Zimbabwe: Agriculture Sector Disaster Risk Assessment MARCH 2019 Zimbabwe: Agriculture Sector Disaster Risk Assessment Zimbabwe: Agriculture Sector Disaster Risk Assessment ACKNOWLEDGEMENTS This report was developed by a team led by Azeb donor community, who ensured fruitful and produc- Fissha Mekonnen, Agricultural Specialist of the tive discussions of the findings of this assessment at Agriculture Global Practice at the World Bank. The various workshops. activities were supported by Carlos E. Arce and Jorge The team also extends its appreciation to the Caballero as lead consultants for the overall study. stakeholders from major agricultural supply chains Eija Peju, Alejandra Campero, Ramiro Iturrioz, and that participated at various stages of the fieldwork Easther Chigumira provided specialized analysis and in workshops to discuss the findings. The team is during the risk assessment. indebted to them for encouraging the team to main- Special thanks are due to the team of Zimbabwean tain a realistic and practical perspective. consultants who carried out field assessments for Valuable and constructive comments were various agricultural commodities and livestock received from Diego Arias Carballo, Lead Agriculture and for the valuable and practical comments and Economist (World Bank); John Luke Plevin, contributions they provided to this study: Carren Financial Sector Specialist (World Bank); and Niels Pindiriri, Jacqeline Mutambara, Jacob Gusha, and Balzer, Deputy Country Director (World Food Vimbai Audrey Chandiwana. Willem Janssen, Lead Programme, Zimbabwe), acting as peer reviewers Agricultural Specialist, and Julie Dana, Practice for the study. Mukami Karuki, Country Manager Manager for the Global Facility for Disaster Reduction for Zimbabwe; Mark Cackler, Practice Manager, and Recovery, Barry Patrick Maher, Senior Financial Agriculture Global Practice; and Holger Kray, Lead Sector Specialist, of the World Bank, provided advice Agriculture Economist, Agriculture Global Practice at various stages. provided important guidance and support. This The team is grateful to the Ministry of Land, activity is financed by the EU-funded ACP-EU Agriculture, Water, Climate and Rural Resettlement Africa Disaster Risk Financing Program, managed for leadership and support, as well as to the par- by the Global Facility for Disaster Reduction and ticipants from the private sector, government, and Recovery. This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments 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. CONTENTS Acknowledgements v Abbreviations and Acronyms xi Executive Summary 1 Chapter 1 Introduction 5 1.1  Overview of the Agricultural Sector 5 1.2  Methodology Used for This Risk Assessment 6 Chapter 2 Agricultural Risk Exposure 9 2.1  Production Losses 9 2.2  Production Losses for Different Return Periods 10 Chapter 3 Sources of Risk 13 3.1  Agro-ecological Regions 13 3.2  Risk Identification 14 3.3  Risk Prioritization 16 Chapter 4 Capacity to Manage Agricultural Risk 21 4.1  Vulnerability Factors 21 4.2  Agricultural Services for Strengthening Resilience 22 4.3  Assessing Disaster Risk Management Capacity 23 Chapter 5 Toward a Risk Management Strategy 31 5.1  Risk Mitigation 32 5.2  Agricultural Insurance 38 5.3  Sovereign Disaster Risk Management 40 Chapter 6 Concluding Remarks 45 Bibliography 47 vii viii Contents Appendix A Description of Agricultural Supply Chains Used in the Risk Assessment 49 Appendix B Methodology Used to Estimate the Value of Crop Losses 57 Tobacco Crop Loss Estimates 57 Maize Crop Loss Estimates 59 Appendix C Methodology for the Risk Assessment for Various Return Periods 63 Materials and Methods 63 Main Findings and Discussion 66 Appendix D Risks for Crop and Livestock Supply Chains 83 Crops 83 Cattle Production 89 Poultry Industry 91 Tables 1 Land Holdings Before and After the Introduction of Fast Track Land Reform, Zimbabwe 6 2 Zimbabwe: Main Agricultural Commodities Included in the Risk Assessment 7 3 Estimated Volume and Value of Crop Losses in Zimbabwe (Average of 30 Years, 1986-2016) 10 4 Estimated Loss at Risk (LaR) for a Portfolio of Crops at Different Return Periods 10 5 Production Risks and Current Risk Management Practices by Stakeholder Group, Zimbabwe 17 6 Basic Legislation for Disaster Risk Management 24 7 Additional Legislation Supporting Disaster Risk Management 25 8 Policy and Strategies Related to Disaster Risk Management and Financing 25 9  Allocations from the Ministry of Finance and Economic Development Contingency Fund   to the Department of Civil Protection for Disaster Operations 28 10 Budget allocations (in US$) from the Civil Protection Fund 28 11 Maize Imports and Their Value, 2015–18 29 12 Annual Budget for Agricultural Services 33 13 Estimated Financing Gap in Disaster Risk Management in Zimbabwe 41 Appendix Tables B.1 Data for Estimating Tobacco Crop Losses and Their Value 57 B.2 Data for estimating Maize Crop Losses and Their Value 59 C.1 Zimbabwe Crop Portfolio: Exposure–Expected Gross Value of Production (US$) 68 C.2 Zimbabwe: Whole Crop Complex: Expected LaR Values for Different Return Periods 69 C.3 Zimbabwe: Contribution of each Crop to the Portfolio Risk 70 C.4 Zimbabwe: Expected LaR Values for Maize for Different Return Periods 71 C.5 Zimbabwe: Expected LaR Values for Tobacco for Different Return Periods 72 C.6 Zimbabwe: Expected LaR Values for Sugarcane for Different Return Periods 74 C.7 Zimbabwe: Expected LaR Values for Groundnuts for Different Return Periods 76 C.8 Zimbabwe: Expected LaR Values for Cotton for Different Return Periods 78 C.9 Zimbabwe: Expected LaR Values for Sorghum for Different Return Periods 80 C.10 Zimbabwe: Expected LaR Values for Wheat for Different Return Periods 81 C.11 Zimbabwe: Expected LaR Values for Soybeans for Different Return Periods 82 D.1 Variation in Maize Yields by Province, Zimbabwe 85 Figures 1 Relationship Between Agriculture Value Added and Overall GDP Growth 7 2 Agro-Ecological Regions and Soil Map 14 3 Correlation Between Rainfall and Crop Yields, Zimbabwe 15 4 Prioritized Agricultural Risks, Zimbabwe 18 5 Structure of Zimbabwe’s emergency Management System 26 Contents ix 6 Zimbabwe Humanitarian Response Plan, 2008–16 30 7 A Risk Layering Strategy for the Government of Zimbabwe 32 8 Grouping Farmers to Develop Different Strategies for Agricultural Innovation 34 9 Zimbabwe: Financial Risk Layering to Respond to Natural Disasters 41 Appendix Figures A.1 Tobacco Growers and Area 50 A.2 Tobacco Supply Chain Actors 50 A.3 Cotton Supply Chain 51 A.4 Sugarcane Supply Chain 52 A.5 Wheat Supply Chain 53 A.6 Horticulture Supply Chain 55 A.7 Beef Supply Chain 56 B.1 Yield Trends and Crop Losses, Tobacco 58 B.2 Yield Trends and Crop Losses in Relation to Severe Weather Events, Maize 60 B.3 Maize Yields Over Time (National Yields and Yields at Different Scales of Production) 61 C.1 Zimbabwe: Historic Evolution of Maize Yields 64 C.2 Zimbabwe: Maize Yield Deviations from Expected Yields 65 C.3 Zimbabwe: Detrended Maize Yields 65 C.4 Zimbabwe Maize: Best Fit Probability Distribution Function Over Detrended Yields 65 C.5 Methodology Used for the Assessment 67 C.6 Zimbabwe: Production of Main Crops (t), 1986–2015 68 C.7 Zimbabwe: Yields of Main Crops (kg/ha), 1986–2015 69 C.8 Expected LaR for Different Recurrence Periods for the Whole Crop Portfolio in Zimbabwe 69 C.9 Zimbabwe: Maize Production (t), 1986–2015 70 C.10 Zimbabwe: Maize Yields (kg/ha), 1986–2015 70 C.11 Zimbabwe: Expected LaR for Maize for Different Recurrence Periods 71 C.12 Zimbabwe: Tobacco Production (t), 1986–2015 71 C.13 Zimbabwe: Tobacco Yields (kg/ha), 1986–2015 72 C.14 Expected LaR for Different Recurrence Periods for Tobacco in Zimbabwe 72 C.15 Zimbabwe: Sugarcane Production (t), 1986–2015 73 C.16 Zimbabwe: Sugarcane Yields (kg/ha), 1986–2015 74 C.17 Zimbabwe: Expected LaR for Sugarcane for Different Recurrence Periods 74 C.18 Zimbabwe: Groundnut Production (t), 1986–2015 75 C.19 Zimbabwe: Groundnut Yields (kg/ha), 1986–2015 75 C.20 Zimbabwe: Expected LaR for Groundnuts for Different Recurrence Periods 76 C.21 Zimbabwe: Cotton Production (t), 1986–2015 77 C.22 Zimbabwe: Cotton Yields, 1986-2015 77 C.23 Zimbabwe: Expected LaR for Cotton for Different Recurrence Periods 78 C.24 Zimbabwe: Coffee Production (t), 1986–2015 78 C.25 Zimbabwe: Coffee Yields (kg/ha), 1986-2015 79 C.26 Zimbabwe: Sorghum Production (t), 1986–2015 79 C.27 Zimbabwe: Sorghum Yields (kg/ha), 1986-2015 80 C.28 Zimbabwe: Wheat Production (t), 1986–2015 81 C.29 Zimbabwe: Wheat Yields (kg/ha), 1986–2015 81 C.30 Zimbabwe: Soybean Production (t), 1986–2015 82 C.31 Zimbabwe: Soybean Yields (kg/ha), 1986–2015 82 D.1 Maize: Estimated Annual Yield Losses 84 D.2 Groundnuts: Estimated Annual Yield Losses 84 D.3 National Retail Maize Prices in Zimbabwe 85 D.4 Seed Cotton: Estimated Annual Yield Losses 86 x Contents D.5 Wheat: Estimated Annual Losses 86 D.6 Tobacco: Estimated Annual Yield Losses 87 D.7 Sugarcane: Estimated Annual Yield Losses 88 D.8 Effect of Weather-Related Risks on Livestock Numbers (Headcount), 2009–17 90 D.9 Effect of Weather-Related Risks on Milk Yields, 1987–2017 90 D.10 Poultry Meat Supply and Stocking Capacity, 2015–17 91 Boxes 1 Salient Characteristics and Diagnosis of the Early Warning System in Zimbabwe 37 2 Recent Innovations in Weather Index Insurance in Africa 39 ABBREVIATIONS AND ACRONYMS AGRITEX Agricultural Technical and Extension Services AIS Agricultural Innovation System ARC African Risk Capacity ARDA-DDP Agriculture and Rural Development Authority – Dairy Development Program ASP Adaptive Social Protection CADRI Capacity for Disaster Reduction Initiative of the United Nations CERF Central Emergency Response Fund of the United Nations CIAT International Center for Tropical Agriculture (Centro Internacional de Agricultura Tropical) CIMMYT International Maize and Wheat Improvement Center (Centro Internacional del Mejoramiento de Maíz y Trigo) CPO Civil Protection Organization DCP Department of Civil Protection DR-SS Department of Research and Specialists Services DRM Disaster risk management ENSO El Niño Southern Oscillation EWS Early warning system FAO Food and Agriculture Organization of the United Nations FAOSTAT Food and Agriculture Organization statistical database FNC Food and Nutrition Council GDP Gross domestic product GFDDR Disaster Risk Financing Facility GMB Grain Marketing Board GoZ Government of Zimbabwe GVP Gross value of production ICT Information and communication technology LaR Loss at risk MAMID Ministry of Agriculture, Mechanization and Irrigation Development MLACWRR Ministry of Lands, Agriculture, Climate, Water and Rural Resettlement MLARR Ministry of Lands, Agriculture and Rural Resettlement MoFED Ministry of Finance and Economic Development xi xii Abbreviations and Acronyms NADF National Association of Dairy Farmers NCCRS National Climate Change Response Strategy NCPCC National Civil Protection Coordination Committee NDVI Normalized difference vegetation index NGO Non-governmental organization OCHA Office for the Coordination of Humanitarian Affairs (United Nations) PML Probable Maximum Loss PQS Plant Quarantine Service R&D Research and development SGR Strategic Grain Reserve SPS Sanitary and phytosanitary UN United Nations UNDP United Nations Development Programme UNICEF United Nations Children’s Fund USAID United States Agency for International Development WII Weather Index Insurance ZFU Zimbabwe Farmers Union ZCFU Zimbabwe Commercial Farmers Union ZDFA Zimbabwe Dairy Farmers Association ZIMSTAT Zimbabwe National Statistical Agency ZimVAC Zimbabwe Vulnerability Assessment Committee EXECUTIVE SUMMARY This report presents an assessment of Zimbabwe’s losses represent 7.3 percent of agricultural GDP. agriculture sector disaster risk and management Moreover, the losses change from year to year, capacity. The findings indicate that Zimbabwe is for example, the value of crop losses was esti- highly exposed to agricultural risks and has limited mated at US$321 million in the drought year capacity to manage risk at various levels. The report of 2001 and reached a virtually catastrophic shows that disaster-related shocks along Zimbabwe’s level—US$513 million—in 2008. agricultural supply chains directly translate to vola- 2. The most important agricultural risk in Zim- tility in agricultural GDP. Such shocks have a sub- babwe is drought. It affects agricultural pro- stantial impact on economic growth, food security, duction and food security. For example, the and fiscal balance. 2015/16 drought induced by ENSO caused When catastrophic disasters occur, the economy agricultural output to fall and an estimated absorbs the shocks, without benefiting from any 4 million people were food insecure in 2016. instruments that transfer the risk to markets and Increasingly frequent and severe droughts in coping ability. The increasing prevalence of “shock- southern and western Zimbabwe are making recovery-shock” cycles impairs Zimbabwe’s ability these areas unsuitable for rainfed maize pro- to plan and pursue a sustainable development path. duction highlighting the need to reconsider The findings presented here confirm that it is highly the boundaries and crop suitability patterns pertinent for Zimbabwe to strengthen the capacity to established for Zimbabwe’s agro-ecological manage risk at various levels, from the smallholder zones. Other weather-related risks are also farmer, to other participants along the supply chain, frequent but have lower and more isolated to consumers (who require a reliable, safe food sup- impacts. Sanitary and phytosanitary risks are ply), and ultimately to the government to manage common among all crops and livestock; they natural disasters. are usually managed if agrochemicals and vac- The assessment provides the following evidence cines are available, and farmers have resources on sources of risks and plausible risk management to purchase them. Price volatility has also been solutions. It is our hope that the report contributes reported as a risk, and price stabilization poli- to action by the Government of Zimbabwe to adopt cies are perceived to have added more uncer- a proactive and integrated risk management strategy tainty to the market. appropriate to the current structure of the agricul- 3. Poor farmers are most exposed to produc- tural sector. tion and market risks. Not only are they more exposed to risks, but their initial vulnerability is higher, and they have a lower capacity to man- Agricultural Risk Exposure age agricultural risks. In sum, managing a great 1. Zimbabwe loses approximately US$126 mil- portion of agricultural risks in Zimbabwe will lion each year due to production risks. These require that special attention is given to building 1 2 Zimbabwe: Agriculture Sector Disaster Risk Assessment resilient, diversified agriculture for smallholder reduce risk and improve resilience at the farm farmers in these high risk exposed areas. level. The report identifies three areas in which 4. To a great extent, the management of agricul- the Agriculture Innovation System will benefit tural risk in Zimbabwe depends on the capac- from leap-frogging: ity of smallholders, who dominate agricultural – Shifting the paradigm from conventional, production. Yet, smallholders have limited ability high-input agriculture to knowledge-intensive to reduce such risks on their own. In this context, sustainable intensification of agriculture; the public sector can play an important role in – Leaping from uniform production patterns supporting the agricultural sector by (1) assisting to more specialized production in a spatial farmers in strengthening their resilience at the development framework; and farm level to improve production and productiv- – Pursuing the digitalization of agriculture. ity and (2) by providing timely disaster response If the Agriculture Innovation System is to support after high-impact events. The capacity undertake these strategic approaches, it will of the public sector to provide the necessary require policy support and investments to support appears weak, however. strengthen the public and private institutions 5. Important gender asymmetries in asset that support smallholder farmers and bring ownership and access to services leave women about change. farmers more exposed to risks compared to 8. Risk transfer mechanisms have the potential to male farmers, and render them less able to miti- transfer the residual risk to the capital markets. gate the risks or cope with them as they occur. Agricultural insurance is one potential risk trans- In Zimbabwe, women constitute 54 percent of fer tool that farmers and other stakeholders can the agricultural labor force, but men have better use to manage risks that cannot be mitigated at access to land than women. Currently, 18 percent the farm level. Insurance instruments transfer of A1 farmers and 12 percent of A2 farmers are part of that risk to another party in return for female farmers; collectively they have access to a fee (or premium). Where it is available and 10 percent of the land redistributed under the affordable, agricultural insurance (for crops and/ Fast Track Land Reform. Women own 1,900 of or livestock) under public-private partnership the 18,000 farms in the A2 zone. In the commer- arrangements can greatly benefit large groups of cial farming sector, 80 percent of cattle are owned farming households: by men and 20 percent by women, while on com- – Insurance can (and should) be used to comple- munal farms only 35 percent of cattle are owned ment other risk management approaches. In by women. the event of a major weather shock, insurance can be designed to protect against revenue or consumption losses, enabling households Towards a Risk Management Strategy to avoid selling critical livelihood assets or 6. An integrated risk management strategy is drawing on savings. needed. The report recommends the need to – Insurance can assist farmers in accessing new start the process towards an integrated risk opportunities by improving their ability to management strategy with a package of inter- either borrow money or in-kind credits. In ventions related to mitigating risk at farm doing so, farm households may potentially level, possibilities to transfer agricultural risk experience higher returns. to financial markets, and adopting a proactive – Innovative risk transfer programs such as disaster risk financing strategy to cope with risk agricultural insurance at the sovereign and at catastrophic levels. farmer level are being implemented in vari- 7. Risk mitigation at farm level as a priority. An ous sub-Saharan countries. These programs integrated agricultural risk management strat- could provide valuable lessons on how to egy for the current context in Zimbabwe must transfer agricultural risk to capital markets in promote risk mitigation measures at the farm Zimbabwe. level as a priority. This requires a leap-frogging 9. Disaster risk financing needs to be proactive. approach in Agriculture Innovation Systems. Disaster risk financing in Zimbabwe has relied The approach can strengthen the capacity to on humanitarian funds owing to the severe Executive Summary 3 fiscal constraints of the government balance gency funds to natural disasters and emergen- sheet, and this situation is likely to continue cies, it would not cover the estimated average until the economy revives. The government annual gap. There are no financial margins to earmarks around US$35 million a year for con- cushion the effects of agricultural risk and natu- tingencies. In times of tight budgeting, that ral disasters, and the country absorbs the shocks fund is meant to cover many other contingen- without transferring any of the risks to markets. cies arising from all sectors of the economy and 10. In summary, Zimbabwe should start transition- not only those arising from natural disasters. In ing away from its current reactive strategy for practice, the government makes relatively small managing disaster and agricultural risks and allocations on a yearly basis for immediate res- move toward a proactive integrated risk man- cue and emergency operations following disas- agement strategy that combines improvements ters, recognizing that humanitarian assistance for managing risk at the farm level, risks trans- takes time to approve and disburse. Yet even if fer mechanisms and effective catastrophic risk the government were to allocate all of its contin- management strategy. Chapter 1 INTRODUCTION Agricultural risk has been a continuing concern is to inform GoZ decisions on agricultural risk man- of the Government of Zimbabwe (GoZ), owing to agement and risk financing strategies. The assess- agriculture’s pivotal role in Zimbabwe’s economy ment provides analytical evidence on sources of with respect to jobs, incomes, exports, and poverty risks and plausible risk management solutions. The reduction. Agriculture accounts for 11 percent of findings and recommendations emerging from this gross domestic product (GDP) and is the main source assessment are intended to contribute to the adop- of livelihood, employment, and income for around tion of a proactive and integrated risk management 67 percent of the population.1 Agricultural output strategy that is appropriate to the current structure through the years has shown considerable volatility, of the agricultural sector. resulting in high losses in the agricultural sector. As With those objectives in mind, this report is orga- this assessment will demonstrate, Zimbabwe loses nized as follows. This first chapter provides contex- approximately US$126 million each year on average tual information on Zimbabwe’s agricultural sector due to production risks that could be better managed. and the methodology used to conduct the disaster These losses represent 7.3 percent of agricultural risk assessment. Chapter 2 presents estimates of the GDP. Losses in years when production risks are high annual average losses incurred in the agricultural can escalate to virtually catastrophic levels. For exam- sector due to production risk at different levels of ple, losses in the drought year of 2001 were estimated intensity. Risk profiles developed for key agricultural at US$321 million, and in 2008, when agriculture was commodities and livestock are described in Chapter 3 seriously affected by drought and financial restric- and used to prioritize agricultural risks for the sec- tions, losses escalated to US$513 million. Such losses tor as a whole. Chapter 4 presents an evaluation of have a direct impact on growth, food security, and fis- the capacity to manage agricultural risks at different cal balances. Recently, effects of the El Niño Southern levels. Recommendations for an improved risk man- Oscillation (ENSO) during the 2015/16 cropping agement system are discussed in Chapter 5. Chapter 6 season produced low rainfall and drought, which led offers concluding remarks. to large food deficits. At the peak of the lean season prior to the subsequent harvest in 2017, an estimated 1.1  Overview of the Agricultural Sector 4 million people needed temporary food assistance.2 Strengthening Zimbabwe’s resilience to agricul- Aside from contributing 11 percent of GDP and sup- tural risk, particularly the resilience of its small-scale porting the livelihoods of approximately 67 percent producers, is becoming a key development priority of the population, agriculture also serves as the back- as a way of making agricultural investments more bone of Zimbabwe’s largely agro-based industrial sustainable, strengthening food security, and reduc- sector. Agriculture-related employment supports ing rural poverty. The objective of this risk assess- one-third of the formal labor force. ment, which was undertaken by the World Bank at Most agricultural land has been worked by small- the request of the Government of Zimbabwe (GoZ),3 scale producers since the Fast Track Land Reform 5 6 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table 1.  Land Holdings Before and After the Introduction of Fast Track Land Reform, Zimbabwe Area (million ha) Land category 1980 2000 2009 Communal area 16.4 16.4 16.4 Old resettlement 0.0 3.5 3.5 New resettlement A1 0.0 0.0 4.1 New resettlement A2 0.0 0.0 3.5 Small-scale commercial farms 1.4 1.4 1.4 Large-scale commercial farms 15.5 11.7 3.4 State farms 0.5 0.7 0.7 Urban land 0.2 0.3 0.3 National parks and forest land 5.1 5.1 5.1 Unallocated land 0.0 0.0 0.7 Source: Kasiyano 2017. Note: Inhabitants of communal areas do not possess title to the land, which is communally owned and allocated to families for arable farming and settlement. “Old resettlement” refers to areas where land was redistributed in the 1980s. The two main groups benefiting from land redistribution during and after the 1990s are smallholders (the “new resettlement A1” group) and medium-scale producers (the “new resettlement A2” group). was introduced after 2000. Table 1 shows that the cotton production offers the main link to markets area dedicated to large commercial farms decreased and is a key component of livelihood strategies from 15.5 million hectares in 1980 to 3.4 million among isolated and vulnerable rural households. hectares in 2009. The fact that agriculture is domi- After tobacco, cotton is Zimbabwe’s second or nated by small-scale production presents important third (together with sugar) largest agricultural challenges to the government with respect to increas- foreign exchange earner, contributing 12.6 percent ing productivity, linking producers to markets, and to agricultural GDP. managing risk. Zimbabwean agriculture is widely diversified, Methodology Used for This Risk 1.2  owing to diverse agro-climatic conditions that make Assessment it possible to produce over 20 types of food and cash crops as well as poultry, pigs, and dairy and To undertake this assessment, the team used an estab- beef cattle. The most important agricultural com- lished participatory methodology developed by the modities are the staple food grains that constitute World Bank that prioritizes agricultural risks across the basis of local diets—maize, wheat, small grains a set of representative agricultural commodities.5 (millet and sorghum), groundnuts, and beans— A comprehensive framework is used to assess and and export and cash crops (mainly tobacco, cotton, effectively start the process of managing systemic sugarcane, and horticultural crops). Appendix A risks to the agricultural sector by assessing the fre- contains detailed descriptions of the importance, quency and intensity of observed risks, which makes performance, and governing structure of each of it possible to estimate the value of their impacts. An these supply chains. understanding of agricultural stakeholders’ capacity Four of these commodities play particularly to manage risk in specific supply chains also helps critical roles. Maize is the main staple food crop to prioritize the most important risk management and therefore at the center of national food secu- investments. rity. Groundnuts are critical for household nutri- The agricultural supply chains selected for this tion. Tobacco is the major agricultural export assessment (Table 2) represent 90 percent of the total commodity, contributing 25.2 percent of agricul- agricultural value added, measured as an average of tural GDP in 2016, accounting for over 50 percent the last five years, and they use most of the agricul- of agricultural exports, and representing an aver- tural land.6 Within this group of products, maize age of 29 percent of the country’s total exports in (as noted) is the main staple food, and tobacco is a 2016 and 2017.4 Cotton is a crop of strategic impor- major export product (second only to minerals as tance for promoting inclusive economic growth, a source of export proceeds). Risks identified along poverty alleviation, rural development, and food the supply chains for these commodities can help to security in Zimbabwe, because in various regions reveal the drivers of volatility in agricultural growth, Introduction 7 Table 2.  Zimbabwe: Main Agricultural Commodities Included in the Risk Assessment Percentage of Exports agricultural GDP Area (average 2012–16, Agricultural commodity (average 2012–16)* (average 2012–16) (ha)** US$ 000s)*** Tobacco 36.08% 91,816 879,198 Cattle production (beef and dairy) 10.81% 34,059 Maize 10.07% 1,460,810 790 Cotton 9.89% 224,923 103,214 Sugarcane 6.59% 45,961 102,848 Horticulture (fruits, vegetables, etc.) 6.59% 69,612 18,072 Poultry 6.23% Wheat 2.38% 12,497 Groundnuts 1.83% 258,597 Total 90.48% 3,452,125 3,197,974 Source: * Ministry of Agriculture of Zimbabwe, Agricultural Statistical Bulletin, 2016; ** Ministry of Agriculture of Zimbabwe, Agricultural Statistical Bulletin, 2016 and FAOSTAT (sugar, horticulture, total); *** International Trade Center. Figure 1.  Relationship Between Agriculture Value Added and Overall GDP Growth Zimbabwe - Value added constant (annual % growth) 40 30 20 10 0 Moderate drought Severe –10 Moderate drought in drought some –20 Severe regions drought Severe Land reform, –30 drought land invasions and drought –40 Drought, shortage of inputs, hyper inflation, –50 cash shortage 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Agriculture, value added (annual % growth) GDP growth (annual %) Source: Based on WDI 2017. the sources of food insecurity in the country, and the Notes need for risk financing at different levels of intensity. 1. WDI (2017). Shocks along these supply chains directly translate 2. ZimVAC (2016). into volatility in agricultural GDP, aside from having 3. In this report, “agricultural risk” is understood as impacts on overall economic growth. Figure 1 illustrates an unexpected and sudden event that has the potential to the correlation between agricultural GDP and overall cause losses to stakeholders in the agricultural sector. 4. Reserve Bank of Zimbabwe (2017). GDP growth. More specifically, it shows the impact of 5. World Bank (2016a). severe and moderate drought in slowing growth both 6. Appendix D presents a detailed description and analy- within the sector and the economy as a whole. sis of risk and production losses for food and cash crops. Chapter 2 AGRICULTURAL RISK EXPOSURE An agricultural risk assessment aims to arrive at and sugarcane—the six crops that account for a short list of risks that are key priorities because 68 percent of agricultural GDP—and thus captures they are the main drivers of agricultural volatility. all production risks such as drought, floods, and pest That process begins by estimating the overall losses and disease outbreaks. incurred by the agricultural sector as a result of Based on this methodology (described in Appen­ production risks at the farm level;1 those losses will dix B), the aggregate value of production losses in reveal the relative magnitude of the impact of agri- those six crops, arising from production risks realized cultural risk. Note that the figures presented here do between 1986 and 2016 and monetized at 2016–17 not include post-harvest losses and losses incurred prices, is shown in Table 3. In brief, over 1986–2016, by the sector due to price volatility and market- production risks led to losses in crop production related risk. valued at approximately US$126 million per year on average, representing an annual average loss of around 7.3 percent of agricultural GDP. As noted 2.1  Production Losses in Chapter 1, the value of crop losses was esti- Since agricultural production is exposed to normal mated at US$321 million in the drought year of 2001 inter-annual variations and occasional shocks caused and reached a virtually catastrophic level—US$513 by weather, disease, and factors related to markets million—in 2008, when drought as well as finan- and policy, it is pertinent to identify the main sys- cial restrictions seriously affected agriculture. The temic shocks that affect output beyond manageable sources of risk that are driving this output volatility thresholds.2 The data available on actual losses are are described in the following chapter, and a more not always accurate or consistent enough to facili- detailed explanation is available in Appendix D. tate comparisons and to rank the costs of adverse Maize and sugarcane accrued the largest average events. The analysis presented here is thus based on losses, though maize was produced on 1.2 million estimates of the “indicative” value of potential losses hectares against 45,000 hectares of sugarcane (in over the longest period that historical data allow. 