The Union of the Comoros: Jumpstarting Agricultural Transformation Agriculture Sector Review Public Disclosure Authorized

Home , Anjouan, La Grille, Mount Karthala

THE WORLD BANK GROUP Public Disclosure Authorized

Public Disclosure Authorized

Report No: P164195

Comoros Agriculture Sector Review (P164195)

Public Disclosure Authorized

Standard Disclaimer:

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.

The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly.

For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522- 2422, e-mail [email protected].

Contents Acknowledgements ...... 8 Acronyms and Abbreviations ...... 9 1 Executive Summary ...... 10 1.1 Pillar 1: Public Actions to Boost Agricultural Transformation...... 15 1.2 Pillar 2: Boosting Food Production and Smallholder Access to Markets ...... 18 1.3 Pillar 3: Export Promotion of Established and New Products ...... 21 2 Introduction ...... 23 2.1 Country Context ...... 25 2.1.1 Geography and Demography ...... 25 2.1.2 Structural Features of the State and Economy ...... 26 2.2 Sector Context ...... 30 3 Enabling Environment for Agriculture ...... 33 3.1 Complementary Infrastructure ...... 34 3.1.1 Road Network ...... 34 3.1.2 Inter-island Maritime and Air Transport...... 35 3.1.3 Market Infrastructures ...... 36 3.2 Investment Climate and Trade ...... 36 3.2.1 Access to Finance ...... 37 3.2.2 Access to Energy...... 39 3.2.3 Access to Land ...... 41 3.2.4 Trade Policies and Business Environment for Agriculture...... 41 4 Sources of Demand ...... 44 4.1 Size of the Food and Beverage Market ...... 44 4.2 Export Market Trends ...... 47 4.3 Demand from Hospitality Sector...... 49 5 Sources of Supply – I: Subsistence Production Systems ...... 51 5.1 Typology of Production ...... 51 5.2 Production Performance – Snapshot and Trends ...... 53 5.2.1 Cereals...... 55 5.2.2 Other food crops ...... 58 5.2.3 Fisheries ...... 59 5.3 Priority Actions for Subsistence Production Systems ...... 61 6 Sources of Supply – II: Agricultural Value Chains ...... 63

6.1 Introduction ...... 63 6.2 Established Value Chains ...... 65 6.2.1 Cloves ...... 65 6.2.2 Vanilla ...... 67 6.2.3 Ylang-ylang...... 70 6.3 Growing Value Chains ...... 72 6.3.1 Poultry ...... 73 6.3.2 Dairy ...... 77 6.4 High Potential Value Chains ...... 82 6.4.1 Horticulture ...... 82 6.5 Priority Actions for Value Chains ...... 86 7 Institutional Environment for Agriculture ...... 88 7.1 Strategic Framework ...... 89 7.2 Institutional Structure at the National Level ...... 92 7.3 Governorates ...... 93 7.4 Agriculture Institutions ...... 93 7.5 Non-Governmental Organizations ...... 95 8 Summary of Recommended Actions and Expected Outcomes ...... 95 References ...... 102 Annex A: Supplementary Figures ...... 105 Annex B: Supplementary Tables ...... 114 Endnotes ...... 124

List of Tables

Table 1-1: Public Actions to Boost Agricultural Transformation ...... 16 Table 1-2: Boosting Food Production and Access to Markets for Smallholders ...... 19 Table 1-3: Export Promotion of Established Value Chains and New Products ...... 22 Table 2-1: Top 5 Agricultural Import Commodities, 2015 ...... 29 Table 2-2: Top 3 Agricultural Export Commodities, 2015 ...... 29 Table 2-3: Benchmarking the Size of the Agriculture Sector, 2014 ...... 31 Table 4-1: Size of the Market for Food and Beverages in South-West Indian Ocean ...... 44 Table 4-2: Budget Shares, by Food Segment...... 46 Table 4-3: Market Shares of Domestic Producers, Product-level...... 47 Table 4-4: France & USA import prices, ylang-ylang oil, 2011-2017, US$ & €/kg ...... 49 Table 4-5: New Industrial Export Products and Potential Markets ...... 49 Table 4-6: Evolution of the Price Index (base=100 in 1999)...... 50

Table 5-1: Summary of Agricultural Production Systems, 2004 ...... 51 Table 5-2: Cropping calendar in the Comoros ...... 53 Table 5-3: Crop Production – A Snapshot from Most Recent Data Sources...... 54 Table 5-4: Cereal Yields in Neighboring Comparators, MT/ha (2016)...... 56 Table 5-5: Demand and Sources of Supply for Cereals ...... 56 Table 5-6: Priority Actions for Enhancing Food Security and Incomes ...... 62 Table 6-1: Value Chains at a Glance ...... 64 Table 6-2 : Number of Farms and Animals Per Island ...... 77 Table 6-3: Import of Dairy Products in 2013 (Source FAOSTAT) ...... 78 Table 6-4 : Number of farms, animals and production per island ...... 81 Table 6-5: Vegetable production from 2008 to 2011 (MT) ...... 82 Table 6-6: Inputs, Outputs and Gross Margins of Some Horticultural Productions (KMF/10 acres) ...... 85 Table 6-7 Priority Actions to Develop Value Chains ...... 86 Table 8-1: Public Actions to Boost Agricultural Transformation ...... 97 Table 8-2: Boosting Food Production and Access to Markets for Smallholders ...... 99 Table 8-3: Export Promotion of Established Value Chains and New Products ...... 100 Table 0-1:Trends for Selected Macroeconomic and Poverty Indicators ...... 114 Table 0-2: Screening Indicators: Export Crop Value Chains ...... 115 Table 0-3: Screening Indicators: Livestock Value Chains ...... 117 Table 0-4: Ylang-ylang prices in Comoros, KMF ...... 118 Table 0-5: France & USA import prices, ylang oil, 2011-2017, US$ & €/kg ...... 118 Table 0-6: Ylang grower's production model ...... 118 Table 0-7: Ylang-ylang Oil Distiller’s Production model – wood, unimproved furnace ...... 118 Table 0-8:Ylang-ylang Collector's Production Model ...... 119 Table 0-9: Ylang Collector's Production Model (Improved Technology) ...... 120 Table 0-10: Ylang distiller's Production Model (Improved Technology) ...... 120 Table 0-11: Production Model for a Medium-Sized Poultry Farm...... 121 Table 0-12: Screening Indicators: Fresh Fruit and Vegetables ...... 123

List of Figures

Figure 2-1: The Union of Comoros ...... 25 Figure 2-2: Demographics – Comparing Comoros with the World ...... 26 Figure 2-3: Agriculture-led Economic Growth ...... 27 Figure 2-4: Shares by Major Employment Sectors of the Household’s Head (in Percentage) ..... 27 Figure 2-5: Correlation between Income and Food Imports in SIDS ...... 28 Figure 2-6: Dimensions of Natural Disaster Risk ...... 29 Figure 2-7: Benchmarking the Size of the Agriculture Sector ...... 30 Figure 2-8: Benchmarking the Performance of the Agriculture Sector, 2014 ...... 31 Figure 2-9: Agricultural Trade Trends ...... 32 Figure 2-10: Agriculture and GDP Growth ...... 32 Figure 3-1: Network Condition of The Three Islands ...... 35 Figure 3-2: Average Expenditures on International Transport as a Percentage of the Value of Imports for Small-Island Developing States, 2004-2013 ...... 36

Figure 3-3: Benchmarking Financial Access in Comoros ...... 37 Figure 3-4: Lending in Comoros, by Industry ...... 38 Figure 3-6: Bank Branches and MFIs (Number), 2015 ...... 38 Figure 3-5: Financial Sector Assets & Liabilities by Share of GDP, 2015 ...... 39 Figure 3-7: Distance to Frontier - Getting Electricity ...... 40 Figure 3-8: Solar Potential in the Comoros ...... 40 Figure 3-9: Trading across Borders in Comoros and Comparators - Rank and DTF ...... 42 Figure 3-10: Trading across Borders - Time and Cost ...... 43 Figure 4-1: Comoros is an Outlier to Engel’s Law ...... 45 Figure 4-2: Budget Shares, by Food Segment ...... 46 Figure 4-3: Major Markets for Vanilla Exports, 2008-2016, MT...... 47 Figure 4-4: Major markets for Clove Exports, 2008-2016 ...... 48 Figure 5-1: Shares of Areas Harvested, by Crop Categories ...... 54 Figure 5-2: Agricultural Growth Decomposition in SIDS ...... 55 Figure 5-3: Benchmarking Cereal Yields Figure 5-4:Cereal Production – Comoros ...... 56 Figure 5-5: Benchmarking Rice Yields, hg/ha ...... 57 Figure 5-6: Rice Production Trends – Comoros ...... 57 Figure 5-7: Area Harvest for Major Non-Cereal Crops ...... 59 Figure 5-8: Yields for Major Non-Cereal Crops...... 59 Figure 5-9: Fish Trade Balance, by Volume (MT) ...... 60 Figure 5-10: Fish Trade Balance, by Value (‘1000 USD) ...... 61 Figure 6-1: Clove exports from major origins, 2008-2016, MT ...... 66 Figure 6-2: Vanilla Exports from Major Origins to Major Markets, 1999-2015, MT...... 68 Figure 6-3: Ylang-ylang Exports from Comoros & Mayotte, 2008-2014, MT ...... 70 Figure 6-4: Margins across Ylang-Ylang Value Chain ...... 71 Figure 6-5: Margins across Improved Ylang-ylang Value Chain ...... 72 Figure 6-6: Estimated Quantities of Poultry Meat by Origin (MT per year) ...... 73 Figure 6-7 : Local Poultry Value Chain Map – Main Actors per Level ...... 74 Figure 6-8: Production Cost of a Locally Produced Egg – Distribution by Type of Expenditure 76 Figure 6-9: Animal Ownership, by Island ...... 78 Figure 6-10: Dairy Value Chain Segments ...... 78 Figure 6-11: Retail Price of a Litre of Liquid Milk According to its Origin (KMF) ...... 81 Figure 6-12: Horticulture Value Chain Actors ...... 83 Figure 6-13: Structure of Production Costs ...... 85 Figure 6-14: Break-even Point and Average Selling Price of Horticultural Products ...... 86 Figure 7-1: Strategic Framework for Agriculture ...... 89 Figure 0-1: Benchmarking Trends in Overall CPIA Scores ...... 105 Figure 0-2: Progress in Disaggregated CPIA Scores ...... 105 Figure 0-3: Poverty and Consumption Trends in Comoros ...... 105 Figure 0-4: A snapshot of the job market in Comoros ...... 106 Figure 0-5: Agricultural Trade Deficits in SWIO Countries ...... 106 Figure 0-6: Overall Trade Trends ...... 106 Figure 0-7: Benchmarking Livestock Production per Capita ...... 107 Figure 0-8: Benchmarking Yields of Major Crops ...... 107 Figure 0-9: Comparison of cost to export (US$ per 20-ft container) ...... 107 Figure 0-10: Comparison of port tariffs (US/TEU) ...... 108

Figure 0-11: Road Density per Land Area (km/1,000 km2) ...... 108 Figure 0-12: Road Density per Capita (km/10,000 Inhabitants)...... 108 Figure 0-13: Rankings on Doing Business Topics (2018) ...... 109 Figure 0-14: Distance to Frontier (DTF) on Doing Business Topics (2018) ...... 109 Figure 0-15: Madagascar vanilla prices (nominal), 1997-2017, US$/kg ...... 109 Figure 0-16: Madagascar clove prices (nominal), 1997-2017, US$/MT ...... 110 Figure 0-17: Seasonal Price Volatility for Fresh Fruits and Vegetables ...... 110 Figure 0-18: Export Crop Areas in Moheli ...... 110 Figure 0-19: Export Crop Areas in Anjouan ...... 111 Figure 0-20: Export Crops Areas in Grande Comore ...... 111 Figure 0-21: Main Dairy Production Areas in Comoros ...... 112 Figure 0-22: Main Poultry (Layer and Broilers) Production Areas ...... 112 Figure 0-23: Horticulture Production Areas in Comoros ...... 113

CURRENCY EQUIVALENTS December 2017 1 EUR = 490 KMF 100 KMF = 0.204 EUR 1 USD = 415 KMF 100 KMF = 0.240 USD

Acknowledgements This report has been prepared by a team led by Ashesh Prasann (Agriculture Economist, GFA07). The core World Bank team comprised Norman Piccioni (Senior Agriculture Economist, GFA07), Joanne Gaskell (Senior Agriculture Economist, GFA07), and Lorenzo Bertolini (Senior Private Sector Specialist, GFCAC). The background value-chain analyses for export commodities, livestock products, and fruits and vegetables were conducted by Stephen Caiger (Sibelius Natural Products), Alban Bellinguez (FAO-TCIA), and Jean-Charles Heyd (FAO-TCIA) respectively.

The team gratefully acknowledges the overall guidance of Mark Lundell (Country Director, AFCS2), Dina Umali-Deininger (Practice Manager, GFA07), Rasit Pertev (Resident Representative, AFMKM), and Alberta Mascaretti (Chief, Africa Service, FAO Investment Center). The team would like to thank the peer reviewers Adesimi Freeman (Lead Private Sector Specialist, GFCAC), Eli Weiss (Senior Agriculture Economist, GFA02), and Irina Schuman (Senior Agriculture Economist, GFA07) for their constructive and insightful comments. Finally, we would also like to thank Antoissi Said Ali Said (Operations Officer, AFMKM), Sitti Fatouma Ahmed (Program Assistant, AFMKM), Marie Lolo Sow (Program Assistant, GFA07), and Bodomalala Rabarijohn (Program Assistant, GFA07) for excellent operational support.

We would also like to thank the many stakeholders in Comoros (government authorities, private sector representatives, civil society organizations, other development partners), who have contributed to the preparation of this document with their inputs, knowledge and advice.

The work was made possible through a collaboration between the World Bank and the Investment Center Division of the Food and Agriculture Organization of the United Nations (FAO/TCIA) as part of the FAO/ World Bank Cooperative Program.

Acronyms and Abbreviations

ACEFER Improvement of the competitiveness of exports and income streams AU African Union CAADP Comprehensive Program for The Development of African Agriculture CAPAC Central Purchasing of Agricultural Professionals of The Comoros CET Common External Tariff CFAE Center Administrative Formalities for Businesses CIRAD Agricultural Research for Development International Centre CNH National Horticulture Center COMESA Common Market for Eastern And Southern Africa CRDE Regional Economic Development Center DNSAE National Agricultural Strategy and Livestock Directorate EAC East African Community EU European Union FAO Food And Agriculture Organization FNAC National Federation of Comorian Farmers FTA Free Trade Area IFAD International Fund for Agricultural Development IITA International Institute of Tropical Agriculture INRAPE National Research Institute for Agriculture, Fishing and The Environment KMF Comoros Franc MECK Mutual Savings and Credit Is Komor MAPEEIA Ministry of Agriculture, Fisheries, Environment, Energy, Industry and Handicrafts NAIP National Agricultural Investment Program NEPAD New Partnership for African Development NGO Non-Governmental Organization OP Organization of Producers OPA Agricultural Professional Organization RGA General Agricultural Census SADC Southern African Development Community SCA2D Accelerated Growth of Sustainable Development Strategy SNAC National Union of Comorian Farmers SWOT Strength, Weaknesses, Opportunities, Threats TFP Technical and Financial Partner UNDP UN Development Programme USD United States Dollar VAT Value Added Tax VC Value Chain

1 Executive Summary

This report is the World Bank’s first Agriculture Sector Review (ASR) for the Union of the Comoros (henceforth “Comoros”), the first analytic product in the Bank’s reengagement with the sector after a gap of two decades. In the short-term, the findings from this ASR are intended to launch a participatory and inclusive consultation process that feeds into an integrated sector strategy and investment plan, which then charts the path towards agricultural transformation. In the long term, this review aims to provide the Government of Comoros with the evidence base necessary for strategic policymaking and smart investments critical to jumpstarting agricultural transformation and accelerate progress towards its stated goal of becoming an emerging country by 2030.

