in the historical record. More specifically, short-term climate fluctuations during ENSO event years are compared with recent “neutral” weather years (without ENSO shocks) to identify deviations in rainfall and temperature variables. Second, changes in weather variables during the crop growing season are translated into physical, agricultural productivity outcomes using a combination of statistical and process-based models. Process-based crop models to estimate ENSO-affected seasonal yield deviations of major crops in a grid-based spatial analysis framework are also applied to isolate ENSO impacts from other events. Daily historical weather data (spatially interpolated from weather station data), linked with the correspond- ing ENSO phase, were used as input to the crop modeling framework that estimated crop yield changes for important crops: rice, maize, and tomatoes. Rice and maize are major crops in the country, and tomatoes act as a proxy for a broader array of vege- tables (a standard approach in climate studies). The crop models also estimate how yield responses differ when using improved or traditional seed varieties, and with and without chemical fertilizer, or depending on the water management regime (with and without irrigation infrastructure). Third, non-crop impact channels are also considered, such as livestock and fisheries. In the absence of sophisticated models for these subsectors, we rely on secondary evi- dence compiled from other studies. These studies typically focus on a specific ENSO year, such as the severe 2016 event. The livestock sector analysis is supplemented with   69 70 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture estimated Temperature Humidity Indices to estimate heat stress levels and productivity losses for cattle and poultry. Finally, the estimated impacts of ENSO events on crop, livestock, and fisheries yields are imposed on a dynamic computable general equilibrium (CGE) model. This class of model captures all producers and consumers in an economy, including the government and interactions with the rest of the world (e.g., imports and exports). All sectors and households are disaggregated across major subnational regions. Region- and crop-­ specific productivity shocks thus translate into changes in agricultural and national GDP, employment, and prices. The model reacts to crop- and sector-specific produc- tivity changes by reallocating resources and products between sectors and households to minimize overall losses to the economy (i.e., autonomous adaptation). The model is linked to a survey-based microsimulation module that tracks changes in national and subnational poverty rates. The integrated approach to measure economy-wide impacts of climate shocks is sim- ilar to what is often used for long-term climate change impact studies. The DSSAT and CGE models represent some of the most sophisticated tools available for such analysis, and the high-resolution spatial databases used in both types of models are quite unique, both for Cambodia and developing countries in general. The framework makes it possible to isolate the impacts of ENSO events, as well as to assess outcomes in hypothetical alternative policy environments, such as changes to trade policies or the scaling up of social safety nets. CGE modeling RIAPA is a recursive dynamic computable general equilibrium (CGE) model that sim- ulates the functioning of a market economy, including markets for products and factors (i.e., land, labor, and capital). RIAPA measures how impacts are mediated through prices and resource reallocations, and ensures that resource and macroeconomic con- straints are respected, such as when inputs or foreign exchange are limited. RIAPA provides a consistent “simulation laboratory” for quantitatively examining value-chain interactions and spillovers at national, subnational, and household levels. RIAPA divides the national economy into different sectors and household groups that act as individual economic agents. Producers maximize profits and supply output to national markets, where it may be exported and/or combined with imports depending on relative prices, with foreign prices affected by exchange rate movements. Produc- ers combine factors and intermediate inputs using sector-specific technologies. Maize Annex 1: Methodological Specifics 71 farmers, for example, use a unique combination of land, labor, machinery, fertilizer, and purchased seeds. Workers are divided by education levels, and agricultural capital is separated into crop and livestock categories. Labor and capital are in fixed supply, but less-educated workers are treated as underemployed. Producers and households pay taxes to the government, who uses these and other revenues to finance public services and social transfers. Remaining revenues are added to private savings and foreign capital inflows to finance investment, i.e., investment is driven by levels of sav- ings. RIAPA is dynamic, with past investment determining current capital availability. RIAPA tracks changes in incomes and expenditures for different household groups, including changes in food and nonfood consumption patterns. Poverty impacts are mea- sured using survey-based microsimulation analysis. Individual survey households map to the model’s household groups. Estimated consumption changes in the model are applied proportionally to survey households, and post-simulation consumption values are recalculated and compared to a poverty line to determine households’ poverty status. Simulated interventions In order to understand if and how different policy options can mitigate impacts from ENSO events, different policy options are incorporated into the CGE model. Some of the scenarios reflect existing policies in the country, while others consider