2016). Sugarcane suffered only three severe droughts, For the purpose of this assessment and considering but losses were very high in 1991/92. Tobacco, planted the data available, the period analyzed is 1986–2016, on 108,000 hectares in 2015, suffered through four using national statistics provided by the Ministry of droughts with more modest losses than maize and Agriculture, Mechanization and Irrigation Devel­ sugarcane (30,000–58,000 tons). opment (MAMID). In line with the aims of this disaster risk assessment The indicative value of agricultural output lost for for agriculture, these estimates have demonstrated the a particular year is calculated as the downward yield impact of weather-related risks over the past 30 years, deviation from the historic trend.3 The quantifica- but it is important to consider that climate change will tion of losses presented here is based on yield data have impacts on Zimbabwe’s agricultural sector over for tobacco, maize, wheat, seed cotton, groundnuts, and above the weather-related disasters that are the 9 10 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table 3.  Estimated Volume and Value of Crop Losses in Zimbabwe (Average of 30 Years, 1986–2016) Annual average Annual average Percentage of losses (30 years) losses (30 years) agricultural GDP due to production as percentage of Losses 2001 Losses 2008 Crop (average 2012–16) risks (US$) agricultural GDP (US$) (US$) Tobacco 36.8% 19,804,922 1.15% 0 184,527,678 Maize 10.1% 44,601,263 2.59% 241,748,598 273,905,499 Seed cotton 9.9% 4,151,986 0.24% 22,483,106 23,791,027 Sugarcane 6.6% 47,015,487 2.73% 0 0 Wheat 2.8% 3,393,939 0.20% 0 31,066,372 Groundnuts 1.8% 7,056,972 0.41% 57,746,716 0 Total 6 crops 68% 126,024,570 7.33% 321,978,420 513,290,577 Table 4.  Estimated Loss at Risk (LaR) for a Portfolio of Crops at Different Return Periods Recurrence period (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 27.17% 32.94% 36.28% 39.93% 41.77% 42.92% 44.23% 46.67% LaR (US$ millions) 256.7 311.2 342.8 377.3 394.7 405.6 417.9 441.0 Source: Based on data from MAMID and FAOSTAT. Note: The crops are coffee, cotton, groundnuts, maize, sorghum, soybeans, sugarcane, tobacco, and wheat. focus of this assessment. For example, a recent study Zimbabwe will need to be proactive and holistic, using a Computable General Equilibrium Model has including public support to build risk mitigation/ found that under a dry/hot future climate scenario, adaptation capacity for smallholders as well as ex-ante Zimbabwe may lose about 2.3 percent of its 2030 financing instruments that can be used to protect GDP—or up to US$370 million. smallholders’ livelihoods from disasters of various levels. The following chapter provides a basis for developing such a strategy by prioritizing the major Production Losses for 2.2  production risks. Different Return Periods Severe losses in agricultural production may arise Notes in Zimbabwe as a result of adverse events that recur periodically, sometimes over relatively short spans of 1. The detrending analysis consisted of a weighted time. A production risk assessment was performed average of a lineal detrending and a polynomial of second for a selected portfolio of nine crops in Zimbabwe order detrending. Both show a similar trend. The period considered for the detrending comprised the full series over different periods of recurrence.4 The crop port- (1986–2016). During this period, the land reform and folio may face a loss equivalent to 27.2 percent of changes in property rights at the beginning of 2000 severely the national crop gross value of production (GVP)5 affected agricultural output. Additionally, Zimbabwe expe- equivalent to US$256.7 million in 1 of every 10 years, rienced successive droughts from 2001 to 2008, followed by crop losses of 39.9 percent of the national GVP (or a season of excessive rain, when annual production started US$377.3 million) in 1 of every 100 years, and crop to increase, spurred on by deregulation. Since it is not pos- sible to differentiate the impacts of changes in the produc- losses of 44.23 percent of GVP (US$417.9 million) tion environment from the effects of the drought/excessive in 1 of every 250 years (Table 4). For details on the rainfall seasons, the series was not adjusted, which means calculations, see Appendix C. that these estimates must be interpreted with caution. See The current structure of the agricultural sector Appendix C for technical details. (that is, the dominance of small-scale producers), the 2. The smaller inter-annual variations in yield that can changes in technology arising from the shift from be part of the cost to farmers of doing business. 3. See Appendix B. Yield deviations are calculated with large-scale production to small-scale production, respect to the historic trend line of the yields. Then, those and the effects of climate change all demand a new years in which the negative deviations are greater in abso- approach to managing agricultural risk. Going for- lute value than the standard deviation of the deviations are ward, the agricultural risk management strategy in taken as the years in which significant risk events occur. Agricultural Risk Exposure 11 Then, for those years, the deviations from the trend are 4. The analysis covered coffee, cotton, groundnuts, multiplied by the harvested area. This approach makes it maize, sorghum, soybeans, sugarcane, tobacco, and wheat possible to estimate the volume of production losses. Next, in the aggregate for the whole country. The analysis was losses in value are calculated by multiplying the volume of based on country-level data on crop area, production, and losses by the price of crops. Note that the assessments of yields for the period 1986, up to and including 2015 and market risk and enabling environment risk use a different 2016 prices. methodology based on price series analysis and stakeholder 5. For this report, the GVP measures the total value of interviews. goods produced by the whole portfolio under study. Chapter 3 SOURCES OF RISK To isolate the most important risks that drive volatility suitable for intensive cropping and livestock produc- in agricultural GDP and food insecurity at the national tion. The cropping systems are based on flue-cured level, this chapter identifies and prioritizes production tobacco, maize, cotton, wheat, soybeans, sorghum, risks for each of the agricultural commodities and live- groundnuts, seed maize, and burley tobacco grown stock production activities studied for this assessment.1 under dryland conditions as well as with supplemen- The discussion begins with a review of Zimbabwe’s tary irrigation in the wet months. Irrigated crops agro-ecological regions and then moves on to present include wheat and barley grown in the colder and the results of the risk identification and prioritization drier months (May-September). Region II is suited to analysis. Results of that analysis are used to formulate intensive livestock production based on pastures and the potential agricultural risk management strategies pen-fattening utilizing crop residues and grain. The described in later chapters. main livestock production systems include beef, dairy, pig, and poultry systems. 3.1  Agro-ecological Regions Region III is located mainly in the mid-altitude Zimbabwe is a landlocked country divided into five areas, characterized by annual rainfall of 500– agro-ecological regions defined by their rainfall 750 milli­meters per year, midseason dry spells, and regime, soil quality, and vegetation, among other high temperatures. This semi-intensive farming factors (Figure 2).2 In general, farm households in region is suited for livestock production, together Regions II and III allocate 40–50 percent of the arable with production of fodder crops and cash crops land under cultivation to food crops. The proportion under good farm management. The main crops are rises to 60–70 percent in the dry Regions IV and V.3 maize and cotton, and the region is also suitable for Region I lies in the east and is characterized by rain- producing groundnuts and sunflowers. fall of more than 1,000 millimeters per year (which Region IV, located in the low-lying areas in the falls throughout the year), low temperatures, high alti- north and south, has annual rainfall of 450–650 mil- tude, and steep slopes. The country’s timber produc- limeters per year, severe dry spells during the rainy tion is located in this region. Region I is ideally suited season, and frequent seasonal droughts. Although for intensive diversified agriculture and livestock pro- Region IV is considered unsuitable for dryland crop- duction, mainly dairy farming. Common crops are ping, smallholder farmers grow drought-tolerant vari- tropical crops such as coffee and tea, deciduous fruits eties of maize, sorghum, pearl millet, and finger millet. such as bananas and apples, and horticultural crops, This region is ideally suited for raising cattle in exten- such as potatoes, peas, and other vegetables. sive production systems and for wildlife production. Region II is located in the middle of northern Region V covers the lowland areas below 900 meters Zimbabwe. Rainfall ranges from 750 millimeters to above sea level in the north and south, with highly 1,000 millimeters per year and is fairly reliable, fall- erratic rainfall that is less than 650 millimeters per ing from November to March/April. Because of the year. Although the northern part of Region V along the reliable rainfall and generally good soils, this region is Zambezi River receives reasonable rainfall, its uneven 13 14 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure 2.  Agro-Ecological Regions and Soil Map MOZAMBIQUE V ZAMBIA IV V IV IIA III III I IIB III IV III I BOTSWANA V E BIQU Z AM MO SOUTH AFRICA 100 0 100 200 300 km topography and poor soils make it unsuitable for crop introduced in Chapter 1. At the request of the GoZ, production. Generally, Region V is suitable for exten- this chapter focuses on production risks at the farm sive cattle production and game-ranching, and it is also level, as there is much interest in using the findings appropriate for forestry and wildlife/tourism. to develop options for more proactive agricultural Although both Regions IV and V are too dry for disaster risk management practices. crop production, households on the communal lands The impacts of different types of risks on the dif- in these regions grow grain crops (maize and millet) ferent crops and supply chains vary depending on the for food security and produce some cash crops such severity of the event, the risk exposure, and the capac- as cotton. Crop yields are extremely low, and the risk ity to manage risk. For instance, the yields of ground- of crop failure is high (likely to occur in one out of nuts and maize show a higher and positive correlation three to five years). Cattle and goat production are with rainfall, while tobacco, sugarcane, and lint cotton major sources of cash income. Most of the commu- yields have a lower correlation with rainfall (Figure 3). nal lands are in the marginal agro-ecological regions. These results indicate that groundnut and maize yields They are characterized by low rainfall (averaging are more highly determined by weather-related factors 400–500 millimeters per year), severe dry spells in than yields of tobacco, sugarcane, and cotton. the rainy season, and shallow soils of low fertility. Tobacco and to a lesser extent sugarcane and cotton Such conditions are very marginal for the produc- producers and their respective supply chains have rel- tion of major crops, even drought-resistant grain atively effective production risk management mecha- crops such as sorghum and millet. nisms available (such as irrigation, improved seed, and good agricultural practices), mostly because contract farming dominates production of these commodities. 3.2  Risk Identification In addition, tobacco can be planted early to avoid pos- This section presents Zimbabwe’s exposure to risks sible mid-season dry spells, while the opposite occurs for the various crop and livestock commodities with the other crops. Sources of Risk 15 Figure 3.  Correlation Between Rainfall and Crop Yields, Zimbabwe Rainfall (mm) and Groundnuts yields (tons/ha), Deviation Rainfall (mm) and Maize yields (tons/ha), Deviation from Trend, 1985/6-2015/16 from Trend, 1986/7-2015/16 0.4 0.6 0.3 0.4 R2 = 0.2349 0.2 Groundnuts yields 0.2 2 R = 0.4441 0 Mize yields 0.1 –0.2 0 –0.4 –0.1 –0.6 –0.2 –0.8 –0.3 –1 –0.4 –1.2 –300 –200 –100 0 100 200 300 –300 –200 –100 0 100 200 300 Rainfall Rainfall Rainfall (mm) and Tobacco yields (tons/ha), Deviation Rainfall (mm) and Sugar yields (tons/ha), Deviation from Trend, 1984/5-2015/16 from Trend, 1986/7-2015/16 1 40 0.8 R2 = 0.1449 0.6 20 0.4 0 Mize yields Tobacco yields 0.2 0 –0.2 –0.2 R2 = 0.0131 –0.4 –0.4 –0.6 –0.6 –0.8 –0.8 –1 –1.2 –100 –300 –200 –100 0 100 200 300 –300 –200 –100 0 100 200 300 Rainfall Rainfall Rainfall (mm) and Cotton yields (tons/ha), Deviation from Trend, 1986/7-2015/16 0.4 0.2 Tobacco yields 0 R2 = 0.1657 –0.2 –0.4 –0.6 –0.8 –300 –200 –100 0 100 200 300 Rainfall Source: Based on MAMID data. 16 Zimbabwe: Agriculture Sector Disaster Risk Assessment The risks for the various agricultural supply chains of occurrence, from highly probable (1 in 3 years) to were identified through quantitative analysis, second- less probable (1 in 20 years), and their expected impact ary information, and interviews with numerous stake- (from High to Catastrophic levels of losses for the holders in each commodity supply chain.4 This chapter sector). This figure has been adjusted to reflect the summarizes the risk identification exercise for each capacity to manage risks by stakeholders. commodity; for technical details, see Appendix D. In the figure, the risks plotted in red (weather The most important agricultural risk in Zimbabwe related) and green (pests and diseases) are production is drought. Other weather-related risks are also fre- risks, whereas the circles in yellow show market and quent but have lower and more isolated impacts.5 enabling environment risks. Those risks plotted toward Sanitary and phytosanitary risks are common among the right-hand side of the figure are the most signifi- all crops and livestock; they are usually managed as cant risks in terms of their potential to cause the great- long as agrochemicals and vaccines are available and est losses and their lower capacity to be managed by farmers have resources to purchase them. Price vola- stakeholders or the government. For example, a critical tility has also been reported as a risk, and price sta- risk, occurring approximately every 5 years, is severe bilization policies are perceived to have added more drought with high temperatures in agro-ecological uncertainty to the market. Regions IV and V. This type of drought affects both In general, at a certain manageable degree of crops and animals, and because it causes large losses in intensity the main production risks are only partially fragile agro-ecological regions, it is a high priority that prevented or mitigated by familiar risk mitigation requires effective mitigation measures (to be discussed practices such as irrigation, diversifying crops, using in Chapter 5). appropriate inputs, and so on. For infrequent and Other noteworthy risks are severe drought with intense disasters, ideally households should be able high temperatures, occurring in all regions with a to transfer risk to capital markets. The majority of frequency of 1 in 10 years, and prolonged consecu- Zimbabwean households, however, apply limited risk tive seasons of under-normal rainfall in sugarcane mitigation measures and no transfer mechanisms, so production areas, occurring approximately every they are severely affected when risks materialize. In 10 years. Droughts can greatly affect agricultural pro- those cases, farmers tend to absorb the losses through duction and food security. For example, the 2015/16 coping strategies such as selling assets (which reduces drought induced by ENSO caused agricultural out- their disposable income), reducing meals, and pulling put to fall by 5 percent in 2016 (World Bank 2017a). children out of school, impacting their wellbeing in An estimated 4 million people were food insecure in the short and long run. 2016. Increasingly frequent and severe droughts in Table 5 summarizes results of the technical southern and western Zimbabwe are making these exercise detailed in Appendix D to identify pro- areas increasingly unsuitable for rainfed maize pro- duction risks by category of stakeholder, as well as duction (MAMID/FAO 2017), highlighting the need the risk management practices currently used by to reconsider the boundaries and crop suitability stakeholders. patterns established for Zimbabwe’s agro-ecological zones 60 years ago. Other weather-related events (such as prolonged 3.3  Risk Prioritization mid-season dry spells, erratic rainfall, and hail- This section narrows the list of agricultural production storms), as well as issues related to policy (such as risks presented in Table 5 down to a group of key prior- support prices and input provision to farmers), occur ity risks that are important at the national level because with relatively high frequency (1 in 3 years) and can of their potential to cause agricultural production vola- have highly negative impacts, although less so than tility and food insecurity. To better identify policies and other risks, on the different stakeholders of most allocate scarce resources for agricultural risk manage- supply chains. Two other risks with highly negative ment, the risks identified in Table 5 are prioritized in impacts are the uncertain availability of animal draft terms of: (1) their frequency of occurrence; (2) their power (1 in 10 years) and pests (fall armyworm) in potential to cause losses; and (3) the capacity of stake- maize production, as well as incursions of the Tuta holders to manage the risks. Figure 4 plots the risks that absoluta moth in horticulture and tobacco produc- were identified to be a priority based on the probability tion (1 in 20 years). Sources of Risk 17 Table 5.  Production Risks and Current Risk Management Practices by Stakeholder Group, Zimbabwe Current risk management Stakeholders Production risks (mitigation and transfer) Small-scale maize, • Severe droughts and high temperature in Regions IV • Water harvesting techniques (including digging groundnut, tobacco, and V (1 in 5 years), affecting all crops and animals infiltration pits) for crops cotton, and cattle • Severe droughts and high temperature in all regions • Conservation techniques (including mulching for producers (1 in 10 years), affecting all crops and animals cotton) and zero tillage for maize and • Erratic rainfall distribution affecting non-drought- • Staggering planting dates to spread the weather risk tolerant and non-irrigated crops (e.g., maize, tobacco) • Pest and disease control with chemicals in crops and Medium- to large-scale • Delayed onset of rains affecting mostly tobacco animals is common, including after harvest maize, tobacco, and farmers (1 in 5 years) • Crop rotation to avoid build-up of diseases and pests cattle producers (A2) • Prolonged mid-season dry spells affecting yields of all • Drought-, disease-, and pest-tolerant varieties/breeds crops except cotton and tobacco; tobacco is affected • Smallholder irrigation, mostly for maize (few) and mostly in terms of quality (1 in 3 years) about 30% of medium- and large-scale farmers have • Short rainy season, affecting quality and yields of irrigation facilities crops and grazing (1 in 3–5 years) • Vaccination against anthrax, foot and mouth, and • Floods affecting low-lying areas (1 in 10 years) black leg • Hailstorm for tobacco, which are confined to a limited • Fire guards in large-scale farms area but very destructive • Insurance against hailstorms and drought in tobacco, • Pests and diseases in the field (armyworm for maize required for contract farming and cotton, mealybug for cotton, rust in tobacco, Tuta absoluta in tomatoes, acaricide-resistant ticks in cattle, foot and mouth disease), and post-harvest pests and diseases • Introduction of viral Maize Lethal Necrosis Disease, which is currently present in Kenya and South Africa; it can cause very high losses and there is no control with chemicals Medium- to large-scale • Severe droughts (1 in 10 years), affecting availability • Irrigation farmers growing wheat of irrigation water • Early-maturing varieties under irrigation (A2) • Early onset of rains • Chemical applications • Pests and diseases • Fire guards Medium- to large-scale • Severe droughts (1 in 10 years) • Drilling boreholes to supplement surface irrigation (A2) sugarcane farmers • Prolonged consecutive seasons of under-normal water rainfall (1 in 3–5 years) • Insurance and • Pests and diseases (yellow sugar, Eldana saccharina) • Fire guards Sugar estates • Black maize beetle, affecting A2 farmers • Drought-, disease-, and pest-tolerant varieties • Chemical control of pests and diseases • Rotational use of pesticides and insecticides • Biological control of pests • Guarding fields to protect against theft and wildlife • Buy electricity generators Small-, medium-, • Severe droughts (1 in 10 years) affect production of all • Very exposed and large-scale horticultural products horticultural crop • Prolonged mid-season dry spells affecting yields producers (1 in 3 years) • Frost • Excess rainfall increases incidence of fungal diseases Commercial poultry • Disease outbreak (Newcastle, avian influenza) • Farmers tend to grow own maize and soybeans as a producers complement • Buy electricity generators Maize and wheat • Drop in the supply of raw material • Very exposed millers (Table continues next page) 18 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table 5.  (continued) Current risk management Stakeholders Production risks (mitigation and transfer) Tobacco traders • Drop in the supply of raw material • Contract farming to secure supplies Horticultural • Drop in the supply of raw material • Contract farming to secure supplies processors and traders • Own electricity generators Abattoirs • Drop in the supply of raw material • Keep some herds to supplement feeding and meet their customers’ quality requirements • Own electricity generators Dairy processors • Drop in the supply of raw material • Contract farming to secure supplies Input suppliers • Availability of maize and soybeans as main feed • Contract farming to secure supply of animal feed ingredients, which could become scarce because of drought Source: This table presents findings from field research with stakeholders, plus information obtained from the periodic reports of public and private sector agencies, and from informed opinion of experts. Figure 4.  Prioritized Agricultural Risks, Zimbabwe Tuta 1 in 20) Absoluta (horticulture & tobacco) Severe Army worm drought in maize with high temperatures in regions (1 in 10) Under normal I to III rainfall (sugar cane) Frequency Erratic rainfall Input price in regions (1 in 5) uncertainty IV and V (communal farms) Severe drought in regions IV to V Availability Hailstorm animal in tobacco draught (1 in 3) High losses Very high losses Catastrophic level losses Impace level Note: Production risks are plotted in red (weather related) and green (pests and diseases); market and enabling environment risks are in yellow. Sources of Risk 19 In conclusion, agricultural production and market Notes risks are more likely to impact smallholder farmers 1. The findings are in line with the World Bank (2016a) in Regions IV and V than to impact smallholders methodology for systematically assessing and prioritizing in other agro-ecological regions, and poor soils agricultural risk at the sector level. and deforestation are increasing the vulnerabil- 2. Based on FAO (2006), but note that the GoZ is ity of farmers in Regions IV and V. The group that redrawing the agro-ecological zones to better reflect is most exposed to production and market risks is current conditions. 3. Most crops are planted in November/December at poor farmers, especially farm households headed by the beginning of the rains and harvested between April women and children. Not only are they more exposed and June. Winter wheat, barley, and various horticultural to risks, but their initial vulnerability is higher and products are grown in the dry season under irrigation. they have a lower capacity to manage agricultural Irrigation schemes are also important in supplementing risks. In sum, managing a great portion of agricul- the production of wheat, tobacco, maize, cotton, soybeans, groundnuts, and coffee. tural risks in Zimbabwe will require that special 4. Results were presented at a technical workshop in attention is given to building a resilient, diversified Zimbabwe on May 7, 2018 to validate the interpretation. agriculture for smallholder farmers in these highly 5. With the exception of the most recent incursion of risk exposed areas. fall armyworm. Chapter 4 CAPACITY TO MANAGE AGRICULTURAL RISK The design of agricultural risk management strate- 4.1  Vulnerability Factors gies takes into account the frequency of risk events as well as the potential to cause losses, as seen in High levels of poverty, gender asymmetries, lack of previous chapters. It is equally important to assess assets, and absence of mechanisms to absorb income the capacity of stakeholders at various levels, from shocks all expose the rural population to agricultural smallholder farmers to other participants along the risks. That exposure leads to further vulnerability and supply chains, to manage the risks to which they negative impacts on livelihoods when agricultural are exposed. To a great extent, the management of risks rise to catastrophic levels. 1 agricultural risk in Zimbabwe will depend on the capacity of smallholders, who dominate agricultural 4.1.1 Poverty production, to reduce and curtail risk at the farm level through risk reduction strategies such as those Over 70 percent of Zimbabweans live in rural areas, identified during the field assessment and described and about 67 percent rely on agriculture for their in Chapter 3. Yet as the field assessment has shown, livelihoods. Poverty is very high in rural areas, where smallholders have limited ability to reduce and more than 72 percent of households live in chronic curtail agricultural risks. In this context, the pub- poverty (in comparison, in 2015 poverty rates were lic sector can play an important role in supporting 37.2 percent in Bulawayo and 36.4 percent in Harare). the agricultural sector by (1) assisting farmers in The incidence of poverty varies by province and strengthening their resilience at the farm level and district. Poverty is more prevalent in drier regions (2) by providing timely disaster response support (agro-ecological Regions IV and V). For example, after high-impact events. The capacity of the public the poverty atlas in 2015 indicated that poverty was sector to provide the necessary support appears weak, highest in Matebeleland North (a dry region), with however. a Gini coefficient of 85.7 percent in 2015. A 2003 This chapter starts by describing some of the poverty assessment found that the incidence of factors that make Zimbabwe’s rural households and poverty was higher among female-headed households farmers vulnerable to natural disasters. It then out- (around 72 percent) than male-headed households lines the public sector’s role in managing agricultural (58 percent).2 risks and capacity to do so, focusing first on public Agricultural risks exacerbate existing poverty agricultural services for strengthening the resilience traps and cause volatility in agricultural and eco- of smallholder farmers and second on a diagnosis of nomic growth. Crop failure caused by droughts and public sector capacity to respond to natural disasters. pests is often the biggest shock faced by rural house- The findings inform the suggestions for designing an holds and may also represent the biggest poverty agricultural risk management strategy, presented in trap. Agricultural risks have a profound impact on Chapter 5. poverty, as they undermine rural entrepreneurs’ 21 22 Zimbabwe: Agriculture Sector Disaster Risk Assessment (particularly producers’) possibilities to accumulate Agricultural Services for 4.2  assets, invest in and develop their businesses, and Strengthening Resilience gain access to health and education services. Given the prevailing agricultural risks and consider- able vulnerability of the rural population to those 4.1.2  Gender Asymmetries risks (especially female- and child-headed house- Important gender asymmetries in asset ownership holds in Regions IV and V), the public sector plays an important role in promoting the development and access to services leave women farmers more of innovations to increase resilience to disasters, exposed to risks compared to male farmers, and engage in the distribution of those innovations, and render them less able to mitigate the risks or cope train small-scale producers to use best agricultural with them as they occur. In Zimbabwe, women practices. It is also important for the public sector constitute 54 percent of agricultural labor force, to develop mechanisms that can rapidly control the but men have better access to land than women. spread of pests and diseases. In other words, the Currently, 18 percent of A1 farmers and 12 percent agricultural innovation and (phyto)sanitary systems of A2 farmers are female farmers; collectively they will be especially crucial elements of public sector have access to 10 percent of the land redistributed efforts to reduce small-scale producers’ vulnerability under the Fast Track Land Reform.3 Women own to agricultural risks. 