Endowed with fertile volcanic soil, a gradient of altitudes and micro-climates, and a colonial legacy of export crop plantations and related factories, the economy of the archipelago of Comoros has historically been dominated by agriculture. Since 2000, Comorian agriculture has led slow economic growth of 2.6 percent per annum. Accounting for 38 percent of jobs and 36 percent of GDP, the relative size of the sector ranks in the 94th percentile globally and is an anomaly among Small Island Developing States (SIDS). A narrow base of three export crops (cloves, vanilla, and ylang-ylang) provide 90 percent of export income, with nascent attempts to diversify and revive older export products like copra, coconut oil and coffee.

However, agricultural growth during the last two decades has been driven by an unsustainable strategy of bringing more marginal forest land under cultivation, rather than improvements in farm productivity or value addition. At least 72 percent of farmers are subsistence producers with weak linkages to a rapidly growing domestic market for food and a diet shifting towards meat, vegetables and dairy. The remaining producers form the base of three export-oriented value chains and are highly exposed to volatility in global markets in which Comoros’s price-setting power is low and waning. There are signs of labor exit from primary production, especially among educated youth, but the absence of off-farm and non-farm economic opportunities has translated into a high degree of labor force inactivity and youth unemployment.

Given finite land and freshwater resources, rapid population growth, and decreasing competitiveness against food imports, the limits to Comorian agriculture’s growth pattern are clear. In fact, the ability of Comorian agriculture to continue to provide growth, jobs, food security and critical export revenues is already threatened. Food imports are surging as weak transport infrastructure and unorganized marketing segment a small domestic market and delink producers from growing modern retail. Export production is caught in a pattern of catch-up to price fluctuations and is unable to make strategic investments in value-addition technologies. Nascent producer cooperatives, small and medium agribusinesses, and young entrepreneurs face high barriers to entry, with finance, energy and land constraints locking them out from potentially profitable and job-creating investments. The risks of natural disasters have grown, and it is estimated they reduce the GDP by 1 percent on average every year, exerting a disproportionately negative impact on agricultural producers.

Simultaneously, the potential for a different growth path, focused on agricultural transformation – raising farm productivity and strengthening linkages between farms and the rest of the economy – is also evident. Yield-improving investments in climate-smart inputs, modern production and processing technologies, and sustainable practices can deliver large food security gains and build resilience to environment risks. Investments that facilitate internal trade will enhance the competitiveness of crop and livestock perishables, in which Comoros’s remoteness becomes a buffer

10 against shipped imports, bring local supply in sync with market demand, raise farm incomes and lower the food import bill. Export promotion in the form of downstream industrial development of existing products, facing favorable market conditions after a decade long downturn, can allow Comoros to tap into newer markets and enable producers to capture greater value domestically. Already, this vision is reflected in the National Industrial Strategy of Comoros (2017), in which all of the domestic and export products identified for short-term industrial development are agriculture-derived products.

Given agriculture’s large footprint and the limited set of investment opportunities in Comoros’s economy, it represents the clearest path to economic transformation, the long-term reallocation of labor from low to high-productivity activity. In fact, economic transformation in Comoros does not imply an exit from agriculture but building on it to modernize farming and strengthen efficiency enhancing linkages between agriculture and nascent sectors like manufacturing and tourism. On one hand, agricultural producers will be assured of a domestic market, if they are able to meet its quality and quantity requirements. On the other hand, manufacturing in the form of agro-processing would have guaranteed inputs in an economy with small endowments of non-agricultural raw materials. In the tourism sector, the expected arrival of new airline carriers, foreign tourists and international hotel chains will raise the demand for high-quality, locally sourced perishables like tropical fruits, vegetables and fish.

Consequently, building up agro-processing and ancillary post-harvest services (storage, distribution, transport, logistics, retailing, preparation for restaurants and hotels) has a high likelihood of success in generating economic activity, fostering formal enterprises, and creating remunerative jobs that draw skilled labor away from inactivity in Comoros. In Tanzania and Malawi, 40 percent of manufacturing and services jobs are now generated by the agri-food system. Like Mauritius, Comoros also has the ingredients to leverage its agricultural base to power economic transformation. Unlike Mauritius, the clearest path to economic transformation runs through non-textile based agro- processing. Coupled with climate-smart agriculture investments on farms operated like businesses, a successful agricultural transformation process thus has the potential to deliver on the jobs, food security, competitiveness, and natural resource management objectives identified by the Ministère de l’Agriculture, de la Pêche, de l’Environnement, de l’Aménagement du Territoire et de l’Urbanisme (MAPEAU).

This review casts the analytical net both wide and deep to conduct a comprehensive overview of sector opportunities that can contribute to agricultural transformation, identify the binding constraints that hold back such potential, and outline policy and investment options that can unlock it. Specifically, the ASR is structured to provide the following: i) an outline of structural constraints and the state of play for the enabling environment of agriculture, ii) an assessment of the current and projected sources of demand, supply, and market trends for agricultural products, iii) a synthesis of deep-dive studies of established, growing and high potential agricultural value chains, iv) a description of the strategic framework, institutional structure and other sector actors, and v) a proposal of specific policy options and investments that are likely to have systemic impacts on food security and competitiveness. Some of the major findings from the ASR are:

Sector Context

Agriculture-led Economic Growth

5.0 4.0 3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 -4.0 -5.0

Agriculture Industry Tertiary

Source: INSEED and World Bank staff calculations and estimates

 The fast-growing population of Comoros is young, rural, and out of work. Creating high quality and remunerative rural jobs (farm and off-farm) are critical to ensuring a demographic dividend.

 The agriculture sector has an outsized footprint, ranking in the 94th percentile globally. In fact, agriculture is the second-largest sector, the largest employer, and provides for 90 percent of export income, an outlier among SIDS.

 Since 2000, agriculture has led slow economic growth but agricultural transformation has stagnated. Nearly all growth in the 2001 – 2013 period was driven by input expansion, rather than improvements in yield or TFP. Historical linkages between agriculture and industry (eg. coconuts and soap making) are now weakened and new linkages have not been developed.

 Large current account deficits are chronic, with food imports being a major contributor. Even among import reliant SIDS, Comoros is an outlier, importing nearly 70 percent of its food requirement by volume. By value, food imports comprised 44 percent of merchandise imports and 29 percent of GDP in 2015.

Enabling Environment for Agriculture

Rankings on Doing Business Topics (2018)

 Weak complementary infrastructure segments the internal market. Poor rural connectivity weakens linkages between coastal markets and highland production basins. An informal maritime transport system and lack of market infrastructure segments an already small internal market into three markets.

 Land, credit and energy constraints are more binding for agriculture than other sectors. Land access for commercial agriculture is hindered by insecurity generated by three competing tenure systems. Agricultural lending is limited by collateral requirements and low appraisal capacity among MFIs. Insufficient and unreliable electricity hampers small-scale processing.

 Applied input tariffs are higher than published rates, raising the costs of agricultural inputs and introducing production uncertainty. The time and cost of obtaining tariff exemptions leads to importers paying higher tariffs, passing on costs to producers and delays in input supply. This lowers the use of improved seeds, fertilizers, equipment, and animal feeds, consequently lowering yields and competitiveness of marketed production.

Sources of Demand

 The Comorian food and beverage market is $570 million, almost as large as the Food Expenditure Shares by Segment economy. Relative to SIDS of comparable size 25 and income, consumption and food 25 19

20 17 expenditure are high in Comoros. 16

15 Percent  Diet transformation is already underway, 10 5.3 4 5 3.5 3.2 3.1 with spending on crop and livestock 1.6 .18 perishables now surpassing cereals. The 0 Cereal Vegetables segmentation of Comoros’s food and beverage Meat Fish market now resembles the middle-class diet in Fruit Dairy Condiments Sugar developing countries, with growing domestic Oil Beverage demand for perishables and processed foods. Coffee Source: 1-2-3 Survey (2014)

 Local food production is inadequate and out of sync with market demand. Despite signs of portfolio transition from rice and coconuts to tubers and pulses, food production has been unable to align itself with the demand changes accompanying the diet transformation.

Sources of Supply

 Cassava and maize have displaced coconuts and rice as the main food crops. Crop Yields In national data sources, cassava output is now the highest among all crops, replacing 20 coconuts. Rice, the main staple, is now 15 barely produced in the country and maize is the main cereal. 10 5  Introducing high-yielding yellow

cassava can produce long-lasting food kg/ha 1000 Yields, 0 security gains. Given widespread Comoros Fiji Mauritius Samoa cultivation and consumption of cassava Bananas Coconuts Maize Rice, paddy and its forecasted climate suitability, technology transfer of biofortified yellow cassava has high potential for rapid delivery of food security and nutrition gains.  Cash crop production employs 18 – 28 percent of producers and is expanding in response to surging prices. Strategic investments in quality control, downstream products, and downstream development are required to mitigate against an impending downturn.  Smallholder growers of cloves and ylang-ylang are receiving the majority shares of export prices (78 and 62 percent respectively) in currently favorable market conditions. Vanilla growers receive much smaller shares of export prices, as their bargaining power is reduced by an exporter oligopoly and a government determined price structure.  Egg and dairy value chains are growing and are already competitive. While eggs are now the most competitive food value chain in Comoros, dairy is developing from a small base in Anjouan and is competitive with a segment of the milk market.

 At present, 99 percent of consumed chicken meat is imported, driven by a 50 – 100 percent price differential. Like several African countries, frozen chicken wings – a residual from European demand for lighter chicken meat – are imported at low prices, and local producers are unable to compete.

 Developing a market segment for domestic chicken meat requires public investments in nutrition awareness, feed production and a robust veterinary system. Shifting consumer preferences towards light meat through a sustained nutrition campaign is required to change current demand patterns. Stimulating supply requires the development of local feed production, with linkages to maize producers and fisheries. Lastly, provision of veterinary care and inexpensive vaccination is critical for mitigating against disease risks.

 Domestic fisheries remain artisanal, leading to low catches and rising fish exports. Full-time and subsistence fishers largely use wooden canoes, basic fishing gear and rudimentary methods, leading to annual catches below sustainable levels, even as growing demand is met through fish imports.

 Vegetables and fruits are seasonally competitive in geographically segmented markets, but greater domestic competitiveness is constrained by erratic input supply, poor inter-island connectivity, and weak linkages with a growing retail sector.

Institutional Environment

 At present, the sector lacks a validated strategy that outlines the path toward achieving the government’s commitment to agricultural transformation. Comoros has launched the process of implementation of CAADP but does not have an updated and validated National Agriculture Strategy. Consequently, it does not have a National Agricultural Investment Program (NAIP), has not signed the compact, and its progress towards CAADP goals could not be evaluated for the Africa Agricultural Transformation Scorecard (AATS, February 2018).

 The institutional structure governing agriculture is decentralized and complex, but service delivery is poor due to the absence of sustainable revenue-generating models. Public investment in agriculture has been low for several years, with almost all agricultural support initiatives relying on donor-funded projects. Due to the absence of revenue generating models, agricultural extension and veterinary technicians are underpaid and seek work in donor-funded projects or offer their services privately.

The main forward-looking messages that have emerged from this review are grouped under three pillars around which an integrated sector strategy and investment plan can be designed: i) public actions to boost agricultural transformation, ii) boosting food production and smallholder access to markets, and iii) export promotion of established and new products. The key priorities and recommendations from these pillars are summarized in the sections and tables below:

1.1 Pillar 1: Public Actions to Boost Agricultural Transformation

 Strategic public investments supporting crop and livestock perishables will bring sector in sync with market demand, raise farm incomes and lower the food import bill. Comoros’ remoteness and long shipping times are a buffer for locally produced perishables – eggs, milk, vegetables and fruit – high-value products that are already competitive seasonally and in segmented markets. Strategic investments in extension and animal health services can accelerate adoption of modern inputs, and build resilience to pests and animal diseases, the latter having already caused debilitating effects to the local cattle population in 2009.

 Structuring the regulatory simplification agenda around the Enabling Business of Agriculture (EBA) indicators to accelerate agricultural investments. The establishment of simplified, low-cost regulatory systems for seeds, fertilizers, machinery, finance, markets, transport, information and communication technology (ICT), and water can have a catalyzing effect on the investments and performance of key players across agricultural value chains. The collaboration between ANPI and IFC to improve Doing Business indicatorsi can be built upon, with Enabling the Business of Agriculture (EBA) indicators providing an analogous tool for policy action and quantitatively monitoring regulatory changes over time, and allowing cross-country comparisons.

 Rationalize list of high priority food imports. At present, a list of high priority food imports benefit from lower tariffs because of their high domestic consumption and the disproportionate effect of their prices on consumers. This approach weakens the competitiveness of domestic producers over time and exposes consumers to greater risk of international price shocks. The inclusion of rice – now barely produced in Comoros – on this list is reasonable, given little adverse impacts on domestic producers. However, the inclusion of eggs – extremely competitive with imports – is not. A smart targeting approach, based on an assessment of existing production capacity, domestic competitiveness, and potential pass-through effects of higher tariffs needs to be adopted urgently.

 Strengthening the agriculture advisory and extension system. Equipping the recently established agriculture advisory and extension system (CRDEs) with the human, physical and financial resources to rapidly disseminate inputs and technical assistance related to climate-smart agriculture is critical for the delivery of key services and successful implementation of public investment projects in the sector.  Support to land administration services to promote greater tenure security. Transitioning to a market-based land regime is unlikely in the short-term, given the complexity of three overlapping land tenure systems (customary, Islamic and colonial) and competitive jurisdictional authorities (community leaders, Islamic judges and the land administration). However, greater land tenure security can be achieved if land administration services are strengthened to undertake participatory titling in production basins. This process would consist of consensual demarcation, multi-stakeholder land recognition committees that facilitate a small-scale cadaster census, and land use planning.

 Establishment of an agricultural data unit. The establishment of an agricultural data collection, monitoring and planning unit is critical for making strategic and targeted investments in the sector. The first and last agricultural census in Comoros was conducted in 2004. There are no recent survey based estimates of output, harvested area, yields, number and type of farms, agribusinesses etc., a knowledge gap reflected in large discrepancies across macro estimates of production. Granular, geo-referenced data capturing island-level heterogeneity is unavailable. It is thus difficult to plan, monitor and evaluate investment projects at present. As a first step, a well-designed and validated Agricultural Census is urgently required to update public understanding of the sector.