policies or investments that could be implemented in the future. The scenarios are therefore a combination of current and potential interventions, benchmarked to existing policies and evidence to ensure that the scale of policy change is plausible and relevant. We consider the following five policy scenarios: • Drought-tolerant varieties: It is assumed that drought-resistant varieties for beans, maize, rice, and tomato are adopted by farmers in regions that anticipate the greatest water challenges. The International Rice Research Institute (IRRI), working together with national research institutes, has developed and disseminated drought-tolerant rice varieties. Scaling up the adoption of these seeds would require improvements in both seed and farmer extensive systems. • Additional irrigation: The amount of cultivated land in the model that uses irriga- tion infrastructure is increased. Rice is already irrigated in Cambodia, although cov- erage is far from complete (i.e., 12 percent coverage of rice land in the Plains and 10 percent elsewhere). Further expansion is therefore possible. A 10 to 20 percent increase in irrigation use is simulated, reaching 14 percent in the Plains, 11 percent in the Lake, and 12 percent in the Coastal and Mountain regions. This is a relatively small percentage change in rice irrigation, but may cover a large land area. Maize is rarely grown on irrigated lands, and so while we model a doubling of irrigated 72 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture maize land, the final coverage remains low in all regions (just 2 percent in the Plains and zero elsewhere). The irrigation scenario assumes that there is adequate water to operate additional irrigation systems. Although the focus is on irrigation infrastruc- ture, an alternative option may be to improve water use efficiency. Since the model does not capture potential water supply constraints, the scenario is equivalent to one that maintains yields during ENSO shocks through more efficient water use. • Rice export restrictions: A partial ban on rice exports implemented over the course of year is applied in the model. Cambodia has regularly used this policy instru- ment before to respond to concerns about domestic rice supply shortages, such as in 2001, 2008, and 2011. The goal of export bans is to reduce domestic prices for rice consumers, but it also has the effect of lowering producer prices for would-be exporters. The impact on the poor, who tend to be smallholder farmers, but also net consumers of rice, depends on the complex income and expenditure structures of households living close to the poverty line. Unlike export taxes, export bans do not directly generate revenues for the government. The indirect economic costs of an export ban are internalized in the model, however, via changes to government revenues and fiscal deficits. • Supply 250,000 tons of stored rice from public and private stocks: Depleting stocks addresses short-term supply shortfalls during ENSO events, and offsets some of the price increases caused by production losses. Like export bans, depleting grain stores benefits consumers, but may prevent market forces from limiting farm rev- enue losses via higher prices for agricultural products. The scenario assumes that storage facilities already have or can be expanded to achieve this capacity. Histor- ical evidence indicates that Cambodia has limited capacity to accommodate the scale of drawdown in this scenario. For example, the FAO Food Balance Sheets show that, from the mid-1990s to the mid-2000s, the country regularly depleted its collective grain stocks by more than 250,000 tons. However, in years following this period, stock depletions have not exceeded 100,000 tons (data series ends in 2013). This suggests that the grain storage scenario may not reflect the country's current capacity, and would require new investments in storage facilities and systems. Note that we do not consider the financial cost of restocking public and private grain stores in the years following an ENSO event. Note also that the grain storage sce- nario is equivalent to an alternative scenario in which the government procures grain in foreign markets (financed by foreign borrowing) and distributes the grain in domestic markets. • Provide short-term cash transfers to poorer households (Quintiles 1–3) equal to just over US$4.3 per person: Currently, the government’s various social protection and pension schemes transfer, on average, about US$1.2 per person per year in these poorer households, and so this marks a significant increase in their social Annex 2: National Plans and Agency Roles Related to Climate 73 transfer receipts. That said, social transfers currently account for less than 1 per- cent of these households’ total incomes, and so large transfers are needed to have a discernable effect. Moreover, the increase is small when compared to the level of national social transfers. The government transfers, on average, US$13.7 per person per year to all households. The additional US$1.2 per capita transferred in this scenario is therefore only an 18 percent increase over current spending, and this is only during the El Niño event year. Households in the model can use additional transfers to offset higher food costs, or to purchase nonfood products, whose prices may also rise during ENSO events as economic shocks spill over from agriculture to nonfarm sectors. The fiscal cost of expanding social transfers is internalized through higher direct taxes (e.g., pay-as-you-earn and corporate taxes). The sce- nario assumes that the distribution of new cash transfers occurs through existing social protection systems and does not increase the administrative cost of this sys- tem. This is equivalent to assuming that additional