1,900 of the 18,000 farms in the A2 zone. In the commercial farming sector, 80 percent of cattle are owned by men and 20 percent by woman, while on 4.2.1  Agricultural Innovation System communal farms only 35 percent of cattle are owned The public agricultural innovation system (AIS) in by women. Zimbabwe includes the Department of Research and Land and cattle are critical assets; ownership of Specialist Services (DR-SS), which is responsible for these assets is fundamental for individuals seeking providing research goods; the Agricultural Technical credit to develop an enterprise, since they are used and Extension Services (AGRITEX), charged with as collateral. Access to credit is a constraint for all providing extension services; agricultural education farmers, but only 2 percent of women farmers in for technical training; and the Agricultural Research communal lands have obtained credit compared to Council, whose role is to prioritize and coordinate 9.6 percent of men. Access to financing is directly agricultural research countrywide. Aside from these linked to the use of agricultural inputs and the mech- public research and development (R&D) services, anization of production and processing. Suitable the AIS includes the research programs in aca- farm machinery is needed to reduce the labor burden demia and the private sector, especially crop breed- in smallholder agriculture, especially the labor burden ing research by seed companies. Some of the larger of women farmers. Women’s restricted access to land seed companies do their own variety development makes them more vulnerable to poverty, as they have and agronomy research. The largest is Seed Co, no influence over the land assets and are deprived of a Zimbabwe-based company with a presence in more the water and other natural resources associated with than 10 African countries. Many of the seed compa- access to land. nies use genetic material provide by the International Maize and Wheat Improvement Center (CIMMYT) 4.1.3  Limited Income Buffers in their breeding programs. CIMMYT research includes work on drought-resistant maize varieties Livestock often serve as an income buffer for farmers, and climate-smart maize production systems for helping them to obtain additional resources in times smallholders. of need. When a natural disaster strikes, however, Government agricultural research spending farmers often have to sell their livestock at lower than has increased substantially after the multicurrency average prices, bringing little relief. As an alternative, regime was implemented in 2009–11, reaching many countries are developing agricultural insurance US$43.4 million. Even so, this level of expenditure products and extending them to small-scale pro- is low compared to public sector support for agri- ducers to help them manage agricultural risks, but culture in other countries in the region. Agricultural such products are limited in Zimbabwe. research spending represents 1.4 percent of GDP in Capacity to Manage Agricultural Risk 23 Zimbabwe, compared to 3.1 percent in Namibia and Research Institute based in the Plant Quarantine 2.9 percent in Botswana. Station in Mazowe. PQS is responsible for both the In terms of human resources, Zimbabwe has internal and international plant quarantine regimes. 208 full-time equivalent public sector researchers, It is also responsible for certifying plants and plant 58 percent of whom have graduate degrees. The public products for export and issuing the corresponding extension service (AGRITEX) has about 1,900 exten- phytosanitary certificates. It coordinates plant health sion workers to serve the country’s approximately and laboratory services and inspections of seed crops, two million small-scale farmers (one extension nurseries, warehouses, and other facilities. PQS worker for every 1,053 farmers). Irrigated areas have staff are based in the plant health offices and points a dedicated extension staff (about 150). The chal- of entry across the regions of Zimbabwe and are lenge of effectively reaching the multitude of farmers charged with both inland inspections and inspections is daunting, especially given the limited mobility of of imports. Inspections of cut flowers, fruits, and AGRITEX staff. The relationship between researches vegetables immediately prior to export are performed at DR-SS with extension at AGRITEX is also poor by inspectors based at the Plant Inspection Unit in due to the limited mobility of AGRITEX staff and the cargo area of Harare Airport (EC 2011).4 overall resource constraints. Resources provided by The central animal health institution is the GoZ to the public sector AIS are used mainly for sal- Depart­ ment of Livestock and Veterinary Services. aries versus research or extension operations. The main animal health issues involve straying Ideally, the public AIS would play a key role in animals, illegal movement of animals, spread of strengthening the resilience of small-scale producers, animal diseases, and poor veterinary care. because they have the least capacity to manage risk The European Union (EU) has audited Zimbabwe’s and adapt to changes in climate over time. Three pri- sanitary and phytosanitary (SPS) system and the orities for the public advisory services emerged dur- Food and Agriculture Organization (FAO) of the ing this assessment: (1) the need for staff to receive United Nations (UN) has supported capacity build- technical training in new skills, such as information ing in SPS services. Both the EU and FAO concur in and communication technology (ICT) and the use their general observation that the systems (regulatory of value chain approaches, among others; (2) the and relevant instruments) are in place for seed, plant, need to improve the mobility of extension staff so and animal health, but their implementation is limited that they can reach farmers in their area; and (3) the by the lack of human resource capacity and analytical need to digitalize services (for example, digital media infrastructure. can enable advisory services to reach more farmers Food safety in Zimbabwe is covered by the Food more effectively and can also provide early warn- and Food Safety Standards Act and the Public Health ing information more rapidly). Similar restrictions Act. Responsibility for assuring food safety is shared exist with regard to the public research system: (1) its between several institutions and departments, led by human resources and infrastructure are severely the Ministry of Health and Child Care. Other agen- limited, and (2) its staff need training in research cies include local authorities and several entities on climate-smart agriculture, inter-disciplinary in MAMID. For dairy, the Dairy Act delegates the research, socioeconomic and gender research, facili- authority to conduct food safety inspection to the tating innovation platforms, and other areas. Finally, Ministry of Health. Local authorities have by-laws a major challenge is to better integrate support for to ensure food safety. They collaborate with the key agricultural research, extension, and education. ministries (Codex Alimentarius, 2016).5 4.2.2  Sanitary And Phytosanitary System Assessing Disaster Risk 4.3  Management Capacity The Plant Quarantine Service (PQS) is the national plant protection organization for Zimbabwe, charged In addition to investing in public sector risk reduction with implementing official plant health controls. activities through the public AIS and SPS systems, the Other institutions involved with plant health and crop GoZ has a role in supporting communities affected protection operate under DR-SS: the Horticultural by disasters through various response and recovery Research Institute, the Central Service Research interventions. Legal and institutional frameworks Institute, the Seed Service, and the Plant Protection are crucial to provide timely assistance to respond to 24 Zimbabwe: Agriculture Sector Disaster Risk Assessment disasters. The following sections assess the disaster would normally be incorporated in a legislative response frameworks in place in Zimbabwe and the document of this kind. The review adds that the capacity to provide effective risk response support. 2011 bill as it stands is associated with significant costs, particularly the requirement that a minimum 4.3.1  Legal Framework of 1 percent of the national budget must be appro- priated to address disaster risk management. Due to Legislation approved in Zimbabwe during the last the financial constraints presently facing Zimbabwe, two decades defines the responsibilities and regu- the report suggests that this commitment may need lates the activities of the public sector in relation to to be amended to reflect a desired end-state, with disaster risk management (Table 6). Most notably flexibility to work toward the 1 percent financing goal the Civil Protection Act (1989) sets forth the current as finances allow. legal basis for organizing, coordinating, and planning Two additional pieces of legislation related to the response to natural disasters and emergencies the disaster risk management system are presented occurring in Zimbabwe. The act contains provisions in Table 7. The institutions established through this establishing the Department of Civil Protection additional legislation have some limitations that are (DCP) in the Ministry of Local Government, Public relevant to this disaster risk assessment, especially Works and National Housing and establishing the considering that weather-based disasters are the National Civil Protection Fund to finance civil pro- most common types of disaster in Zimbabwe. For tection activities in the event of a disaster. Given the example, although the Meteorological Services Act recognized shortcomings of this act—which provides (1990) sets up a meteorological services fund that only for civil protection and emergency manage- enables the Meteorological Department to fulfill its ment rather than for a more holistic approach that functions of forecasting and supporting the manage- includes disaster risk reduction, preparedness, and ment of weather-based disasters, the fund is set up risk financing—it is to be replaced by a Disaster Risk only to finance the services of the department and Management Bill (2011), still under discussion in does not serve as a contingency fund for emergencies. Parliament. Also, although the Grain Marketing Act (1996) pro- A recent review by the UN (CADRI 2017) reports vides guidelines to ensure that the GoZ has reserves of that although the Disaster Risk Management Bill is around 500,000 metric tons of grain to be used dur- not aligned with the Sendai Framework for Disaster ing emergencies, the Grain Marketing Board (GMB) Risk Reduction, the bill as it stands provides a rea- faces challenges in meeting this requirement owing to sonable level of detail and covers most aspects that the recent low grain production levels in Zimbabwe. Table 6.  Basic Legislation for Disaster Risk Management Legislation Key provisions Civil Protection Legislates for the coordination of preparedness planning for emergencies and disasters. Provisions Act 10.06 (1989) include the establishment of national, provincial, and district civil protection committees, made up by existing government, civil society, non-governmental, and United Nations organizations. The act establishes civil protection organizations and provides for the operation of civil protection services in times of disaster. The act explicitly states that the Head of State is the only individual who can declare a state of disaster in the country after receiving recommendations from the responsible minister. This arrangement may delay the declaration of a state of disaster and thus delay the mobilization of funds and other types of support. The Disaster Risk This bill, if approved by the legislature, will update and supersede the Civil Protection Act. It provides more Management elaborate mechanisms for disaster risk reduction and addresses structural and organiza­ tional gaps in the (DRM) Bill (2011) Civil Protection Act. It emphasizes localized decision-making in which local authorities take a leading role in disaster preparedness and response. The draft DRM Bill requires a minimum of 1 percent of the national budget to be appropriated to address disaster risk management (Government of Zimbabwe 2011). Additionally, the bill proposes a disaster risk management levy: “The Minister of Agriculture, in consultation with the Working Party and with the approval of the Minister for the time being responsible for Finance, may by notice in a statutory instrument, impose a Disaster Risk Management Levy on any person or class of persons whose activities are a potential hazard that include buildings, roads (tollgate fees), insurance, fire, fuel, carbon tax and tourism.” Capacity to Manage Agricultural Risk 25 Table 7.  Additional Legislation Supporting Disaster Risk Management Legislation related to disaster risk management Key provisions Meteorological Services Act logical Department This act establishes the administration, functions, and powers of the Meteoro­ (1990) in the Ministry of Environment, Water and Climate. The department’s functions include collecting and disseminating meteorological data, issuing weather and climate forecasts and advance warnings on weather conditions, and carrying out meteorological research and investigations. Grain Marketing Act (1966) This act establishes the Grain Marketing Board (GMB) under the Ministry of Agriculture and prescribes its powers, functions, and duties. The main responsibility of the GMB is to regulate and control prices and marketing of agricultural products and their derivatives that are sensitive to food security. The GMB also establishes and administers a trading reserve fund for each controlled product, and if at the close of the financial year a controlled product has a surplus, it is transferred to the reserve fund. The GMB maintains a strategic grain reserve. 4.3.2 Policy Framework for disaster risk financing, given that Zimbabwe is susceptible to climate-related disasters, particularly Table 8 summarizes the key policies and strategies droughts and floods. The NCCRS provides a com- that support disaster risk management and financing. prehensive and strategic approach to disaster risk The National Climate Policy and National Climate management through its first pillar, which focuses Change Response Strategy (NCCRS) provide an on Adaptation and Disaster Risk Management. The opportunity to integrate climate change and climate NCCRS proposes that $519 million be allocated for risk management policy (CADRI 2017). The National disaster risk management and human settlement; Climate Policy (2016) calls on government to establish potential sources of financing are the government, a National Climate Fund (NCF) that is supported development partners, private sector, and local, by an annual allocation from the national budget regional, and multilateral banks. Yet as shown in to finance the climate strategies and implement the Table 8, many of the policies and strategies make few Climate Policy. If established, the NCF could be used or no provisions to address disaster risk financing. Table 8.  Policy and Strategies Related to Disaster Risk Management and Financing Policy/strategy Key provisions National Climate Policy The policy seeks to reduce vulnerability to climate change and variability and strengthen adaptive (2016) capacity in key economic sectors such as health, water, agriculture, forestry, and biodiversity. It commits the government to ensure that mitigation and adaptation measures enhance agriculture- based livelihoods, by promoting food security and poverty alleviation. It also commits the government to strengthen the infrastructural capacity of the National Meteorological and Hydrological Services and Climate Change Management Departments to carry out research on climate change through improved data collection and management, and climate modeling. It looks at the establishment of a National Climate Fund, which is supported by a 10 percent budgetary allocation, to finance climate- related activities and programs. National Climate Change The NCCRS has specific provisions for dealing with climate change issues, understanding the extent Response Strategy of the threat, and putting in place specific actions to manage potential impacts. Financial resources (NCCRS) (2015) will be allocated by the national treasury, private sector, green climate funds, bilateral donors, and international agencies. Zimbabwe National This contingency plan, developed through a contingency planning workshop, presents a hazard profile of Contingency Plan the country and prioritizes the key hazards that are likely to require contingency measures. The purpose of the plan is to inform the disaster preparedness processes of the government and civil society. Disaster Risk This strategic disaster risk management plan looks at the institutional capacity of the National Civil Management Strategic Protection Committee to mitigate, prepare for, respond to, and help the country recover from disasters. Plan 2016–2020 It identifies financing gaps in performing these functions and opportunities for addressing them. National Policy and This policy recognizes the effects of drought on rural communities and encourages strategies that Programme for Drought aid communities in adapting to climate change. These strategies encompass early planting, choosing Mitigation drought-tolerant and early-maturing varieties, adopting water conservation measures, and cross- breeding and selling livestock. 26 Zimbabwe: Agriculture Sector Disaster Risk Assessment For example, the DCP Strategic Plan for 2015–2020 4.3.3 Institutional Framework and Coordination makes no mention of disaster risk financing or strat- egies for financing. The draft National Disaster Risk Figure 5 depicts the structure of Zimbabwe’s emer- Management Strategy of 2012 only mentions the gency management system. The DCP, the center for creation of a localized “disaster fund” to enhance coordinating all disaster management activities, works resilience to disasters at all administrative levels in through a national, multisectoral platform—the Civil the country. Protection Organization (CPO)—which provides In sum, even though the existing and proposed for the operation of civil protection services when policies and legislation regulating disaster risk man- disasters occur. The CPO platform is made up of civil agement in Zimbabwe mention potential sources protection committees with multisectoral represen- of funds for risk financing, the missing piece is the tation from government ministries and departments, design of a clear disaster risk financing strategy. This parastatals, donor partners, non-governmental orga- strategy should identify the sources of contingency nizations (NGOs), and co-opted members from the financing for emergencies, define various financing private sector (Ministry of Local Government 2009). layers and instruments to meet the requirements for The role of this platform is to provide advice and disasters with different levels of severity, establish the coordination related to national disaster risk efforts rules for triggering the use of each source of finance, as well as to make recommendations to the DCP on and determine the channels through which resources risk reduction. will reach beneficiaries in times of crisis—from early Institutional systems for disaster risk manage- warning systems and preparedness to the implemen- ment in the country currently comprise the follow- tation of recovery programs, including support for ing basic structures: the National Civil Protection rehabilitating small-scale production and restoring Committee, DCP, Food and Nutrition Council food security. Public financing sources to respond to (FNC), and Zimbabwe Vulnerability Assessment disasters will be reviewed in detail in the next chapter. Com­mittee (ZimVAC). Other participating structures Figure 5.  Structure of Zimbabwe’s Emergency Management System International Regional Support President's Office Community (SADC) (UN-bodies: INGOs) Ministry of Local Government Rural & National Civil Protection Urban Development Committee (DCP Director (chair); Secretary of Health; Commissioner of Police; Commanders of The Department of Civil various Defense Forces branches; Director Protection of Prisons; Secretary-General of the (DCP) Zimbabwe Red Cross Society; Director of Civil Aviation) Provincial Level Police Forces, Defense Dept. of Civil Protection Forces, and the Private Administrators & Officer Sector (At National; Provincial & District District Level Levels) Department of Civil Protection Administrators & Officer Source: Chikoto and Sadiq 2013. Capacity to Manage Agricultural Risk 27 are the Provincial Civil Protection Committee, rainfall and heat waves. The Meteorological Services District Civil Protection Committee, Emergency Department also produces medium-term forecasts to Services Subcommittee, National Food and Water develop weather advisories for agriculture, disaster Subcommittee, National Epidemics and Zoonotic risk management, and water resources. The effective­ Crisis Subcommittee, and National Resource Mobili­ ness of Zimbabwe’s Meteorological Services Depart­ zation Subcommittee (CADRI 2017). ment and its satellite offices at the district level for These Civil Protection Committees are grouped early warning is largely constrained by inadequate into four functional subcommittees: Food Supply funding, the failure to upgrade and manage equip- and Security; Health, Nutrition, and Welfare; Search, ment (especially at local stations), and staff turnover. Rescue, and Security; and International Cooperation In the agricultural sector, the Ministry of Lands, Assistance (Ministry of Local Government 2009). Agriculture, Climate, Water and Rural Resettlement The DCP and its substructures have been affected (MLACWRR) and the FNC are also key institutions by staff turnover and the government policy of not that play a role in disaster risk management and replacing departing staff, so they have insufficient financing. MLACWRR is the custodian of the over- resources to respond effectively to national disasters. all policy governing agricultural production in The DCP coordinates the National Civil Protection Zimbabwe. It is the arm of the GoZ mandated to Coordination Committee (NCPCC), which derives its provide technical, extension, advisory, regulatory, mandate from section (41) (2) of the Civil Protection and administrative services to the agricultural sector Act and is responsible for the execution of civil pro- to achieve food security and economic development. tection functions. The National Civil Protection The extension officers play a key role in preparing Plan provides the overall framework for promoting, communities for an impending disaster and advising coordinating, and executing emergency and disaster them in the event that it occurs. The FNC, under the management in Zimbabwe. It contains guidelines Office of the President and Cabinet (OPC), works for planning, executing, and preserving civil protec- closely with Meteorological Department and the tion systems and functions in Zimbabwe (Ministry MLACWRR to oversee drought management and of Local Government 2009). Decentralization of the response. The FNC chairs and utilizes ZimVAC as DCP structures to the district level and cross-sector an early warning tool for monitoring food security participation are positive traits for the NCPCC. at the household level in both rural and urban areas. The lean structure of the DCP makes it difficult ZimVAC assessments provide hazard information for the DCP to effectively coordinate disaster risk that helps to identify actual or impending external management in the country. The DCP does not have shocks that may affect livelihood systems. Hazard an emergency operations center that can be activated information details traditional early warning data in the event of a disaster. Moreover, CADRI (2017) such as weather, crop production, price and market has reported that the multisectoral stakeholder plat- form does not necessarily comprise individuals with information, and other shock indicators. This set of expert knowledge in disaster risk management, which information is essential for understanding the nature results in poor or slow decision making. In general, and magnitude of a climatic shock and the specific the disaster risk management system predominantly expression of this shock in a geographical setting at focuses on responding to natural disasters after they the provincial and district levels, which then enables occur and on financing emergency relief activities early action and disaster risk preparedness by institu- from the annual government budget and international tions such as the Civil Protection Unit. humanitarian assistance. The disaster response and preparedness of the 4.3.4 Disaster Financing Mechanisms DCP is informed by the Meteorological Services Department, which systematically observes and This section reviews the effectiveness, gaps, and monitors hydro-meteorological parameters for weaknesses of existing risk financing strategies. This hazard forecasts and near-real-time data for early review forms the basis for the recommendations warnings to support emergency preparedness and presented in the next chapter on how to strengthen response. The use of radar enables the department to the disaster response and risk financing systems release short-term (10-day) weather forecasts, which in Zimbabwe’s agricultural sector. This analysis is in make it possible to monitor extremes such as heavy line with the current approach adopted to strengthen 28 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table 9.  Allocations from the Ministry of Finance the limited budget allocated to DCP. The amount and Economic Development Contingency Fund to the allocated to the fund varies annually. Table 10 shows Department of Civil Protection for Disaster Operations the breakdown of disaster risk related programs Allocations from Treasury that were funded by the Treasury through the Civil Year to DCP (US$) Protection Fund from 2009 until 2016. 2012 1,000,000 The fund caters for all civil protection operations, 2013 Not available including disaster events, throughout the country’s 2014 450,000 2015 300,000 59 districts and 1,200 wards, and is generally con- 2016 300,000 sidered insufficient. The absence of reserved funds at 2017 354,000 the provincial and district level places a lot of pres- 2018 1,220,000 sure on the Civil Protection Fund and is a huge set- Source: DCP. back to efforts to reduce disaster risks in the country. Most local authorities are expected to react to disas- ters first, before requesting external assistance, but financial resilience developed by the Global Facility these institutions are incapacitated financially and in for Disaster Reduction and Recovery.6 turn rely on the DCP or donors and civil society for To finance disaster response, the Ministry of funding. Local authorities’ budgets are separate from Finance and Economic Development (MoFED) uses the central government budget and consist of local mostly ex-post financing options through budget revenue derived primarily from sales, fees, fines, per- reallocation and debt rescheduling arrangements and mits, property rates, and licenses. If local authorities ex-ante financing through budget reserves and con- estimate that their current budgets are insufficient, tingency funds. To channel resources to the affected the Ministry of Local Government, Public Works and population, the DCP requests funds from the MoFED National Housing issues “borrowing limits,” which contingency fund; the fund is replenished annually are calculated depending on the population of the with US$35 million (DCP 2018) for extraordinary specific district. When a disaster occurs at the urban budget requests, which could include disaster response level, the urban council is responsible for disaster funding. Table 9 shows the disaster fund allocations response. A review by the World Bank (2016) shows made by MoFED to the DCP from 2012 to 2018. that total debt across all local urban authorities in The DCP’s operational budget is very low and not Zimbabwe had reached $555 million, however, rep- sufficient to meet its operational mandate effectively. resenting 105 percent of all revenue collected by local In 2017, MoFED allocated an annual operational authorities.7 Consequently, local authorities that do budget of US$286,000 (representing 0.004 percent of not have sufficient revenue to finance disasters sub- the Ministry of Local Government’s allocated budget) mit requests for assistance to the DCP, which can (Government of Zimbabwe 2017c). make use of the Civil Protection Fund or request The National Civil Protection Fund, which is additional funds from MoFED. administered by the DCP and receives financing One of the challenges cited by the DCP in rela- from both MoFED and the public, is an important tion to accessing finance from MoFED is generally instrument for disaster risk response financing, given the lag between the time the ministry announces its Table 10.  Budget Allocations (In US$) from the Civil Protection Fund Food and Harmonized Food deficit Animal diseases Disbursed Nutrition Council cash transfers mitigation and risk management 2009 33,500 1,165,851 3,000 3,348,839 2010 171,000 1,418,692 4,250,000 4,164,087 2011 360,000 1,473,657 1,350,000 4,987,779 2012 510,000 350,000 230,000 2,568,232 2013 4,669,801 900,000 — 1,839,526 2014 240,000 370,000 — 572,506 2015 365,963 1,652,000 274,000 904,322 2016 767,000 3,762,000 3,784,273 685,444 Average 889,658 1,386,525 1,236,409 2,383,842 Source: DCP. Capacity to Manage Agricultural Risk 29 budget allocation to the department and the actual amount needed for the SGR. Table 11 shows maize time that DCP receives the funds. In some instances, imports and their value between 2015 and 2018. the total allocation may take as long as six months to Resources from the SGR are released only when arrive, because it is provided in installments, which the president declares a national disaster. Such a dec- limits the DCP’s ability to respond to a disaster in a laration is often made when the crisis has already timely, adequate manner. For example, the DCP indi- occurred, as in the 1991/92 and 2015/16 droughts. In cated that it could not respond quickly to the 2017 the case of the 1991/92 drought, the government had flooding in Tsholotsho District because the resources already sold its SGR and had to appeal for humani- for disaster management had not yet been allocated tarian aid, but by then most parts of the country were from MoFED. The DCP then used resources from already experiencing chronic food insecurity. The other lines in its budget, which were not enough to slow response to the early warning by decision makers respond the disaster. In line with the provisions of the affects the effectiveness of the SGR as a disaster risk Civil Protection Act, the private sector (in this case financing measure in the case of a drought. The DCP the Bankers Association of Zimbabwe), which was has noted that in terms of disaster risk financing, the approached for assistance, donated money toward SGR is housed inside the GMB, and no structures are disaster relief. Moreover, the weak financial capac- in place to differentiate it from the GMB. ity of the DCP curbs its ability to provide follow-up funds once the disaster has passed. Ex-Post Disaster Risk Financing from Donor 4.3.5  As indicated in Table 7, the Grain Marketing Act and Humanitarian Aid (1966) established the GMB as a parastatal with a commercial and social role. The Grain Marketing Act When a disaster exceeds the national capacity requires the GMB to hold a Strategic Grain Reserve to respond, national authorities request interna- (SGR) of 936,000 tons of maize and other grains to tional assistance. Humanitarian aid organizations, meet food security needs and act as a buffer during Zimbabwe’s development partners, and NGOs mobi- food emergencies, including critical periods when lize disaster response resources from different sources. drought can occur. Of that amount, 500,000 tons For example, humanitarian organizations can access must be held as physical stocks and the remain- funding for life-saving activities from pooled funds that ing 436,000 tons must be backed by the equivalent can disburse resources quickly but that have limited cash reserve (World Bank 2017). The GMB is man- resources. Key humanitarian and development part- dated to procure grain for the reserve from domestic ners involved in providing disaster response assistance sources and across the Southern Africa Development in Zimbabwe include the World Food Programme Community region. Zimbabwe requires 1.5 million of the UN, UNICEF, FAO, the Department for metric tons of cereal grain for human consumption to International Development of the United Kingdom, be self-sufficient (ZimVAC 2016). At the current pro- and the United States Agency for International curement price of US$390 per ton of maize, the GMB Development (USAID), including through major sup- has to hold grain stocks worth about US$195 million. port of the United Nations Development Programme’s The GMB can sell grain commercially, and during (UNDP’s) Zimbabwe Resilience Building Fund crisis the lean season it generally provides grain for govern- modifier mechanism, the EU, Swedish International ment food assistance programs in collaboration with Development Agency, Japan International Cooperation the Ministry of Social Welfare. As noted, however, Agency, and civil society organizations. Another the GMB faces challenges in meeting the targeted resource that has been used in Zimbabwe is the Central volume of grain reserves due to the recent low pro- Emergency Response Fund (CERF) under the UN. duction levels experienced in Zimbabwe. Structural constraints (such as limitations on farmers’ access to inputs, the lack of credit facilities, and challenges Table 11.  