Table 1-1: Public Actions to Boost Agricultural Transformation

PRIORITIES AND MEASURES Direct Role of Public Timeframe Beneficiaries Sector

Long Long Short Short Policy Policy maker maker maker Market Market Medium Medium Financer Financer Convener Convener RANSFORMATION T PRIORITY 1: Improving access to climate-smart technologies, innovations, management practices (TIMPs), and advisory services (i) Join regional agricultural research and technology transfer partnerships to develop, transfer, and Crop and Policy maker Medium GRICULTURAL GRICULTURAL

A disseminate TIMPs that enhance productivity, livestock Financer Medium climate resilience, and nutritive value by: producers Convener Long

OOST OOST  harmonizing policies, laws, and regulations to Convener Short B accelerate transfer of off-the-shelf technologies Export crop  establishing national center of specialization to producers accelerate development in target crop or livestock product(s) Private seed CTIONS TO TO CTIONS

A  enabling private sector involvement in technology and animal development (seed multiplication, animal breeders breeding) UBLIC UBLIC

P  fostering producer-retailer dialogue that enables Retailers farmer organizations to adapt to market demand

(ii) Improved access to crop and livestock extension services by: Crop and Financer Medium  equipping public advisory and extension services livestock Policy maker Short (CRDE) physical and financial resources producers Convener Medium  promoting technical assistance through productive alliances between cooperatives and anchor Export crop investors producers  enabling development of privately managed dairy and poultry hubs to disseminate livestock inputs, Extension animal vaccines and veterinary services agents

Veterinary technicians PRIORITY 2: Enable private investments in agriculture by: i) improving the investment climate for agriculture, ii) strengthening land administration services to promote land tenure security, and iii) building capacity to collect, monitor and disseminate real-time market information. (i) Improving investment climate and matching the supply Producer Policy maker Medium and demand for investment capital organizations Market maker Short  structure the regulatory simplification agenda Market maker Medium around the Enabling Business of Agriculture (EBA) Agribusiness indicators startups and  facilitate supply of agricultural credit by providing SMEs MFIs training on appraisal of agribusiness business plans Microfinance  accelerate the development of an incubation institutions ecosystem with linkages between universities and private sector representatives for preparation of Private sector viable business plans, early-stage mentoring, and representatives linkages to MFIs Universities (ii) Strengthening land administration services to promote Food Financer Short land tenure security by: producers Financer Medium  equipping land administration office with Convener Medium geospatial tools and technical assistance Export crop Policy maker Medium  conducting a participatory titling program in producers production basins  forming multi-stakeholder land recognition Producer committees to facilitate demarcation and small- organizations scale cadaster census  developing land-use framework for areas where Small and both food and export crops are cultivated medium agribusiness (iii) Build capacity to collect, monitor and disseminate real- Food Financer Short time agriculture data by establishing an agriculture producers Financer Medium data unit to: Convener Short  conduct an agricultural census to fill knowledge Export crop gaps, better policy making and enable private producers investments  utilize remote sensing tools to monitor and respond Producer to weather, yields, prices, pests and diseases in real organizations time  organize the development of mobile applications to Small and disseminate real-time weather and market medium information agribusiness PRIORITY 3: Reorient role in determination of import tariffs and distribution of food, to enhance domestic competitiveness and free up public resources for investments

(i) Implement a smart targeting approach by:  assessing pass-through effects of tariffs on Crop and Policy maker Short consumer prices and welfare against existing livestock Policy maker Short competitiveness and supply capacity of domestic producers production  enabling annual reassessment of list by developing Consumers criteria-based determination and monitoring of agricultural production and import trends (ii) Reorient ONICOR to the role of a market coordinator and facilitator Food Policy maker Medium  importing and distribution of rice should be importers Market maker Medium handled by established private enterprises in the rice import business Public sector  activities should be restricted to setting and applying standards for imported rice, market Consumers monitoring, collection of duties and sales taxes, and maintaining a modest emergency stock Key: (1) Direct beneficiaries are identified based on the expected first-order effects. (2) The role of public sector lists the primary role of the public sector, and it is not restrictive. (3) Timeframe is classified as Short-term for actions that are implementable in 0 – 2 years, Medium-term for 3 – 5 years, and Long-term for more than 5 years.

1.2 Pillar 2: Boosting Food Production and Smallholder Access to Markets

 Enforcement of import exemptions on agricultural inputs. Cumbersome tariff exemptions and a cash-strapped importer association introduce uncertainty in input supply through delays, thus reducing smallholder use of modern inputs. Public and private input importers report tariffs of up to 9 percent instead of the exempt rate of 1 percent, raising costs when producers can afford input purchases. Enforcement of import exemptions needs to be coupled with a revamp of the business model used by importers, to ensure greater input access and competitiveness.

 Joining regional agricultural research and technology transfer partnerships. The transfer of off-the-shelf resilient and biofortified varieties, planting material, seeds and breeds from Regional Centers of Excellence (RCoE) supported by the World Bank would enhance productivity, adaptation and food security without requiring Comoros to make duplicative financial and human resource investments in research and development. It would have the added benefit of harmonizing Comoros’ certification policies and regulations, phytosanitary requirements and import control protocol with neighboring countries, thus integrating domestic value chains into a regional market.  No-regret investments in improved, climate-smart varieties of maize, cassava and pulses are critical to ensure food security and resilience to environment shocks. While production data are tentative, they suggest that subsistence-oriented production has shifted towards maize, cassava and pulses over the last decade. To ensure food security, it is critical that reorientation towards market-oriented perishables is coupled with the transfer of climate-smart inputs, production and processing technologies, and sustainable practices to support smallholder producers.

 A broad-reaching product organization and development strategy is necessary to ensure all- year supply of fresh fruits and vegetables. The market for fresh fruit and vegetables is growing, but any one product market is too small to justify investment. For vegetable value chains, the priority is to integrate from production to commercialization by upgrading associations, which are already dynamic, to professional organizations. For fruit value chains, initial interventions should focus on the introduction and multiplication of fruit and seedlings adapted to the local eco-climate context in nurseries, and then building linkages from nurseries to horticulture value chains.

 Public investments in nutrition awareness, feed production and animal health can lead market development and partially substitute imports for domestic poultry meat. Consumer preferences for dark meat drive the demand for cheap, imported frozen chicken wings in Comoros. A public nutrition awareness campaign could shift demand towards more expensive light meat, against which indigenous broiler meat is likely to be more competitive. Simultaneously, the development of local feed production, with backward linkages to domestic maize production systems and fisheries, can substantially lower the cost and uncertainty associated with imported feed, thus stimulating supply. Lastly, supporting the provision of veterinary care and inexpensive vaccination is critical to prevent shocks that can rapidly deplete a small broiler population.

 Upgrade dilapidated feeder roads serving production basins. Dilapidated feeder roads erode the competitiveness of agricultural value chains over overseas suppliers, segment markets, and exacerbate post-harvest losses. The most common mode of crop transport from the field to rural collection centers is being carried by laborers on their heads. This mode of transport is slow, inefficient and costly. Improving the existing rural pathways to even allow the use of non- motorized transport vehicles would be a significant step towards linking farmers to markets. Unorganized maritime transport, and low storage capacity raise transaction costs and segment the internal market. Aggregation, cold storage and logistics can increase value-chain competitiveness across islands, raise farm incomes, and lower the food import bill.

 Organization and regulation of maritime transport services are critical for enhancing competitiveness of perishables and export crops. Maritime transport of perishables like milk, eggs, fruits and vegetables across islands is informal, unpredictable and negligible in volume terms. Introduction of measures to organize and regulate public transport services to improve frequencies, rationalize fares to make them consistent with socio-economic reality, and introduce a contracting system that will establish competition for the market on high-frequency routes are critical to enhancing the competitiveness for local produce and building economies of scale for export crops.

 Private investments in on-farm warehouses and public investments in cold storage can improve food security by reducing price volatility and enhancing market access. Demand- driven access to finance for on-farm storage and warehousing infrastructure can help smoothen seasonality in farmgate and consumer prices, especially when managed by producer associations with well-identified business needs. Complementary port infrastructure, especially cold storage, can reduce food loss in surplus producing islands (Anjouan and Moheli) and reduce price volatility in Grande Comore by lengthening the supply chain and enabling greater coordination between road and inter-island transport operators. Table 1-2: Boosting Food Production and Access to Markets for Smallholders

PRIORITIES AND MEASURES Direct Role of Public Timeframe Beneficiaries Sector

OOD OOD

F Long Long Short Short Policy Policy maker maker maker Market Market Medium Medium Financer Financer Convener Convener OOSTING OOSTING RODUCTION AND AND RODUCTION B P PRIORITY 1: Enforcement of import exemptions on agricultural inputs to reduce input costs and erratic input supply

Increasing input supply to smallholders by:  introducing a one-stop shop at ports to reduce time Crop and Policy maker Short and cost of compliance with procedures required for livestock Market maker Medium importers to obtain exemptions for agricultural producers Convener Short inputs  facilitate access to finance for input importer Producer association and farmer federations, conditioned on organizations preparation of viable business models  coordinate distribution of time-sensitive inputs at the Private start of planting season importers

PRIORITY 2: Strategic public investments in climate-smart agriculture (CSA) that boost yields and build smallholder resilience to environmental shocks Target CSA dissemination to crop and livestock producers by:  piloting improved, biofortified, and resilient Crop and Financer Short varieties of cassava, maize, and pulses in livestock Policy maker Medium production basins producers  providing market incentives to adopt and scale-up tested CSA technologies, innovations, management Consumers practices (TIMPs) for crop and livestock production systems PRIORITY 3: Facilitate internal food trade to raise market participation and food access (i) Linking highland production basins with coastal urban markets by: Crop and Policy maker Short  enable access to non-motorized (bicycles, carts) and livestock Policy maker Short motorized transport (motorcycles, tuk-tuks) through producers Financer Medium temporary and targeted import tariff waivers  coordinate organized collection from remote areas Producer and marketing through national-level farmer organizations organizations  upgrade and maintain dilapidated rural roads Traders and through locally managed public works programs logistics SMEs

Consumers ii) Investing in market access infrastructure to enhance marketability of perishables and reduce price volatility: Crop Market maker Medium  enabling private investments in on-farm warehouses producers Financer Medium and chilling dairy plant (Anjouan) by assisting and Financer Medium cooperatives with preparing viable business cooperatives Policy maker Medium proposals  investing in cold storage at ports to facilitate Dairy maritime transport of perishables (fish, dairy, cooperatives vegetables and fruits) and reduce food loss in surplus markets Fish  operationalizing existing liquid nitrogen facility in cooperatives Moheli to facilitate vaccine storage and artificial insemination for livestock Food  facilitating competition between mobile operators to consumers incentivize wider coverage and increase low-cost access to market information

iii) Breaking down barriers to inter-island movement of agricultural goods to reduce transit costs, accelerate volume of Crop Financer Long internal food traded, and increase competitiveness by: producers Policy maker Short  investing in port capacity and infrastructure in and Policy maker Medium Grande Comore and Moheli to reduce congestion cooperatives and tariffs  organization, regulation and rationalization of fares Dairy for maritime transit cooperatives  introduction of a contracting system to establish market competition on high-frequency routes Fish cooperatives

Food consumers

Key: (1) Direct beneficiaries are identified based on the expected first-order effects. (2) The role of public sector lists the primary role of the public sector, and it is not restrictive. (3) Timeframe is classified as Short-term for actions that are implementable in 0 – 2 years, Medium-term for 3 – 5 years, and Long-term for more than 5 years.

1.3 Pillar 3: Export Promotion of Established and New Products

 New, promising cooperatives in export crops (cloves, ylang-ylang and vanilla) on all three islands are ripe for the Productive Alliances (PA) approach. Cooperatives formed after the introduction of the 2014 Cooperatives Law have nascent experiences with aggregating production to secure higher prices, modernization of curing and distillation technologies, and moving up the value chain to sell processed product directly to exporters and international buyers. The PA model, successfully deployed by the World Bank in several Latin American countries, can enable cooperatives to overcome market barriers and gain stability through consistent, higher prices while buyers receive a consistent, reliable supply of goods meeting their quality standards.

 Investment in vanilla curing facilities would increase the returns to production and can expand the number of producers. Priced by quality at export, good processing facilities and equipment can raise the margins of vanilla curers and smallholder returns in two ways: i) by minimizing the conversion ratio of green to cured beans, and ii) secure higher prices through better quality grades. To enable the establishment of curing facilities by commercial grower-curing cooperatives requires access to finance and technical assistance, both of which could be structured in a PA with major international buyers that are looking to diversify their procurement across several origins to hedge against the price volatility that is characteristic of the vanilla market.

 Incentives nudging quality differentiation in cloves are required to access higher priced niche markets. Since all cloves are mixed into the general supply of minimum quality at the export stage, the current value chain structure does not reward quality improvements. To access higher priced niche markets, established exporters need to identify buyers, and require technical and financial assistance in developing post-harvest traceability systems. Large-scale grower groupings (100+ MT) can then directly supply to exporters without going through collectors, negotiate individual pricing, and prepare batches of qualities that could be marketed separately by exporters for a price premium.

 Water and wood-saving upgrades ylang-ylang distillation units will enhance commercial and environmental sustainability. Access to water is increasingly a binding constraint for distillation during the dry season, restricting oil production. Wood overuse has led to severe environmental

degradation, and improved furnaces have proven difficult to maintain. Since wood will remain the dominant fuel source for the medium term, policies linking woodlot plantations to distillation use must be developed. Access to finance is required for investments in stainless steel stills, which will improve yield, oil quality, and thus margins, relative to galvanized steel still, many of which are still in use despite age and repair requirement. Coupled with installation of improved furnaces, closed water circulation systems, and construction of large storage areas for fresh flowers, there is great scope to move ylang-ylang exports closer to the production possibilities frontier.

Table 1-3: Export Promotion of Established Value Chains and New Products

PRIORITIES AND MEASURES Direct Role of Public Timeframe Beneficiaries Sector

Long Long Short Short Policy Policy maker maker maker Market Market Medium Medium Financer Financer Convener Convener

PRIORITY 1: Raising smallholder returns in established value chains by: i) supporting plantation renewal and expansion, ii) enabling processing yield and quality improvement, iii) promoting productive alliances between cooperatives and buyers

(i) Supporting plantation renewal and expansion by:  consolidation of bulking plots for supply of high Export crop Convener Short quality planting material to replace aging vanilla producers Financer Short RODUCTS

P vines and Financer Short

EW EW  providing technical assistance for management of cooperatives N ylang-ylang plantations  establish well-managed central nurseries for cloves Exporters and ylang-ylang trees (ii) Enabling processing yield and quality improvements by: Export crop Convener Short  enabling upgrades and investments in vanilla curing producers Policy maker Short facilities by grower-curer cooperatives and Policy maker Short STABLISHED AND STABLISHED

E  incentivizing distillery upgrades to water and wood- cooperatives Convener Short saving structures and technologies for sustainable ylang-ylang production Curers,  standardizing low-cost field tests to reduce dryers and fraudulent manipulation of ylang-ylang oil quality distillers

ROMOTION OF ROMOTION  promote investment in drying and storage facilities P to minimize on-farm losses for cloves Exporters

XPORT XPORT (iii) Promoting productive alliances between cooperatives and Export crop Market maker Short E buyers by: producers Market maker Short  facilitating formation of producer cooperatives and Convener Short through decentralized extension cooperatives Convener Medium  supporting development of niche qualities through certification (eg. Fair Trade, Organic) to attract new Curers, buyers and capture price premiums dryers and  convening and structuring dialogue with distillers international anchor investors to assess mutual needs and identify scope for collaboration Exporters  ensure long-term support and mentoring of cooperatives through input financing and technical Anchor assistance focused on management and operational investors capacity

(iv) Improve regulatory environment by: Export crop Convener Medium  ensuring greater representation of smallholders in producers Policy maker Medium bodies representing each value chain and Policy maker Short  monitoring non-competitive behavior in cooperatives Financer Medium concentrated value-chain segments  removing government involvement in price-setting Curers, in vanilla value chain dryers and  mitigate effects of price volatility on smallholders distillers through targeted income support Exporters PRIORITY 2: Downstream diversification and development of new products by assessing and enhancing commercial viability of new products Assessing and enhancing commercial viability of new products by: Export crop Financer Short  conducting technical and financial assessments of producers Convener Medium commercial viability of derivatives of existing and Financer Short products (clove oil, vanilla powder, vanilla extract, cooperatives Market maker Medium aromatic soap etc.) and new products (ground spices, other essential oils) Agribusiness  establishment of technical university courses aimed startups and at development of derivatives SMEs  multiplication of planting material for crops will small production bases (coffee, pepper etc.)  support establishment of small-scale spice grinding and artisanal extraction facilities Key: (1) Direct beneficiaries are identified based on the expected first-order effects. (2) The role of public sector lists the primary role of the public sector, and it is not restrictive. (3) Timeframe is classified as Short-term for actions that are implementable in 0 – 2 years, Medium-term for 3 – 5 years, and Long-term for more than 5 years.