administrative costs (not actual transfers) are borne by foreign development partners. • Combined: All the above policies implemented concurrently. While these scenarios are incomplete, they provide an insight into the effect of pos- sible measures available for the government to better manage the impacts of future ENSO. These scenarios do not capture all possible government responses to ENSO shocks. Nor are the scenarios designed to reflect the complexities of specific policies. Such detailed analysis is beyond the scope of this study. Instead, the scenarios are purposefully selected to reflect the range of policy instruments available to the gov- ernment, including investments in farm production (i.e., seeds and irrigation); trade and price policies (i.e., export bans); and standard emergency responses (i.e., grain stock management and social safety nets). Within each type of instrument there are further options to be considered, such as the targeting and distribution mechanisms for emergency cash transfers. Annex 2: National plans and agency roles related to climate International agreements Cambodia has been a signatory to five international agreements on climate change in a span of 20 years. Under the Paris Agreement, Cambodia committed to reduce the total amount of Greenhouse Gas (GHG) emission by 27 percent and to implement vari- ous activities in the agricultural sector. These activities strengthen climate information and early warning systems; develop crop varieties that are suitable to different Agro 74 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Ecological Zones (AEZs) and are resilient to climate change; and develop climate adaptive agriculture and aquaculture production systems.165 National legal framework Law on Disaster Management, signed into law on July 10, 2015 (Royal Decree NS/ RKM/0715/007), is the overarching law regulating all activities related to disaster risk reduction and management in Cambodia. This legal document regulates the roles and responsibilities of different institutions, and manages the resource allocation and coor- dination mechanisms for disaster risk reduction and natural disaster prevention and relief. National Strategy Development Plans (NSDP) 2009–2013 and 2014–2018 are the key policy documents containing the major development goals of Cambodia, the strat- egies to achieve these goals, and the potential obstacles hindering or slowing down the process. In these plans, natural disaster is considered as one of the most pressing concerns. The NSDP 2009–2013 identified floods, droughts, typhoons, and epidemic diseases as the major natural disasters in Cambodia. The impacts of these natural calamities on human health, crop and livestock production, infrastructure and properties, and the national economy in general were also identified (General Directorate of Agriculture, 2013). Disaster risk reduction challenges and measures were included in the NSDP 2014–2018. Financial resources were also determined to fund the different activities outlined in the plan (IFRC, 2017). The National Action Plan for Disaster Risk Reduction (NAP-DRR) for the period 2014–2018 was developed and endorsed by NCDM in 2014 to replace the completed SNAP-DRR 2008–2013 (Hou, 2016). Taking into consideration the implementation results of SNAP-DRR in the previous period, NAP-DRR identified the activities that would address major gaps in existing legal framework, coordination, planning, and other challenges, including capacity building and resource allocation for the subna- tional committees for disaster management (IFRC, 2017). The Strategic National Action Plan for Disaster Risk Reduction (SNAP-DRR) 2008–2013 was launched in 2008 to address the implementation of the Hyogo Framework for Action in Cambodia. An interinstitutional task force to implement 165 http://www4.unfccc.int/submissions/INDC/Published%20Documents/Cambodia/1/Cambodia’s%20 INDC%20to%20the%20UNFCCC.pdf Annex 2: National Plans and Agency Roles Related to Climate 75 the prioritized activities for disaster risk reduction was established and co-led by the National Committee for Disaster Management in cooperation with different ministries and agencies (General Directorate of Agriculture, 2013). The National Adaptation Program of Action to Climate Change (NAPA) was developed by the Ministry of Environment, and was endorsed by the government in 2006. NAPA identified the key adaptation needs in human health, agriculture, water resources, and coastal zones (General Directorate of Agriculture, 2013), and priori- tized the 20 highest priority NAPA activities for implementation. Community-Based Disaster Risk Management (CBDRM) is considered by the government of Cambodia as an integrated component of the poverty reduction plan in rural areas, focusing on the protection of life, property, and rural livelihoods (Leng, 2014). CBDRM is a strategy to build appropriate and feasible disaster risk reduction committees within communities. These committees were expected to prepare plans for the communities based on their vulnerabilities, existing capacities, and coping mechanisms. The vulnerability mapping was the initial input of the CBDRM strategy where vulnerable communities and local actors worked together to identify, classify, and evaluate the disaster risks, and to map the potential affected areas. Based on the vulnerabilities, appropriate activities to enhance the local capacity for long-term risk preparedness were identified and prioritized in accordance to the local situation. To support the effective implementation of the above mentioned laws, national strat- egies, and action plans, the government of Cambodia also issued different instructive and directive legal documents, namely Sub-Decree No. 30 ANKR.BK, dated April 09, 2002, on