Maize Imports and Their Value, 2015–18 in accessing cash), aggravated by the impact of Year Metric tons Value (at US$390/t) droughts and floods, have contributed to the decline 2015–2016 49,689.91 19,379,064.51 in national production capacity (CADRI 2017). Two 2016–2017 447,229.63 174,414,554.14 sets of crop assessments conducted bi-annually are 2017–2018 128,971.00 50,298,691.17 used to project crop output and determine whether Total 625,890.538 244,097,309.82 grain must be imported or exported as well as the Source: MOFED (Dzenga and Nyaruwanga, personal communication). 30 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure 6.  Zimbabwe Humanitarian Response Plan, 2008–16 $700 $600 Total (US$m) $500 69% 63% $400 $300 47% 46% 87% $200 47% $100 52% 0 0 0 0 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Response plan/appeal funding Unmet requirements Source: OCHA Financial Tracking Service, https://fts.unocha.org/appeals/519/summary, accessed September 2018. The CERF tends to be one of the first ports of call for Notes small grants not exceeding US$250,000 for projects 1. Vulnerability, a common term in the poverty and implemented within six months or less after a disaster. food security literature, is defined as “the likelihood The CERF provides larger grants to UN agencies that that at a given time in the future, an individual will have are expected to implement them through national actors a level of welfare below some norm or benchmark.” (World Bank Group 2017). Common welfare indicators include poverty measurements, Currently, several partners provide international household expenditures, savings levels, and food security cooperation funding through various programs, and nutrition measures. Though vulnerability depends on the severity of external shocks like climate-related projects, and initiatives at the national and local levels. events, the likelihood of a drop in welfare depends both According to the Financial Tracking Service of the on people’s context and on their capacity to act and react. UN Office for the Coordination of Humanitarian Socioeconomic assets and institutions play an important Affairs (OCHA), in 2016 Zimbabwe received inter­ role in the extent of people’s vulnerability. national donor assistance of almost US$166.9 million 2. See UNDP (2017), UNOCHA (2016), and ZIMSTAT out of the total $352.3 million required that year. This (2015). 3. MAMID (2012). amount included funding for all disaster events and 4. See EC (2011), the final report of a November 2010 related sectors, such as water, sanitation, and hygiene; mission to evaluate Zimbabwe’s system of official controls food security and agriculture; and nutrition and health. and the certification of plants for export to the EU. Figure 6 shows the Zimbabwe Humanitarian 5. See also http://www.zimcodex.gov.zw/food-control- Response Plan from 2008 to 2016. The blue shading in-zimbabwe/. 6. GFDRR (2017). represents funding obtained from appeals; the yellow 7. For that review, debt was defined as the total shading represents the deficit that remained unmet (in amount the local authority owed to all creditors, including other words, the funding gap between the total appeal bank debt, payment arrears, and salary arrears, among and the amount mobilized). others. Chapter 5 TOWARD A RISK MANAGEMENT STRATEGY The findings of this assessment suggest that agricul- Mitigation strategies that strengthen producers’ tural risk is a principal cause of transient food inse- resilience by reducing risks that occur at a low sever- curity and disruption to agricultural supply chains. ity and high frequency at the farm level. In Layer 2, Agricultural risks deepen poverty by preventing rural where risky events occur at a medium frequency and entrepreneurs (particularly producers) from building medium level of severity, the government can use their assets, investing in developing their businesses, Risk Retention strategies such as a contingency fund and paying for healthcare and education. Agricultural or contingent credit to finance a response to these risks also exacerbate existing poverty traps in vulnera- events. Finally, Layer 3 is a Risk Transfer window for ble populations and lead to uneven growth in agricul- transferring risk to capital markets for risks that are ture and the economy as a whole. In rural Zimbabwe, very infrequent and cause high losses that the gov- crop failure induced by drought is the biggest shock ernment cannot limit. By undertaking comprehen- and may also be the biggest poverty trap. The increas- sive risk assessments for the main commodity supply ing prevalence of “shock-recovery-shock” cycles in chains and by adopting a holistic layered approach Zimbabwe, where the economy depends so heavily to identified risk, the government will be in a better on agriculture, is reducing the government’s ability position to start managing the main drivers of agri- to plan and pursue a sustainable development path. cultural volatility and food insecurity. Zimbabwe stands to benefit considerably from tran- The findings reported here emphasize that an sitioning toward a more proactive risk management integrated agricultural risk management strategy strategy. for the current context in Zimbabwe must pro- Agricultural risk management is a dynamic, chang- mote risk mitigation measures at the farm level— ing process that requires periodic assessments of risk, in other words, it must strengthen the capacity to of agricultural stakeholders’ risk management capac- reduce risk and improve resilience at the farm level. ity, of the strength of public and private institutions Investing in risk mitigation at the farm level will go involved in managing agricultural risk, and of fiscal a long way to reduce agricultural volatility, man- constraints. In this sense, all agricultural risk man- age food insecurity, and help smallholders adapt to agement is a context-specific endeavor. This chapter climate change. presents potential components of an integrated agri- Of equal importance for the government is the cultural risk management strategy that can serve as the need to make ex-ante provisions and strengthen its basis for GoZ to improve the mitigation of identified capacity to finance and deliver timely assistance when risks, promote risk transfer mechanisms, and/or make disasters of medium/low frequency and medium/high ex-ante financial provisions for coping with disasters. severity occur. This step is the key to avoiding future The conceptual framework for these actions is food insecurity, particularly in rural Zimbabwe. presented in Figure 7. The figure shows a “layering Innovative risk transfer programs such as agricul- approach” to risk management. In Layer 1, the gov- tural insurance at the sovereign and farmer level are ernment invests in adopting and implementing Risk being implemented in various sub-Saharan countries. 31 32 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure 7.  A Risk Layering Strategy for the Government of Zimbabwe PROBABILITY LAYER 3 Low frequency, High losses LAYER 2 Risk Mitigation Medium frequency, + Risk Retention Medium losses + Risk Transfer Risk Mitigation LAYER 1 + Risk Retention High frequency, Low losses Risk Mitigation SEVERITY These programs could provide valuable lessons on 5.1.1  Moving Toward an Enabling Policy Framework how to transfer agricultural risk to capital markets in Zimbabwe. This review identifies three areas in which the AIS The rest of this chapter provides more specific will benefit from leap-frogging: (1) shifting the para- suggestions for policy makers in Zimbabwe to start digm from conventional, high-input agriculture to identifying risk management strategies that are best knowledge-intensive sustainable intensification of aligned with national development policies, priori- agriculture; (2) leaping from uniform production ties, and fiscal constraints. patterns to more specialized production in a spatial development framework; and (3) pursuing the digita- lization of agriculture. If the AIS is to undertake these 5.1  Risk Mitigation strategic approaches, it will require policy support Zimbabwean agriculture is turning the corner toward and investments to strengthen the public and private higher growth and productivity, presenting a valu- institutions behind the AIS that support smallholder able opportunity to look forward and establish the farmers and bring about change. basis for a sustainable and resilient agricultural sec- tor that reflects the reality of the emerging produc- From Conventional Agriculture to Sustainable tive matrix, which mostly comprises large numbers Intensification of Agriculture of small-scale producers. At this juncture, rather A significant shift is needed to move away from the than opting to re-create the past or to adopt a piece- conventional agriculture paradigm to a climate-smart, meal approach, Zimbabwe has an opportunity to sustainable agriculture orientation that focuses on see where an evolutionary approach can be replaced production systems capable of supporting the ecologi- by leap-frogging—by jumping over a generation cal reconstruction of degraded areas and landscapes of technical and/or social development and into a through (for example) rehabilitating soil structure and new era. Zimbabwe can do this because it possesses fertility, reforestation, water resource management, plenty of human capital and knowledge. Many edu- sustainable water harvesting and optimized use in cated professionals have left the country during the agriculture, protected production in horticulture, and past decades, and economic difficulties have reduced the diversification of crop production (supplement- the strength of agricultural innovation institu- ing cereals with pulses and high-value crops) and crop tions, yet the core of those institutions and exper- rotations, to cite some of the options.1 It is important tise remains and is ready to engage and expand as to note that the context for this sustainable produc- resources become available. tion paradigm is Zimbabwe’s agricultural systems and Toward a Risk Management Strategy 33 landscapes. Disciplinary research (plant breeding, soil information, and for farmers to get better prices for science, agronomy, and so on) will retain a central their produce and conclude transactions. Beyond role, but its goals and research questions will emerge the farmer level, all agricultural institutions will from the sustainable production paradigm. benefit greatly from using ICTs to share informa- tion, ease communication, increase the effectiveness From National Production Goals to Regional of administration, and so on. Specialization Maize is the staple crop in the food system, and An Enabling Policy, Regulatory, and Incentivizing the current policy is to extend maize production Environment throughout the country. For that reason, maize For all of three of the “leaps,” an enabling policy, production is being promoted to some extent even regulatory, and incentivizing environment is essen- where the climatic risks are too high to justify its tial. The AIS also needs to engage creatively to “oil” production. Zimbabwe could benefit from revisit- the machine for these transformative changes. One ing the idea of regional and agro-ecological special- option that could be considered is a strategic initia- ization, and from identifying the geographical areas tive to facilitate south-south learning and innovation, that are most suitable for certain crop or livestock which could be resourced by a competitive grant systems—for example, pasture and livestock pro- facility to support consortia ready to contribute to duction areas, horticultural crop production areas, leap-frogging. maize production systems, and so on—with their The most important priority in the short to associated market clusters and supply chains, with medium term is the following public investments the relevant stakeholders (traders, regulators, proces- to strengthen the resilience of smallholder farm- sors, financial agents, and others). A first step could ing systems: strengthening the AIS and SPS system be to revisit the agro-ecological zones and suitability in public-private partnerships, irrigation develop- of different crops in these zones and identify areas ment, and improving the agricultural early warning of high dynamism for further cluster development. systems. These services need strong support from Linkages between rural towns and adjacent rural and the government, as the vast majority of smallholder peri-urban areas could be developed jointly in a spa- farmers have limited capacity to manage agricultural tial development framework. risk, and they need interventions to provide the pub- lic goods that can accompany them in leap-frogging Digitalization of Agriculture toward a more sustainable and resilient agriculture Massive expansion of 3G mobile broadband services as the country moves forward. across the country—mobile broadband penetration The reality, however, is that resources to strengthen increased more than seven-fold within four years risk mitigation measures among small-scale produc- and by 2017 had exceeded 100 percent—means that ers still show a downward trend across all categories half of the population now has access to the inter- of spending (Table 12). Once this trend is reversed, net (POTRAZ 2017). Most agricultural researchers several suggestions for the priority services that need and extension workers have access to mobile phones strengthening can be pursued, as discussed in the and internet, but they are only just beginning to sections that follow. use these digital tools for sharing information and managing knowledge (Mugwisi et al. 2015). The tre- mendous growth in mobile and internet access in Table 12.  Annual Budget for Agricultural Services Zimbabwe provides the foundation to move toward Budget (US$ 000s) e-agriculture across subsectors and services, includ- Sector 2014 2015 2016 ing extension, GIS for land management, agro- Meteorological Services 3,838 2,926 2,330 weather services, governance in institutions, market Water Resources Management and Development 79,227 40,439 28,564 information, communication among producers and Agricultural and Extension Services 36,183 35,110 19,599 trade organizations, and livestock traceability, to Irrigation and Development 15,218 10,703 7,147 Livestock Production and Development Division 6,697 5,082 5,558 mention a few (World Bank 2017). Embracing the Veterinary Services Division 22,109 19,008 12,850 ICT opportunities in agriculture can make a huge Agricultural Engineering and Mechanization 4,760 5,285 3,531 difference in reaching the large number of small- Civil Protection Unit 550 350 360 holders with technical support and agro-weather Source: Government of Zimbabwe, 2014–2016 National Budget Statement. 34 Zimbabwe: Agriculture Sector Disaster Risk Assessment Strengthening the Agricultural 5.1.2  extension, agribusiness, and so on). To think through Innovation System the associated issues, it may be helpful to consider the typology of farmers developed by Berdegué and After the land reform, Zimbabwe has 300 large-scale Escobar (2002).3 The typology is based on producers’ farmers (farms averaging 2,200 hectares or more); asset ownership (high or low) and production envi- 16,386 resettled A2 farmers (∼318 hectares); 145,775 ronment (unfavorable or favorable). It encompasses resettled A1 farmers (∼37 hectares); 76,000 old- three major types of small farming communities, resettled farmers (∼46 hectares); and about 1,300,000 which would benefit from different types of policy communal farmers (∼4 hectares). Such a dramatic and innovation support (Figure 8): change in the structure of the farming sector continues to pose an enormous challenge for the AIS. Until the 1. Where assets favor the development of competi- late 1990s, DR-SS and AGRITEX focused on provid- tive agriculture, particular emphasis should be ing services for large-scale farmers. The small-scale given to commercial initiatives and private sector farmers in the communal lands were largely left to contributions. their own means and innovations. Since the land 2. Where farmers have the potential to embark on reform, this dichotomy has disappeared, and the market-oriented agriculture but are constrained research and extension system must now serve close by their asset base, public (and private) efforts to 2 million small-scale communal, peri-urban, and should aim to provide resources and experience to resettlement farmers, most of whom depend on develop a vibrant small-farm sector. rainfed agriculture.2 3. Where rural households lack many of the assets that might allow them to profit from commer- Different Resource and Services cial agriculture, more broad-based rural poverty for Different Subgroups reduction policies must be pursued, often in col- Subgroups within these 2 million smallholders will laboration with local organizations, the UN sys- require different types of resources and services from tem, and NGOs that can facilitate building link- the key innovation system stakeholders (research, ages and institutions. Figure 8.  Grouping Farmers to Develop Different Strategies for Agricultural Innovation Figure 1 Differential strategies for the development of agriculture knowledge and information systems HIGH ASSET POSITION A UNFAVOURABLE PRODUCTION FAVOURABLE PRODUCTION ENVIRONMENTS ENVIRONMENT B C LOW ASSET POSITION Source: Berdegué and Escobar 2002. Toward a Risk Management Strategy 35 The typology is presented merely to illustrate one likely that cooperation will emerge, although way of grouping small-scale farmers to start devel- consistent leadership and a clear strategy for oping differentiated agricultural development and cooperation will be important. Competitive innovation strategies. grant systems designed for inter-disciplinary and multistakeholder cooperation will further What is Needed for the AIS To Leap-Frog? incentivize cooperation. The AIS in Zimbabwe has an institutional basis: DR-SS provides public research goods, AGRITEX Given the core human capacity in the country provides public extension services, and Agricultural and limited resources, every opportunity for regional Research Council of the Research Council of and international R&D interaction should be seized. Zimbabwe facilitates the identification of national The World Bank–financed Agricultural Productivity priorities, conducts stakeholder consultations, and Program for Southern Africa could be one such can manage competitive grant funding systems. opportunity. These public R&D services are supplemented by A final point is that for leap-frogging to be suc- research in academia and private companies, espe- cessful, the AIS should more clearly reflect the wider cially crop breeding by seed companies. Regional and perspective on innovation systems and practices that international agricultural research institutions com- characterizes the AIS approach. An AIS approach plement and collaborate with the national AIS. All of looks at the multiple conditions and relationships the public institutions have a core of capable human that promote innovation in agriculture. Compared resources. But the main constraint of the AIS is its to traditional, linear agricultural research and exten- inadequate fit to agriculture as currently practiced sion efforts, the AIS approach may offer a more flex- in Zimbabwe, its severely limited resources, and the ible means of dealing with the varied conditions and lack of cooperation across the AIS. contexts in which innovation must occur. It consid- ers the diverse actors involved, their potential inter- •• Both DR-SS and AGRITEX, the key public sector actions, the role of informal practices in promoting agricultural R&D entities, would benefit from a innovation, and the agricultural policy context. fresh look at their focus and strategies in service The AIS principles of analysis and action inte- of the 2 million small-scale farmers. This process grate the more traditional interventions (support for must recognize the different groups among these research, extension, and education and the creation of farmers (defined according to their asset base and links among research, extension, and farmers) with the environmental production conditions) and their other complementary interventions needed for inno- different requirements for R&D support, as sug- vation to take place. Such interventions include pro- gested in the exercise conducted by Berdegué and viding the professional skills, incentives, and resources Escobar (2002), referenced earlier. to develop partnerships and businesses, improving •• While the core human resources are present in knowledge flows, and ensuring that the conditions Zimbabwe’s public research, extension, and devel- that enable actors to innovate are in place.4 opment institutions, they remain severely limited, along with resources for operations and capital Improving Sanitary and Phytosanitary 5.1.3  investments for infrastructure. These institutions Measures need to increase the number of researchers and extension agents, but they also need to train and Aside from the weather-related risks, fieldwork re-train staff in the modes of operation and skills for this risk assessment identified important pest required today (climate-smart agriculture, ICTs, and disease risks for plants and animals, which inter-disciplinary research, socioeconomic and were included in the risk prioritization assessment gender research, facilitating innovation platforms, in Chapter 4. Although this assessment did not and so forth). undertake a separate assessment of the SPS system •• The resource constraints of prior decades have in Zimbabwe, this section advances some potential left few opportunities and incentives for inter- measures to strengthen the SPS system, which can agency cooperation and in some cases have serve as the basis for prioritizing further research and led to a coping strategy of isolationist behav- investments designed to strengthen the public insti- ior. Once resources become more available, it is tutions and universities that address these issues. 36 Zimbabwe: Agriculture Sector Disaster Risk Assessment Phytosanitary Systems The bill calls for the establishment of a Food Control Facilities at the Mazowe Plant Quarantine Station Authority of Zimbabwe with a comprehensive man- include entomology, nematology, and phytopathol- date to ensure food safety. The emphasis in recent ogy laboratories and the quarantine facilities for years on food production, along with limitations in imported plant materials, including a small quaran- public resources, have delayed upgrading of the food tine glasshouse. Detection of harmful organisms is control system. based mainly on visual examination. The equipment available for bacteriology, mycology, and virology is 5.1.4  Developing Irrigation fairly basic, and there are no regular evaluations of the standards of laboratory work (EC 2011). Given that drought is the main driver of agricultural Participants in the national stakeholder workshop risk in Zimbabwe, and given the projected impact of stressed the need to strengthen the capacity of DR-SS a drier future climate, irrigation development is vital and AGRITEX to develop and disseminate pest- and for building resilience. Limited availability of water disease-tolerant crop varieties and animal breeds and is a key constraint for small-scale producers. Much to test the efficacy of indigenous knowledge systems irrigation infrastructure is in poor condition, so and medicinal native plants in controlling diseases farmers in irrigated cropping systems are producing and pests. In addition, collaboration between pub- below the yield potential. Key priorities for irrigation lic and private agricultural research, education, and are to rehabilitate existing infrastructure; adopt more extension institutions, which are currently working modern, efficient technologies; and expand into under- in isolation, should be strengthened. irrigated areas with good production potential. The EU has audited the phytosanitary system (EC 2011), and FAO has supported capacity build- 5.1.5  Improving Early Warning Systems ing in this area. The main recommendations ema- nating from the EC (2011) report are: (1) review the The framework established under the UN International regulatory framework; (2) assess the capacity for Strategy for Disaster Reduction and best practices on implementation (institutional mandates, expertise, early warning systems suggest that four elements are infrastructure, operational funds); and (3) phase important: the collection and assessment of data on investments to improve the system. Participants in risks, establishment of monitoring and warning ser- the stakeholder workshop for this assessment noted vices, communication, and the capacity to respond to the need for the Plant Protection Unit and the a risk or hazard. Department of Livestock and Veterinary Services to Zimbabwe’s long history of drought and climate improve public biosecurity measures within the coun- vulnerability has led to the progressive establish- try and at international borders to prevent the trans- ment of more effective early warning systems, yet a mission of pests and diseases (such as foot and mouth number of opportunities exist to strengthen these disease, full armyworm, and Tuta absoluta) from systems, improve the coordination of early warning endemic areas into disease- and pest-free zones. They efforts, and use increasingly sophisticated technology.5 also emphasized the need to make the required pub- Zimbabwe’s well-established hazard and monitor- lic sector investments to produce animal vaccines (for ing systems are supported by institutional structures Newcastle disease, anthrax, rabies, foot and mouth, at all administrative levels, including the DCP and and tick-borne diseases). Meteorological Services Department, which are the key institutions for disaster risk and preparedness. Even so, Food Safety there is a need to strengthen cross-institutional coor- Pswarai et al. (2014), in a review of the food safety dination for early warning activities that are comple- system in Zimbabwe, find it to be fragmented, with mentary or require improved coordination across no clear mechanisms for coordinating the activi- institutions. Investments that develop early warning ties of the entities involved. This fragmentation and infrastructure are also needed for Zimbabwe to take lack of coordination make it difficult to ensure a advantage of new technology and data sources, which safe national food supply. The main reason for these will be particularly important for rapidly triggering problems is the lack of resources allocated to the financial preparedness and funding responses. food control system, which had been noted previ- It would also be interesting for Zimbabwe to ously and led to the draft Food Control Bill (2011). develop agroclimatic zoning that can help stakeholders Toward a Risk Management Strategy 37 make decisions based on historical probability esti- access to source data by all stakeholders and more mates, which can help to forecast the level of risks for robust analysis. specific crops in specific areas. These efforts could be •• Contingency planning should be expanded in part of the proposed agro-ecological rezoning men- response to increasingly regular ENSO events. tioned earlier. •• Early warning products and findings should be In addition, early warning systems should be more accessible to end users and context specific, strengthened to increase the dissemination of more particularly for women farmers. location-specific weather and seasonal forecasts to •• Institutional coordination should be strength- rural communities through digital platforms and ened, both nationally (across the wide range of the government agricultural extension service. While departments and organizations involved in early mobile penetration rates are high in Zimbabwe, warning data collection) and regionally. some remote areas such as Umguza, Siyakova, and •• Technical capacity building and training is needed Umzingwane have poor or no mobile network cover- at the national and local levels to upgrade skills age, making dissemination of early warning or market and fill technical staffing gaps. information products by SMS or social media plat- •• Data collection equipment and technology should forms ineffective. Box 1 summarizes a recent diagnosis be upgraded, with greater utilization of digital ICT. of the early warning system (World Bank 2018). •• Consolidate the policy framework across climate, The following 10 key recommendations for disaster risk management, and early warning for strengthening Zimbabwe’s early warning system are consistency, and move forward with the draft designed to build a stronger system that can manage DRM Bill. the effects of a likely increase in climate variability: •• Expand the inclusion of market information into early warning systems for better decision-making. •• The early warning system should incorporate •• Explore options for a sustainable financing mech- a wider range of data and analysis. For example, anism. The ultimate effectiveness of Zimbabwe’s it should expand data collection to enable better early warning system is largely constrained by baseline comparisons, assess vulnerability in urban inadequate funding and increasingly obsolete areas, include more market information data, and equipment. support other sectors that are also impacted by cli- mate variability (health, infrastructure). As agricultural risk management needs intense •• Early warning data and information systems inter-institutional coordination, Zimbabwe could should be integrated into one platform and an consider developing a Technical Support Unit (TSU) open data approach adopted, to allow greater to coordinate issues related to agricultural risk Box 1.  