In summary, this review points to a programmatic approach spanning agricultural policy reforms, strengthening of institutional capacity and enabling environment investments in infrastructure and business climate improvements. By implementing strategic evidence-based policies and investments, the country can build on its agricultural base and link it to agribusinesses and industrial production, thus generating modern, professional jobs that absorb its young and rapidly expanding workforce. In fact, accelerating structural transformation in Comoros is an urgent, attainable goal with broad implications for the country’s ambitious objective of achieving “emergence” by 2030.

2 Introduction

This report is the World Bank’s first Agriculture Sector Review (ASR) for the Union of the Comoros (henceforth “Comoros”), the first analytic product in the Bank’s reengagement with the sector after a gap of two decades. This review is a stocktaking exercise that aims to provide the Government of Comoros with the evidence base necessary for strategic policymaking and smart investments critical to jumpstarting agricultural transformation in the country. In the short-term, the findings from this ASR are intended to launch an evidence-based, participatory, and inclusive consultation process that feeds into an updated sector strategy and investment plan that charts the path towards agricultural transformation.

This ASR follows the African Center for Economic Transformation (2017) and refer to agricultural transformation as the process of i) boosting farm productivity through modernization, and ii) strengthening of linkages between farms, manufacturing and other sectors of the economy through

23 agro-processing, other agriculture-based manufacturing, finance, logistics, and other upstream and downstream services.ii We argue that the process of agricultural transformation produces gains in food security and competitiveness, which we take to be the twin objectives for the agriculture sector in Comoros. Through the ASR, we report direct and indirect indicators for food security and competitiveness, in order to establish a sector baseline, against which progress can be monitored. We conclude with a summary of recommendations and their expected impacts on food security and competitiveness. While jobs and natural resource management are other objectives flagged by the Government of Comoros, the unavailability of data on jobs and natural resource use at the sub-sector level does not allow us to report them as consistently.

In line with World Food Program (WFP), we take food security to mean the consistent availability and adequate access to sufficient, safe, nutritious food to maintain a healthy and active life.iii Since Comoros is unranked in IFPRI’s Global Hunger Index and the Economist Intelligence Unit’s Food Security Index due to paucity of data, we are unable to make international comparisons and this report restricts its analysis to known pathways from agricultural production to food security. Lastly, we refer to agricultural competitiveness as the ability to face competition successfully, to sell products that meet demand requirements and, simultaneously, ensure profits over time, or the aptitude to gain market shares.iv Our approach to assessing competitiveness is driven by data availability at the national and sector levels, and through primary data collection for agricultural value chains. Farm productivity, generally considered an important indicator of agricultural competitiveness, is evaluated against comparator countries in Chapter 5. In Chapter 4, we refer to international and domestic market shares to benchmark the international competitiveness of export products and domestic competitiveness of livestock and horticultural products. In Chapter 6, we supplement these metrics with firm-level indicators like the profitability of representative poultry and horticulture production units.

Chapter 3 delineates state of the enabling environment for agriculture, focusing on complementary infrastructure, trade policy and investment climate bottlenecks that require urgent policy action for the sector to thrive.

Chapter 4 assesses on the demand side of agriculture and estimates the size of the domestic food and beverage market, also outlining market trends for export crops and the potential markets for new semi- industrial products derived from agriculture.

In Chapter 5, we evaluate the changing production profile of Comoros, and the supply-side features of the main subsistence crops and artisanal fisheries.

Chapter 6 presents a synthesis of deep-dive background studies of six established, growing and high potential agricultural value chains, zooming in to map the structure, margins and geography associated with each of them.

Chapter 7 undertakes an assessment of the institutional environment, with a selected review of the recent national strategies pertaining to agriculture and industry, institutional structure governing agriculture and a discussion of other prominent sector actors.

Lastly, this review concludes with proposed policy actions and investments customized to agricultural value chains interventions that are likely to have systemic impacts on food security and competitiveness. The recommendations are grouped under three pillars around which an integrated sector strategy and investment plan can be designed: i) public actions to boost agricultural transformation, ii) boosting food production and access to markets for smallholders, and iii) export promotion of established value chains and new products. Under each pillar, specific actions, expected beneficiaries, role of the public sector, and the implementation timeframe are identified to enable clear prioritization by the government.

2.1 Country Context Key features

 Remote archipelago with a small population  Fragility has delayed economic development  Population is young, predominantly rural, poor and out of work  Low-income country but household expenditures are high  Structural dependence on food imports and remittances

2.1.1 Geography and Demography

1. Comoros is located in the Indian Ocean, near the Northern tip Figure 2-1: Comoros Archipelago of Madagascar. It is an archipelago composed of four islands - Anjouan, Moheli and Grande Comore, where the capital Moroni is located, and the island of Mayotte (under French administration). Covering a territory of about 2,200 km with a coastline of 340 kmv, Comoros is densely populated with an average of nearly 360 people per km2. The population of 795,000 (excluding Mayotte) is distributed across three islands – 52 percent in Grande Comore, 42 percent in Anjouan, and 6 percent in Mohéli – with Anjouan being the most densely populated.

2. The productive capacity of the Comoros islands is a direct function of their age. The elements have had the most time to erode the rocks and render the land relatively flat in Moheli and

25 then Anjouan, covering them with a thick layer of fertile soil and enabling an abundance of fresh water through permanent rivers, which in turn has led to high productive capacities on both islands.vi However, given its relatively small size and dense forest cover, the population in Moheli is much lower than Anjouan, which contributes the largest share of agricultural production in the Union. On the other hand, the youngest island of Grand Comore is still being reshaped by an active volcano, Mount Karthala. Due to its young age, the island has a shallow layer of soil and rainfall has little chance of seeping into the groundwater and there are no permanent rivers.vii

3. The fast-growing population of Comoros is young, rural, poor, and out of work. Creating high quality rural jobs are critical to ensuring a demographic dividend. The median Comorian is just 19.8 years old and the population is experiencing a youth bulge. Largely rural, the country’s also exhibits one of the fastest rates of rural population growth globally (Fig. 2.2).viii. Even more unusually for a low-income countryix with high povertyx, only 45 percent of the population is employed, one of the highest rates in SSA.xi While high labor force inactivity is expected given that 42.2 percent of the working age population is in school,xii a significant disincentive to work is also linked to remittances received by families of emigrants – close to 40 percent of Comorian households received foreign remittances in 2014, with amounts representing 22 percent of recipients’ revenues.xiii To ensure that Comoros reaps a demographic dividend in a context of shrinking emigration opportunities, structural transformation and the creation of high quality rural jobs will be essential for attracting and absorbing new entrants into the workforce.

Figure 2-2: Demographics – Comparing Comoros with the World Comoros's WDI outcomes and percentile comparison to the World SSA average Comoros (2014-2016) Percent Low income average Percentile in World (RHS) Percentile

80 90% 87% 100% 70 78% 60 75% 50 40 50% 30 20 17% 25% 10 2.3 2.2 71.7 46.0 0 0% Population growth (annual %) Rural population growth (annual Rural population (% of total Employment to population ratio, %) population) 15+, total (%) (modeled ILO estimate)

2.1.2 Structural Features of the State and Economy

4. Agriculture has traditionally led the economy, which has shown weak signs of economic transformation. The classic path to economic transformation—a declining agricultural share (of production and employment) and a rising manufacturing share – has eluded Comoros. Agriculture has

26 led slow economic growth and its share of production has only recently slipped below trade and services, declining to an estimated 39 percent in 2016. While employment indicators show a diversification of the household head’s activities over the 2004 – 2014 period, it is not clear if this reflects labor exit from agriculture, intra-household labor reallocation and feminization of farm labor, or simply more off-farm work by men during the slack season. Further, the returns to diversification are low. Available evidence suggests that consumption gains during the 2004 – 2014 period are largely explained by asset ownership, and to a much smaller extent to returns to employment.xiv

Figure 2-3: Agriculture-led Economic Growth

5.0 4.0 3.0 2.0 1.0 0.0 -1.0 -2.0 -3.0 -4.0 -5.0

Agriculture Industry Tertiary

Source: INSEED and World Bank staff calculations and estimates

Figure 2-4: Shares by Major Employment Sectors of the Household’s Head (in Percentage)

100 15 19 22 27 80 6 12 8 15 12 60 13 15 15 40 63 56 54 20 45 0 2004 2014 2004 2014 Poor Total population

Agriculture Industry Trade Service

Source: EIM 2004 and EESIC 2014.

5. Comoros is a low-income country, but poverty is relatively low and private consumption is strong, underpinned by private remittance flows. Economic growth has averaged 2.6 percent since 2000 in Comoros, but GDP/capita has grown much slower due to the population boom (Annex A, Table 1). Despite slow growth in living standards, private consumption has remained strong, underpinned by private remittance flows from the diaspora, which in turn is supported by improvements in the French economy (Annex A, Fig. 3). In fact, Comorian revenues are highly correlated with remittances flows, and the economy is narrowly bound to international prices shocks. Based on the international threshold of US $ 1.9 per day, poverty is 19.1 percent, well below the Sub- Saharan Africa average of 42.7 per cent.xv It is evident that strategies aimed at reducing aid dependency

27 vis-à-vis migration revenues and increasing resilience to international price fluctuations need to be prioritized for significant poverty reduction and shared prosperity. The enhancement of the agriculture productivity and the promotion of productive employment through diversification and enhanced competitiveness are key priority actions.xvi

6. Like other small island developing states (SIDS), Comoros faces challenges linked to geographic isolation, limited resources and a small domestic market. International and internal connectivity (air and maritime) is weakxvii and transport costs are high, fragmenting Comoros’s already small internal market and limiting inter-island sales opportunities. Further, high transport costs also raise the cost of intermediate inputs and imports, making it difficult to export to regional and international markets. In combination with low economic diversification, these characteristics contribute to large and persistent trade deficitsxviii, a narrow export base, and structural dependence on food imports and remittancesxix. However, unlike other SIDS, tourism, fishing and marine trade sectors remain nascent and underdeveloped in Comoros.

7. Large trade deficits are chronic in the Comorian economy, with imports of processed foods being a major contributor. Even among import reliant SIDS, it is an outlier.xx Comoros currently imports nearly 70 percent of its food requirement by volumexxi, and an uptick in food imports represented large shares of imports (44 percent) and GDP (29 percent) in 2015 (Table 2.2). Diet staples – rice, chicken and beef – frequently top commodity import tables, and traders often dump cheap imports to restrict competition. The current dynamic places a growing burden on foreign exchange reserves, exposes consumers to food insecurity due to price shocks like 2008, and discourages private investment in production.

Figure 2-5: Correlation between Income and Food Imports in SIDS

Food Imports & GDP/capita in Small Island 40 States 35 KIR COMSTP CPV 30 VCT 25 VUT DMA PLW SLB LCA GRD SYC KNA TUV MDV ATG BRB 20 JAM BHS 15 FJI TTO 10

0 Food imports share of Merchandise Imports Merchandise of share imports Food 0 5000 10000 15000 20000 25000 GDP/capita

In recent years, prepared vegetables and fruit juice imports have also surged, reflecting the dominance of imported non-perishable and processed products in local retail. Growing urban demand coupled with the absence of domestic food processing imply that this upward trend is likely to continue.xxii On the other side of the ledger, exports, consisting of only cloves, vanilla and ylang-ylang, have been

28 relatively stagnant over the last two decades. Structurally, the pegging the Comorian franc to the euro is a feature which helps keep inflation low, but has also made exports less competitive.

Table 2-1: Top 5 Agricultural Import Commodities, 2015

Commodity Value, million US$ Share of Imports, % Origin Prepared Vegetables 86.9 24 Tanzania Rice 13.2 3.6 Pakistan Fruit Juice 12.4 3.4 Tanzania Chicken Meat 10.6 2.9 Argentina Bovine meat 6.4 1.9 India Agricultural Imports 162 44.3

Table 2-2: Top 3 Agricultural Export Commodities, 2015

Commodity Value, million US$ Share of Exports, % Destination Cloves 22.2 57 India Ylang-Ylang 6.6 17 France Vanilla 6.3 16 Germany Agricultural Imports 35.1 90 Source: MIT Observatory of Economic Complexity

8. Comoros is highly susceptible to natural disasters, suffering on average, direct losses of $5.7 million – nearly 1 percent of GDP - annually.xxiii Natural disasters like tropical cyclones, floods, fires and volcanic eruptionsxxiv have a disproportionately negative impact on the economy given its relatively small size, and high dependence on agriculture for export revenues. By some metrics, the country is one of the most vulnerable and susceptible to natural disasters, with rising sea levels threatening erosion, agricultural failures and displacement of up to 10 percent of the coastal populationxxv. Since natural disasters are also likely to become more frequent and intense with climate change, the absence of risk mitigation tools like early warning systems are conspicuous by their absence in Comoros.

Figure 2-6: Dimensions of Natural Disaster Risk

Comoros's outcomes and percentile comparison to the World Unit based on Comoros Percentile in World (RHS) Percentile indicator 94% 100 87% 82% 82% 100% 80 61% 75% 60 35% 50% 40 20 25% 7.3 11.0 66.4 58.7 84.5 56.2 0 0% Natural Disaster Natural Disaster Natural Disaster Natural Disaster Natural Disaster Natural Disaster Risk (2016) (Exposure) (Vulnerability) (Susceptibility) (Coping (Adaptive (2016) (2016) (2016) capacities) capacities) (2016) (2016) Source: UN University Database

2.2 Sector Context

Key Features

 Comoros is a global outlier in terms of relative size of agriculture sector  Moderate growth has been driven by extensification, while yields remain poor  Three cash crops generate more than 90 percent of export income  Highly exposed to international price shocks and natural disasters  Weak inter-island transport links between farmers and markets  Virtual absence of agro-processing due to enabling environment constraints

9. Agriculture is the second-largest sector, the largest employer, and provides for 90 percent of export income. Endowed with fertile volcanic soil, a gradient of altitudes and micro-climates, and a moderate growing period, the economy of Comoros has traditionally been dominated by agriculture. Growing at 2 percent in 2015, the sector has often led economic growth in Comoros, but has recently been overtaken by the trade and services sectorxxvi. Accounting for 38% of jobs and 36% of GDP, the relative size of Comorian agriculture ranked in the 94th percentile over the 2014 – 16 period, an anomaly among Small Island Developing States (SIDS).xxvii A narrow base of three export crops (cloves, vanilla, and ylang-ylang) provide 90% of export income, with nascent attempts to diversify and revive older export products like copra, coconut oil and coffee. In fact, agriculture in Comoros dwarfs comparator countries on every metric of size except the output of the livestock sub- sector (Table 1). Mauritius and Fiji produce nearly 80 and 32 times more poultry meat despite having 1/3rd and 1/33rd the agricultural population, a differential that is an artefact of the miniscule poultry value chain in Comoros and not significantly lower yield.xxviii

Figure 2-7: Benchmarking the Size of the Agriculture Sector

Comoros's Agriculture, value added (% of 70.0 GDP) comparison to the World: 2014-2016 60.0 50.0 35.9 40.0

30.0

20.0 All Countries

10.0 0.0 Palau China Latvia Belize Liberia Jordan Algeria Cyprus Zambia Albania Belarus Burundi Lesotho Belgium Sweden Bulgaria Senegal Ecuador Hungary Uruguay Denmark PDR Lao Tanzania Singapore Nicaragua Zimbabwe Mauritania Timor-Leste South Africa South Mozambique Leone Sierra Czech Republic Czech Egypt, Arab Rep. Arab Egypt, Trinidad and Tobago and Trinidad Source: World DevelopmentChina Prov.of Taiwan Indicators

Table 2-3: Benchmarking the Size of the Agriculture Sector, 2014

Comoros Mauritius Fiji Samoa Share of GDP (value-added), % 35.9 3.5 11.3 Share of Employment, % 38 7.5 1 5.4 Population in Agriculture, 1000 288.6 94.4 8.9 10.4 Rural share of population, % 71.6 60.4 45.9 80.9 Share of Land, % 71.5 42.4 23.3 12.4 Land under cereal production, 1000 ha 26.9 3 0.4 Poultry meat output, 1000 MT 0.6 47.6 19.2 0.5 Source: World Development Indicators, FAOSTAT. Note: Population in agriculture is calculated using the WDI indicator for share of employment.