the Organization and Functioning of the National and Sub-National Commit- tees for Disaster Management; Sub-Decree No. 61 ANKR.BK, dated June 29, 2006, on the establishment of the Commune Committee for Disaster Management (CCDM); and Direction No. 315 NCDM, dated July 21, 2010, on the establishment of the Village Disaster Management Team (VDMT) for the implementation of Community-Based Disaster Risk Management (Leng, 2014; Hou, 2016). As shown earlier, Figure 23 summarizes the major international treaties and agree- ments, and national documents that created the legal foundation for the disaster risk management in Cambodia. The figure also shows the establishment of the National Committee for Disaster Management (NCDM) as the coordinating agency for disas- ter risk management and reduction in Cambodia. NCDM takes the leading role and cooperates with different ministries such as the Ministry of Environment (MoE), 76 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture FIGURE 23: Timeline of Cambodian Legal Framework for Disaster Risk Reduction and Climate Change Adaptation. Cambodia signed Cambodia participated in NCDM set up the Hyogo SNAP-DRR the Sendai Framework under sub-decree framework for 2008–2013 for Disaster Risk No. 54 ANKR-BK action was launched Reduction 2015–2030 1995 2002 2005 2008 2014 2015 Cambodia NCDM’s mandate Cambodia NAP-DRR Law on Cambodia signed the and structure ratified the 2014–2018 Disaster ratified the UNFCCC clearly defined in AADMER was launched Management Paris sub-decree enacted Agreement NS/RKT/0202/040 Source: Leng, 2014. Ministry of Agriculture, Forestry and Fisheries (MAFF), Ministry of Water Resources and Meteorology (MOWRAM), as well as international organizations and the private sector (Leng, 2014). Department functions The NCDM Secretariat is composed of five departments with different functions in the disaster prevention, response, and recovery (Leng 2014). The departments and their functions are described below: • The Department of Information and Relations coordinates the information collection on potential disasters with foreign countries and appropriate ministries (in particular MOWRAM). This department is also responsible for reporting the hydrometeorological condition, circulating the forecast among concerned institu- tions, and disseminating warnings through different media channels. • The Department of Emergency Response and Rehabilitation receives the fore- cast from the Department of Information and Relations to prepare the emer- gency response plans and to conduct damage assessment in an affected area. • The Department of Preparedness and Training is in charge of equipment prepa- ration for the emergency response plans and human resource capability building to improve preparedness at the commune and village levels. • The Department of Search and Rescue coordinates search and rescue operations with the civil aviation, armed forces, and other agencies (Leng 2014). The organizational structure of NCDM is decentralized to the commune level. This means that there is a Committee for Disaster Management at the provincial, district, Annex 3: Government Initiatives for Agriculture and Farming Households 77 and commune levels in all regions. Through Direction No. 315 NCDM, dated July 21, 2010, the establishment of the Village Disaster Management Team (VDMT) was approved. It was then placed under the CBDRM, which is the grassroot-level actor for disaster management. The Provincial Committee for Disaster Management (PCDM) is chaired by a provin- cial governor with the participation of a deputy provincial governor, heads of relevant provincial departments (e.g., Department of Water Resource, Department of Agricul- Provincial/Municipal Red Cross. ture, Forestry and Fisheries), and a Director of the ­ Similarly, the District Committee for Disaster Management (DCDM) is comprised of a district governor as president, a deputy district governor, chiefs of different offices, and a chief of the Cambodian Red Cross district subbranch. The PCDM and DCDM have similar responsibilities, which are as follows: • Implementing the National Policy for Disaster Management at the provincial or district levels; • Organizing and conducting training and selecting appropriate officers for training on disaster management; • Presiding, instructing, and leading the emergency relief operation during the occur- rence of a disaster; and • Writing and submitting reports to the higher level Committee for Disaster Manage- ment on damage assessment, needs assessment, budget and equipment provision proposal, and human resource allocation plans. Annex 3: Government initiatives for agriculture and farming households The drought in 2016 occurred in many parts of the country and its impacts were very diverse. The government of Cambodia was able to partly respond to the emergency needs of the affected population (Caritas Cambodia, 2016). This section presents the initiatives of the government for agriculture during the drought in 2016 such as the provision of input/equipment, agroclimatic forecasts and early warning systems, irri- gation management shifts, structural measures, and land use and crop pattern advice. 