Salient Characteristics and Diagnosis of the Early Warning System in Zimbabwe • Drought monitoring and warnings are akin to monitoring the  Early warning products tend to be available in English and impacts of drought rather than climate forecasting. Shona and are not translated into all 16 official languages; • Drought policy has a disaster rather than a drought mitigation as a result, the information cannot be read, heard, and orientation. understood by everyone in all regions. • Zimbabwe has quick response mechanisms for long-term  The spatial disadvantage of remote areas with poor access shocks such as drought that are highly effective, while it fails to mobile networks makes dissemination of early warning to reproduce the same mechanism for short-term shocks like products by SMS or social media platforms ineffective. floods or pest and disease epidemics. • A silo mentality and lack of information sharing prevail among • The limitation in long-range forecasting constrains the poten- government departments, ministries, NGOS, and academia. tial to transform early warning systems into action that guides • Funding is released only when the president declares a state of drought/flood preparedness. disaster—often after a crisis has already occurred. A disaster • Poor utilization of early warning products: risk/early warning systems fund is needed to support mitiga-  Early warning products are released to the public in a tion, preparedness, response, and recovery. generic form that is not specific to the need to trigger a response (in the form of a particular action) among users. Source: World Bank 2018b. 38 Zimbabwe: Agriculture Sector Disaster Risk Assessment management. The TSU could be housed in the Ministry index products can be relevant for Zimbabwe. Because of Agriculture or in the Disaster Department. The the discussion in this assessment emphasizes particu- role of the TSU would be to coordinate risk manage- lar types of crop insurance products corresponding ment in agriculture, to perform the related research to a specific set of risks, a broad discussion of agri- (risk assessments, maps, support information, data- cultural insurance (including livestock insurance) is bases, and so on), develop the training programs, and not undertaken here.6 Instead, the following sections develop proposals for high-level decision-makers, review the differences between a subset of traditional including active participation in the early warning sys- and non-traditional crop insurance products. tem. The TSU could be guided by a high-level steer- Smallholder farmers’ uptake of agricultural insur- ing committee (constituted by representatives of the ance has been historically very low in Zimbabwe, Ministry of Agriculture, MoFED, disaster risk man- and at the same time Zimbabwe has seen limited agement office, planning office, and so on) that will be investment toward the provision of market-based responsible for defining the strategy. The TSU should insurance services for this sector. Smallholders be supported by a technical committee comprising often suffer the full impact of poor weather on their representatives of public and private agencies, such crops because, unlike commercial farmers, they are as insurance companies, universities, extension agen- unlikely to take out insurance to transfer the risk, cies, research institutions, the Meteorological Services opting instead to rely mainly on the limited risk miti- Department, and other concerned institutions. gation strategies they can implement themselves at the farm level. With the recent and continuous pilot- ing and implementation of index insurance projects 5.2  Agricultural Insurance in various countries in sub-Saharan Africa, however, As mentioned at the beginning of this chapter, agri- some opportunities may be emerging for Zimbabwe. cultural insurance is one potential risk transfer tool Weather index insurance (WII) is an innovation that farmers and other stakeholders can use to man- with advantages beyond traditional agricultural insur- age risks that cannot be mitigated at the farm level. ance, which bases indemnity payments on verifiable Insurance instruments transfer part of that risk to losses. Index insurance pays out benefits based on a another party in return for a fee (or premium). Where predetermined index—such as the average yields in it is available and affordable, agricultural insurance an area, rainfall, temperature, or a normalized dif- (for crops and/or livestock) can greatly benefit farm ference vegetation index (NDVI)—that is corre- households: lated with actual loss of assets and investments. An indemnity is paid whenever the realized value of the •• Insurance can (and should) be used to comple- index exceeds or falls short of a previously specified ment other risk management approaches. Farmers threshold, without requiring the traditional services can rely on informal household- and community- of insurance claims assessors. Because field loss assess- level strategies such as crop and labor diversifica- ments are not needed, administrative costs can be tion to manage small to moderate risks. In the drastically reduced. Index insurance can therefore be event of a major weather shock, insurance can be sold at lower prices and pay out claims more rapidly. designed to protect against revenue or consump- Index insurance also reduces moral hazard7 and has tion losses, enabling households to avoid selling low adverse selection,8 given that an objective trigger is livelihood assets or drawing on savings. used for claim payments. This combination of factors •• Insurance can assist farmers in accessing new improves the potential for the insurance product to be opportunities by improving their ability to either commercially sustainable. borrow money or in-kind credits. In doing so, farm Four innovative models of WII have been imple- households may potentially experience higher mented recently in sub-Saharan Africa, including returns. one type piloted in Zimbabwe by the World Food Programme under its R4 Rural Resilience Programme. Crop and livestock insurance products are widely Box 2 describes these efforts and the lessons they used in high-income countries. Markets are large, and present. These examples (among many others) indi- there is long experience in finding ways to insure agri- cate the potential for innovation in this field, and it is culture with traditional insurance products. Recent worthwhile to examine them more closely and evalu- experiences in insuring smallholders with innovative ate their feasibility for Zimbabwe. Implementing such Toward a Risk Management Strategy 39 Box 2. Recent Innovations in Weather Index Insurance in Africa The World Food Programme’s R4 Rural Resilience Initiative an asset protection approach and covers financial losses incurred (Zimbabwe). This initiative supports the provision of weather when animals become stressed by drought. Satellite observations index insurance (WII) to 500 beneficiaries in Chebvute, Masvingo are used to derive a normalized difference vegetation index (NDVI) District. Coverage begins only after enough rain has fallen to that proxies pasture availability. A model is used to correlate the permit a farmer to plant. If this “effective rainfall” is not experienced NDVI with livestock mortality or animal stress occurring when for- by December 5, the policy automatically starts on December 6. age is unavailable. The product acts as an early detection mecha- Farmers will be covered for the following 85 days, encompassing nism that triggers payouts before animals die, and it should help the germination, vegetative, and flowering periods of the growing herders to keep their animals alive. Pastoralists pay the risk pre- season. Rainfall is monitored using satellite information specific mium, which is set at the actuarially fair price, and donors pay the to the geo-location of each farmer’s village, and monitoring is sup- subsidy premium of 40 percent, which allows the insurance com- ported by manual rain gauges. If a rainfall deficit is experienced, pany to cover costs and earn a profit. Commercial premium rates payouts are triggered automatically. The sum insured is based on the are around 8 percent of the insured amount. The program started value of the agricultural inputs. If triggered during the germination in 2010 and is now available in several regions of northern Kenya. period, payouts will cover 20 percent of the sum insured; a payout of A number of technical lessons have been learned through this 40 percent of the sum insured will be disbursed during the vegeta- experience. For example, it is necessary to calibrate the index to tive period; and 60 percent will be triggered during the flowering specify precisely whether a particular level of greenness reflects phase. A farmer can receive only one payout per policy. vegetation that is edible or palatable to cattle, by adapting the filters used to interpret the satellite observations in any particu- Seed protection (Rwanda, Tanzania, Kenya). This product lar context. The product evolved into an asset protection product covers the loss of a bag of seed; payouts occur when there is not because the historic data on mortality required to trigger payouts enough rain for seed to germinate during a 21-day planting window was not available everywhere, and because herders and insurers for a given location. The product aims to enable farmers to afford were not so interested in the initial mortality contract. a second purchase of seed and to replant during the same season. In 2015, with technical support from the World Bank, the Govern- The seed company introduces a voucher (a scratch card with a ment of Kenya started buying the product to protect 5,000 vulnerable code) in each 2-pound bag of seed. The farmer needs to dial a pre- households from drought as another component of the government determined number of the mobile operator to register and enter social safety net. the voucher code. When the farmer activates the code, the mobile network picks up the location where the phone call is made, and Insuring input loans in integrated supply chains (Zambia). the farmer is insured as a beneficiary of the policy for that location. This WII product is designed to protect the value of inputs obtained Technical support is provided by the Agriculture and Climate Risk by credit extended to cotton farmers. It uses satellite weather data Enterprise Ltd to the insurer for designing and pricing the contracts estimates to price the risk and for monitoring the contract, protecting for maize at the different locations using satellite weather data farmers through the different phases of the crop cycle. It measures (past and current), with the assistance of the International Research cumulative rainfall over 20 days for drought-prone periods during the Institute of Climate and Society at Columbia University. crop cycle to trigger payments when precipitation is less than an The seed company pays a percentage of the insurance cost and agreed percentage of normal. The index also measures cumulative donors pay the remainder based on a previously agreed subsidy rainfall over any 10 consecutive days for some periods of the crop model. Whereas the insurance pricing is done by location, the seed cycle with risk of excess precipitation and triggers payments when company and donors pay upfront an aggregated risk premium (of precipitation is some agreed percentage over normal rainfall levels. around 10 percent) to the insurer. The compensation is sent to The product is promoted by MUSIKA (an NGO) with the participa- the farmer via mobile money. The farmer can then purchase seed, tion of the input supplier NWK and the Mayfair and Focus insurance replant, and potentially harvest in the same season, or obtain companies as underwriters. Farmers are charged an unsubsidized cash in compensation. In participating in this project, Seed Co premium rate of around 8 percent, advanced by NWK at the begin- has incurred the added costs of repackaging seed to include the ning of the season as part of the input credit. Farmers’ enrolment in insurance voucher and paying their share of the premium. These this insurance scheme is voluntary. NWK advances the premiums costs have been justified in terms of market differentiation, secur- upfront to the insurer at the beginning of the season and recovers ing customer loyalty, and increasing their market share in Kenya those premiums from the ginneries that receive and process the (5 percent) by promoting their drought-resistant seed. seed cotton farmers deliver at the end of the season. The strategy is to identify an insurance distributor that has Livestock insurance as social protection (Kenya). Developed wide outreach to farmers, and NWK presented the opportunity, by the International Livestock Research Institute and sold to individ- as all cotton farmers use improved seed and fertilizers. NWK has ual pastoralist households in various districts of Kenya at subsidized positioned itself in the market as a supplier of improved seed with levels (around 40 percent of commercial premiums), this livestock a market share of around 25 percent, and it saw WII as an oppor- insurance product has evolved rapidly. Originally covering mortal- tunity to protect its credit and retain clients. This type of insurance ity levels following catastrophic events, the product now reflects program could work in integrated supply chains. 40 Zimbabwe: Agriculture Sector Disaster Risk Assessment models would require the establishment of public- As noted, the costs of these losses have been met private partnerships, funding from donors, and the by limited domestic fiscal resources and interna- support and participation of universities, specialized tional humanitarian assistance. The failure to use consultants, Zimbabwe weather and information sys- appropriate risk financing instruments to transfer tems, and potentially a risk pool of local insurers to risk to international markets has left wide financial underwrite the contracts. gaps to fund disaster costs. While many of the risk mitigation strategies and risk transfer products discussed here can help house- 5.3.1  Financial Gap and Sources of Funds holds cope with the impact of low and moderate weather risks, they need to be complemented by gov- In the absence of consistent annual evaluations of ernment support when larger, more severe weather the total costs of natural disasters in Zimbabwe, the shocks or natural disasters occur. In such cases, most figure used for this analysis is taken from the total countries trigger disaster emergency programs that pledge requirements that the GoZ has submitted include providing support to the poorest households to UN organizations and other donors since 2008. to ensure food security, assisting small-scale produc- These data are used as a proxy to estimate the financ- ers to return to productive activities, and rehabilitat- ing requirements for disaster risk management on a ing infrastructure. yearly basis (Table 13).9 These figures include expenses for direct emer- gency support as well as programs related to humani- 5.3  Sovereign Disaster Risk Management tarian crises. The total yearly average requirement (2008–16) for disaster risk management has been Investing in risk mitigation at the farm level can around US$333.3 million. Such requirements have go a long way to strengthen the resilience of agri- been met with MoFED allocations to disaster risk man- cultural systems. Risk mitigation at the farm level is agement programs and to the Civil Protection Fund not enough, however, considering that Zimbabwean for an estimated annual average of US$5.2 million (or agriculture remains highly vulnerable to severe nat- 1.6 percent of the average annual requirements). ural disasters. Given the high costs of these events The largest source of disaster financing has come in terms of losses in agricultural output, reduced from humanitarian sources, including UN and non- incomes, and negative effects on food security, the UN agencies, with an estimated annual average of country also needs to have the institutions, pro- US$246.8 million (74.1 percent of total requirements). cesses, and financing in place to reach affected These allocations from MoFED and humanitarian households in a timely way through ready-made responses still do not cover the total requirements. programs that help them recover from shocks and They leave an estimated annual average funding gap avoid cyclical poverty traps, especially the most of US$81.1 million (24.3 percent of average annual vulnerable households. In other words, the GoZ requirements). requires the capacity to respond rapidly through Disaster risk financing in Zimbabwe has relied ready access to a package of various sources of on humanitarian funds owing to the severe fiscal finance that can be used for meeting the costs of constraints of the government balance sheet, and different levels of impacts. this situation is likely to continue until the economy Zimbabwe is susceptible to extreme weather revives. MoFED earmarks around US$35 million a events such as droughts, heat waves, heavy rains, year for contingencies. In times of tight budgeting, flash floods, strong winds, and hailstorms. The that fund is meant to cover many other contingen- country has experienced a total of 18 ENSO events cies arising from all sectors of the economy and not since 1951, and historical records show that 62 per- only those arising from natural disasters. In practice, cent of ENSO episodes since 1970 have been char- MoFED makes relatively small allocations on a yearly acterized by low and erratic rainfall (UNDP 2017). basis to devote to immediate rescue and emergency Six of these recorded ENSO events have been cat- operations following disasters, recognizing that egorized as either strong or severe. Combined with humanitarian assistance takes time to approve and limited adaptive capacities, these events have caused disburse. Yet even if MoFED were to allocate all of its food insecurity to peak every four to five years in contingency fund to natural disasters and emergen- Zimbabwe and across southern Africa, with each epi- cies, it would not cover the estimated average annual sode lasting 9–18 months. gap. There are no financial margins to cushion the Toward a Risk Management Strategy 41 Table 13.  Estimated Financing Gap in Disaster Risk Management in Zimbabwe Humanitarian Response Total GOZ requirements allocatlions to GOZ allocatiions Inside UN Outside UN for DR financing DRM programs to civil protection response/ response/ Year (US$) (US$) fund (US$) appeals (US$) appeals (US$) Financial gap * ** ** *** **** ***** 2008 583,447,922 — — 400,468,563 72,079,089 110,900,270 2009 722,198,333 4,551,190 131,848 456,361,623 185,877,162 75,276,510 2010 478,399,290 10,003,779 172,976 226,189,188 90,042,159 151,991,188 2011 478,582,358 8,171,436 236,000 221,723,553 9,086,250 239,365,119 2012 238,444,169 3,658,232 439,328 206,902,892 27,395,486 48,231 2013 146,971,839 7,409,327 113,600 76,494,116 18,015,731 44,939,065 2014 — 1,182,50 241,080 — — (1,182,506) 2015 — 3,196,285 176,000 — — (3,196,285) 2016 352,318,995 8,998,717 110,000 166,855,915 64,434,960 111,919,403 Average 333,373,656 5,241,275 180,092 194,999,539 51,881,204 81,117,888 % 100.0% 1.6% 0.1% 58.5% 15.6% 24.3% Sources: * UN OCHA Financial Tracking Service (FT5) as Sep 2018 ** Department of Civil Protection (2018) *** Response from the UN system **** Response outside the UN system ***** Estimated financial gap effects of agricultural risk and natural disasters, and and proactive financial risk management strategy. In the country absorbs the shocks without transferring other words, not only does Zimbabwe need to invest any of the risk to international markets. in risk mitigation at the farm level, but—as circum- In this context, a key challenges for Zimbabwe is stances allow and the country enjoys more financial that apart from the contingency funds and humani- flexibility—it should start planning to improve its tarian assistance, it has limited financial instruments financial preparedness to manage risk and respond to facilitate rapid access to significant sources of to disasters. Like many other countries, Zimbabwe financing when a disaster happens. Zimbabwe needs could start by implementing a risk layering approach to begin transitioning away from its current reactive that combines the use of different financial instru- strategy of managing agricultural risks and natural ments depending on the frequency and intensity disasters after the fact and moving toward a planned of the risks the country faces. Figure 9 illustrates Figure 9.  Zimbabwe: Financial Risk Layering to Respond to Natural Disasters HAZARD FINANCING TYPE INSTRUMENT Risk Transfer Risk transfer for assets such as property insurance or Low Frequency/ Market-Based High Severity International Assistance (uncertain) Instruments agricultural insurance and risk transfer for budget management like parametric insurance, cat bonds/swaps Contingent Financial instruments that provide liquidity immediately after Contingent a shock Credit Financing High Frequency/ Low Severity Budget Reserve funds specifically designated for financing disaster Budgetary Reserves/ related expenditures, general contingency budgets, or Instruments Reallocations diverted spending from other programs Source: Disaster Risk Financing and Insurance Program, World Bank Group. 42 Zimbabwe: Agriculture Sector Disaster Risk Assessment layered risk financing strategies that Zimbabwe could insurance contracts, and catastrophe bonds—can be identify and implement as circumstances allow. used to transfer the risk of specific meteorological or The objective would be to move toward a pro- geological events (droughts, hurricanes, earthquakes, active (and more cost-effective) approach to finan- and floods) to actors in the market (insurance com- cial planning to protect national budgets, as well as panies, reinsurance companies, banks, and investors) to shield the lives and livelihoods of rural people who are willing to accept them at a price. from the impacts of disasters, like those experienced At the highest levels of intensity of natural disas- during severe ENSO episodes and flash floods. This ters, there is a need to transfer risk to international approach would help the government to consider markets through some type of reinsurance or insur- climate shocks as part of its fiscal risk management ance scheme. In this regard, Zimbabwe signed a mem- strategies. It would also complement other elements orandum of understanding with the African Risk of a comprehensive disaster risk management strat- Capacity (ARC) in 2012. ARC is a Specialized Agency egy, ranging from investments in risk reduction to of the African Union that provides risk-pooling ser- designing shock-responsive social safety nets. vices in the form of catastrophic indexed insurance Financial protection involves planning ahead to to governments against severe drought. Zimbabwe better manage the cost of disasters, ensure predictable is making the final arrangements to subscribe to this and timely access to needed resources, and ultimately transfer instrument. mitigate long-term fiscal impacts. By combining Another type of instrument that transfers agricul- various financial instruments—such as contingency tural production risk to international markets is agri- budget, contingent loans and grants, and risk transfer cultural insurance. In this regard, lessons from other solutions—financial protection allows governments countries (Box 2) suggest that a sound approach is to to manage the full range of disaster impacts. Different establish a public-private partnership with agricul- instruments help address different risks (ranging from tural stakeholders, including the insurance compa- recurrent to more rare events) and different funding nies. This partnership could identify the constraints needs (ranging from short-term emergency relief to for insurance market development, adopt policies recovery and reconstruction) (World Bank 2016b). and investments tending to develop the agricultural In many countries, contingency/reserve funds are insurance market, and identify the models and scenar- used to finance relief, rehabilitation, reconstruction, ios where agricultural insurance makes sense. Some and prevention activities for national emergencies. initiatives already exist. For example, in 2016, Blue Sovereign funds specifically dedicated to disas- Marble, a consortium of nine insurance companies, ter response exist in Colombia, Costa Rica, India, launched a pilot for agricultural index insurance in Indonesia, the Lao People’s Democratic Republic, Zimbabwe. The piloted products aimed to protect the Marshall Islands, Mexico, the Philippines, and maize, paprika, and other crops grown by small- Vietnam, among others. A number of other countries holders against drought. are working to establish similar funds. In Kenya, for example, the government is in the final stages of oper- 5.3.2 Adaptive Social Protection ationalizing a national contingency fund dedicated to drought emergencies. One use of the funds made available by risk financ- Contingent loans are financial instruments ing strategies is to support vulnerable house- designed to give countries access to liquidity imme- holds through social protection interventions. The diately following an exogenous shock, such as a Adaptive Social Protection (ASP) approach that has natural disaster. They are typically offered by multi­ emerged in recent years was first conceived as a series lateral development banks and international finan- of measures to build resilience to climate change cial institutions (including the World Bank, Asian among the poorest and most vulnerable people by Development Bank, Inter-American Development combining elements of social protection, disaster Bank, and International Monetary Fund). risk reduction, and climate change. Since then, the Market-based risk transfer solutions are used in term “adaptive” has come to be understood by social every sector of the economy and have growing rel- protection policy makers and practitioners as the evance in development due to increased exposure need to ensure that the social protection system can to risks that result in economic loss. A broad menu adapt safety net programs to respond to all types of of underlying instruments—derivative contracts, shocks (World Bank 2018). Toward a Risk Management Strategy 43 ASP typically encompasses two types of mea- deliver assistance to households in the most affected sures. The first type is deployed before shocks occur areas. Recently this approach reached existing benefi- and aims to boost the resilience of the most vulner- ciaries who were affected by disasters in Fiji and the able households. This resilience-building approach Philippines, as well as households affected by climate seeks to break the cycle of poverty and vulnerability. shocks in Senegal and Mauritania. For example, a household is better at withstanding ENSO shocks if it has more human capital and can Notes access job opportunities, accumulate physical capital, and diversify livelihoods. 1. The Tropical Soil Biology and Fertility program of The second type of ASP measure focuses on the International Center for Tropical Agriculture (CIAT) has developed the concept of “non-responsive” soils, increasing the capability of social safety nets to which have lost most of their organic carbon and nutrient- respond to shocks just before, during, and after a holding capacity. Application of mineral fertilizer will not disaster has occurred by introducing flexibility and lead to a yield response, as all fertilizer will wash out of scalability in program design. Such design features the soil and cannot be retained. Such soils already exist enable faster adjustment to meet post-shock needs. in many communal areas in Zimbabwe, where excessive tillage-based agriculture on sandy soils of granitic origin Conceptually, the safety net system becomes capable has degraded the organic matter content. Most of these of “scaling out” to households affected by shocks soils need to be brought back into production by increas- beyond its regular beneficiaries and/or “scaling up” ing the organic matter content first, before fertilization to increase benefit amounts or frequency of trans- will make any difference. Options for doing so include fers to existing social safety net beneficiaries at a (1) an increased biomass input in which large quantities time of acute need. For slow-onset shocks, programs of animal or green manure are applied, (2) conservation agriculture, and (3) agro-forestry (Vanlauwe et al. 2010; can ideally reach the most vulnerable households C. Thierfelder, pers.com.). before the shock leads them to adopt depletive 2. See https://gain.fas.usda.gov/Recent%20GAIN%20 coping strategies (such as reducing consumption, Publications/Zimbabwe%20Agricultural%20Economic% foregoing care, selling assets, pulling children from 20Fact%20Sheet_?Pretoria_Zimbabwe_9-22-2015.pdf. school, and so on). 