10. Although sectoral growth is moderate, it is characterized by extensification, erratic input supply, and a lagged response to international prices. Farm productivity is low and stagnant, and Comoros ranks 41st out of 54 African countries in terms of agribusiness competitiveness.xxix Growth is driven by land extensification, and consequently one of the fastest rates of deforestation in Africa, especially in Anjouan.xxx The market share of local poultry and bovine meat is miniscule (less than 1%), due to disorganized supply of imported feeds and breeds coupled with weak veterinary and surveillance services. Fruit and vegetable supply is competitive only during the rainy season, due to lack of access to varieties with multiple annual production cycles. Cloves has become the largest source of declining export revenue, with vanilla and ylang-ylang unable to respond quickly to ongoing price spikes after a decade of low prices, which have led to a dynamic of reduced planting area and lower quality differentiation.xxxi

Figure 2-8: Benchmarking the Performance of the Agriculture Sector, 2014

Comoros Mauritius Fiji Samoa Growth rate (value-added), % 2 -0.3 -0.3 6.4 Value-added per worker (constant 982 9,653 2,800 4,255 2010$) Cereal yield, 1000 kg/ha 1.4 3.8 2.4 Vegetables yield, 1000 kg/ha 6.6 9.3 11.5 1.2 Fresh Fruit yield, 1000 kg/ha 4.8 8.9 5.5 4.9 Source: World Development Indicators, FAOSTAT

Figure 2-10: Agriculture and GDP Growth Figure 2-9: Agricultural Trade Trends 12 120 10 100 8 80 6 60 4 2 40 0 20 -2 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 -4 1997 1999 2001 2003 2005 2007 2009 2011 2013 Agriculture, annual growth Food Imports as % of Merchandise Imports GDP, annual growth Food Exports as % of Merchandise Exports Source: WB staff calculations using WDI, ReSAKSS Database Source: World Development Indicators (WDI). Note: Agricultural imports and exports are classified as food imports and exports in WDI.

11. Subsistence agriculture continues to dominate, and is linked to failures in the land and credit markets. Subsistence agriculture using poor inputs on small plots dominates on the archipelago. Commercial agriculture is limited to three export crops, with 80 percent of agricultural production intended for self-consumptionxxxii. Aggregate land availability – more than 2.2 ha of agricultural land per producerxxxiii – coexists with three overlapping land tenure systems that impose costly barriers to land market activity, even as agricultural financing remains contingent on land titles, gold or large cash guarantees. Producer associations have been unable to link smallholders to the credit market and consolidate high quality output consistently, leading to high transaction costs, weak forward and backward linkages with buyers and input suppliers, and little or no value addition.

12. Missing transport linkages and value-chain gaps are binding supply constraints to greater domestic competitiveness of perishables. On one hand, highland production basins in Anjouan and Grande Comore are hampered by poor feeder roads and the absence of aggregators equipped with transport services, segmenting surplus production areas from growing urban markets. On the other hand, erratic air and maritime connectivity and missing cold storage facilities at ports and collection points, make it difficult for food produced on one island to be competitive in a different island. In particular, these missing linkages impact the supply chain for locally competitive perishable produce like milk, fruit, vegetables and fish, resulting in low prices and food loss at the farm gate, even as consumer choice is restricted to imported substitutes at modern retail stores, which pay a premium for reliable food supplies.

13. Established private sector players are reluctant to enter food processing, mainly due to a weak enabling environment. Despite the opportunity to capture a larger share of the growing segment of processed food, the established Comorian private sector with access to capital – traders of imported food and exported cash crops, small manufacturers – displays little appetite for investing in domestic food processing and beverage production. In addition to the weak contract enforcement environment faced by any business – Comoros is ranked 180th on this dimension of Doing Business indicatorsxxxiv – interviewed private sector players cited three constraints specific to food processing: i) inconsistent aggregation and supply of throughputs (quantity, quality and timing), ii) unpredictable customs duties associated with inter-island maritime and air transport, and iv) high transaction costs associated with direct sourcing from poorly connected production basins. However, in contrast with the established

32 private sector, young startups being currently incubated by industry associations have a high share of agribusiness plans that could be supported.

14. Agricultural exports have been stagnant for a decade, but vanilla and ylang-ylang prices are currently booming, increasing the risk of quality deterioration. In contrast to food crops, the production systems for cloves, ylang-ylang and vanilla are characterized by contract farming, well- developed value chain linkages, and established relationships between exporters and international buyers. However, a combination of worsening quality, low international prices, and reduced output have led to a slowdown in exports over the last decade, which now account for less than one-fourth of just food imports, compounding the growing trade deficit. While cloves are now the Comoros’ biggest export, current government strategy is focused on vanilla, the prices of which are experiencing a 13-year high. However, there remains a high risk of a crash given widespread speculation and premature harvest of green beans in Madagascar, the global price-setter. Similarly, prices for ylang-ylang flowers are 3x their 2015 prices and have incentivized a manipulation of the quality of the essential oil distilled from them, consequently raising the likelihood of a collapse in future prices. The absence of domestic standards, and reliable and inexpensive field testing has contributed to the lack of quality control in both products.

15. Slow climate changes have disrupted local production systems and natural disasters pose a major threat to the sector. Climate change has contributed to the rainy season becoming shorter and more unpredictable in recent years.xxxv Disturbances in the hydrological cycle have also led to the drying up of rivers and water sources in Grand Comore, and a lowering of the water table. Given that agriculture accounted for 47 percent of freshwater withdrawals when data was last reportedxxxvi, water shortage has become the binding agro-ecological constraint on the island. Simultaneously, accelerated soil erosion due to deforestation has increased runoff and led to flooding, reducing soil fertility and agricultural yields in Anjouan as well.xxxvii 3 Enabling Environment for Agriculture

Main Messages

 Poor rural connectivity weakens linkages between coastal urban markets and production basins in rugged highlands.  An unpredictable, informal maritime system limits inter-island mobility of perishables and effectively segments an already small internal market.  Access to credit for producer organizations is limited by collateral requirements, high interest rates, perceptions about high risks associated with agriculture lending, and low appraisal capacity among MFIs, even as traditional banks remain severely constrained.  Access to energy for agri-processing is hindered by insufficient and unreliable electricity; untapped solar potential near production areas has the potential to transform the sector.  Access to land for large-scale commercial agriculture is low due to high transaction costs and insecurity associated with market activity  Tariff exemptions are not applied consistently, raising the costs of agricultural inputs and exports, and introducing production uncertainty.

3.1 Complementary Infrastructure

The paucity of complementary infrastructure constrains farm and rural non-farm market linkages and create gaps in agricultural value chains in Comoros. The poor state of rural roads, ports and communications infrastructure segments urban consumer demand from rural producers. Grid electrification or mini solar-grid solutions can kickstart small-scale food processing and value addition, but at present, biomass remains the primary source of energy in highland production basins. On-farm storage and warehousing infrastructure can help smoothen seasonality in producer prices and food consumption, but are currently limited to sporadic donor-financed projects. Port infrastructure, in particular cold storage, can reduce food loss and price volatility by lengthening the supply chain spatio- temporally. Infrastructure investments in all of these cases can help enhance food security, competitiveness and generate jobs for youth, in both farm and non-farm activities.

3.1.1 Road Network

16. Comoros has an extensive road network that is rapidly deteriorating, reinforcing the geographical isolation of highland production basins. Road transport is the main mode for movement of goods (intra-islands) and passengers in the country.xxxviii Given the small size of the archipelago (about 1,861 km2), the density of paved roads (412 km per thousand km2) is impressive compared to a sub-Saharan Africa average of 31 (Annex A Fig. 7.1, 7.2). However, lack of maintenance during the past 15 years has resulted in extensive sections of the paved road network being seriously dilapidated, with an additional 240 km of roads and dirt tracks being nearly impassable. Estimates indicate that more than half of the road network is currently in poor or very poor condition due to the absence of qualified personnel and financial resources for maintenance.xxxix

17. Poor rural connectivity especially limits agriculture because of the challenge posed by the rugged terrain and the location of food production basins in highlands. The fragmented provision of services prevents capture of economies of scale and drives up intra-island transport costs, which are quite high compared to other Small Island Developing States (SIDS), and result in sharp price differentials between production and consumption centers.xl This erodes the competitive advantage of agricultural value chains over overseas suppliers, segments markets across islands, and exacerbates post-harvest losses. The most common mode of transport of crops from the field to the collection center is rudimentary, loads being carried by laborers on their heads over several kilometers.xli This mode of transport is slow, inefficient and costly. Improving the existing rural pathways to even allow the use of non-motorized transport vehicles would be a significant step towards linking farmers to markets.

Figure 3-1: Network Condition of The Three Islands

70%

60%

50%

40%

30%

20%

10%

0% Gde Comore Anjouan Mohéli

Good Fair Poor Very Poor

Source: PADDST 2014

3.1.2 Inter-island Maritime and Air Transport

18. Low port capacities and high tariffs leads to high trade costs which hinder domestic and export competitiveness of agriculture. As in most SIDS, Comoros depends heavily on maritime transport, through which the bulk of international and inter-island trade takes place. Despite having a high-traffic port in Anjouan and a smaller port in Moroni,xlii their low capacities result in high port tariffsxliii – five times higher than in Mauritius and three time higher than in Mombasa (Annex A, Fig. 9). In fact, Comoros spends significantly more than the SIDS and global averages on international transport costs as a share of imports (2004–2013 average) (Fig. 3.2). On one hand, additional import costs are passed on to importers and agricultural producers, limiting access to and use of yield- enhancing improved seeds, fertilizers and mechanized equipment. On the other hand, exporters also pay higher than they would to export a standard container from neighboring SIDS (Annex A, Fig. 10), reducing export competitiveness.

19. The unpredictable informal inter-island maritime system limits the mobility of perishables and effectively segments an already small internal economy. Despite the islands being separated by short distances, almost all inter-island passenger traffic is by air because of the lack of scheduled sea passenger transport. Similarly, sea transport of perishables like milk, eggs, fruits and vegetables across islands is informal and negligible in terms of volume. The introduction of measures to organize and regulate public transport services in order to improve frequencies, rationalize fares to make them consistent with socio-economic reality, and introduce a contracting system that will establish competition for the market on high-frequency routes are critical to building economies of scale and enhancing the competitiveness for local produce.

Figure 3-2: Average Expenditures on International Transport as a Percentage of the Value of Imports for Small-Island Developing States, 2004-2013

Source: UNCTAD estimates, Review of Maritime Transport 2014, chapter 6.

20. Air transport between the three islands is particularly expensive and discourages inter- island movement of passengers and goods. Small airplanes play an important part in connecting islands but the passenger air fare is about four to five times higher than the sea rate. Air fares on the domestic network are particularly expensive because of: (a) the lack of air transport capacity; (b) the difficulty of controlling the abnormal sales practices of certain companies; and (c) the impossibility of purchasing tickets on the domestic network from abroad as sales are not available on the most widely used electronic distribution networks.

3.1.3 Market Infrastructures

21. Nascent vanilla cooperatives and older producer associations are constrained by the inability to access market access infrastructure. Demand-driven on-farm storage and warehousing infrastructure can help smoothen seasonality in producer prices and food consumption, especially when managed by producer associations with well-identified business needs. In field interviews with vanilla cooperatives supported by ITC and UNDP, it was evident that on-farm warehousing coupled with management oriented technical assistance would enable the groups to aggregate, store, cure and potentially sell cured vanilla without being subject to the bargaining power of the three existing exporters, which dominate a highly consolidated value chain segment. Similarly, port infrastructure, especially cold storage, can reduce food loss in surplus producing islands (Anjouan and Moheli) and reduce price volatility in Grande Comore by lengthening the supply chain and enabling greater coordination between road and inter-island transport operators.

3.2 Investment Climate and Trade

Investment climate affects the pace of enterprise and job creation worldwide. Global evidence shows that 9 out of 10 jobs are created by the private sector,xliv in stark contrast to the dominance of public

36 sector in formal employment in Comoros.xlv Evidence also shows that the investment climate impacts private enterprises and associated job creation.xlvi The formal private sector in Comoros faces many challenges to doing business, with contract enforcement and payment of taxes being especially difficult (Annex A, Fig. 10, 11). However, the specific binding constraints faced by the large share of informal agricultural enterprises are: i) access to finance for producer associations and emerging agribusinesses, ii) predictable and low-cost access to energy, and iii) establishing a predictable and business-friendly investment climate and trade policy, which are the three main components critical for a policy framework that supports agricultural development and creates food system jobs.

3.2.1 Access to Finance

22. Limited geographical coverage and legacy lending issues limit the traditional banking sector’s lending for productive uses, in particular commercial agriculture. The institutional landscape for financing agribusiness investments is underdeveloped and Comoros ranks 122nd globally in terms of ‘Getting Credit’ on the Doing Business indicators.xlvii The financial sector – comprising 4 commercial banks, 3 microfinance networks and 3 financial intermediariesxlviii – has low geographic penetration in terms of brick-and-mortar commercial bank branches (3.1 per 100,000 adults) and ATMs (5 per 100,000 adults),xlix reflected in low financial access (Fig. 3.3). Mobile money is non- existent, increasing reliance on cash and negating the possibility of digital credit. In addition, the sector is undergoing rehabilitation, with the privatization of the Development Bank of Comoros and the postal bank (SNPSF), and expansion of the savings and microfinance networks. In the context, lending for productive uses, particularly agriculture, has been low (Fig. 3.4). Non-performing loans comprised 20.5 percent of all lending in 2015 and continue to act as a drag on the sector.l Historically the largest lender in the country, the Bank for Industry and Commerce (BIC), is restrained in its support to agribusiness due to unresolved debts stemming from the collapse of vanilla exports in early 2000’s.li

Figure 3-3: Benchmarking Financial Access in Comoros

SSA average 400 Comoros: Period 1 (2014-2016)

300

200

100 116 10 34 21 0 Bank accounts per Borrowers from Borrowers from all Remittance inflows to 1,000 adults commercial banks per microfinance GDP (%) 1000 adults, Number institutions (MFIs) per 1000 adults, Number Source: IMF Financial Access Survey and Global Financial Development

Figure 3-4: Lending in Comoros, by Industry

Source: Central Bank of Comoros and IMF Staff Calculations

23. Microfinance institutions Figure 3-5: Bank Branches and MFIs (Number), 2015 supported by AFD are now playing a leading role in urban lending,lii but agricultural lending is constrained by low capacity to appraise sector investments. Together, three microfinance institutions – U-MECK, Sanduk and USM – account for 67 of the 123 banking outlets in Comoros.liii The largest, U-MECK, grew to more than 40,000 members and $25 million in deposits in 2016,liv gradually displacing market share from larger commercial banks, and providing microcredit to urban artisans and small Source: International Monetary Fund (December 2016) commerce. In Anjouan, Sanduk plays a similar role, with deposits of $22 million and a $17 million gross loan portfolio in 2016.lv However, agricultural financing provided by the institutions is largely short-term, with interest rates ranging from 9 – 12 percent, a requirement of 20 percent cash as guarantee, or collateral in the form of gold or land,lvi due to weak coverage of the credit registry.lvii These requirements are stringent for investments by young entrepreneurs or producer associations, which often lack access to collateral, and are unable to provide guarantees. In interviews with the team, the management of both institutions strongly indicated that technical assistance to modify product offerings, undertake due diligence and appraise agricultural investment proposals is a key supply constraint in terms of credit for agriculture.