78 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Input/equipment provision In response to the emergency needs of the affected communities, Prime Minister Hun- sen called on the armed forces, civil servants, and the Cambodian Red Cross to ensure adequate fresh water distribution and provision. Under the coordination of NCDM, the local authorities (e.g., Siem Reap Provincial Department of Environment and Battam- bang Provincial Red Cross) transported and distributed water to 331,505 most vulner- able families in 18 provinces (ACAPS, 2016; Cambodia Humanitarian Forum, 2016; Caritas Cambodia, 2016; Kunthear, 2016). Agroclimatic forecasts and early warning systems In November 2015, MOWRAM released the first statement about the El Niño situation in Cambodia, warning the residents about a “dry and hot” climate that would occur from December 2015 to May 2016 (Sotheary, 2015). In line with this, MOWRAM allocated more resources for updates on the situation with forecasts aired once or twice a month. Database/information was also received from Japan, France, and other for- eign countries covering the precipitation, temperature, and prediction of the end of the dry and wet seasons. The specific information on ENSO came from the Japan Meteorological Agency every month, and the El Niño forecast was given three months before its occurrence. After collection and analysis by ­ MOWRAM, the El Niño forecast was sent to MAFF via official letters for use in agriculture preparation and response activities. The infor- mation on weather forecasts was also generated for public dissemination (interview with MOWRAM representative on October 9, 2017). MOWRAM and its provincial departments, as well as the relevant institutions and local authorities also advised the people to conserve water for daily use and for agriculture production. An example of MOWRAM’s forecast information for the fishery and transportation sectors is shown in Figure 24. Irrigation management shifts Before the El Niño occurrence, MOWRAM requested the Department of Water Resources and Meterology to encourage all stakeholders to store as much water as possible. The stakeholders were encouraged to properly use water and were advised to save drinking water first before using it for domestic and other uses. MOWRAM also installed a pumping station in high vulnerability areas for pump- ing and storing water to prepare for droughts. The Minister of MOWRAM mobilized Annex 3: Government Initiatives for Agriculture and Farming Households 79 FIGURE 24: MOWRAM’s Weather Forecast Products for the Fishery (left) and Transportation (right) Sectors. Source: Cambodia Meteo (http://www.cambodiameteo.com) human resources in the central and provincial levels to implement water saving and emergency pumping measures. The drought in 2016 showed that MOWRAM does not have a protocol to respond to such extreme event in terms of water resource management, in general, and in irriga- tion, in particular. The measures of MOWRAM in irrigation management are being applied on a case-by-case basis. Structural measures Representatives from MOWRAM stated that droughts are their biggest concern. To reduce the impacts of drought on agriculture, MOWRAM constructed a number of small reservoirs and ponds to improve the irrigation capacity of drought-affected regions. This was especially true in in Kampong Speu and Prey Veng provinces, which were considered as the most prone areas to both droughts and flash floods. Although MOWRAM constructed the reservoirs, community ponds, wells, and canals ahead of the drought occurrence, these water sources, except for a few deep wells, 80 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture dried up during the drought. Consequently, water scarcity was experienced by local communities (Cambodia Humanitarian Forum, 2016). The emergency strategy of the government during the drought in 2016 was to dig new wells and to rehabilitate the ponds and canals to ease water shortages (ACAPS, 2016). For instance, the Kratie Provincial Department of Rural Development assisted the rehabilitation of community ponds in Kratie, while MOWRAM rehabilitated and pumped water into a canal in Prey Veng province (Cambodia Humanitarian Forum, 2016). This strategy resulted in approximately 965 new dug wells, 684 rehabilitated old wells, 240 new ponds as well as 125 restored old ponds, and 18 restored canals in 2016 (Kunthear, 2016). According to MAFF, finding the groundwater source, digging a well, and installing a pumping system to transport water to a drip irrigation system was a good short-term strategy. An effective long-term strategy requires the development of a set of system- atic and comprehensive solutions for infrastructure, irrigation systems, agricultural production, and markets. Land use and crop pattern advice In an article in Phnom Penh Post published in September 2015, the spokesperson of MAFF stated that the authorities have been developing a training program for aware- ness raising on climate change and for the production of stress-tolerant crops. The Cambodian Centre for Study and Development in Agriculture recommended the mit- igation of drought impacts by convincing farmers to shift to crops like sweet potatoes and beans (David and Harfenist, 2015). MAFF also recommended shifting from crops to livestock and aquaculture as a potential measure. MAFF faces difficulties in agriculture transformation toward climate resilience. It is done with an initial phase of ecological zoning activities and analysis of suitable zones for different crops and livestock. More technical and financial support in the ecologi- cal zoning and analysis of the suitable agricultural production system are needed. The information that will be generated can help identify the crops requiring less water and that have high market value. Annex 4: Summary of El Niño–Southern Oscillation (ENSO) Forecasts Made Available by Technical Agencies 81 Annex 4: Summary of El Niño–Southern Oscillation (ENSO) forecasts made available by technical agencies166 Typical forecast Forecast information Frequency of the ENSO Dynamical Agencies/ product forecast or Single-model institutes released format statistical or MME ENSO Others APEC Climate Monthly Graphical Dynamical MME http://www.apcc21 http://www.apcc21 Center (APCC) products, (multiagency167) .org/ser/enso .org/ser/outlook outlook .do?lang=en .do?lang=en • Air temperature • Precipitation • 500 hPA geopotential height • SST International Monthly Graphical Dynamical MME http://iri.columbia http://iri.columbia Research products, and (multiagency168) .edu/our-expertise/ .edu/our-expertise/ Institute for interpretation statistical climate/forecasts/ climate/forecasts/ Climate and enso/current/ seasonal-climate- Society (IRI) forecasts/ • Air temperature • Precipitation World Regular Bulletins Expert opinion based on http://www.wmo Meteorological seasonal issued with assessment of all important .int/pages/prog/ Organization press briefing global sources (from about wcp/wcasp/enso_ (WMO) 28 climate centers) of ENSO update_latest predictions .html Climate Monthly Graphical Dynamical MME http://www.cpc http://www.cpc Prediction products, and (multiagency169) .ncep.noaa.gov/ .ncep.noaa.gov/ Center (CPC)/ diagnostic statistical products/analysis_ products/NMME/ National discussions monitoring/ seasanom.shtml Centers for enso_advisory/ • Air temperature Environmental • Precipitation Prediction • SST (NCEP)/National Weather Service (NWS) European Monthly Graphical Dynamical Single-model http://www Centre for products (SEAS2011 cycle .ecmwf.int/ Medium-Range 36r4) en/forecasts/ Weather charts/seasonal/ Forecasting Niper_centF1o- (ECMWF) plumes-public- charts-long-range- forecast UK Monthly Graphical Dynamical Single-model http://www Meteorological products, (GloSea5) .metoffice.gov Office interpretation .uk/research/ (UKMO) climate/seasonal- to-decadal/gpc- outlooks/el-Niño- la-Nina (continued) 166 https://reliefweb.int/sites/reliefweb.int/files/resources/El%20Nino%20report-%20finalized%20ESCAP07082017.pdf 167 Climate forecasting information from 16 climate centers. 168 Climate forecasting information from 22 climate centers. 169 The official outlook is produced jointly with IRI. 82 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Typical forecast Forecast information Frequency of the ENSO Dynamical Agencies/ product forecast or Single-model institutes released format statistical or MME ENSO Others Australia 15 days Graphical Dynamical Single-model http://www.bom http://www Bureau of products, (POAMA) + .gov.au/climate/ .bom.gov.au/ Meteorology interpretation multiagency enso/ climate/model- (BOM) survey170 summary/#tabs= Indian-Ocean • Indian Ocean Dipole Japan Monthly Graphical Dynamical Single-model http://ds.data http://ds.data Meteorological products, (JMA/MRI- .jma.go.jp/tcc/tcc/ .jma.go.jp/tcc/tcc/ Agency (JMA) outlooks CGCM2) products/elNiño/ products/model/ outlook.htm index.html • Air temperature • Precipitation • 500 hPa geopotential height • Sea level pressure • 200 hPa and 850 hPa stream function • Velocity potential • SST Korea Monthly Graphical Dynamical Single-model https://web Meteorological products (Glosea5) .kma.go.kr/eng/ Administration weather/forecast/ (KMA) ella_forecast.isp Beijing Climate Monthly Graphical Dynamical Single-model http://cmdp http://cmdpo.ncc Center products (BCC-CGCM) .ncc.cma.gov .cma.gov.cn/pred/ .cn/Monitoring/ en_cs.phop en_enso.php • Air temperature • Precipitation • 500 hPa geopotential height • 200 hPa and 850 hPa winds • SST India Monthly Graphical Dynamical Single-model http://www • Indian Ocean Meteorological products in (CFS V2 from .imdpune.gov.in/ Dipole Department bulletin NCEP) Clim_RCC_LRF/ Products.html Pacific ENSO Quarterly, Graphical Climate model information https://www • Precipitation Applications with products is not provided .weather.gov/ • Tropical cyclone Climate Center supplemental in bulletin peac/update activity (PEAC) updates as (Pacific ENSO • Sea level required update for the Pacific region) 170 Climate forecasting information from eight climate centers. References ACAPS. 2016. Cambodia Drought. Briefing Note, May 3, 2016. https://www.acaps .org/special-report/cambodia-drought Adamson, P., and J. Bird. 2010. The Mekong: a drought-prone tropical environment? International Journal of Water Resources Development 26 (4): 579–594. Asian Development Bank (ADB). 2014. Additional Financing of Flood Damage Emergency Reconstruction Project (RRP CAM 46009-003): Disaster Management in Cambodia. https://www.adb.org/sites/default/files/linked-­documents/46009- 003-sd-02.pdf Baran, E., N. Schwartz, and Y. Kura. 2009. Climate change and fisheries: vulnerability and adaptation in Cambodia. WorldFish Center Issue Brief 20008. Barnston, A. 2014. How Good Have ENSO Forecasts Been Lately? NOAA-Climate. Science and Information for a Climate-Smart Nation. https://www.climate.gov/ news-features/blogs/enso/how-good-have-enso-forecasts-been-lately Basak R. 2016. Benefits and costs of climate change mitigation technologies in paddy rice: Focus on Bangladesh and Vietnam. CCAFS Working Paper no. 160. Copen- hagen, Denmark: CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS). https://cgspace.cgiar.org/handle/10568/75662 Cambodian Humanitarian Forum (CHF). 2016. Cambodia Drought Rapid Assess- ment Report. https://www.humanitarianresponse.info/system/files/documents/ files/assessment_report_2016_chf_10_provinces.pdf Caritas Cambodia. 2016, June 2. Cambodia: drought emergency appeal. Retrieved from Caritas Cambodia: http://www.caritascambodia.org/index.php/ cambodia-drought-emergency-appeal-2016 Chakrya, Khouth Sophak and Alessandro Marazzi Sassoon. 2016, Apr 22. Water short- age fears grow amid record heat wave. Retrieved from The Phnom Penh Post: http://www.phnompenhpost.com/national/water-shortage-fears-grow-amid-record- heat-wave Chey Tech 2012. The Kingdom of Cambodia: Updating and Improving the Social Protection Index. https://www.adb.org/sites/default/files/project-document/76061/ 44152-012-reg-tacr-32.pdf   83 84 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture Chhinh, N. 2015. Drought, drought vulnerability and adaptation policy in Cambodia. https://www.humanitarianresponse.info/en/system/files/documents/files/drought_ drought_vulnerability_and_adaptation_policy_in_cambodia_with_reference_to_ the_farming_sector_nyda_thesis_0.pdf Chum, S., S. Tep, and J. Nevill. 