3. See https://www.odi.org/sites/odi.org.uk/files/ Increasing support to existing social safety net odi-assets/publications-opinion-files/5208.pdf. 4. World Bank (2012). beneficiaries during a severe climate shock and tem- 5. World Bank (2018b). porarily supporting other households affected by 6. For a detailed discussion of the development of agri- shocks can be an efficient mechanism to mitigate cultural insurance and the role of governments in that impacts. Indeed, it can leverage existing public works effort, see Mahul and Stutley (2010). or cash transfer programs and their instruments, 7. Moral hazard is the perverse incentive to assume a including targeting mechanisms, outreach staff, and higher level of risk because somebody else is bearing the costs of that risk. For example, a farmer may not adequately payment systems. This approach offers an opportu- look after the crop to prevent damage from drought, because nity to build on existing platforms to horizontally it is insured. and vertically scale in the wake of a shock. It is also 8. Adverse selection occurs when an individual’s demand fundamental that those platforms are functional and for insurance is positively correlated with the individual’s have the trust of external donors. Electronic registries, risk of losses. In agriculture, this situation encourages a high proportion of farmers to take insurance, which in turn raises for example, can guarantee transparency, account- premiums. ability, and efficiency in assisting vulnerable house- 9. These preliminary estimates are based on informa- holds in the aftermath of a shock. Existing programs tion from the GoZ, international agencies, and qualified can then serve as conduits to rapidly and efficiently informants during the development of this report. Chapter 6 CONCLUDING REMARKS This report has presented findings of the assessment for this assessment, it is increasingly apparent that of agricultural risk and diagnosis of the disaster risk the public sector has a key role to play, not only in financing currently in place in Zimbabwe, under- correcting the present asymmetries in risk manage- taken by the World Bank at the request of the GoZ. ment but in assisting the large number of rural poor In general, this report shows that shocks along and vulnerable households to become more resilient. Zimbabwe’s agricultural supply chains directly This report emphasizes that an important com- translate to volatility in agricultural GDP as well as ponent of an integrated agricultural risk manage- impacts on overall economic growth, food security, ment strategy for Zimbabwe’s current context is to and the fiscal balance. Zimbabwe is highly exposed invest in risk mitigation measures that help small- to agricultural risks and has limited capacity to man- scale producers to strengthen the natural capacity for age risk at various levels. The resulting annual losses resilience and reduce risks at the farm level. Investing represent 7 percent of agricultural GDP on average, in risk mitigation at the farm level can go a long way severely affect vulnerable rural communities, and to reduce agricultural volatility, manage food inse- force the government to shift limited fiscal resources curity, and assist smallholders to adapt to climate away from development to cope with the effects of change. To meet these needs, the report suggests stra- agricultural risk. When catastrophic disasters occur, tegic steps for Zimbabwean agricultural systems— the economy absorbs the shocks, without benefit- and equally important, the public sector services ing from any instruments that transfer the risk to that support them—to leap-frog beyond current international markets. The increasing prevalence of perspectives and approaches to meet the needs of an “shock-recovery-shock” cycles impairs Zimbabwe’s’ agricultural economy largely defined by the needs of ability to plan and pursue a sustainable development small-scale producers whose realities are changing in path. Going forward, Zimbabwe will benefit from concert with the climate. For example, innovations moving toward a more proactive risk management in commercial agricultural insurance in various sub- strategy as it turns the corner toward recovery in its Saharan and other countries are worth studying for agriculture-based economy, which relies on a vast their potential to transfer risk in Zimbabwe when- number of smallholders who are highly exposed to ever possible. agricultural risks and have limited capacity to man- Another important point is that the public sector age them. must lead the way in strengthening the govern- The findings presented here confirm that it is highly ment’s capacity to provide timely assistance with the pertinent for Zimbabwe to strengthen the capacity to most severe and catastrophic disasters. Specifically, manage risk at various levels, from the smallholder Zimbabwe should start transitioning away from its farmer, to other participants along the supply chain, to current reactive strategy for managing agricultural risk consumers (who require a reliable, safe food supply), and move toward a proactive financial risk manage- and ultimately to the government to manage natural ment strategy, in which financial resources and deliv- disasters. Based on the fieldwork and analytical work ery mechanisms are in place before agricultural risks 45 46 Zimbabwe: Agriculture Sector Disaster Risk Assessment materialize. In other words, Zimbabwe should struc- It is hoped that the results and recommendations ture financial risk management strategies to improve of this assessment will serve as a basis and spring- financial preparedness for disaster response. Like board for an analysis of the options for Zimbabwe to many countries, Zimbabwe could start implementing implement a range of measures to reduce volatility in a risk layering approach combining different financial agriculture and the wider economy, strengthen food instruments that can be used depending on the fre- security, and minimize the fiscal risk derived from quency and intensity of the risks the country faces. natural disasters. BIBLIOGRAPHY Berdegué, J.A., and G. Escobar. 2002. “Rural Diversity, ———. 2017a. “Zimbabwe Drought Risk Management Agricultural Innovation Policies,and Poverty Reduction.” Strategy and Action Plan 2017–2015.” Ministry of Agricultural Research and Extension Network Agriculture, Mechanization and Irrigation Development/ (AGREN) Paper No. 122. Overseas Development Food and Agriculture Organization. Harare. Institute (ODI), London. https://www.odi.org/sites/ ———. 2017b. “Abridged Postal and Telecommunications odi.org.uk/files/odi-assets/publications-opinion-files/ Sector Performance Report, 3rd Quarter.” Postal 5208.pdf. and Telecommunications Regulatory Authority of CADRI (Capacity for Disaster Reduction Imitative). 2017. Zimbabwe. https://t3n9sm.c2.acecdn.net/?wp-content/ “Capacity Assessment of the Disaster Risk Management uploads/2017/12/3rd_Quarter_2017.pdf. System in Zimbabwe.” Harare. https://www.cadri.net/ ———. 2017c. “Rural Livelihoods Assessment Report.” sites/default/files/Zimbabwe-Report-May-2017.pdf. Zimbabwe Vulnerability Assessment Committee Chikoto and Sadiq (2013). Disaster Mitigation and (ZimVAC). Harare. Preparedness: Comparison of Nonprofit, Public, and Government of Zimbabwe and World Bank Zimbabwe Private Organizations. Nonprofit and Voluntary Sector Climate Change Technical Assistance (zccta) Programme. Quarterly, 42(2), 391–410. 2017. “Zimbabwe Agroecological Rezoning. A Draft Codex Alimentarius. 2016. http://www.zimcodex.gov.zw/ Concept Note.” Harare. food-control-in-zimbabwe. IndexMundi. 2018. Zimbabwe Corn Production by Year. EC (European Commission). 2011. “Phytosanitary Audit.” https://www.indexmundi.com/agriculture/?country= Final Report of a mission carried out in Zimbabwe zw&commodity=corn&graph=production. from 22 to 26 November 2010. Harare. Kasiyano, I. 2017. “Smallholder Agriculture Productivity and FAO (Food and Agriculture Organization). 2006. “Fertilizer Growth in Zimbabwe: Constraints and Oopportunities.” Use by Crop in Zimbabwe.” Rome. Paper prepared for the World Bank. October. Harare. GAIA Environmental Information System. Undated. “Zimbabwe: Land Use in Tropical Dry Savannahs. Soils Mugwisi, T., J. Mostert, and D.N. Ocholla. 2015. “Access to and Soil Erosion.” http://www.ess.co.at/GAIA/CASES/ and Utilization of Information and Communication ZIM/soils.html. Technologies by Agricultural Researcher and Extension GFDRR (Global Facility for Disaster Reduction and Workers in Zimbabwe.” Information Technology for Recovery.) 2017. Fundamentals. Disaster Risk Finance. Development Vol 21: 67–84. https://doi.org/10.1080/ Presentation. 02681102.2013.874317. Government of Zimbabwe. 2012. “Department of Murisa, T., and K. Mujeyi. 2015. “Land and Agrarian Agricultural Education Report.”, Ministry of Agriculture, Policy Reforms post 2000: New Trends, Insights and Mechanization and Irrigation Development. Harare. Knowledge.” In Beyond the Crises: Zimbabwe’s Prospects Government of Zimbabwe. 2014. “Zimbabwe National for Transformation. Harare: Weaver Press. Nutrition Strategy 2014–2018.” Ministry of Health Pswarayi, F/, A.N. Mutukumira, B. Chipurura, B. Gabi, and Child Care. Harare. and D.J. Jukes. 2014. “Food Control in Zimbabwe: Government of Zimbabwe. 2016a. “Agricultural Information A Situational Analysis.” Food Control 46: 143–151. Bulletin.” June. Ministry of Agriculture, Mechanization Reserve Bank of Zimbabwe. 2017. “Balance of Payments and Irrigation Development. Harare. Statistics.” RBZ Official Documents. Harare. ———. 2016b. “Understanding Gender Equality in SNV.2016. “The Groundnut Sub-sector Study.” Harare. Zimbabwe.” Women and Men Report. Harare. www.snvworld.org/Zimbabwe. 47 48 Zimbabwe: Agriculture Sector Disaster Risk Assessment Tyson, J.E. 2017. “Zimbabwe: A Roadmap for Economic World Bank. 2016a. Agricultural Sector Risk assess- Transformation.” Overseas Development Institute ment: Methodological guidance for practitioners. and UK Aid, London. https://set.odi.org/wp-content/ Agriculture Global Practice. Discussion Paper 10. uploads/2017/08/Zimbabwe_Roadmap-to- Washington, D.C. transformation_Aug2017.pdf. World Bank. 2016b. Disaster Risk Finance as a Tool for UNDP (United Nations Development Programme.). Development. A Summary of Findings from the 2017. “Zimbabwe Human Development Report 2017: Disaster Risk Finance Impacts Analytics Project. Climate Change and Human Development: Towards Global Facility for Disaster Reduction and Recovery. Building a Climate Resilient Nation.” Harare. Washington, D.C. UNOCHA (United Nations Office for the Coordination World Bank. 2017a. World Development Indicators: of Humanitarian Affairs). 2016. “Humanitarian Needs Zimbabwe. Washington, D.C Overview, Zimbabwe.” Harare. World Bank. 2017b. ICT in Agriculture. Updated Edition. Workman, D. 2017. “Zimbabwe’s Top 10 Imports.” World’s World Bank. 2018a. Zimbabwe Climate Smart Agriculture Top Exports, http://www.worldstopexports.com/ Profile. zimbabwes-top-10-imports/. World Bank. 2018b. “Strengthening Zimbabwe’s Early World Bank. 2007. Assessment of the economic impacts of Warning Systems: A Scoping Study.” Prepared by climate change on agriculture in Zimbabwe, A Ricardian Easther Chigumira. April. Harare. approach. Policy Research Working Paper 4292. ZIMSTAT. 2015. “Zimbabwe Poverty Atlas, Small Area Washington, D.C. Poverty Estimation.” Zimbabwe National Statistical World Bank. 2012. Agricultural Innovation Systems: An Agency, Harare. Investment Sourcebook. Washington, D.C. ZimVAC. 2016. Rural Livelihoods Assessment. APPENDIX A DESCRIPTION OF AGRICULTURAL SUPPLY CHAINS USED IN THE RISK ASSESSMENT Maize subsistence and also sell to farm-gate buyers. Consolidated produce from large-scale farmers is Maize is produced by both small-scale, mainly com- sold to GMB and processors such as Lions, Agri- munal, and larger-scale farmers. In resettled areas, seeds, Agricom, Reapers, and small-scale processors, small-scale farmers are in A1 resettlements while among others. Other groundnut value chain actors large scale farmers are in A2. The degree of mechani- include NGOs, farmers’ associations, banks, and gov- zation in maize farms has declined tremendously as a ernment extension. The marketing of groundnuts is result of the Fast Track Land Reform. Highly mecha- not highly controlled as in maize. The Agricultural nized large commercial maize farms were converted Marketing Authority is involved in export-import into farms on small holdings during the reform. regulation of the crop. Farmers are represented by various farming associations—for example, the Zimbabwe Farmers Union (ZFU) and Zimbabwe Commercial Farmers Tobacco Union (ZCFU). Farmers get inputs from agro- Tobacco is grown in Regions II and III by both com- dealers and the key agribusiness players in Zimbabwe munal and commercial farmers, but today there are are Seed Co, Pannar, Pioneer, ZFC, and Windmill, more small-scale tobacco growers than larger com- among others. The other key players in the maize mercial farmers. From the early 2000s, the number value chain are processors such as National Foods, of tobacco growers has increased exponentially (see Grain Millers Association of Zimbabwe (GMAZ), Figure A.1). Besides farmers, traders and processors and Blue Ribbon, and the government, through the (cigarette manufacturers), there are a number of GMB. The GMB is the main regulator of maize institutional players in the tobacco supply chain (see marketing in the country. Figure A.2). The Tobacco Industry Marketing Board is the tobacco regulatory board in Zimbabwe. The tobacco supply chain is well organized. Price Groundnuts discovery works reasonably well. Price volatility in A majority of communal maize farmers also produce tobacco is not a serious risk for tobacco farmers, as groundnuts. About 75 percent of groundnut farmers international price exhibits moderate variability. are in communal and resettlement areas, while only Contract farming, which is common among tobacco 25 percent are in commercial, A2, and other agricul- farmers, contributes to the stability of the supply tural areas.1 chain. Farmers are provided with inputs on credit Because of the lack of scale in production, ground- and allowed to pay back at the end of the season nuts produced by commercial and A2 farmers are after harvesting. Farmers under contract farming consolidated and traded through rural agro-dealers arrangements are required to take insurance against such as Mbare traders, Inter-grain, and Peak hold- natural risks. The contract is offered as a package, ings. Many smallholder farmers produce just for including the farmer’s insurance. 49 50 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure A.1.  Tobacco Growers and Area 120000 Tobacco growers and area 100000 80000 GROWERS 60000 AREA (HA) 40000 Expon. (GROWERS) Poly. (AREA (HA)) 20000 0 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 14 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 Source: MAMID. Figure A.2.  Tobacco Supply Chain Actors Curing Grading and Buying Auction Floors Zimbabwe Tobacco Association Contractors Zimbabwe Association of Tobacco Growers Tobacco Industry and Marketing board (TIMB) The key players are: The key players are: Commercial and small scale tobacco growers Premier Tobacco Auction Floors Boka Tobacco Auction Floors Tobacco Sales Floor Limited Farmers Stemming and Drying Zimbabwe Tobacco Association Zimbabwe Association of Tobacco Growers Tobacco Processors Zimbabwe (TPZ) Commercial Farmers Union Zimbabwe Tobacco Seed Association The key players are: Tobacco Processors Zimbabwe (TPZ) The key tobacco growing areas are: Chidziva Tobacco Processors Eastern Highlands British American Tobacco North and East of Harare Highveld region Retail Outlets Cigarette Manufacturing Retailers Association of Zimbabwe Tobacco Processors Zimbabwe (TPZ) British American Tobacco (BAT) The key players are: Savannah Tobacco Supermarkets Chain stores The key players are: British American Tobacco (BAT) Savannah Tobacco Cutrag Processors Fodya Private Limited Appendix A: Description of Agricultural Supply Chains Used in the Risk Assessment 51 Seed cotton and bio-electricity, molasses, and carbon dioxide (among other byproducts) that are of value to con- Cotton farming has shifted from large-scale commer- sumers in food industry and energy sectors. About cial farmers to smallholder farmers since the mid-1990s, 65 percent of the sugar produced is for the domestic following continued decline in international lint prices market and the remainder is exported to the region, that caused a downward trend in local prices. Cotton USA, and EU. is currently grown predominantly by smallholder Sugarcane in Zimbabwe is grown under full irriga- farmers (communal, old resettlement, A1, and small- tion in the South-East Lowveld areas of Masvingo and scale commercial) in marginalized and dry rural areas. Manicaland Provinces on the basis of a dual structure: Many smallholder farmers are women and youths. large scale corporate estates and small- to medium- The vast majority of these smallholders use basic equipment, such as animal-drawn implements to scale resettlement farms (A2). Hippo Valley, Triangle, prepare land and knapsack sprayers to control pests. and Chisumbanje are the estates.1 The three estates Most of the labor is from the family, although hired are vertically integrated into processing, whereby labor can be engaged for cotton picking and weed- sugarcane is milled into the two main products ing. Cotton production is structured almost entirely (sugar and ethanol) and several byproducts. About around contract farming (98–99 percent of farmers), 80 percent of Zimbabwe`s sugarcane crop is produced driven by contracting companies registered as ginners. by these three large estates, with the remainder pro- There are several private ginners and a govern- duced by private smallholder farmers. Figure A.4 ment one (Cottco). Suppliers of agricultural inputs illustrates the sugarcane supply chain in Zimbabwe. (fertilizers, agro-chemicals, and seed) used by Among the A2 farm owners in old and new reset- farmers are regulated by the Ministry of Agricul­ tlement areas, about 50 percent are female-headed ture, Mechanization, and Irrigation Development households, with only 10 percent of females owning (MAMID) through DR-SS. Figure A.3 shows the farms in their own names. An estimated 10 percent of cotton supply chain in Zimbabwe. farm owners are households headed by children who inherited the farms after the deaths of their parents. Women, who make up an estimated 10 percent of the Sugarcane labor force on sugarcane farms, are involved in light Sugarcane is an important crop in Zimbabwe. The duties such as office administration, weeding, fertil- industry provides sugar and ethanol as main products ization, and trashing. Figure A.3.  Cotton Supply Chain 52 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure A.4.  Sugarcane Supply Chain Some A2 farmers produce their sugarcane and The wheat supply chain comprises several actors sell to the company, whereas numerous indepen- (producers, processors, traders, retailers, and con- dent A2 farmers depend on the services of the sumers). Figure A.5 highlights how they are linked. sugar milling plants to mill their sugarcane at a fee The wheat supply chain starts with the input (18 percent), with the value of the cane determined by suppliers who provide the seed, fertilizer, and crop its sugar content (in other words, a monopsony chemicals that are needed by the producers. During market organization exists in the sugar milling recent years the supply of inputs has been crippled business). Extension services for large estates are by the liquidity crisis, which makes it difficult to entirely private, while among the A2 farms, both the procure the raw materials needed for manufacture. milling company and government provide exten- The main input suppliers in the country are ZFC, sion support. Windmill, and National Tested Seeds. The main grain trader in the country is GMB, which purchases wheat from commercial farm- Wheat ers based on a predetermined government market Wheat became popular as a staple crop due to high rate. Apart from being a trader, the GMB is a demand for flour and bread. Wheat is grown during national storage facility. GMB occasionally rents the winter, as it requires low temperatures for high out its silo storage facilities to private traders such productivity and successful crop development. It also as CropLink, Staywell, and Intergrain Enterprises. requires irrigation, which limits wheat production There are several milling companies (from small to areas close to dams, reservoirs, boreholes, sprin- scale to large scale) that are involved in the process- klers, and electricity. As a result, wheat is associated ing of wheat. The major player is National Foods. with high production costs, and most producers are Other wheat processors are GMB, Blue Ribbon, medium- to large-scale commercial farmers (A2) and Premier Milling. The main wholesalers include located in Mashonaland and Manicaland Provinces. Mahommad Mussa, OK Mart, N Richards, and These farmers are affiliated with organizations such Metro & Peech, which are found extensively in as ZFU and ZCFU. the cities and towns in Zimbabwe. The retailers Appendix A: Description of Agricultural Supply Chains Used in the Risk Assessment 53 Figure A.5.  Wheat Supply Chain Retailers and Wholesalers Wheat, fluor and Supermarkets wheat commodity Retailers imports Processors Flour and Bran Millers Traders Bakeries National Traders Private Sector Traders Producers Medium to large scale commercial farmers (A2) Large-scale commercial farmers Input suppliers Seed Houses Fertilizer manufacturers Pesticide manufacturers are wholesalers, retail supermarkets, bakeries, and is varied, producing vegetables, fruits, and flow- restaurants. ers on large farms or estates. The produce mainly goes to the export market, local retailers (chiefly supermarkets), and food processing companies. Fruits Horticulture that are produced for export include citrus (oranges, Horticulture is a specialized subsector that is cur- grapefruit, lemons), subtropical fruits (bananas, rently Zimbabwe’s fifth-largest agricultural export mangoes, passionfruit), deciduous fruits (peaches, earner. It contributes 6.5 percent to agricultural GDP. apricots, plums, other stone fruit, apples, and Horticultural production occurs mainly in the coun- pears), and strawberries. Vegetables include cherry try’s agro-ecological Regions I and II. Production is tomatoes, sweet corn, chilies, peas, and fine beans. conducted within close proximity to major urban These are processed and sold as packs of pre-washed centers, along the roads that link urban settlements. mixed vegetable that are ready to cook. Horticulture is performed at large and small scale. Small-scale horticultural production consists Large-scale commercial horticultural production of communal, resettlement A1, old resettlement, 54 Zimbabwe: Agriculture Sector Disaster Risk Assessment small-scale commercial farms, and peri-urban and collection, bulk tank supply, and sample collection urban producers that practice horticulture in the for laboratory testing. The other important players garden or backyards of residential stands. Smallholder in Zimbabwe’s dairy industry are the input suppliers farmers who have access to irrigation facilities and and the dairy processors such as Dairibord Zimbabwe, have sufficient water supplies during the dry season Kefallos, Alpha and Omega, Nestlé, and Den Dairy, to produce for the market. Most of the produce from name just a few. The dairy sector also has company these farmers is sold through the informal sector, dairy farmers who are both producers and processors. while a few are contracted to supply formal markets, agro-food processing companies, and export markets. Poultry Contract farming is another prominent produc- tion system under which both large-scale producers The poultry industry can be divided into two main and small-scale farmers are engaged. Contract farming categories: commercial and indigenous free-range is viewed as a more profitable venture, as the returns chicken farming. Indigenous chicken production sys- are usually higher than selling on the local market, tems are mostly based on the local scavenging domes- and there is a guaranteed market for the farmer. The tic fowl (Gallus domesticus) predominant in African companies that provide contracts are Cairns Foods, villages. Local chicken breeds are the most abundant FAVCO, and Interfresh Limited, as well as retail- livestock species in Zimbabwe. Indigenous chickens ers such as TM/Pick N Pay and Food Lovers Market. are mostly raised in a free-range system in small flocks Figure A.6 shows the horticulture supply chain. of less than 30. They are more adapted to local condi- tions than the hybrid breeds but have lower produc- tivity. The indigenous chicken production system is Cattle normally found at the subsistence level but it is very Beef cattle are produced in all provinces, with the high- important for these household as a source of eggs and est numbers produced in the dry areas. The players are meat and a ready source of cash during a family crisis. large commercial farmers, small-scale commercial Commercialization of indigenous chicken production farmers, and communal households. Smallholder is on the rise with the supply of the Boschveld breed and communal farmers are mostly subsistence and of road runners supplied by Charles Stewart day-old place greater emphasis on the social importance chicks. However, the contribution of this sector in of livestock. Besides farmers, the beef supply chain poultry production is still very low. is integrated by input suppliers and abattoirs. See Commercial poultry production is an intensive Figure A.7. business mainly done by individual households, Dairy farmers can be grouped into two categories small-scale producers, and large-scale commercial based on their production levels and the resources operators. These producers either produce broilers for they own. The first category includes the dairy farm- meat or layers for table eggs. Irvine’s, Charles Stewart, ers that are registered with the National Association and Masvingo are some of the main suppliers of day- of Dairy Farmers (NADF) and Zimbabwe Dairy old chicks for both layers and broilers. There are many Farmers Association (ZDFA). The other group abattoirs and processors that sometimes add value by is the smallholder dairy farmers, made up of com- processing and producing chicken byproducts. Feed munal, small-scale, and resettlement farmers under suppliers and veterinary distributors are very critical the Agriculture and Rural Development Authority – in chicken production. Dairy Development Program (ARDA-DDP) and individual smallholder dairy farmers who are not in Notes the ARDA-DDP register. There are many associations in the dairy indus- 1. SNV (2016). try which provide technical and extension services 2. Tongaat-Hulett, a South African sugar company, owns 100 percent of the Triangle Sugar Estate and about to dairy farmers, such as the Ministry of Lands, 50.3 percent of the Hippo Valley Estate. Hippo Valley Agriculture, and Rural Resettlement, NADF, ZDFA, Estate is a public company listed on the Zimbabwe Stock ZFU, Zimbabwe Dairy Industry Trust, and the Exchange, and other shareholders include Tate & Lyle and National Dairy cooperative responsible for milk the British Agro-Company. Appendix A: Description of Agricultural Supply Chains Used in the Risk Assessment 55 Figure A.6.  Horticulture Supply Chain Exported fruits and Imported fruits, Formal Wholesalers vegetables vegetables and Informal national Processed horticultural processed markets products horticultural products Global Markets Retail Markets Wholesalers Processors Traders Small-scale and large-scale processors Producers Small-scale producers Large-scale producers Smallholder plots Commercial farmers Irrigation schemes Estates Gardens Input suppliers Seed/seeding houses Fertilizer suppliers 56 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure A.7.  Beef Supply Chain STRUCTURE OF THE BEEF INDUSTRY IN ZIMBABWE LARGE SCALE COMMERCIAL HERO By-products-hides, tallow, meals REGION 1 & 2 FARMS Export: high grade beef Pen Ranches REGION 3, 4 & 5 fattening FARMS Feedlots Finished off Wildlife grass Cattle finance scheme COLD STORAGE Domestic sales IMPORTED INPUTS, COMMISSION FEEDS, CHEMICALS, ABBATOIRS DRUGS, SEMEN Research, Dip tanks, Animal health PRIVATE ABBATOIRS COMMUNAL HERBS AGRITEX FARM SLAUGHTERINGS DRSS RESETTLEMENT HERBS DLVS GVT of Zimbabwe Command livestock SMALL-SCALE COMMERCIAL FMD Fencing Development & Tsetse control partners PROCESSING MARKETING PRODUCTION APPENDIX B METHODOLOGY USED TO ESTIMATE THE VALUE OF CROP LOSSES Yield deviations are calculated with respect to the multiplied by the area. This algorithm make it possible trend line of the yields; those years in which the nega- to estimate the volume of production losses. Losses in tive deviations are greater in absolute value than the value are then calculated by multiplying the volume of standard deviation of the deviations are taken as the losses by the price of crops. The sections that follow years in which significant risk events occur. Then, present examples of these calculations for tobacco and for those years, the deviations from the trend are maize. See also Tables B.1 and B.2 and Figures B.1-B.3. Tobacco Crop Loss Estimates Table B.1.  Data for Estimating Tobacco Crop Losses and Their Value Deviation of yields Trend of Trend of with respect to D - Standard Losses Yields yields yield trend - D deviation of (US$ at Year Area (ha) (t/ha) US$/kg (t/ha) (t/ha) D (t/ha) Losses (t) 2014 prices) 1985 52,464 2.012 2.34 –0.33 –0.54 0 0 1986 57,349 1.993 2.31 –0.32 –0.54 0 0 1987 63,536 2.015 2.28 –0.27 –0.54 0 0 1988 59,178 2.026 2.26 –0.23 –0.54 0 0 1989 57,660 2.254 2.23 0.03 –0.54 0 0 1990 59,425 2.253 2.20 0.05 –0.54 0 0 1991 66,927 2.542 2.17 0.37 –0.54 0 0 1992 80,070 2.512 1.62 2.14 0.37 –0.54 0 0 1993 82,900 2.634 1.24 2.12 0.52 –0.54 0 0 1994 67,416 2.51 1.73 2.09 0.42 –0.54 0 0 1995 74,550 2.666 2.12 2.06 0.61 –0.54 0 0 1996 81,231 0.481 2.94 2.03 –1.55 –0.54 126,035 399,472,981 1997 90,630 1.893 2.33 2.00 –0.11 –0.54 0 0 1998 91,905 2.349 1.72 1.98 0.37 –0.54 0 0 1999 84,762 2.267 1.74 1.95 0.32 –0.54 0 0 2000 84,857 2.792 1.69 1.92 0.87 –0.54 0 0 2001 76,017 2.664 1.75 1.89 0.77 –0.54 0 0 2002 74,295 2.213 2.27 1.87 0.35 –0.54 0 0 2003 49,571 1.673 2.25 1.84 –0.16 –0.54 0 0 2004 44,025 1.565 2 1.81 –0.24 –0.54 0 0 2005 57,511 1.3 1.61 1.78 –0.48 –0.54 0 0 2006 58,808 0.943 2 1.75 –0.81 –0.54 47,683 151,131,657 2007 54,551 1.339 2.31602 1.73 –0.39 –0.54 0 0 2008 61,622 0.792 3.21196 1.70 –0.91 –0.54 55,834 176,967,160 2009 62,737 0.934 2.97859 1.67 –0.74 –0.54 46,187 146,390,111 2010 67,054 1.842 2.87904 1.64 0.20 –0.54 0 0 (Table continues next page) 57 58 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table B.1.  (continued) Deviation of yields Trend of Trend of with respect to D - Standard Losses Yields yields yield trend - D deviation of (US$ at Year Area (ha) (t/ha) US$/kg (t/ha) (t/ha) D (t/ha) Losses (t) 2014 prices) 2011 78,415 1.689 2.72932 1.61 0.07 –0.54 0 0 2012 76,359 1.893 3.65099 1.59 0.31 –0.54 0 0 2013 88,627 1.852 3.6749 1.56 0.29 –0.54 0 0 2014 107,371 2.104 3.1695 1.53 0.57 –0.54 0 0 2015 108,307 1.551 4.29746 1.50 0.05 –0.54 0 0 Source: Tobacco Industry Marketing Board (2015); Ministry of Agriculture, Mechanization and Irrigation Development, Zimbabwe, Agricultural Statistical Bulletin (2016). Figure B.1.  Yield Trends and Crop Losses, Tobacco 1.00 Zimbabwe - Tobacco 140,000 0.50 120,000 100,000 0.00 80,000 –0.50 60,000 –1.00 40,000 –1.50 20,000 –2.00 0 07 09 11 13 15 05 95 97 99 01 03 91 93 85 87 89 20 20 20 20 19 20 20 20 20 19 19 19 19 19 19 19 Losses (Tons) Deviation of yields with respect to yield trend - D (Tons/Ha) Trend of D - Standard deviation of D (Tons/Ha) Appendix B: Methodology Used to Estimate the Value of Crop Losses 59 Maize Crop Loss Estimates Table B.2.  Data for Estimating Maize Crop Losses and Their Value NATIONAL - Deviation of yields NATIONAL - Trend Trend of with respect of D - Standard Losses yields to yield trend - D deviation of D Losses (US$ at Year Production (t) Area (ha) (t/ha) (t/ha) (t/ha) (t) 2016 prices) (1) 1986/87 1,530,000 774,800 1.61 0.367410753 –0.332851207 0 0 1987/88 2,253,100 1,299,500 1.57 0.166536448 –0.332851207 0 0 1988/89 1,931,200 1,198,300 1.53 0.084662143 –0.332851207 0 0 1989/90 1,993,800 1,149,800 1.49 0.246787838 –0.332851207 0 0 1990/91 1,585,800 1,101,200 1.45 –0.007086466 –0.332851207 0 0 1991/92 361,000 881,000 1.41 –0.996960771 –0.332851207 878,322 342,545,751 1992/93 2,011,850 1,238,000 1.37 0.258164924 –0.332851207 0 0 1993/94 2,326,000 1,401,200 1.33 0.333290619 –0.332851207 0 0 1994/95 839,600 1,397,900 1.29 –0.685583686 –0.332851207 958,377 373,767,199 1995/96 2,609,000 1,535,000 1.25 0.45354201 –0.332851207 0 0 1996/97 2,191,370 1,640,100 1.21 0.129667705 –0.332851207 0 0 1997/98 1,418,030 1,223,800 1.17 –0.0072066 –0.332851207 0 0 1998/99 1,519,560 1,446,400 1.13 –0.075080905 –0.332851207 0 0 1999/00 1,619,651 1,373,117 1.09 0.094044791 –0.332851207 0 0 2000/01 1,526,328 1,239,988 1.05 0.185170486 –0.332851207 0 0 2001/02 604,758 1,327,854 1.01 –0.550703819 –0.332851207 731,254 285,189,165 2002/03 1,058,786 1,352,368 0.97 –0.182578124 –0.332851207 0 0 2003/04 1,686,151 1,493,810 0.93 0.203547571 –0.332851207 0 0 2004/05 915,366 1,729,867 0.89 –0.356326733 –0.332851207 616,398 240,395,164 2005/06 1,484,839 1,712,999 0.85 0.021798962 –0.332851207 0 0 2006/07 952,600 1,445,800 0.81 –0.146075343 –0.332851207 0 0 2007/08 470,700 1,722,322 0.76 –0.494949648 –0.332851207 852,463 332,460,440 2008/09 1,242,566 1,521,780 0.72 0.085176047 –0.332851207 0 0 2009/10 1,327,572 1,803,542 0.68 0.015301743 –0.332851207 0 0 2010/11 1,451,629 2,096,034 0.64 0.048427438 –0.332851207 0 0 2011/12 968,041 1,689,786 0.60 0.395553133 –0.332851207 0 0 2012/13 798,596 1,265,236 0.56 0.065678828 –0.332851207 0 0 2013/14 1,456,153 1,655,366 0.52 0.355804524 –0.332851207 0 0 2014/15 742,225 1,531,663 0.48 –0.004069781 –0.332851207 0 0 2015/16 511,816 1,161,997 0.44 –0.003944086 –0.332851207 0 0 Source: MAMID Agricultural Statistical Bulletin, 2016. Note 1: Resettlement Areas included in Communal Area Totals from 1980/81 onwards, Small-scale A2 Totals in A2 Area Totals from 1985/2013. (1) Maize retail price, national average, June 2016 US$/kg 0.39. Source: FAO/GIEWS. 60 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure B.2.  