Figure 3-6: Financial Sector Assets & Liabilities by Share of GDP, 2015

Source: International Monetary Fund (December 2016)

24. Despite an emerging incubation ecosystem, there remains high unmet demand for services for agribusiness startups, which comprise the large majority of new enterprises. At present, the Union of Chamber of Commerce, Industry, and Agriculture (UCCIA), designed to assist SMEs with business development is the interface between the private sector and the government.lviii The UCCIA, with support from the US$3 Million Kuwait-Financed SME financing scheme (AMIE) for agribusiness and fisheries, and U-MECK, launched an ambitious program which supported 32 of 167 MSME startups with seed capital and mentoring services in 2014.lix Following its success, AfDB helped UCCIA launch an incubator Innov’Lab in 2017, which selected and is supporting 15 of nearly 100 young entrepreneurs annually by providing office space, legal services, and business plan development and connections to the microcredit institutions. The incubator has identified target sectors and 10 of the current 15 projects are from the agri-food and livestock sectors.lx

3.2.2 Access to Energy

25. Insufficient and unreliable electricity are one of the most severe constraints to the development of agricultural value chains and rural services. Estimates suggest that biomass (wood, charcoal) accounts for 70 percent of energy use, while the remainder is generated using imported diesel fuel.lxi Underinvestment in the production capacity of public operators has led to a prolonged period of shortages – 1 ktoe per annum – since 2010.lxii In 2018, Comoros was ranked 135th globally on the ‘Getting Electricity’ indicator in Doing Business.lxiii In particular, the time taken to obtain an electricity connection – 120 days – remains higher than the SSA average and the cost associated with it – 2,050.5 percent of per capita income – is a high deterrent for SMEs and potential entrants in the agro-processing space. Even at this cost, the reliability of supply by the public network is scored zero by this sub- indicator, reflecting the frequency of power outages and the absence of monitoring and restoration mechanisms. Consequently, most companies own their own generators and negotiate pricing conditions for fuel supply, a burdensome business cost that has contributed to the shutdown of previous donor-financed attempts at small-scale processing of vegetables and fruits.lxiv

Figure 3-7: Distance to Frontier - Getting Electricity

26. Changing the energy mix to tap into solar potential could dramatically change the viability of small-scale agricultural processing. In 2016, the Comorian government affirmed its support for a heavy-fuel plant financed with a concessional loan from the EXIM Bank of India, which is expected to provide a firm supply foundation in the medium term. The government also continues to work with the World Bank and the African Development Bank to strengthen the capacity of the national utility MAMWE and to diversify to solar, wind, and geothermal energy.lxv In particular, the proximity of high solar potential geographies to production basins on all three islands (Fig. 3.8) could be transformational if mini solar grids are established to power small-scale processing like grinding, crushing, juicing, milling and chilling close to producers. In Haiti, a USAID supported company has successfully deployed mini solar grids for milling and processing of maize and breadfruit, agricultural products common to Comoros, at the scale of a community of 30,000.lxvi Innovative off-grids solutions like milk coolers piloted in a remote Kenyan community can also be adapted to Comorian highlands.lxvii

Figure 3-8: Solar Potential in the Comoros

3.2.3 Access to Land 27. As arable land has grown scarce, the co-existence of three overlapping land tenure systems – customary, Islamic and colonial – has become a source of dispute and conflict. In the customary system, land is largely communal at the village, clan or family level, and belongs to the tiller; i.e. land use rights are conditional on cultivation and can be claimed by another community member if idle or abandoned.lxviii On non-communal land, private property can be established by a father for his daughter – manyahouli land – via purchase or land (forest, bush) clearing.lxix Islamic law also allows for both community (usoyezi) and privately owned land (milk), with the principle for acquisition of exclusive but not absolute ownership being vivification of abandoned land, i.e. being put to productive use.lxx Lastly, the colonial system recognizes community tenure, privately owned titled properties and state-owned public land.lxxi In recent years, land conflicts and disputes have amplified within communities and in courts, with population growth, plot fragmentation, and continued dependence on agriculture and fishinglxxii triggering a competitive occupation of ‘idling’ agricultural land (often already owned), the competition between official administration agents (notaries, judge) and kadis (Muslim judges) in issuing land succession acts and rendering land – related judgements.

28. Land governance is particularly weak in rural areas, characterized by non-existent titling, poor land administration services and an unclear land policy. The post – independence revolution of 1975 – 1978 overturned the situation of strongly enforced colonial legislation over state owned land.lxxiii With an estimated 1,700 land titles issued and 5,000 applications pending in Grande Comore, the Service des Domaines (land registration system) effectively controls only cities suburbs but is scarcely present in rural areas.lxxiv The paucity of rural land titles is coupled with an absence of land use planning and management policy (definition of land statuses, institutional arrangement, land taxation, land valuation), and poorly-resourced land administration services which are unable to conduct domain and topographical surveys.lxxv Rated below the SSA average in terms of quality of land administration,lxxvi the Service des Domaines has no clear vision of its estate, and state domain lands are almost customarily occupied.lxxvii

29. Investments in large-scale commercial agriculture will remain low if the high transaction costs and insecurity associated with land transactions remain unaddressed. The land sale market is constrained by: i) the high market price of land, especially in Anjouan where most agricultural production takes place, ii) the transaction costs associated with purchase of land held at the community and family levels, which requires community consent, and iii) high administrative costs associated with land titling for individuals outside the community. The average price of land - $250 - $600/m2 is 4 to 10 times higher than in Madagascar. Land titling also requires the payment of a flat fee, a 3% conservation fee on the declared land value, and topographical demarcation fees. Cumulatively, purchasing and securing a land with a surface area of 100 meter2 is an extremely time and cost-intensive transaction that is unaffordable for most agribusinesses, especially in the context of a weak credit market and poor administrative capacity.lxxviii. Other land market activity comprises of verbal contracts for sharecropping and leasing between farmers from the same village, community, or family and the land remains inalienable to outsiders. The transaction, built on mutual trust, is neither registered in the Commune nor in the land administration service, exposing it to future disputes that are ambiguously arbitrated.

3.2.4 Trade Policies and Business Environment for Agriculture

30. In principle, the trade regime in Comoros has been liberalized, with the rationalization of tariffs and the elimination of all non-tariff barriers, except for rice.lxxix Given Comoros’s

41 membership of the Common Market for Eastern and Southern Africa (COMESA) Free Trade Area, the COMESA Common External Tariff (CET) determines most of the tariff and trade policy.lxxx Comoros’s weighted average MFN tariff for 2013 was 6.7 percent, lower than the corresponding SSA (8.4 percent) and global (7.6 percent) averages.lxxxi While vegetables and tubers face a 35 – 40 per cent ad valorem duty, strategically important products like poultry meat and eggs attract a lower total duty of 19 percent, contributing to lower domestic competitiveness.lxxxii Other priority products like fish and basic rice imports face a customs duty of 150 KMF and 50 KMF per kg respectively.lxxxiii Basic rice is imported solely by the State-trading enterprise (SOE) ONICOR, which is mandated with maintaining security of supply and price stability. The low ad valorem duty on rice (equivalent of 33 percent) is intended to maintain the local consumer price. Comoros is also expected to complete its WTO accession process in 2018 and in preparation, has elaborated a comprehensive action plan to ensure the consistency of national laws with phytosanitary, animal health and food safety regulations. As a LDC, has full duty free and quota-free access to the EU for all exports except arms and armaments.

31. In practice, tariff exemptions are not applied consistently, raising the costs of agricultural inputs and exports, and introducing production uncertainty. In principle, imported agricultural inputs have been tariff exempt since 2012 and are entitled to a 1 percent ad valorem rate. In practice, the burdensome process of obtaining exemptions leads to input importers paying tariffs between 3 – 9 percent and supplying inputs after the planting season,lxxxiv lowering smallholders’ use of improved seeds, fertilizers, and mechanized equipment, consequently lowering yields and competitiveness of marketed products like sweet potatoes and vegetables.lxxxv The average time and costs associated with border compliance are higher than neighboring ports in the region (Fig. 3.10), and perhaps more importantly, are inconsistently applied. Official export tariffs on the three export crops (cloves, vanilla, and ylang-ylang) are revised annually and in principle, are determined by international prices. However, an ambiguity about which prices form the base for tariffs, and exporters report paying highly varying tariff rates at the customs office.

Figure 3-9: Trading across Borders in Comoros and Comparators - Rank and DTF

Source: Doing Business (2018)

Figure 3-10: Trading across Borders - Time and Cost

32. The Doing Business Report ranked Comoros 158 out of 190 economies in 2018, reflecting a poor environment for attracting investments in all sectors. Comoros ranked regional neighbors, where Mauritius positioned at 25 is leader, and only Madagascar, at 162, is close to Comoros. It also ranks lower than benchmark countries such as Cape Verde (127), Fiji (101) and Samoa (87). From this vantage point, the country is relatively less attractive than the competition for attracting investment in any sector. Burdensome regulations adversely affect market entry, exit, trade, and competition. Enforcing contracts appears to be the most challenging area for Comoros (with a ranking of 180th), as the average time required to resolve a commercial dispute between two domestic investors taking a little over 500 days. Another problematic area is also the paying taxes indicator. The country stands out with an unusually high total tax rate (216.5 % of total profit and 4 times the regional average) — signaling burdensome and difficult to comply-with tax administration processes. Despite generous incentives and exemptions, the 2007 Investment Code has had limited impact on attracting new investment. The Code is currently being revised to enhance selectivity, predictability and transparent application of its provisions, strengthen the legal protection of investors and support small and medium-sized enterprises while rationalizing tax and customs incentives.

33. Structuring the agenda for regulatory simplification around the Enabling Business of Agriculture (EBA) indicators can accelerate investments along the agricultural value chains. The establishment of simplified, low-cost regulatory systems for seeds, fertilizers, machinery, finance, markets, transport, information and communication technology (ICT), and water can have a catalyzing effect on the investments and performance of key players across agricultural value chains—from smallholder farmers to existing agribusinesses, producer organizations and young entrepreneurs seeking to enter the sector. The template set by the collaboration between the National Investment Promotion Agency (ANPI) and IFC to improve Doing Business indicatorslxxxvi can be built upon, with Enabling the Business of Agriculture (EBA) indicators providing an analogous tool for policy action and quantitatively monitoring regulatory changes over time, and compared to other countries.lxxxvii

34. The industrial development agenda for SMEs significantly overlaps with the enabling environment needs for an agribusiness sector, reflecting the need for a unified strategy. In the draft National Industrial Strategy (2017) prepared by the MAPEAU, all of the domestic and export products identified for short-term industrial development are agriculture-derived products. A large share of the policy actions outlined for medium and long-term development of MSMEs and the private sector – improvement in business climate, transport infrastructure, access to land etc. – are also the binding constraints faced by agricultural value chain actors. This convergence is a recognition that in the Comorian economy, there is a dearth of non-agricultural raw materials to power a diversified manufacturing sector. Given this convergence, it is critical that private sector bodies like the UCCIA,

43 which represented both industry and agriculture until a recent split, are reunified to ensure coordinated lobbying for regulatory simplification and policy action.

4 Sources of Demand

Main Messages

 The Comorian food and beverage market is almost as large as the economy.  Diets are diversifying and the majority of food expenditure is now on perishables like vegetables, meat, and fish.  Local producers show signs of competition in the growing domestic market for perishables, but require strategic linkages with transport and retail for greater competitiveness.  Cloves, ylang-ylang and vanilla face favorable market conditions after a decade long downturn, and new cooperatives have opportunities to move up the value chain.  Diversification of export revenues is possible with the downstream development of existing export products and niche products.  An emerging hospitality and tourism sector represents an untapped opportunity for agricultural producers.

4.1 Size of the Food and Beverage Market

35. The Comorian food and beverage market is almost as large as the economy. Utilizing the latest household consumption data (2014), we computed national-level spending on food and beverages. At $572 million, the estimated domestic expenditure on food and beverages is nearly as much as all of estimated economic activity – $617 million in 2017 (Table 4.1). If GDP estimates are revised upwards by 40 – 100 percent, as is expected after a revision of national accounts this year, the estimated size of the food and beverage market will be far less surprisinglxxxviii. However, Comoros will remain an outlier to Engel’s Law, which states that as income increase, the budget share of food declines. Taking consumption per capita to be a proxy for income – since income is not typically well- measured – we find that while Engel’s Law generally holds for SIDS, it does not hold for Comoros. In fact, despite having nearly equal consumption/capita, Comoros (61.4) spends 12 percentage points more than Cape Verde (49.7 percent) on food (Fig 4.1).

Table 4-1: Size of the Market for Food and Beverages in South-West Indian Ocean

(1) (2) (3) (4) Country Population, Food and Beverage Market Size thousand, 2017 Consumption/capita estimate, estimate, USD 2017 millions (USD 2017) Cabo Verde 540 615.1 332.2 Fiji 899 573.1 515.2 Mauritius 1,263 1224.7 1,546.8

Sao Tome and Principe 200 409.9 82.0 Comoros 796 719 572.3 Data Sources: World Development Indicators, Global Consumption Database, Comoros EDMC 1-2-3 Survey (2014) Note: Column (3) is computed by extrapolating from the last data point for food and beverages expenditure per capita, using the country’s average GDP/capita growth rate between that year and 2016. Since the last data point for Comoros was taken from 2014 survey round, the exchange rate used is 1 USD = 405 KMF (December 31, 2014). The market size estimate in column (4) is a product of the columns (2) and (3).

Figure 4-1: Comoros is an Outlier to Engel’s Law

Food Share of Consumption 0.8 0.7 Sao Tome and Principe 0.6 Comoros 0.5 Cabo Verde 0.4 Sao Time Fiji 0.3 0.2

Food & Beverage Share & BeverageFood 0.1 0 0 500 1000 1500 2000 2500 Consumption per Capita

Source: Global Consumption Database. Data for Comoros is based on EDMC 1-2-3 Survey (2014).

36. The majority of food expenditure is on perishables like vegetables, meat and fish. Across Eastern and Southern Africa, the evolution of dietary choices with income growth shows three trends – i) 61– 83 percent of consumed food is purchased by the middle class, ii) 70 – 80 percent of purchased food is processed, and iii) 44 – 55 percent of purchased food is perishable.lxxxix Comoros reflects the second and third trends (Fig. 4.2) and while there is no quantitative evidence on the first, the growth of modern retail suggests that it might be subject to the first trend as well. In particular, the large budget shares for perishables – vegetables, meat and eggs, fish, fruit, and dairy – are a salient market signal for domestic producers (Table 4.2). Cumulatively, these segments aggregate to a $257 million sub- market, which is larger than the entire agriculture sector at present. Investments targeted at enhancing competitiveness in this sub-market have the potential to raise farm incomes, align supply with future sources of demand, and professionalize the sector with the entry of agribusinesses and young entrepreneurs.

Figure 4-2: Budget Shares, by Food Segment

Food Expenditure Shares by Segment 25 25

20 17 16 15 Percent 10 5.3 4 5 3.5 3.2 3.1 1.6 .18 0 Cereal Vegetables Meat Fish Fruit Dairy Condiments Sugar Oil Beverage Coffee

Source: 1-2-3 Survey (2014)

Table 4-2: Budget Shares, by Food Segment

(1) (2) (3) (4) Food Segment Categorization Household Estimated Market Budget Segment Size, Share, % millions (USD) Cereals (including bread) Non-perishable/Perishable 24.5 140.3 (bread), processed low-value Vegetables Perishable, unprocessed 19.2 110.1 Meat and eggs Perishable, processed low- 17.4 99.5 value Fish Perishable, unprocessed 16.3 93.5 Fruit Perishable, unprocessed 5.3 30.6 Dairy Perishable, processed high- 4.0 23.0 value Condiments Non-perishable, processed 3.5 20.0 high-value Sugar Non-perishable, processed 3.2 18.5 low-value Oil Non-perishable, processed 3.1 17.6 high-value Beverages Non-perishable, processed 1.6 9.1 high-value Coffee Non-perishable high-value 0.2 1.0 Note: Column (3) categorization is derived from Tschirley et al (2015). Column (4) is computed by multiplying column (3) with estimated size of food market – $572.3 million and divided by 100.