2016. Rapid Assessment of the Drought in Koh Kong Province. Final Report. https://www.humanitarianresponse.info/en/operations/ cambodia/document/final-report-rapid-assessment-drought-koh-kong-province- april-2016 DANIDA. 2015. https://commons.wikimedia.org/wiki/File:Map_rainfall_Cambodia .jpg (original source: DANIDA). David, S. and E. Harfenist. 2015, September 1. Drought batters 13 provinces. Retrieved from Phnom Penh Post: http://www.phnompenhpost.com/national/ drought-batters-13-provinces ESA. 2018. https://www.esa-landcover-cci.org/ FAO. 2010. Cambodia—Food and Nutrition Security Profiles 2010. Available at: http://www.fao.org/3/a-at706e.pdf FAO. 2018. FAOSTAT statistics database. Rome: FAO. Gaulter, S. 2016, April 26. Sixty-five tonnes of fish perish in Cambodian lake. Retrieved from Al Jazeera English: http://www.aljazeera.com/news/2016/04/sixty-tonnes- fish-perish-cambodian-lake-160426102049630.html General Directorate of Agriculture. 2013. Plan of Action for Disaster Risk Reduction in Agriculture 2014–2018. Retrieved from FAO: https://coin.fao.org/coin-static/ cms/media/21/14109270844140/etfinal_national_poa_drr_gda_main_eng.pdf Hing, P. 2016. Women in the Face of Climate Change: The Driving Force for Any Solution. http://www.kh.undp.org/content/cambodia/en/home/presscenter/articles/ 2016/03/08/women-in-the-face-of-climate-change-the-driving-force-for-any- solution.html Hou, T. E. 2016. Country Report of Cambodia. Disaster Risk Reduction. http://www .unescap.org/sites/default/files/Cambodia%20Country%20Presentation.pdf HRF. 2015. Humanitarian Response Forum Contingency Plan for Cambodia. HRF. Phnom Penh, Cambodia. HRF. 2016. Synthesis report of NGO drought assessments in Cambodia: by participant organisations of the Humanitarian Response Forum (HRF). HRF. Phnom Penh, Cambodia. IFRC. 2017. Relief Web: Implementing the Law on Disaster Management in Cambo- dia Developing Subsidiary Legislation. Retrieved from International Federation of Red Cross and Red Crescent Societies: http://reliefweb.int/sites/reliefweb Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J. Woollen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, References 85 J. Janowiak, K. C. Mo, C. Ropelewski, J.  Wang, A. Leetmaa, R. Reynolds, R. Jenne, and D. Joseph. 1996. The NCEP/NCAR 40-year reanalysis project, Bull. Amer. Meteor. Soc. 77: 437-470 https://journals.ametsoc.org/doi/10.1175/ 1520-0477%281996%29077%3C0437%3ATNYRP%3E2.0.CO%3B2 Kreft, S., Eckstein, D., and Melchior, I. 2017. Global Climate Risk Index 2017: Who suffer most from extreme weather events? German Watch. Bonn, Germany. Kunthear, M. 2016, December 23. Overview of 2016 Natural Disasters. Retrieved from Khmer Times: http://www.khmertimeskh.com/news/33516/ overview-of-2016-natural-disasters/ Lara, L. J. and M. H. Rostagno. 2013. Impact of Heat Stress on Poultry Production. Animals 3 (2): 356–369; doi:10.3390/ani3020356 Leng, H. A. 2014. Country Report of Cambodia. Disaster Management. Ly, S. 2016. Cambodia—Economic update: improving macroeconomic and financial resilience. Cambodia economic update. Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/en/403961467989523939/Cambodia- Economic-update-improving-macroeconomic-and-financial-resilience Ministry of Environment (MoE) and National Council for Sustainable Develop- ment (NCSD), Kingdom of Cambodia. 2016. Cambodia’s Intended Nationally Determined Contribution Report. http://www4.unfccc.int/submissions/INDC/ Published%20Documents/Cambodia/1/Cambodia%27s%20INDC%20to%20 the%20UNFCCC.pdf Miyan, M. A. 2015. Droughts in Asian Least Developed Countries: Vulnerability and sustainability. Weather and Climate Extremes, Vol 7. Pages 8–23. Myer, R. and R. Bucklin. 2001. Influence of Hot-Humid Environment on Growth Per- formance and Reproduction of Swine. Pub. AN107 UF/IFAS Extension. http://edis .ifas.ufl.edu/an107 NASA. 2016. Fires across Thailand and Cambodia. Image available at: https://www .nasa.gov/image-feature/goddard/2016/fires-across-thailand-and-cambodia National Adaptation Program of Action to Climate Change (NAPA). 2006. Royal Gov- ernment of Cambodia. Phnom Penh, Cambodia. National Institute of Disaster Management (NIDM). 2014. Cambodia. http://nidm .gov.in/easindia2014/err/pdf/country_profile/cambodia.pdf National Institute of Statistics, Ministry of Planning. 2015. Cambodia Socio- Economic Survey 2014. Phnom Penh, Cambodia. https://www.nis.gov.kh/nis/ CSES/Final%20Report%20CSES%202014.pdf NCDM and UNDP. 2013. CamDi: Cambodia Disaster Loss and Damage Information System. NEDA. 2016, April. Statement of Socioeconomic Planning Secretary Emman- uel F. Esguerra on the Status of the Roadmap to Address the Impact of 86 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture El Niño. Retrieved October 2017, from NEDA: http://www.neda.gov .ph/2016/04/05/statement-of-socioeconomic-planning-secretary-emmanuel-f- esguerra-on-the-status-of-the-roadmap-to-address-the-impact-of-el-nino-rain/ NOAA. 2018. El Nino/Southern Oscillation (ENSO) Diagnostic Discussion. Issued by Climate Prediction Center/NCEP/NWS and the International Research Institute for Climate and Society. http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ enso_advisory/ensodisc.shtml Open Development. 