Yield Trends and Crop Losses in Relation to Severe Weather Events, Maize Zimbabwe - Maize 0.6 1,200,000 0.4 1,000,000 0.2 0 800,000 Drought –0.2 2015/16 not affecting 600,000 –0.4 Mashonaland Moderate province that drought is main maize –0.6 Severe drought, 400,000 producing Severe inputs area –0.8 Moderate drought shortages, drought, removal hyper inflation, 200,000 –1 of subsidies cash shortages Severe following ESAP drought –1.2 0 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Losses (Tons) NATIONAL - Deviation of yields with respect to yield trend - D (Tons/Ha) NATIONAL - Trend of D - Standard deviation of D (Tons/Ha) Appendix B: Methodology Used to Estimate the Value of Crop Losses 61 Figure B.3.  Maize Yields Over Time (National Yields and Yields at Different Scales of Production) 2 Zimbabwe - Maize yields (Tons/Ha) 1.5 1 0.5 0 –0.5 –1 –1.5 –2 –2.5 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 NATIONAL - Deviation of yields with respect to yield trend - D (Tons/Ha) NATIONAL - Trend of D - Standard deviation of D (Tons/Ha) COMUNAL - Deviation of yields with respect to yield trend - D (Tons/Ha) COMUNAL - Trend of D - Standard deviation of D (Tons/Ha) COMMERCIAL - Deviation of yields with respect to yield trend - D (Tons/Ha) COMMERCIAL - Trend of D - Standard deviation of D (Tons/Ha) → Approximately US$81.6 million at 2015/16 annually. For the moment, the calculation involves prices, or 4.9 percent of the agricultural GDP, was tobacco and maize, but these two crops account for estimated as the value of the average production loss about 47 percent of total agricultural GDP. APPENDIX C METHODOLOGY FOR THE RISK ASSESSMENT FOR VARIOUS RETURN PERIODS Materials and Methods Expected GVP associated with a given probability that is related to a return period. The crop risk assessment is performed for coffee, Cropped area, total production, and yield records cotton, groundnuts, maize, sorghum, soybeans, sug- obtained from MAMID and FAOSTAT are treated arcane, tobacco, and wheat in the aggregate for all prior to use in the analysis. Crop area and yield of Zimbabwe. The analysis covers the period from records are treated for outliers. Outliers were identi- 1986 to 2015. The food crop risk assessment was fied by setting an upper bound and a lower bound largely based on official records from MAMID and for the records, and all records that fell outside of FAOSTAT. This information was complemented with these boundaries were removed. For this analysis, the data on the average price at the province level for 2016 upper bound was set at average value for the variable for the main food crops, also provided by MAMID. plus 3.5 times its standard deviation, and the lower The main outputs of the crop risk assessment bound was set at the average value for the variable model are the Expected Average Loss and the Loss less 3.5 times its standard deviation. at Risk. The Expected Average Loss for each crop is Crop yield series are detrended to remove the calculated based on the deviation of Monte Carlo- effects of shifts in yields along the series. For the generated actual gross value of production (GVP) purposes of detrending yields, a combination based from expected GVP. If the Monte Carlo-generated on the average between a linear detrending and a actual GVP falls short of the expected GVP, then polynomial of second-order function is used for there is a loss proportional to the size of the shortfall. each crop. For the purposes of this model, the Expected Average The correlations among the different crops were Loss for a given unit is determined by the average of considered in the simulation model for the aggregate the Monte Carlo GVP shortfalls with respect to the portfolio, using a Pearson model. The correlations expected GVP. were added to the model by using the RiskCorrmat The Loss at Risk (LaR) or Probable Maximum Function; RiskCorrmat functions are added to each Loss (PML) is a key measure used to infer the poten- of the input distribution functions that are included tial losses in the portfolio. The LaR is a percentile of in defined correlation matrix. The historical burning the loss distribution, calculated in function of the cost (HBCA) was calculated for the different coverage probability of occurrence of a catastrophic event. For levels from 100 percent down to 30 percent for each example, the LaR for an Exceedance Probability “p” crop. The HBCA for each unit area of insurance (UAI), of 1 percent (or return period “RP” of 1 in 100 years), and coverage levels were then fit to a set of continuous is the value of the loss distribution that accumu- probability density functions conformed by a Normal, lates 99 percent of probability, i.e., the 99th percen- Lognormal, Weibull, Loglogistic, Gamma, and inverse tile. For the purposes of this model, the Loss at Risk Gaussian probability density function. The first selec- for a given Unit is determined by the percentile of tion criterion was based on the minimization of the the Monte Carlo GVP shortfalls with respect to the root mean square error (RMSE). In addition to the 63 64 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure C.1.  Zimbabwe: Historic Evolution of Maize Yields 2,500 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 2,000 Linear; y = –41.913x + 1662.8 Yield (kg/ha) 1,500 1,000 500 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg/he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg/he.) RMSE, probability density functions were also tested average between a polynomial of second order against other critical criteria. detrending function and a linear detrending function. The detrending analysis comes from a weighted As a result, an adjusted detrended yield is obtained. average of a linear detrending and a polynomial of Figure C.3 shows the detrended yields for maize in second order detrending. Both show a similar trend. Zimbabwe, alongside EY for 2016. The period considered for the detrending comprised In the next step the stochastic variable of the the full series (1986–2016). The land reform and the model (i.e., YPD) is fitted to a parametric probability changes in property rights at the begging of the 2000s distribution function (PDF). This procedure is done had a severe impact on agriculture output. The land for each crop in the portfolio. Each fit is then tested reform issue and its impact on the interpretation of against three criteria: (1) Chi-squared, (2) Anderson- the results of this analysis are highlighted in the main Darling, and (3) Kolmogorov-Smirnov. Figure C.4 text of this report as a factor to be considered when shows the resulting (detrended) yield PDF for maize interpreting the results, along with the additional in Zimbabwe. consideration that during 2001-08 Zimbabwe experi- Once the PDFs are fitted to the historical YPD data enced successive seasons of drought followed by one for each crop and the correlation matrix among YPD season of excessive rainfall, as well as deregulation samples is calculated, the simulation can be performed. that led to production increases. It is not possible to In this case, a Monte Carlo simulation was used to distinguish the impacts of each of these circumstances generate simulated samples of 10,000 hypo­ thetical on yields, so it was decided that the analysis would years of detrended yields for the crops included in the use the entire data series but that the results must be portfolio. interpreted with caution. Figure C.1 shows the evolu- The valuation of the crops for the calculation of tion and trends in maize yields in Zimbabwe. the GVP is based on the average price for 2016 for the The second phase of the detrending analysis is main crops in Zimbabwe (provided by MAMID). The to estimate the yield variability. For that purpose, stochastic GVP used as basis for the calculation of the percentage deviations (YPD) between the actual the Expected Average Loss, and the Loss at Risk is yields with respect to the corresponding expected the result of the multiplication of each of the Monte yield according to the trend line is also calculated. Carlo-generated deviation from expected crop area Figure C.2 presents the time series of YPD for maize (EHA), times the Monte Carlo-generated deviation in Zimbabwe. from yield (YPD) times the corresponding Expected YPDs are then applied to the expected yield (EY). Yield for 2017, times the average price for the crop The EY is calculated as the projection of the of the provided by MAMID. Figure C.2.  Zimbabwe: Maize Yield Deviations from Expected Yields 80.00% 60.00% Yield Deviations (% of Expected Yield) 40.00% 20.00% 0.00% –20.00% –40.00% –60.00% –80.00% 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Figure C.3.  Zimbabwe: Detrended Maize Yields 800 700 600 500 Yield (Kg/he.) 400 300 200 100 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 “D-Trended Yield” “Expected Yield” Figure C.4.  Zimbabwe Maize: Best Fit Probability Distribution Function over Detrended Yields 1.2000 1.0000 0.8000 F [x] 0.6000 0.4000 0.2000 0.0000 0 100 200 300 400 500 600 700 800 900 1000 1100 Yield (kg/ha) “Historic Actual Yields” “MCO Simulated Yields” 66 Zimbabwe: Agriculture Sector Disaster Risk Assessment By using this methodology , 10,000 hypothetical GVP (US$417.9 million) once in 250 years. Table C.2 years of detrended GVP for the crops and provinces and Figure C.8 present the expected LaR values for included in the portfolio were obtained. These values the crop portfolio in Zimbabwe. were used as the main underlying data for the risk Soybeans make the highest contribution to risk in assessment model used in the study. Figure C.5 pre­ the food crop portfolio in Zimbabwe. The share of sents the schematic description of the methodology soybeans in the portfolio’s average loss cost increased followed for obtaining the 10,000 hypothetical years by 1.49 points for each additional point of increase of detrended GVP for maize. The figure conceptu- of its share in the portfolio. On the other hand, sug- ally shows how detrended Monte Carlo yields are arcane is the crop that shows the lowest contribution generated and how the 10,000 hypothetical years of to the risk in the portfolio. For each basis point the detrended GVP are determined. share of sugarcane crop is increased, its contribution to the portfolio’s average loss cost is increased only by Main Findings and Discussion 0.52 basis points. Table C.3 shows the contribution of each crop to the average loss cost in the portfolio. This section presents the aggregate risk assessment for the whole crop portfolio in Zimbabwe. The find- Maize Crop Risk Assessment ings are the result of the Crop Risk Assessment Tool that was specifically designed for this study. Maize is the staple crop in Zimbabwe, used for household consumption and income generation. Data from MAMID show that maize production Crop Portfolio Risk Assessment and yields in Zimbabwe steadily declined over the Total area planted to the main crops in Zimbabwe is last 30 years. Currently there are 1.45 million hect- 2.25 million hectares. The exposure of main crops in ares sown with maize in Zimbabwe. The GVP for terms of GVP amounts to US$945 million. Tobacco maize, according to the assumptions of this risk and maize, accounting for 38 percent and 28 percent assessment, is US$261.4 million. Figures C.9 and C.10 of the total exposure, are the main crops in the port­ show the decline in maize production and yields folio. Sugarcane with US$177 million in exposure is for the period from 1986 to 2015. also a very important crop that accounts for 19 per- Maize seems to be a medium-risk crop in cent of the total crop exposure. Table C.1 presents data Zimbabwe. The average loss cost for maize crops is on the area and GVP of the crops in the portfolio ana- 11.38 percent of its GVP or US$29.7 million per year. lyzed for Zimbabwe. The LaR analysis indicates that maize may face a Agricultural production has shown a steady down- loss equivalent to 66.4 percent of the national GVP trend over the last 30 years in Zimbabwe. Market ana- (or US$173.5 million) once in 100 years or a loss of lysts and academic researchers often attribute this 75.3 percent of national GVP (or US$197 millions) decline in agricultural output to the 2000 reform that in a 250-year return period. Table C.4 and Fig­ - resulted in a significant number of smallholder farms ure C.11 show the expected LaR values for maize without the skills and ability to efficiently produce in Zimbabwe. agricultural crops compared to the previously large- scale commercial farms. Figures C.6 and C.7 depict Tobacco Crop Risk Assessment the declines in production and yields. Agricultural production in Zimbabwe is very vol- Zimbabwe is the largest grower of tobacco in Africa, atile. The average value of losses for the main crops and the sixth-largest grower in the world. Three in Zimbabwe is calculated at US$125.7 million per types of tobacco have traditionally been grown in year and accounts for 13.30 percent of the portfolio the country: Virginia flue-cured, burley, and orien- GVP. Severe agricultural production losses may recur tal tobacco. Over 95 percent of Zimbabwe’s tobacco at relatively short intervals in Zimbabwe. The LaR consists of flue-cured tobacco, which is renowned for indicates that the food crop portfolio may face a loss its flavor. In 2005, 54 percent of Zimbabwe’s tobacco equivalent to 27.2 percent of the national crop GVP was exported to China. (or US$256.7 million) once in 10 years, crop losses of Land reform in Zimbabwe after 2000 redistributed 39.9 percent of national GVP (or US$377.3 million) land to farmers unskilled in growing tobacco. These once in 100 years, and crop losses of 44.23 percent of farmers held no title to the land, so they lacked the Figure C.5.  Methodology Used for the Assessment MAIZE-YIELD DEVIATIONS –0.473 0.507 5.0% 90.0% 5.0% 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0 0.5 1.0 1.5 –1.5 –1.0 –0.5 Yield Deviations from Expected Yields MAIZE-EXPECTED YIELD 244 697 5.0% 90.0% 5.0% 0.0030 0.0025 0.0020 0.0015 0.0010 0.0005 0.0000 0 200 400 600 800 1000 1200 Yield (kg./ha) MAIZE-EXPECTED PRODUCTION 0.353 1.010 5.0% 90.0% 5.0% 2.0 1.8 MAIZE-EXPECTED YIELD: 467 KG/HA 1.6 1.4 1.2 1.0 0.8 Values x 10–6 0.6 0.4 0.2 0.0 0.2667 0.5333 0.8000 1.0667 1.3333 1.6000 0.0000 Production (Tons) Values in Millions MAIZE-EXPECTED AREA: 1,449,675 HA MAIZE-EXPECTED GVP 138 394 5.0% 90.0% 5.0% 6 5 4 3 Values x 10–9 2 1 Appendix C: Methodology for the Risk Assessment for Various Return Periods 0 0 100 200 300 400 500 600 –100 Values in Millions MAIZE-EXPECTED PRICE: 390 USD/TN 67 68 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table C.1.  Zimbabwe Crop Portfolio: Exposure–Expected Gross Value of Production (US$) Expected crop Expected yield Expected Crop price Exposure - Unit area (ha) (kg/ha) production (t) (US$/kg) GVP (US$) Cotton 149,955 501 75,111 0.500 37,555,732.20 Coffee 2,225 28 63 3.500 218,839.16 Groundnuts 229,510 379 87,097 0.500 43,548,624.55 Maize 1,449,675 462 670,159 0.390 261,362,150.45 Sorghum 224,420 226 50,723 0.379 19,223,947.97 Soybean 47,919 995 47,701 0.480 22,896,717.10 Sugarcane 44,273 72,737 3,220,327 0.055 177,117,978.76 Tobacco 101,435 1,450 147,067 2.440 358,844,578.46 Wheat 14,552 3,249 47,280 0.510 24,112,738.10 Total 2,263,966 944,881,306.75 collateral to obtain bank loans .Much of Zimbabwe’s international players in the local market improved farmland went out of cultivation, and the tobacco contract terms and drove up sales prices. Tobacco crop bottomed out at 48 million kilograms in 2008, production recovered under the contract system. just 21 percent of the 2000 crop. A contract system The 2014 tobacco crop of 217 million kilograms for tobacco farming was introduced in Zimbabwe in was the third-largest crop on record, amounting to 2005. Buyers like British American Tobacco began to 104 percent of the average crop grown from 1991 contract with tobacco farmers to buy their entire crop to 2000. The structure of the industry has been at the end of the season. In return, the buyer would transformed: in 2000, 1,500 large-scale tobacco supply the farmer with all necessary inputs. Buyers farmers grew 97 percent of the crop, and in 2013, also took greater responsibility for the crop, send- 110,000 small-scale farmers grew 65 percent of ing agronomists to the contracted fields to advise the crop. farmers on agricultural techniques and make sure Currently more than 101,000 hectares are planted that tobacco workers were paid on time. China also with tobacco, which is the crop with the highest value began to invest in the tobacco industry. The entry of in Zimbabwe. The GVP of tobacco in Zimbabwe is Figure C.6.  Zimbabwe: Production of Main Crops (t), 1986–2015 3,000,000 2,500,000 2,000,000 Production (Tons) 1,500,000 1,000,000 500,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 COTTON-WHOLE COUNTRY COFFEE-WHOLE COUNTRY GROUNDNUTS-WHOLE COUNTRY MAIZE-WHOLE COUNTRY SORGHUM-WHOLE COUNTRY SOYBEAN-WHOLE COUNTRY TOBACCO-WHOLE COUNTRY Appendix C: Methodology for the Risk Assessment for Various Return Periods 69 Figure C.7.  Zimbabwe: Yields of Main Crops (kg/ha), 1986–2015 3,000 2,500 2,000 Yield (kg/ha) 1,500 1,000 500 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 COTTON-WHOLE COUNTRY COFFEE-WHOLE COUNTRY GROUNDNUTS-WHOLE COUNTRY MAIZE-WHOLE COUNTRY SORGHUM-WHOLE COUNTRY SOYBEAN-WHOLE COUNTRY TOBACCO-WHOLE COUNTRY Table C.2.  Zimbabwe: Whole Crop Complex: Expected LaR Values for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 27.17% 32.94% 36.28% 39.93% 41.77% 42.92% 44.23% 46.67% LaR (US$ millions) 256.7 311.2 342.8 377.3 394.7 405.6 417.9 441.0 Figure C.8.  Expected LaR for Different Recurrence Periods for the Whole Crop Portfolio in Zimbabwe 50% 45% 40% LAR (% Total Exposure) 35% 30% 25% 20% 15% 10% 5% 0% 1 100 200 300 400 500 Recurrence Period (Years) 70 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table C.3.  Zimbabwe: Contribution of Each Crop to the Portfolio Risk Exposure – GVP Average loss cost Portfolio (US$) % (US$ millions) % Contribution index Cotton 37,555,732.20 3.97% 3,664,559.08 2.92% 0.73 Coffee 218,839.16 0.02% 44,584.49 0.04% 1.53 Groundnuts 43,548,624.55 4.61% 3,855,844.93 3.07% 0.67 Maize 261,362,150.45 27.66% 29,734,159.05 23.66% 0.86 Sorghum 19,223,947.97 2.03% 2,593,313.49 2.06% 1.01 Soybeans 22,896,717.10 2.42% 4,539,173.82 3.61% 1.49 Sugarcane 177,117,978.76 18.74% 12,304,014.98 9.79% 0.52 Tobacco 358,844,578.46 37.98% 65,377,385.611 52.02% 1.37 Wheat 24,112,738.10 2.55% 3,554,433.49 2.83% 1.11 Whole portfolio 944,881,306.75 100.0% 125,667,468.94 100.0% 1.00 Figure C.9.  Zimbabwe: Maize Production (t), 1986–2015 3,000,000 2,500,000 2,000,000 Production (Tons) y = –36593x + 2E+06 1,500,000 1,000,000 500,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) Figure C.10.  Zimbabwe: Maize Yields (kg/ha), 1986–2015 2,500 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 2,000 Linear; y = –41.913x + 1662.8 Yield (kg/ha) 1,500 1,000 500 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg/he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg/he.) Appendix C: Methodology for the Risk Assessment for Various Return Periods 71 Table C.4.  Zimbabwe: Expected LaR Values for Maize for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 36.87% 50.26% 58.78% 66.40% 70.47% 73.32% 75.34% 81.58% LaR (US$ millions) 96.4 131.4 153.6 173.5 184.2 191.6 196.9 213.2 Figure C.11.  Zimbabwe: Expected LaR for Maize for Different Recurrence Periods 90% 80% 70% LAR (% Total Exposure) 60% 50% 40% 30% 20% 10% 0% 1 100 200 300 400 500 Recurrence Period (Years) estimated at US$359 million. Tobacco risk exposure Tobacco production is a high-risk endeavor in accounts for 38 percent of the total crop risk expo- Zimbabwe. The expected average loss for tobacco in sure in the country. Figures C.12 and C.13 show the Zimbabwe accounts for 18.22 percent of its GVP or evolution of tobacco production and yields over US$65.3 million per year. The expected LaR for this 1986–2015. crop is 66.7 percent of the GVP (US$239.4 million) Tobacco in Zimbabwe is affected by drought and for a recurrence period of 100 years and 70.38 percent hailstorms. Drought can be especially pervasive of GVP (US$252.6 million) for a recurrence period during ENSO years. Hailstorms are quite frequent in of 250 years. Table C.5 and Figure C.14 show the Zimbabwe, and hail damage affects tobacco produc- expected LaR values for tobacco for different return tion as well as the quality of the product. periods. Figure C.12.  Zimbabwe: Tobacco Production (t), 1986–2015 250,000 200,000 y = –944.93x + 155553 Production (Tons.) 150,000 100,000 50,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) 72 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure C.13.  Zimbabwe: Tobacco Yields (kg/ha), 1986–2015 3,000 2,500 2,000 Yield (kg/ha) 1,500 1,000 Linear; y = –41.913x + 1662.8 500 Polynomial 2nd Order; y = 1.195x – 78.958x + 1860.4 2 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg/he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg/he.) Table C.5.  Zimbabwe: Expected LaR Values for Tobacco for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 49.55% 58.43% 63.11% 66.72% 68.48% 69.62% 70.38% 72.61% LaR (US$ millions) 177.8 209.7 226.5 239.4 245.7 249.8 252.6 260.5 Figure C.14.  Expected LaR for Different Recurrence Periods for Tobacco in Zimbabwe 80% 70% 60% LAR (% Total Exposure) 50% 40% 30% 20% 10% 0% 1 100 200 300 400 500 Recurrence Period (Years) Appendix C: Methodology for the Risk Assessment for Various Return Periods 73 Sugarcane Crop Risk Assessment sugarcane is 6.95 percent of GVP or US$12.3 million. The LaR for sugarcane indicates that this crop may Sugarcane is an important crop in Zimbabwe. It face an aggregate loss equivalent to 24 percent of covers 44,273 hectares and is essential for distilla- the national crop GVP (or US$42.5 million) once tion and ethanol production, providing sweeteners in 100 years or a loss of 25.13 percent of national for industry, making molasses for cattle feed, earn- crop GVP (US$44.5 million) once in 250 years. ing foreign exchange, and generating electricity. Table C.6 and Figure C.17 show the expected LaR Sugarcane GVP amounts to US$177 million, which values for sugarcane in Zimbabwe for different makes this crop the third most important in the return periods. country from an economic standpoint. Sugarcane production takes place in northwestern Zimbabwe in Groundnut Crop Risk Assessment the Lowveld, which has been identified as one of the best places in the world to produce sugar at competi- Groundnuts are currently planted on 230,000 hect- tive costs. The climate is ideal for sugarcane and the ares in Zimbabwe. Groundnuts are grown by a large distances from the mill are quite manageable. proportion of smallholder farmers (36 percent), but Sugarcane yields have shown a downward trend, despite the crop’s importance, production and pro- falling from 110 tons per hectare in the 1980s to ductivity have remained low and stagnant at less 77 tons per hectare in 2015, for three reasons. The first than 500 kilograms per hectare, for three reasons. reason is farmers’ limited access to inputs. Fertilization First, poor access to quality seed of improved vari- rates are far from optimal, and most farmers never eties make farmers rely on retained seed of landra- plow out the cane due to lack of resources. The ces. Second, farmers lack knowledge and skills in second reason is that farmers do not use appropri- groundnut production. Third, farmers have poor ate crop management practices. The third reason market access. Figures C.18 and C.19 summarize the is the lack of capital equipment for sugarcane pro- evolution of groundnut production and yields over duction. Most farmers in Zimbabwe do not even 1986–2015. have a tractor, which is essential machinery for land The GVP of groundnuts is estimated at US$43.5 mil­ preparation and hauling cane. Figures C.15 and C.16 lion. Groundnut exposures accounts for 4.5 percent of show the evolution of sugarcane production and the crop exposures in the country. yields over the period 1986–2015. The main peril affecting groundnuts in Zimbabwe Sugarcane is produced under irrigation and is drought. Groundnut production shortfalls in rela- hence is a low-risk crop. The average loss cost for tion to expected yields were particularly bad during the Figure C.15.  Zimbabwe: Sugarcane Production (t), 1986–2015 6,000,000 y = 8266.5x + 3E+06 5,000,000 Production (Tons.) 4,000,000 3,000,000 2,000,000 1,000,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) 74 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure C.16.  Zimbabwe: Sugarcane Yields (kg/ha), 1986–2015 140,000 120,000 100,000 Yield (kg/ha) 80,000 60,000 40,000 Linear; y = –41.913x + 1662.8 20,000 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg./he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg./he.) Table C.6.  Zimbabwe: Expected LaR Values for Sugarcane for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 18.48% 21.41% 22.90% 24.02% 24.55% 24.87% 25.13% 25.77% LaR (US$ millions) 32.7 37.9 40.6 42.5 43.5 44.1 44.5 45.6 Figure C.17.  Zimbabwe: Expected LaR for Sugarcane for Different Recurrence Periods 30% 25% LAR (% Total Exposure) 20% 15% 10% 5% 0% 1 100 200 300 400 500 Recurrence Period (Years) Appendix C: Methodology for the Risk Assessment for Various Return Periods 75 Figure C.18.  Zimbabwe: Groundnut Production (t), 1986–2015 250,000 200,000 y = 1447.3x + 82619 Production (Tons.) 150,000 100,000 50,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) 1991, 2001, and 2015 droughts (60 percent, 55 percent, Cotton Crop Risk Assessment and 35 percent, respectively). The expected average loss for groundnuts in the country accounts for 8.9 percent Cotton was once a strategic crop for poverty allevia- of its GVP (US$3.9 million). The expected LaR for this tion in Zimbabwe. Cotton contributed sustainably crop is 43.6 percent of the GVP (US$18.9 million) for a to rural income, rural development, employment, recurrence period of 100 years and 48.3 percent of GVP and export earnings. The sector was a major source (US$21 million) for a recurrence period of 250 years. of livelihood for over one million people, includ- Table C.7 and Figure C.20 present the expected LaR ing farmers, farm workers and the textile indus- values for groundnuts for different return periods. try, as it once contributed about 19 percent of the Figure C.19.  Zimbabwe: Groundnut Yields (kg/ha), 1986–2015 900 800 700 600 Yield (kg/ha) 500 400 300 200 Linear; y = –41.913x + 1662.8 100 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg/he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg/he.) 76 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table C.7.  Zimbabwe: Expected LaR Values for Groundnuts for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 26.72% 34.61% 39.42% 43.60% 45.81% 47.23% 48.31% 51.57% LaR (US$ millions) 11.6 15.1 17.2 19.0 19.9 20.6 21.0 22.5 country’s agricultural export earnings. It was the and 2015, when production shortfalls in relation mainstay of rural communities, resulting in the to expected yields were 65 percent, 36 percent, and development of areas like Gokwe, Sanyati, Rushinga, 37 percent, respectively. The expected average loss Checheche, Muzarabani, Matepatepa in Bindura, for this crop accounts for 9.8 percent of its GVP. and Muzarabani. The expected LaR for this crop is 50.7 percent of The number of farmers cultivating this crop has the GVP for a recurrence period of 100 years and recently suffered a sharp decline. In 2012, cotton was 56.2 percent of GVP for a recurrence period of cultivated by about 200,000 smallholders, while in 250 years. Table C.8 and Figure C.23 present the 2014 an estimated 170,000 small-scale cotton produc- expected LaR values for cotton for different return ers grew the crop, representing an average 15 percent periods. decline in two years. The cotton subsector started to experience serious trouble because many farmers Coffee Risk Assessment failed to access adequate inputs from contractors, and some contractors (such as Cottco and Cargill) ceased Zimbabwe’s coffee belt has perfect conditions for operations. The loss of cotton profits has destroyed growing the beans: high mountain peaks and cool livelihoods in rural areas where peoples’ existence was climates. The country was once famous for the intricately linked to growing the crop. Distortions in “super-high quality” and flavor of its beans. In the producer price have also had a negative effect on the 1990s it produced some of the best coffee in production, as farmers have abandoned cotton pro- the world, alongside South America and Kenya, duction in favor of crops such as tobacco, maize, and generating crucial foreign currency and a liveli- soybeans. Currently the area planted with cotton is hood for many laborers and small-scale farmers, as 150,000 hectares, and the GVP for this crop is esti- well as the big commercial farms. mated at US$37.5 million. Figures C.21 and C.22 show Today the industry is in decline: many mills are the evolution of cotton production and yields over abandoned, and farmers are in debt. The country 1986–2015. produced 500 tons of coffee in 2017 compared to The main peril affecting cotton in Zimbabwe is 15,000 tons in 1989. The number of commercial drought, which was especially serious in 1991, 2001, producers has fallen from 120 before the land reform Figure C.20.  Zimbabwe: Expected LaR for Groundnuts for Different Recurrence Periods 60% 50% LAR (% Total Exposure) 40% 30% 20% 10% 0% 1 100 200 300 400 500 Recurrence Period (Years) Appendix C: Methodology for the Risk Assessment for Various Return Periods 77 Figure C.21.  Zimbabwe: Cotton Production (t), 1986–2015 400,000 350,000 y = –3607.2x + 262184 300,000 Production (Tons.) 250,000 200,000 150,000 100,000 50,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) Figure C.22.  Zimbabwe: Cotton Yields, 1986–2015 1,400 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 1,200 Linear; y = –41.913x + 1662.8 1,000 Yield (kg/ha) 800 600 400 200 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg./he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg./he.) 78 Zimbabwe: Agriculture Sector Disaster Risk Assessment Table C.8.  Zimbabwe: Expected LaR Values for Cotton for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 30.60% 40.05% 45.77% 50.71% 53.27% 55.02% 56.27% 60.05% LaR (US$ millions) 11.5 15.0 17.2 19.0 20.0 20.7 21.1 22.6 Figure C.23.  Zimbabwe: Expected LaR for Cotton for Different Recurrence Periods 70% 60% LAR (% Total Exposure) 50% 40% 30% 20% 10% 0% 1 100 200 300 400 500 Recurrence Period (Years) program to just 3 today. Coffee plantations in Sorghum Crop Risk Assessment Zimbabwe occupy only 2,225 hectares. Figures C.24 and C.25 show the evolution of coffee production Sorghum is an important crop in the driest regions and yields over 1986–2015. of the county. Being drought tolerant, it has a strong Coffee production is a high-risk endeavor in adaptive advantage and lower risk of failure than Zimbabwe. The expected average loss for this other cereals in such environments. The current area crop accounts for 20.37 percent of its GVP, while planted with sorghum in Zimbabwe is 250,000 hect- the expected LaR for a recurrence period of one in ares. Like other crops in Zimbabwe, sorghum shows 100 years is 37.7 percent. a downward yield trend. The GVP for this crop in Figure C.24.  Zimbabwe: Coffee Production (t), 1986–2015 16,000 14,000 12,000 10,000 Production (Tons.) 8,000 6,000 y = –477.68x + 14074 4,000 2,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 –2,000 Production (tons) Linear (Production (tons)) Appendix C: Methodology for the Risk Assessment for Various Return Periods 79 Figure C.25.  