37. With strategic investments, Comorian producers could compete with imports in the growing domestic market for perishables. Since the food market is currently dominated by imported produce that is dried, frozen, canned or preserved in other ways to survive the shipping time to Comoros, there exists an opening for their fresh, local substitutes if the latter are reasonably competitive, or if domestic preferences change enough to pay a mark-up for local perishables. At

46 present, eggs and juice concentrate produced domestically occupy significant market shares, signaling their current competitiveness (Table 4.3). In addition, domestically produced liquid milk and fruits are competitive in segmented markets (Anjouan) and seasonally (June – November). It is conceivable that with investments in marketing channels, input supply and a nutrition awareness campaign, processed high-value products from these value chains could also compete with their imported counterparts.

Table 4-3: Market Shares of Domestic Producers, Product-level

Product Estimated Domestic Estimated Domestic Market Share of Demand (volume) Supply (volume) Domestic Producers, % Rice 81,539 MT 1,873 MT 2.3 Chickena 11,000 MT 76 MT 0.7 Eggsa 24 million 20 million 83.33 Milk & milk productsa 2.8 million MT 24,850 MT 0.9 Yoghurtb 115 MT 0.07 MT 0.07 Juice concentrateb 322 MT 110 MT 34.2 Animal Feedb 20,000 MT 120 MT 0.6 Source: aWorld Bank Staff Estimates (2017). bStrategie Industrielle Nationale Basee Sur La Transformation Par Les Mpme/Pme Des Ressources Locales Aux Comores (2017), based on ITC (2013) data.

4.2 Export Market Trends

38. International buyers are looking to diversify procurement of vanilla, an opportunity for new cooperatives in Comoros. The global market for vanilla has continued to expand even as Comorian production has declined over the last decade. Export data for the major origins over the period 1999 to 2015 shows that volumes have increased from around 2,000 MT to the 3,000 – 3,500 MT range. Given the global reliance on Madagascar, whose market share is 80%, international buyers are actively encouraging diversification by supporting vanilla producer organizations in minor origins. Comorian vanilla cooperatives, which have recently started operating as part of a UNDP – ITC financed project and have the advantage of being located close to Madagascar, can capitalize on this opportunity to secure inputs, financing and technical assistance targeted at certifications like Organic and FairTrade, which can help secure price premiums when the inevitable market downturn occurs.

Figure 4-3: Major Markets for Vanilla Exports, 2008-2016, MT

Source: Eurostats; USDA FAS; Central Bank of Comorosxc xci

39. The demand for Comorian cloves has grown steadily in India, mainly due to a duty advantage accorded to LDCs. Internationally traded volumes for cloves typically range between 30,000 and 45,000 MT, depending on crop size and stocks at origin, with the principal markets in the Middle East and Asia, where the spice features prominently in many cuisines. The domestic demand in Indonesia – principally for kretek cigarettes – is estimated at 80,000 to 100,000 MT, and is central to market movements given that it is a major producer as well.xcii In the current market, Madagascar is the largest export origin, exporting 10,000 to 20,000 MT. Tanzania (Zanzibar), Comoros, Brazil and Sri Lanka are other major origins, all exporting in the range 2,000 – 6,000 MT. While Comoros exports to all the major markets, its exports have become increasingly focused on India in recent years, where it shares a significant duty advantage with Madagascar and other LDC origins. In fact, a significant number of Indian, Arab and Chinese traders come on a seasonal basis to establish clove buying and export operations in Anjouan. If Indian or Indonesian production ramps up significantly, exports from Comoros are likely to suffer.

Figure 4-4: Major markets for Clove Exports, 2008-2016

Source: ITC Trade Mapxciii

40. Comoros is the market leader for ylang-ylang but growing demand has led to emerging competition from new origins. Market demand for ylang-ylang essential oil has grown and prices have risen by a factor of 5x in the last decade, with high quality fractions continuing to be used in fine fragrance, and lower quality fractions increasingly finding use in aromatherapy, soaps and other cosmetics.xciv Despite these trends, production in Comoros had been declining until very recently. Madagascar is the only other significant producer, with current production estimated at around 18 MT, but none of this output is the high-quality fractions produced in Comoros.xcv Extensive plantings of ylang-ylang have been made recently in Ecuador, and plantings are also reported in Ghana and Mozambique.xcvi Indonesia and Philippines also produce Cananga oil from a different species, that is used as an alternative in years when supply from Comoros is restricted.xcvii If these origins can produce an oil that can compete with some of the fractions offered by Comoros, then prices will fall, and Comoros will likely have to increase its focus on the production of the highest quality fractions where competition is likely to be weakest.

Table 4-4: France & USA import prices, ylang-ylang oil, 2011-2017, US$ & €/kg

Unit 2011 2012 2013 2014 2015 2016 2017 France/Comoros €/kg 74.81 84.01 80.68 106.27 124.75 172.68 208.07 $/kg 103.99 108.37 107.30 141.34 138.47 191.67 235.12

USA/Madagascar $/kg 176.99 148.75 161.84 104.64 121.55 133.92 140.39

OECD exchange 0.719 0.778 0.753 0.754 0.901 0.904 0.887 rates $/€ Source: Eurostats; USDA FAS; OECD

41. Diversification of export revenues is possible with the downstream development of existing export products and niche products. The value addition of existing export products at origin is technically feasible and financially viable for multi-purpose processing facilities. As an example, cured vanilla beans can be transformed to high-value vanilla powder at a grinding facility, that could also be utilized for grinding cloves, black pepper and coffee for local and regional markets. In addition, extraction of essential oils from clove leaves, stems and buds – treated as waste currently – by adapting the ylang-ylang distillation process could lead to Comoros’ entry into established markets.xcviii The revival of older export value chains like coconut oil, which contributed a large share of revenues until the 1980s, can be supplemented with new products like coconut water and coconut sugar, which have growing global markets. Traditional and quintessentially Comorian products like sago flour are now primed for export to regional markets like Mayotte and the diaspora based in Marseille.xcix As part of the National Industrial Strategy drafted in 2017, other niche products and markets were also identified, with the fruits and fisheries segments also being represented (Table 4.5). Industry associations, incubators and the higher education sector has a key role in the development of these business operations which require the integration of commercial and agricultural skills and expertise.

Table 4-5: New Industrial Export Products and Potential Markets

Product Potential Markets Ylang-ylang essential oil Tourists, new international markets Orange essential oil EU, China Lemon essential oil EU, USA Juice concentrates France, South Africa Vanilla powder and extract EU, USA, China Clove bud, stem and leaf oils Asia, EU Coconut oil, coconut water and coconut sugar Philippines, EU, USA, India Black pepper, Clove powder India, Indonesia Smoked Fish Mayotte, Reunion, France Live crustaceans Mayotte, Reunion, France Source: Strategie Industrielle Nationale Basee Sur La Transformation Par Les Mpme/Pme Des Ressources Locales Aux Comores (2017), Field interviews by World Bank Staff

4.3 Demand from Hospitality Sector

42. An emerging hospitality sector also represents an untapped opportunity for agriculture. Among all expenditure categories, spending on hotels and restaurants grew the fastest over the 2007 – 2014 period (Table 4.6). This is a strong signal of the growing demand for food

49 prepared outside the home, an inference supported by the number of new pizzerias and bakeries in Grande Comore. However, nearly all the ingredients sourced by these establishments are imported because of the missing food processing sector in Comoros. The government strategy of stimulating tourism by attracting more airline carriers and building large hotels could bear fruit for domestic agriculture producers, as an influx of foreign tourists will raise the demand for high-quality, locally sourced perishables like tropical fruits, vegetables and fish. However, if the sector remains unable to meet this demand through production of safe, durable food products that comply with international phytosanitary standards, the current upward trend of food exports will grow unabated. In the Eastern Caribbean island countries where tourism is more developed, fresh products (fruits and vegetables, seafood etc.) account for about 60 percent of food purchased by hotels; however, most continue to be imported.c

Table 4-6: Evolution of the Price Index (base=100 in 1999)

2000 2007 2013 2014 Grains 101.4 130.1 148,6 148.3 Flours 104.5 135 Meat 102.1 103.4 124.1 124.7 Fish 103.5 125.8 149.5 150.1 Milk and dairy products 136.1 120.7 115.0 115.4 Fruit 103.1 246.1 Vegetables 102.7 160.5 Potatoes, cassava, tuber 106.6 219.1 206.1 234.2 Clothing and shoe 109.2 132.6 145.1 146.9 Health care 97.5 113 158.7 162.0 Transport 120.0 138.3 166.2 163.6 Leisure 85.9 139.8 148.3 148.3 Education 100.6 112.2 112.1 112.1 Hotels and restaurants 137.8 196.9 359.1 359.1 Total 105.9 139.9 159.4 163.2 Source: Department of statistics, Commissariat général au plan (CGP), April 20Sources of Supply: Subsistence production systems

5 Sources of Supply – I: Subsistence Production Systems

Main Messages

 Agricultural productivity is low and has remained flat for major crops in Comoros.  Soil quality and water availability are binding constraints for traditional farming in Grande Comore, but enable fertile production basins in Anjouan and Moheli.  Coconuts and rice no longer dominate cultivation, but paucity of data constrains policy and investment planning from responding to new production trends.  The cultivation of tubers, pulses and vegetables has grown but requires transfer of modern inputs, technologies and techniques to stimulate off-season supply, enhance resilience to pests and diseases, and grow smallholder incomes.  Despite immense potential, a fisheries value chain does not exist, mainly due to absence of on-shore landing and processing facilities.

5.1 Typology of Production 43. Agricultural production in the Comoros is organized around four main axes: i) cash crops for export, ii) food crops for subsistence and domestic consumption, iii) livestock production, and iv) fisheries. This chapter presents an overview of subsistence-oriented food crop production systems and fisheries, which are predominantly artisanal and serve the domestic market. The next chapter discusses existing and high-potential agricultural value chains in the cash crops, perishables (fruit and vegetables) and livestock segments. 44. The first and last agricultural census in Comoros was conducted in 2004.ci Since there are no other recent microdata sources on number, type and geographical distribution of farms, these relatively slow-changing statistics are presented below (Table 5.1). This census was instructive in that it established the following facts, which have not changed significantly in the interim period:  The majority of farming units were located in Anjouan, with Grande Comore next and Moheli last, accounting for a tiny share of units in keeping with its population share.  Women operated nearly a third of all farming units; cooperatives were nascent at the time.  The vast majority – 83.6 percent – of farming units cultivated 3 crops or less.  Extension coverage was weak, with only 8.2 percent farms being visited by agents. Table 5-1: Summary of Agricultural Production Systems, 2004

Metric Value, 2004

Number of Parcels 95,695 Number of Holdings 52,464 Farm employment 78,995 Share of holdings in Grande Comore, % 43.8 Share of holdings in Anjouan, % 52.4 Share of holdings in Moheli, % 3.8

Share of holdings operated by women, % 32.6 Share of holdings operated by cooperatives, % 1.4 Share of holdings with more than 3 crops 16.4 Share of holdings visited by extension last year, % 8.2 Source: INSEED, FAO

Given that the data on production outcomes like output, harvested area and yields is more volatile and rapidly changing, it is extracted from Central Bank and FAOSTAT databases, which contain more recent information. Unfortunately, the discrepancy between these sources mean that this data is highly tentative, particularly for food crops. It is thus difficult at present to plan, monitor and evaluate investment projects. Updated and more granular knowledge of farming systems and production is urgently required, and establishment of a data collection, monitoring and planning unit is critical for making strategic and targeted investments in the sector. 45. Crop choice and cultivation system are largely determined by altitude of production basins. In general, low altitude areas, with high temperatures and very high humidity, are suitable for food crop cultivation while at higher altitudes (above 600 meters), the climate becomes wetter, cooler and more conducive to livestock rearing, including with European cattle breeds.cii However, producers on all three islands have exhibited an extensification pattern in recent decades, shifting from fertile coastal plains to more marginal, upland, mountainous forest areas in search of new agricultural land. At these higher altitudes, polyculture systems dominate, with staple and staple, staple and cash crop, and staple and vegetable combinations being most prevalent. Monocropping is concentrated on the coastal plains between the heights of 0 and 400 meters, where it is restricted to cash crops and coconuts.ciii At higher altitudes (above 400 meters), three cropping systems can be distinguished:  Annual crops: annual crops are cultivated where land is available, on plots that might have not been kept fallow, and with very limited input of organic matter on soil that is vulnerable to erosion;  Traditional agroforestry: intercropped systems of crops and fruit/forest trees, which increase resilience to pests and help maintain soil fertility;  Plantation in natural forests: banana plantations are often installed on mountainous forest land with progressive elimination of forest trees, as demographic pressure increases.

46. Food crops are cultivated with little use of mechanization or modern inputs; marketing channels are largely informal. In all systems, the agricultural tools are simple, with simple metal tipped sticks and hoes being the main agricultural implement, and mechanization is nearly absent. The use of improved seeds, fertilizer and insecticides is rare. The bush fallow system is practiced but fallow periods have had to be reduced or abandoned as population pressure has increased. Agricultural extension institutions like CRDE have recently been established, but are under-resourced in terms of both financial and human resources. An estimated 20 – 30 percent of harvest is marketed, and this share is larger for tubers and fresh vegetables. Foods are generally sold in very small quantities. Most marketing is carried out by women and takes one of three forms: i) self-transport by taxi to urban wet markets, ii) sale to intermediary saleswomen, who then transport it to major markets, and iii) a limited barter of upland foods (bananas, yams, sweet potatoes etc.) for coconuts or fish at the coast.

47. Soil quality and water availability are binding constraints for traditional farming in Grande Comore, but enable fertile production basins in Anjouan and Moheli. The soils in Comoros are of volcanic origin, young and very sensitive to erosion, that is accelerated by the steep, rugged terrain and inappropriate agricultural practices. While the red ferralitic soils of Grande Comore are younger, deeper and have low fertility, the brown soils of Anjouan and Moheli are older, shallower and are rich in organic matter, a critical determinant of fertility. In addition, the extremely permeable soil in Grand Comore does not retain moisture and consequently, there are no fresh groundwater (rivers, lakes) sources on the island, making rainwater storage and harvesting is critical for crop cultivation and livestock farming.civ In Anjouan and Mohéli, surface and shallow groundwater is more available with the presence of seasonal rivers. However, rapid deforestation threatens the drying up of these sources as well. Table 5-2: Cropping calendar in the Comoros

5.2 Production Performance – Snapshot and Trends

48. National data shows the rise of tubers, vegetables and pulses, while international data sources systematically overrepresent coconuts and rice. The portfolio of Comorian crops, when evaluated by their corresponding land allocations in FAOSTAT data, shows a dominance of two crops – coconuts and rice (Fig. 5.1 and Table 5.3). However, trends in this data – imputed and extrapolated for several years – also show that land allocated to both crops has shown a markedly upward trend (Fig. 5.4), which is difficult to reconcile with all other data sources. As an example, the large share of paddy rice and its output – 30.3 MT – is puzzling since it cannot be reconciled with the 0.5 MT recorded in the most recent Central Bank data (2011), anecdotal evidence on the collapse of rice production, nor with customs data, which reflect steadily increasing rice imports; a similar pattern extends to coconuts as well.cv For eight other food crops, the Central Bank output is more than 100 percent larger than the corresponding FAOSTAT output, suggesting that the increases attributed to rice and coconuts are actually derived from tubers, vegetables and pulses (Table 5.3). Given these discrepancies even in

53 the same year, our assessment of the sector gives greater weight to FAOSTAT trends, value-chain studies and customs data, which is better measured because of its relative ease of collection and linkage to revenues. Since tubers and pulses did not feature prominently in these sources, this overview is restricted in its discussion of the aforementioned crop categories.