2016, July 4. Drought (last update on 4 July 2016). Retrieved from Open Development Cambodia: https://opendevelopmentcambodia.net/topics/ drought/ Press, A. 2016. PM Hun Sen says all Cambodians must help deal with drought. http:// www.dailymail.co.uk/wires/ap/article-3559140/PM-Hun-Sen-says-Cambodians- help-deal-drought.html Raksmey, U. and Kossov, I. 2015. First Agricultural Insurance Tested in Cambodia. Retrieved February 2018, from Khmer Time: http://www.khmertimeskh.com/ news/13276/first-agricultural-insurance-tested-in-cambodia/ Rayner N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan. 2003. Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res. 108 (D14), 4407, doi:10.1029/2002JD002670. Dataset available at: https:// www.esrl.noaa.gov/psd/gcos_wgsp/Timeseries/Nino34 Robinson, J. 2015. Gender Sensitivity in Disaster Management. Project ‘Building Disas- ter Resilient Communities II’. Retrieved February 2018, from https://reliefweb.int/ sites/reliefweb.int/files/resources/dmkh_finalgendersensitivity_2015-09-01b.pdf Samean, Lay and Sassoon, M. Alessandro. 2016, May 26. Drought relief campaign over: PM. Retrieved from The Phnom Penh Post: http://www.phnompenhpost.com/ national/drought-relief-campaign-over-pm Save the Children. 2016. El Niño-induced drought in Cambodia: Rapid Assessment Report. https://www.humanitarianresponse.info/system/files/documents/files/may_ 2016_multi-province_rapid_assessment_save_the_children.pdf Sokhorng, C. 2016. Crop insurance takes root. Retrieved February 2018, from The Phnom Penh Post: http://www.phnompenhpost.com/business/crop-insurance-takes- root Sotheary, P. 2015. Another drought year in store for Cambodia. http://www .phnompenhpost.com/national/another-drought-year-store-cambodia Stockdale, T., M. Balmaseda, and L. Ferranti. 2017. The 2015/2016 El Niño and beyond. European Centre for Medium-Range Weather Forecasts Newsletter Num- ber 151, Spring 2017, April 2017. https://www.ecmwf.int/en/newsletter/151/ meteorology/2015-2016-el-nino-and-beyond References 87 Suy, R., L. Chhay, D. Bekbauvoa, A. Islamjanova, and I. Iddrisu. 2018. Review on Social Protection for Vulnerable Group in Cambodia to Poverty Reduction. Macrothink Institute. Journal of Social Science Studies 5 (1): 1–22. http://www .macrothink.org/journal/index.php/jsss/article/viewFile/11458/9128 Thul, P. C., Bangkok Post. 2016, April 26. Hun Sen pledges support amid worst drought in 50 years. Retrieved from Bangkok Post: https://www.bangkokpost.com/ archive/hun-sen-pledges-support-amid-worst-drought-in-50-years/948785 Trotsenburg, A. V. 2015. We must prepare now for another major El Niño. http://blogs .worldbank.org/eastasiapacific/we-must-prepare-now-another-major-el-nino UN-FAO. 2014. Cambodia—Food and Nutrition Security Profiles. http://www.fao .org/3/a-at706e.pdf United Nations-Economic and Social Commission for Asia and Pacific (UN-ESCAP). 2016. El Niño 2015/2016: Impact Outlook and Policy Implications. Advisory Note. http://www.unescap.org/sites/default/files/El%20Nino%20Advisory%20Note%20 Dec%202015%20Final.pdf United Nations University (UNU) 2016. 2016 World Risk Report. http://www .welthungerhilfe.de/fileadmin/user_upload/Themen/Konflikte_Katastrophen/ We l t r i s i ko b e r i c h t - 2 0 1 6 / Wo r l d R i s k R e p o r t - 2 0 1 6 - d i s a s t e r- r i s k - B E H - Welthungerhilfe.pdf United States Department of Agriculture. 2013, November 21. Cambodia: Seasonal Flooding Impacts Wet Season Rice Production in 2013. Commodity Intelligence Report. USDA Foreign Agricultural Service. Washington DC, United States. Retrieved from USDA: https://ipad.fas.usda.gov/highlights/2013/11/Cambodia/ Vathana, Sann, Oum, Sothea, Kan, Ponhrith, and Colas, Chervier. 2013. Impact of Disas- ters and Role of Social Protection in Natural Disaster Risk Management in Cambo- dia, ERIA Discussion Paper Series. Vidal, J. 2016. El Niño is over—but it leaves nearly 100 million people short of food. Retrieved July  31, 2017 from https://www.theguardian.com/global- development/2016/may/30/el-nino-is-over-but-it-leaves-nearly-100-million- people-short-of-food Wegner, K., Lambertz, C., Das, G., Reiner, G., and Gauly, M. 2016. Effects of tem- perature and ­ temperature-humidity index on the reproductive performance of sows during summer months under a temperate climate. Anim Sci J, 87: 1334–1339. doi: 10.1111/asj.12569 West, J. W. 2003. Effects of Heat-Stress on Production in Dairy Cattle. J. Dairy Sci. 85: 2131–2144. World Bank. 2010. Designing and Implementing a Rural Safety Net in a Low Income Setting: Lessons Learned from Ethiopia’s Productive Safety Net Program 2005–2009. Washington D.C. 88 Striking a Balance: Managing El Niño and La Niña in Cambodia’s Agriculture World Bank. 2017. The World Bank in Cambodia: Overview. http://www.worldbank .org/en/country/cambodia/overview World Bank-World Development Indicators (WB-WDI) database 2017. http://data .worldbank.org/data-catalog/world-development-indicators World Bank-World Development Indicators (WB-WDI) database. 2018. http://data .worldbank.org/data-catalog/world-development-indicators Xin, H. and J. Harmon. 1998. Heat Stress Indices for Livestock. Iowa State University. https://www.ipic.iastate.edu/info/HeatStressIndicesLivestock.pdf Environmental Benefits Statement The World Bank is committed to reducing its environmental footprint. In sup- port of this commitment, the Publishing and Knowledge Division leverages electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consump- tion, chemical use, greenhouse gas emissions, and waste. The Publishing and Knowledge Division follows the recommended standards for paper use set by the Green Press Initiative. Whenever possible, books are printed on 50 percent to 100 percent postconsumer recycled paper, and at least 50 percent of the fiber in our book paper is either unbleached or bleached using Totally Chlorine Free (TCF), Processed Chlorine Free (PCF), or Enhanced Elemental Chlorine Free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://crinfo.worldbank.org/wbcrinfo/node/4.