Zimbabwe: Coffee Yields (kg/ha), 1986–2015 2,500 2,000 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 Linear; y = –41.913x + 1662.8 1,500 Yield (kg/ha) 1,000 500 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg./he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg./he.) Zimbabwe amounts to US$19.2 million. Figures C.26 Wheat Crop Risk Assessment and C.27 show the evolution of sorghum production and yields over 1986–2015. Wheat production in Zimbabwe is small and way The expected average loss for this crop accounts below the level of consumption. Wheat production for 13.5 percent of its GVP, while the expected LaR has been declining since 2001, when Zimbabwe pro- for this crop is 51 percent of the GVP for a recur- duced more than 300,000 tons. Several constraints, rence period of 100 years, and 54.4 percent of the such as unreliable power supplies for irrigating the GVP for 250 years. Table C.9 shows the expected crop, dilapidated irrigation infrastructure, and late LaR values for sorghum for different return periods. payments by the GMB have contributed to declining Figure C.26.  Zimbabwe: Sorghum Production (t), 1986–2015 200,000 180,000 y = –436.41x + 87464 160,000 140,000 Production (Tons.) 120,000 100,000 80,000 60,000 40,000 20,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) 80 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure C.27.  Zimbabwe: Sorghum Yields (kg/ha), 1986–2015 900 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 800 Linear; y = –41.913x + 1662.8 700 600 Yield (kg/ha) 500 400 300 200 100 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg./he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg./he.) wheat production. Currently wheat is produced on from soybeans). Soybean cake, a byproduct of oil only 14,500 hectares, and production is approxi- extraction, is sold to feed manufacturers. Soybeans mately 47,000 tons per year. Wheat GVP amounts were originally produced by large-scale farmers. Since to US$24.1 million. Figures C.28 and C.29 show the land reform in 2000, the share of production from the evolution of wheat production and yields over small-scale farms has increased. National output has 1986–2015. dropped in recent years to about 50,000 tons per year, Wheat production was severely affected by the produced on 48,000 hectares. This reduction is attrib- droughts of 1991, 2001 and 2007, falling by 61 percent uted to a shrinking producer base and loss of produc- in relation to the expected yield in 1991, 88 percent tivity on small- and large-scale farms. Low output in 2001, and 69 percent in 2007. The expected aver- has caused considerable shortages of raw materials age loss for this crop accounts for 14.77 percent of its for cooking oil and feed. Currently large-scale com- GVP, while the expected LaR for a recurrence period mercial farmers account for 65 percent of national of one in 100 years is 84 percent of the GVP, and the soybean production, and smallholders account for LaR for a 250-year return period is 95 percent of the 35 percent. Soybean GVP amounts to US$3 million. GVP. Table C.10 shows the expected LaR values for Figures C.30 and C.31 show the evolution of soybean wheat for different return periods. production and yields over 1986–2015. Droughts in 1991 and 1994 caused soybean yields to drop by 35 percent each time. The expected aver- Soybean Crop Risk Assessment age loss for this crop accounts for 19.8 percent of its Soybeans are one of Zimbabwe’s high-value crops, GVP, while the expected LaR for a recurrence period and soybean production has strong industry linkages of one in 100 years is 61 percent of GVP, and the LaR because it the crop be processed into such value-added for a 250-year return period is 63 percent of GVP. products as soybean cake, soymilk, and soybean oil Table C.11 shows the expected LaR values for cotton (30 percent of the cooking oil in the country is made for different return periods. Table C.9.  Zimbabwe: Expected LaR Values for Sorghum for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 36.89% 44.09% 48.02% 51.13% 52.70% 53.66% 54.40% 56.36% LaR (US$ millions) 7.1 8.5 9.2 9.8 10.1 10.3 10.5 10.8 Appendix C: Methodology for the Risk Assessment for Various Return Periods 81 Figure C.28.  Zimbabwe: Wheat Production (t), 1986–2015 350,000 300,000 250,000 y = –8842.5x + 308615 Production (Tons.) 200,000 150,000 100,000 5,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) Figure C.29.  Zimbabwe: Wheat Yields (kg/ha), 1986–2015 8,000 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 7,000 Linear; y = –41.913x + 1662.8 6,000 5,000 Yield (kg/ha) 4,000 3,000 2,000 1,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg./he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg./he.) Table C.10.  Zimbabwe: Expected LaR Values for Wheat for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 47.46% 64.19% 74.79% 84.16% 89.32% 92.64% 95.35% 100.000% LaR (USS millions) 11.4 15.5 18.0 20.3 21.5 22.3 23.0 24.1 82 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure C.30.  Zimbabwe: Soybean Production (t), 1986–2015 160,000 140,000 y = –1629.7x + 111296 120,000 Production (Tons.) 100,000 80,000 60,000 40,000 20,000 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Production (tons) Linear (Production (tons)) Figure C.31.  Zimbabwe: Soybean Yields (kg/ha), 1986–2015 2,500 Linear; y = –41.913x + 1662.8 Polynomial 2nd Order; y = 1.195x2 – 78.958x + 1860.4 2,000 Yield (kg/ha) 1,500 1,000 500 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Yield (kg/he.) Yield Polynomial Trend (kg./he.) Yield Linear Trend (Kg/he.) Yield Linear-Polynomial Trend (Kg./he.) Table C.11.  Zimbabwe: Expected LaR Values for Soybeans for Different Return Periods Recurrence (years) 10 25 50 100 150 200 250 500 LaR (% exposure) 49.30% 55.60% 58.66% 60.90% 61.94% 62.58% 63.01% 64.22% LaR (US$ millions) 11.3 12.7 13.4 13.9 14.2 14.3 14.4 14.7 APPENDIX D RISKS FOR CROP AND LIVESTOCK SUPPLY CHAINS Crops incomes, and/or consumption. The impact of drought and other risk events is not only perceived by Weather-Related Risks farmers in terms of production losses but by con- The major historical risk in the agricultural sector of sumers as higher retail prices when market shortages Zimbabwe is drought, usually accompanied by high of basic food staples occur. For example, maize retail temperatures. Substantial drops in maize produc- prices increased in 2008/09, 2014/15, and 2016/17 fol- tion were caused by droughts in 1991/92, 1994/95, lowing droughts, and in 2008 as a result of the short- 2001/02, 2004/05, 2007/08, 2012/13, and 2015/16. age of cash and other enabling environment problems Groundnut production was affected by the droughts (Figure D.3). of 1982/83, 1991/92, 1997/98 (Manicaland), 2001/02, The southern part of the country (corresponding 2004/05, 2007/08, 2012/13, and 2015/16. As both to the dry agro-ecological Regions IV and V) is par- crops are considered food staples, those droughts ticularly exposed to drought risk. Drought in those reportedly affected food security in those years. For regions occurs on average every 3-5 years. For exam- building the risk profiles of crops, a quantitative assess- ple, Masvingo experienced severe drought in 2009/10 ment of losses was made by estimating the variation while no drought occurred in Mashonaland prov- in yields away from the historical trend line and mul- inces during that period. Severe droughts are not that tiplying the output losses by the average price of the frequent in other regions, occurring every 8-10 years last three years (for more details on the methodology, and inducing temporary shortages and price hikes see Appendix B). Secondary data and interviews with (for instance, in horticultural produce). An effec- stakeholders completed the sources of information. tive strategy based on the prevention and mitigation Figures D.1 and D.2 show the reconstructed timeline of drought risk with conservation technologies and of events and the annual estimated losses in maize livelihood diversification (among others) was dis- and groundnuts. cussed with various stakeholders as a priority to face For maize, there were 5 years in which drought severe droughts in a sustainable manner in the most caused yields to deviate more than one standard devi- exposed agro-ecological zones (IV and V). ation from the trend line; for groundnuts, a similar Weather risks have different impacts in Zimbabwe’s deviation is observed for 6 years within the 30-year different agro-ecological environments (regions). For period. In Chapter 4, these deviations are monetized example, maize is produced in each of the 10 prov- to estimate the value of losses, permit comparisons inces of Zimbabwe, and the regional differences in among crops, and prioritize risk. terms of weather and soils determine the potential to The vast majority of smallholders cultivate land grow maize, as well as the impact of weather-related under rainfed conditions and have limited means risks. The major maize-producing provinces are the of protecting themselves from the effects of drought Mashonaland provinces in the North, where weather and/or prolonged dry spells, so they ultimately absorb variability is less extreme, whereas the dry south- the effects of droughts in their productive systems, ern provinces (Masvingo, Matebeleland South, and 83 Figure D.1.  Maize: Estimated Annual Yield Losses Zimbabwe - Maize 0.6 1,200,000 0.4 1,000,000 0.2 0 800,000 Drought –0.2 2015/16 not affecting 600,000 –0.4 Mashonaland Moderate province that drought is main maize –0.6 Severe drought, producing area 400,000 Severe inputs shortages, –0.8 drought hyper inflation, Moderate cash shortages drought, removal 200,000 –1 of subsidies Severe following ESAP drought –1.2 0 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Losses (Tons) NATIONAL - Deviation of yields with respect to yield trend - D (Tons/Ha) NATIONAL - Trend of D - Standard deviation of D (Tons/Ha) Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe. Figure D.2.  Groundnuts: Estimated Annual Yield Losses 0.4 Zimbabwe - Groundnuts 70,000 0.3 60,000 0.2 50,000 0.1 40,000 0 30,000 –0.1 20,000 –0.2 Moderate Mid Moderate Severe drought season drought regional dry spell Severe 10,000 –0.3 Severe drought drought drought –0.4 0 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Losses (Tons) Deviation of yields with respect to yield trend - D (Tons/Ha) Trend of D - Standard deviation of D (Tons/Ha) Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe. Appendix D: Risks for Crop and Livestock Supply Chains 85 Figure D.3.  National Retail Maize Prices in Zimbabwe 3m 6m 1y 2y 3y All Monthly Prices From Oct-08 To Jan-18 0.7 drought regional enabling 0.6 drought environments US Dollar/kg issues 0.5 0.4 0.3 0.2 0.1 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Zimbabwe, Retail, National Average, Maize Grain US Dollar/kg Source: FAO/GIEWS. Matebeleland North)—agro-ecological Regions IV Drought also has potential to be a significant and V—are typically less productive. Dry spells are hazard to wheat production when it is severe and more frequent in the southern regions and sometimes widespread, as in 1992 and 2008 (Figure D.5). become consecutive and prolonged. And in effect, Drought remains a serious hazard for wheat despite maize production in Masvingo and Matebeleland has the fact that the crop is grown during the winter the highest yield coefficient of variation (Table D.1).1 under irrigation, mostly by medium- to large-scale Like groundnuts, cotton, despite being another commercial farmers. drought-tolerant crop, was also adversely affected by Severe droughts affect water availability for irriga- the severe droughts of 1991/92, 2001/02, and 2007/08 tion, as water reservoirs will be filled to low capac- (Figure D.4). Moderate droughts experienced in ity. The effect is more crippling to smallholder and 1994/95 and 2015/16 also negatively affected cotton commercial wheat farmers who do not have secure production, but to a lesser extent. All cotton is gen- financial resources and lack access to high-capacity erally produced under rainfed conditions, mostly in reservoirs and extensive irrigation infrastructure. Midland and Mashonaland Central Provinces. Horticulture, like wheat production, is carried out under intensive irrigation. It suffers heavily from severe drought as in 1992, when reservoirs had water Table D.1.  Variation in Maize Yields by Province, deficits, rivers dried up, and soils experienced mois- Zimbabwe ture stress. Farmers usually try to manage the drought Yield coefficient of risk by reducing the area planted to horticultural Province variation(%) crops, replacing them with more drought-tolerant Mashonaland West 37.0 crops, and focusing on irrigation of high-value Mashonaland Central 33.3 horticultural crops destined for export. Mashonaland East 34.1 Frost affects almost all horticultural produce, Manicaland 34.5 including fruits and vegetables. It occurs frequently Midlands 52.9 Masvingo 58.8 in Zimbabwe, particularly in Regions I and II, where Matebeleland North 55.5 temperatures are cooler and drop dramatically Matebeleland South 64.2 between late April and August. Frost differs in inten- Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe sity and is more severe during the cold winter years. 86 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure D.4.  Seed Cotton: Estimated Annual Yield Losses 0.4 Zimbabwe - Seed cotton 160,000 140,000 0.2 120,000 0 100,000 –0.2 80,000 Combination of Drought side marketing, 60,000 Severe drought, input diversion, –0.4 localised shortage of inputs, drought floods, and hyper inflation, Moderate land reform 40,000 drought cash shortage –0.6 Severe 20,000 drought –0.8 0 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 Losses (Tons) NATIONAL - Deviation of yields with respect to yield trend - D (Tons/Ha) NATIONAL - Trend of D - Standard deviation of D (Tons/Ha) Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe Figure D.5.  Wheat: Estimated Annual Losses 2 Zimbabwe - Wheat (Tons/Ha) 120,000 1 100,000 0 80,000 Dry spell –1 Drought, 60,000 erratic electricity –2 40,000 Drought, macroeconomic –3 instability 20,000 Drought –4 0 85 88 90 92 94 96 98 00 02 04 06 08 10 12 14 16 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 Losses (Tons) Deviation of yields with respect to yield trend - D (Tons/Ha) Trend of D - Standard deviation of D (Tons/Ha) Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe Appendix D: Risks for Crop and Livestock Supply Chains 87 Frost disrupts the growth and flowering stages of capacity to cope with hailstorms and usually resort to plants; tissues blacken and become necrotic, trigger- replanting, especially if the hailstorms occur early in to ing early senescence in the crop. Fruits and vegetables the rainy season when the crops are immature. that would have matured for harvest develop blisters Climate variability in Zimbabwe also manifests as and have decaying, water-soaked tissues. To combat an early onset of the rainy season. An early onset of the effects of frost, producers plant early before the rainfall that coincides with wheat planting and germi- onset of frost, increase sprinkler irrigation, and prac- nation has a detrimental effect on the development of tice mulching to trap heat. the crop and its eventual quality. Farmers counteract Hailstorms are also common in Regions I and II, this risk by planting as early as acceptable. Additionally, where most of the horticulture occurs during the rainy excess rains that fall when wheat has reached maturity season. Hailstorms damage plant leaves and stems, can significantly reduce yield and quality. reducing the photosynthetic capacity of the crop and Prolonged dry spells and erratic rainfall are frequent thus resulting in delayed maturity of the produce in every province at least once in every three years. Mid- and reduced yields. Fruits and vegetables suffer from season dry spells mainly occur in January and February, mechanical damage, which leaves them exposed to when maize and groundnuts are flowering, and affect invasion by pathogens and pests in the field and after yields of those crops. Prolonged mid-season dry spells harvest. The mechanical damage reduces the quality of also threaten the availability of water to sustain the irri- the crop, resulting in reduced market prices for the pro- gation that supports horticulture. Producers practice duce. During the onset of the rainy season, hailstorms conservation methods and mulching that minimally can also affect tobacco by destroying the leaves (which disturbs the soil to maintain soil moisture and fertility are the final product for tobacco farmers). Hailstorms so that crops do not suffer from water stress and are are an idiosyncratic risk, however, not a widespread able to cope through the duration of the dry spell. one. The typical way to manage the impacts of a hail- Dry spells have no significant effect on tobacco storm is to buy insurance. Producers have very low yields but negatively affect quality. Figure D.6 shows Figure D.6.  Tobacco: Estimated Annual Yield Losses 1.0000000 Zimbabwe - Tobacco 70,000 0.8000000 60,000 0.6000000 0.4000000 50,000 0.2000000 40,000 0.0000000 –0.2000000 30,000 –0.4000000 –0.6000000 20,000 Hyper inflation, Moderate cash shortage –0.8000000 drought Severe 2008/09 10,000 2004/05 drought –1.0000000 2007/08 –1.2000000 Land reform and land invasions 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Losses (Tons) Deviation of yields with respect to yield trend - D (Tons/Ha) Trend of D - Standard deviation of D (Tons/Ha) Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe 88 Zimbabwe: Agriculture Sector Disaster Risk Assessment historic drought events and their impact on tobacco All of these risks directly affect farmers but also yields. The dry spells normally occur when the affect processors and traders when the raw materials tobacco crop is already established. for their operations become scarce and most likely Early cessation of rains before the end of the more costly. growing season is also reported, affecting all crops. Marked negative changes in sugarcane produc- Phytosanitary Risks tion associated with droughts were noted in 1973/74, 1992/93, and 2008/09 (Figure D.7). The 1973/74 Pests and diseases in all crops are not a major risk if drought, which occurred between two above- controlled with agrochemicals, although the cost of normal seasons (1972/73 and 1974/75) resulted only in these chemicals is high, and most smallholder pro- moderate losses in yield and output, as water reserves ducers cannot afford them. In horticulture, the inci- from other seasons could be used to irrigate the crop. dence of pests and diseases is exacerbated by excess The 1991/92 severe drought that succeeded three con- rain, as excess moisture provided appropriate condi- secutive years of moderate drought curtailed the water tions for pathogens to breed and for pathogens and supply for irrigation and caused high losses in sugar- pests to disperse from one plant to another. cane yield and output. In 2008/09, drought and other Fungicide resistance is acknowledged to be grow- macro-economic hardships jointly contributed to a ing in wheat and horticulture. Adoption of genetically marked decline in performance in the sugarcane sector. resistant varieties and rotations of agrochemicals are Flooding is another risk, but it mainly affects promoted to combat fungal diseases. Some large- maize yields in a few areas of the country, particu- scale horticultural farms implement integrated larly Muzarabani in Mashonaland Central and some pest management systems, which use a combina- parts of Masvingo and Midlands. Farmers practic- tion of biological, cultural, and chemical means to ing stream bank cultivation to mitigate drought are eliminate pests. more exposed to floods. Cotton is grown in low- In addition to pests and diseases, crops can be lying areas prone to floods owing to their flat terrain susceptible to plagues such as quelea birds (Quelea and low altitude, but cotton can withstand the effects quelea) in wheat. These birds are a problem every of moderate flooding, because it is generally a deep- season, and farmers tend to team up to physically rooted, strongly established crop. ward them off from their fields. Figure D.7.  Sugarcane: Estimated Annual Yield Losses 40 Zimbabwe - Sugar (Tons/Ha) 1,200,000 20 1,000,000 0 800,000 –20 600,000 –40 400,000 –60 200,000 –80 –100 0 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Losses (Tons) Deviation of yields with respect to yield trend - D (Tons/Ha) Trend of D - Standard deviation of D (Tons/Ha) Source: Zimbabwe Sugar Association (ZSA). Appendix D: Risks for Crop and Livestock Supply Chains 89 Various new pest outbreaks have been reported. a noted pest in Mozambique, attacks sugarcane. It One relatively new pest, fall armyworm (Spodoptera severely affects plant growth as it kills the growing frugiperda), invaded Zimbabwe in 2016 and poses an point of the plant, and all affected plants have to be immediate risk to wheat and maize production. The plowed out. The moth is transmitted through cross- fall armyworm is known to cause extensive damage. border movement of chewing sugarcane. The poten- In the field it is often identified late, as it is difficult tial for the moth to arrive in Zimbabwe is very high, to differentiate from other caterpillars and burrows given the frequent informal cross-border movement into the stem, where it is shielded from pesticides. between Zimbabwe and Mozambique. The govern- The current capacity to manage fall armyworm is ment has already set pheromone traps to monitor inadequate. This pest is not fully researched in movement of the moth into the country. Zimbabwe, and the extension services have little knowl- edge of its management. Another important pest, the Cattle Production maize grain borer, destroys maize after harvesting. In cotton, emerging pests such as mealybug and army- Weather-Related Risks worm have been seen in 1998/99, 2016/17 and 2015/16. There is no proven remedy for mealybug or Heliothis As in crop production, in cattle production the main moths, and research continues in public institutions to weather-related risk is persistent drought, especially understand and develop treatments for these pests. severe droughts and high temperatures in Regions IV The leaf miner Tuta absoluta, introduced in and V every one in five years and in the rest of Zimbabwe in 2016, has resulted in huge losses in the the country every one in ten years. Drought can cause field. Most lossess have occurred in horticultural crops high livestock mortality, as water sources dwindle (solanaceous), for which it appears to have a preference, while grazing capacity of rangelands declines, mainly although it can prey on non-horticultural crops such in Matabeleland South and North, Masvingo (espe- as tobacco. Little research has focused on combating cially Gutu, Chivi, and Mwenezi Districts), and in this pest in Zimbabwe, and the extension services have Manicaland Province. been of little assistance to farmers. All of the chemical Dairy cows in particular are very sensitive to tem- formulations currently manufactured in the coun- perature changes, and an increase in temperature is try have lost effectiveness after each spraying, as Tuta associated with stress and high somatic cell counts. absoluta develops resistance at a rapid rate. Producers When drought occurs, the amount of feed pro- are therefore reluctant to invest in planting tomatoes duced at the farm level declines. Drought also affects and are resorting to other horticultural crops such the quality of silage, hay, and available forage, thus as leafy green vegetables, causing shortages of tomatoes increasing production costs at the farm level. Since in the domestic market. False codling moth is another feed costs represent over 70 percent of the variable new horticultural pest that is difficult to detect and has cost at the farm level, droughts severely affect dairy restricted the export of produce from Zimbabwe. farming returns to farmers. Yellow sugar aphid emerged in 2018, with wide- Other less severe weather-related risks for cattle are: spread occurrence and no registered pesticides, (1) erratic rainfall distribution, affecting the quality of although losses are still noted to be minimal at grazing but rarely causing high mortality if the amount 4 percent. Other emerging pests and diseases caus- received is sufficient for dams to be replenished and ing damage in sugarcane are the African sugarcane grasses to grow; (2) delayed onset of rains, affecting borer (Eldana saccharina), ratoon stunting disease, mostly quality and quantity of grazing pastures, being sugarcane smart, and black maize brittle, whose pop- more critical to cattle in arid regions; and (3) a short ulation is starting to increase due to mixed farming rainy season, affecting quality and yields of grazing systems among A2 farmers. (one in three to five years), with detrimental effects Other new pests and diseases may eventually on animal condition. In dairy production, rainfall arrive in Zimbabwe from neighboring countries. irregularities affect silage production. Maize and forage The most relevant threat is Maize Lethal Necrosis sorghum are major silage crops planted by dairy farm- Disease, which can cause very high losses in maize ers. If rainfall distribution is not optimal, these crops production and is present in Kenya and South Africa. will not do well and that will increase production costs. The viruses causing the disease cannot be controlled Since animals will be in poor condition because of using chemicals. Chilo saccariphagus, a moth that is inappropriate feeding caused by drought, they may 90 Zimbabwe: Agriculture Sector Disaster Risk Assessment Figure D.8.  Effect of Weather-Related Risks on Livestock Numbers (Headcount), 2009–17  2007/8 drought, Cattle stock 5.6 2015/16 severe drought hyper inflation and cause livestock death high death rate. 5.5 No drugs Cattle numbers in Millions 5.4 5.3 5.2 5.1 5 Disease outbreaks Tick borne disease, FMD and 4.9 droughts in region 4 and 5 4.8 2009 2010 2011 2012 2013 2014 2015 2016 2017 Period from 2009 to 2017 Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe. fail to cope with any disease outbreak if these events The dairy industry recovered in 1994/95 with a are associated with high disease incidence. Figures D.8 high yield of 223 million liters but was then severely and D.9 show how production risks over the years affected by the 1995/96 drought, in addition to the have affected livestock numbers and milk yields. effects of the adjustment program. From 1996, the A severe drought, coupled with the introduction downward trend was due to macroeconomic poli- of an economic and structural adjustment program cies, drought in 1997, and hyperinflation, which in 1991, were seen as the major events that caused the drove most dairy farmers out of business. The down- decline in milk production in 1991 to 1994 (Figure D.9). ward trend continued as the industry was decimated Figure D.9.  Effect of Weather-Related Risks on Milk Yields, 1987–2017 Dairy sector: Milk yield 235 Another drought in 1996-1997, hyper inflation issues cause by the $50000 which were given to war veterans by government 185 Milk yield in (Millions) Severe drought that occurred in - severe drought 135 2007 Land invasion and - shortage of redistribution caused foreign currency, many farmers to lose cash and farms and sell the dairy hyperinflation 85 cows. Some die due series of disease outbreaks and mismanagement by new farms 35 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: Ministry of Land Agriculture Rural Resettlement of Zimbabwe. Appendix D: Risks for Crop and Livestock Supply Chains 91 by a series of disease outbreaks and farm upheavals at the farm level means major losses in terms of milk following land redistribution in 2001 and 2002. quality, yield, and revenue. Since about 90 percent of the cattle herd is now found in communal areas in Zimbabwe, smallholder Poultry Industry farmers are most affected by the above risks. Weather-Related Risks Sanitary Risks Drought, disease outbreaks, price volatility, and limited Many diseases cause major losses in cattle produc- availability of drugs, remedies, and feeds are the main tion in Zimbabwe. Outbreaks of tick-borne diseases risks affecting the poultry industry. Because poultry are becoming more common, especially in November production is very intensive, with a high demand for through January. Both commercial and communal feed, drought is the major risk affecting the supply of farmers report that ticks are becoming more resistant feed ingredients (chiefly maize and soybeans). The to local dipping chemicals. An estimated 26,000 cattle risks listed below affect both indigenous communal died during the 2017/18 rainy season. Heartwater chicken production and commercial producers. (Ehrlichia ruminantium) and January disease (T.p. Production of day-old chicks has increased bovis) were the most common tick-borne diseases steadily over the years, with both minor and sharp that killed animals in previous seasons. drops occurring in drought years. There was a decline Foot and mouth disease is now a perennial chal- in poultry production in 1991/92, 1997/98, and lenge, and no vaccines are produced locally. Because of 2000-02, and a sharp decline in 2006-09. From 2009 the uncontrolled movement of cattle, the chance of a the industry resumed steady growth, with minor set- foot and mouth outbreak are very high, and it will have backs in 2011 and 2013. A major shock occurred in devastating effects on dairy farming if it occurs. Because 2015, however, when production of day-old broiler of the increase in stray dogs, rabies outbreaks also chicks dropped from above 70 million to less than have a high chance of occurring. Rabies is a zoonotic 40 million in response to a drought that reduced feed disease, meaning that it can be transmitted to humans. availability. Broilers are very sensitive to temperature A high rate of beef measles was reported by meat changes, and an increase in temperature is associ- inspectors, especially in animals from areas with ated with high mortality. Variations in the poultry poor sanitation. The incidence of carcass contamina- meat price are clearly related to changes in stocks of tion has increased, and farmers risk losing payment animals (Figure D.10). for the whole carcass once the disease is detected. As in the other subsectors, uncertainty about In dairy production, mastitis is a major challenge. foreign exchange availability is an enabling envi- There are no vaccines for this disease, and its spread ronment risk. Figure D.10.  Poultry Meat Supply and Stocking Capacity, 2015–17 3,500 $3.70 3,000 $3.50 2,500 Wholesale price Metric tonnes 2,000 $3.60 1,500 $3.10 1,000 $2.90 500 – $2.70 Jan-15 Mar-15 May-15 Jul-15 Sep-15 Nov-15 Jan-16 Mar-16 May-16 Jul-16 Sep-16 Nov-16 Jan-17 Mar-17 May-17 Jul-17 Sep-17 Nov-17 Stockholding Broiler meat Wholesale price Source: Zimbabwe Poultry Producers Association. 92 Zimbabwe: Agriculture Sector Disaster Risk Assessment Notes usually as a percentage. A greater value of the coefficient of variation indicates that greater heterogeneity is present in 1. The coefficient of variation measures the relation- the values of the variable, and the smaller the coefficient ship between the variability of the variable (standard of variation, the more homogeneous are the values of the deviation) and the size of the arithmetic average; expressed variable.