Figure 5-1: Shares of Areas Harvested, by Crop Categories Production Profile of SIDS ) 100 90 80 70 60 50 40 30 20 10 0 Cabo Verde Comoros Mauritius Sao Tome and Seychelles Share of Area Harvest, 2010 (3 year MA 2010 Areayear (3 ShareHarvest, of Principe

Cereals Oils Roots & Tubers Pulses Fruits & Veg. Cash Crops

Source: FAOSTAT. Shares do not sum up to 100 because of missing data for some categories

Table 5-3: Crop Production – A Snapshot from Most Recent Data Sources

Crop FAOSTAT Area FAOSTAT FAOSTAT BCC Output, Harvested, 2016 Yield, 2016 Output, 2016 (1000 MT) (1000 ha) (MT/ha) (1000 MT) Oilcrops 35 3.4 Coconut 33.9 3.0 101.1 40.9b Groundnuts 1.1 0.8 1.0 3.0a Cereals 26.9 1.5 Paddy Rice 23.8 1.3 30.3 0.5a Maize 3.1 2.1 6.5 12.5a Pulses 17.4 8.6 Pigeon Pea 0.5 0.8 0.4 13.2 Roots and Tubers 16.5 5.8 Cassava 11.1 6.2 68.7 48.4a Sweet Potatoes 3.1 2.4 7.4 18.8a Taro 1.5 7.3 10.8 22.1 Fruits, fresh 9.0 4.8 Bananas 7.8 5.7 44.6 53.7a Cash Crops Cloves 9.4 0.3 2.7 3.0a Coffee 1.0 0.1 0.1 Vanilla 0.1 0.2 0.02 0.04a

Vegetables 0.8 6.3 5.0 Tomatoes 0.07 9.1 0.7 7.1a Onions 0.09 2.5 0.2 1.1a Sources: FAOSTAT. aBanque Central des Comores (2011), bBanque Central des Comores (2007). Highlighted cells show a discrepancy of more than 100 percent.

49. Agricultural productivity is low and has remained flat for subsistence crops. Since 2000, the trends in yields for both food and cash crops has largely remained flat, except for cassava (Fig. 5.8). Nearly all the output growth has come about due to expansion in inputs, rather than improvements in Total Factor Productivity (Fig. 5.2). Land expansion is the input growth that has driven production increases. Even as crop choices have changed, the continued use of poor varieties and outdated techniques on marginal land and fragmented plots has contributed to a poor production base. This is compounded by the absence or retreat of the public and private sectors from backward and forward linkages – input supply, extension, processing and marketing – that can incentivize yield-enhancing investments. Given growing demand and the inability of domestic production to meet it, the low and flat yields for all crops have directly translated to an increased food share of imports, as discussed in chapter 4.

Figure 5-2: Agricultural Growth Decomposition in SIDS

Decomposing Agricultural Growth (2001 - 13) 0.0200

0.0150

0.0100

0.0050

0.0000 Comoros Fiji Mauritius Cape Verde Sao Tome and -0.0050 Principe

-0.0100

-0.0150

TFP Growth Input Growth

Source: USDA Economic Research Service

5.2.1 Cereals

50. Cereals comprise a large share of dietcvi, but yields in Comoros are lowest among comparators (Fig. 5.4). Cereals – rice and maize in the Comoros context – are traditionally cultivated by smallholders on family farms for auto-consumption, with only an estimated 20 – 30 percent of output reaching markets. While FAO data suggest that cereal production has doubled over this period, nearly all of it is explained by a doubling of land under rice cultivation, a statistic that could not be validated through other sources or field interviews (Table 5.5)cvii. The low yields exhibited by rice and maize is attributable to poor varieties, rudimentary agricultural techniques, and most importantly, the shift in cultivation to marginal, steeply sloping mountainous land with thin and degraded soilscviii. Sub- optimal fertilizer use is also one of the candidates for low soil fertility – while data on fertilizer use in

Comoros is missing since 2001, it was less than half the corresponding per hectare amount in Eastern Africa during the previous decade (Appendix Fig. 1). Lastly, the absence of linkages between maize and feed production systems reduces the profitability of both harvested maize and animal products.

Table 5-4: Cereal Yields in Neighboring Comparators, MT/ha (2016)

Crop Comoros Madagascar Mauritius Fiji Rice, paddy 1.3 4.4 2.2 3.5 Maize 2.1 1.7 6.8 1.5 Source: FAOSTAT. Yield data is calculated using imputed harvest area and aggregate (combination of estimates, official and semi-official data) production quantities.

Figure 5-3: Benchmarking Cereal Yields Figure 5-4:Cereal Production – Comoros

10000 40000

8000 30000

6000 20000

Hg/ha 4000 10000 2000 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Production (tonnes) Yield (hg/ha) Comoros Madagascar Mauritius Area Harvested (ha)

Source: FAOSTAT, ReSAKSS Database

Table 5-5: Demand and Sources of Supply for Cereals

(1) (2) (3) (4) (5) Estimated Estimated Quantity Annual Estimated Consumption Annual imported, Imports, Domestic per capita, Consumption, MT MT (2013 – production, grams/daya MTa (2016)b 2015) a MT Milled Rice 281 81,539 89,089 76,666 1,873

Wheat 74 21,473 13,833 Flour Note: Column (2) is estimated by multiplying column (1) by current population. Column (5) is computed by subtracting (4) and Japan’s 3000 MT from (2). Sources: aUSAID (2016) bBanque Central Comorienne

51. Rice, Comoros's major staple food, is barely produced in the country. Comorians consume an average of 281 grams of rice every day, translating to an annual requirement of just over 81,000 MT. More than 95% of this requirement was met through imports from Pakistan between 2013 and 2015, with Japan donating 3,000 MT of rice from Southeast Asia.cix Combining consumption and customs data sources, it can be deduced that domestic production is no more than 2,000 MT at most. This is in line with a decline from 2,900 MT, the output recorded by the government in 2007, the last year rice production data was reported. Anecdotal reports suggest that in previous decades, a domestic

56 variety was cultivated at higher elevations in Anjouan and Moheli, but urbanization and substitution toward cash crops has led to this cultivation being abandoned. In the 2000s, China too had invested in a project to re-activate rice cultivation, but production stopped at the end of the project period.

Figure 5-5: Benchmarking Rice Yields, hg/ha Figure 5-6: Rice Production Trends – Comoros

50000 35000 30000 40000 25000 30000 20000 20000 15000 10000 10000 5000 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Madagascar United Republic of Tanzania Area (ha) Production (tonnes) Comoros Yield (hg/ha) Mauritius Source: FAOSTAT Source: FAOSTAT

52. The state import monopoly, ONICOR, imports more than 95% of the rice consumed in the country. The National Office for Rice Imports (ONICOR) is the largest rice importer and the sole buyer of standard white rice 15% broken in Comoros. In recent years, two Pakistani suppliers have won the tender to supply 60,000 MT of white rice annually, at prices set by ONICOR. Private importers dealing only in more expensive basmati rice, also from Pakistan, supply the remainder of the rice imports to supermarkets or act as wholesale distributors themselves.

Office National D'importation et de Commercialisation du Riz (ONICOR)

Given Comoros’s vulnerability to high inflation and external price shocks, rice prices are critical for the economic and social stability of the country. Recognizing the relationship between rice prices and political fragility, the government entrusted the sole responsibility for its import and distribution to a parastatal organization, ONICOR, in 1982. Created by presidential decree, ONICOR’s mandate is to ensure that sufficient and affordable rice stocks are available on the market.

Since its creation, ONICOR has been undermined by difficulties. It has limited financial autonomy, as it is obliged to request funds from the Central Bank to finance purchases on the international market, which has contributed to its unstable performance. Unable to fulfill its mandate, ONICOR has frequently resorted to the granting of import licenses to private traders. The disappointing performance of ONICOR is not surprising as it reflects the experience of similar parastatal organizations in several other countries.

The performance of the rice market in Comoros can be improved by reorienting ONICOR to the role of a market coordinator and facilitator, while leaving the import and actual distribution of rice to private enterprises. In such a scenario, ONICOR would focus on the following activities: (i) setting standards for grade and quality for the import of rice; (ii) development and application of standards for private companies interested in importing rice; (iii) market monitoring activities and publication information on imports, sales, stocks and prices; (iv) collection of customs duties and sales taxes; and (v) maintaining a modest level of stock that can be sold on the market in an emergency.

The rules governing the management of the reserve stock should be transparent, with information available to the public on the size of the reserve stock, the manner of accumulation, the circumstances in which the reserve stock rice will be sold on the market, and the method to dispose of it. Opportunities should be explored to involve producers and private traders in reserve stock operations.

5.2.2 Other food crops

53. Cassava is a major source of calories and the most resilient of local food crops; introducing high-yielding and biofortified varieties can produce long-lasting food security gains. Historically a major crop, cassava output is now possibly higher than coconuts, if BCC estimates are more accurate than FAO (Fig. 5.3). Production of low-yielding varieties is currently concentrated in the coastal and inland areas as cassava suitability declines at an elevation of 500 m above sea level. Spatially explicit crop models show that the climatic suitability for cassava is expected to increase, even as the conditions become less suitable for other crops.cx Given the widespread cultivation and consumption of low-yielding cassavacxi and its forecasted climate suitability, it has high potential for rapid delivery of food security and nutrition gains. In particular, technology transfer of biofortified yellow cassava when intercropped with different grasses can be a potential solution to expand cultivation in upland areas, stabilize soil quality, increase yields and address vitamin deficiencies, an cxii approach which has been successfully developed in mountainous Vietnam .

Figure 5-8: Yields for Major Non-Cereal Crops Figure 5-7: Area Harvest for Major Non-Cereal Crops

70000 40000 60000 35000 50000 30000 25000 40000 20000 Ha

hg/ha 30000 15000 20000 10000 10000 5000 0 0 2000 2002 2004 2006 2008 2010 2012 2014 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Coconuts Cassava Coconuts Cassava Bananas Cloves Bananas Cloves

Source: FAOSTAT 2014 Source: FAOSTAT 2014

54. Potatoes are a profitable crop, but production is highly seasonal and extremely reliant on seed imports. A profitable crop which grows easily on highland production basins, potatoes comprise an increasing share of Comoros’ production profile. In an average year, 45 – 50 tons of potato seeds are imported, but this quantity fluctuates widely (upto 90 tons), driven by fluctuations in access to credit for producers and relatedly, working capital for importers.cxiii Further, only one variety of seeds are typically cultivated, with replanting and off-season cultivation of seeds being rare.cxiv Previous projects that have attempted to import potato seeds and boost production through one-time injections of capital have suffered from inadequate storage infrastructure, which in turn has led to seed spoilage due to mildew and insects.

55. Coconuts have historically been an export as well as food crop, but value addition and linkage to industrial uses is now non-existent. Once the most produced crop in Comoros, it is possible that coconut output is now lower than cassava (BCC estimates, Table 5.3). The coconut palm is planted along the entire coastal belt of each of the three islands, often inter-planted with other food crops. There are virtually no new plantations, with more than half the trees being planted shortly after the great cyclone of 1950; the rest are even older. While processed copra was used in local soap industry and exported until the 1980s, it is neither processed nor exported today.cxv Fertilizer use and plant disease control are also not prevalent, despite harvesting practices that increase the risk of infection. Pests like rodents were a serious threat in the 1980s and it is not clear if any pest-management programs are in operation today. A revival of the coconuts value chain, linked to industry and exports, requires a technical assessment of the global market demand, and specific investments needed for competitive production and export readiness.

5.2.3 Fisheries

56. Domestic artisanal fishing is below sustainable level, contributing to rising fish imports to meet growing demand. With an Exclusive Economic Zone (EEZ) of 160,000 km2, or 70 times the size of the country’s land area, the fishery sub-sector is an important source of employment, subsistence livelihoods, and animal protein among poor, coastal communities in Comoros. In 2009, this sub-sector accounted for 10 – 12 percent of GDP,cxvi employed nearly 13.5 percent of the labor force (32,500

59 jobs), and provided for nearly half the animal protein intakecxvii. Conducted by full-time and subsistence fishers alternating between farming and fishing, the activity is largely artisanal and often uses basic fishing gear and methods.cxviii. Among local fishers, traditional wooden canoes account for two-thirds of all boats and small, motorized fiberglass boats are restricted to the remaining one-third. The wooden canoes stay within a circumscribed area of 9,000 km2 on the continental shelfcxix, where fishing resources are limited, while motorized boats equipped with fish aggregating devices enable some artisanal fishers to travel further offshore and increase their catch. Consequently, while the sustainable level of annual fishing is estimated to be 33,000 MT in this area, average annual domestic production is estimated to be 20,000 – 25,000 MTcxx. Almost none of this catch is exported and in fact, fish imports to Comoros are growing (Fig. 5.9, 5.10).

Figure 5-9: Fish Trade Balance, by Volume (MT)

Source: FAO Country Review – Smartfish Program (2014)

Figure 5-10: Fish Trade Balance, by Value (‘1000 USD)

Source: FAO Country Review – Smartfish Program (2014) 57. Industrial fishing by EU vessels, a source of revenue, has halted due to a red card triggered by illegal fishing. The Comoros has signed several renewable agreements on fishing rights with the EU since 1988, which in exchange for an annual fee, allow a specific number of European boats to fish off its coast up to a maximum tonnage. Conducted by mainly purse-seine vessels from France, Portugal and Spain which fish offshore for several varieties of tuna and swordfish, these agreements have also earmarked funds for the development of a sustainable fisheries policy.cxxi. However, the EU has terminated the latest agreement in March 2018, citing i) the Comorian authorities’ flag of convenience policy; ii) evidence of illegal fishing activities by the Comorian fleet; iii) poor or non-existent monitoring and control capacities; and iv) an outdated legal framework for fisheries.cxxii

58. Processing and infrastructure facilities need to be built to capture greater value domestically. Even if the EU and Comoros were to sign a new agreement on fishing rights, industrial fishing vessels will continue to land in neighboring countries but not the Comoros due to its lack of infrastructure and processing facilities. A modern fisheries value chain, with cutting, gleaning, smoking, curing and storage facilities has high export and domestic potential, given that Comoros currently imports 500 – 1,000 MT of fish annually.cxxiii Reportedly, a tuna-processing facility is under construction, under a mixed public–private venture by Qatar and Sri Lanka.cxxiv The Government has envisioned several other steps to promote additional local processing, including the construction of larger boats, a quality-control laboratory and a new fishing school.cxxv

5.3 Priority Actions for Subsistence Production Systems

Taking a value chain approach, Table 5.6 lists targeted investments and policy reforms that are critical for enhancing farm productivity and marketability of largely subsistence products, and consequently raising food security and incomes of smallholders:

Table 5-6: Priority Actions for Enhancing Food Security and Incomes

Intervention Maize Oilcrops Tubers Legumes Rice Fisheries (Coconuts, (Cassava, (Pigeon Groundnuts) Potatoes, Peas, Yams, Taro) Cowpeas, Mung beans, Groundnuts) Input One-stop window to simplify time and cost of compliance with Facilitate Supply import exemptions access to Facilitate access to working capital for seed importers and modern distributors (CAPAC) by conditioning financing to viable fishing gear business models and Ensure delivery of time-sensitive inputs by distributors and farmer equipment federations at the start of planting season Develop small-scale rainwater harvesting and conservation infrastructure

On-farm Facilitate Promote new Facilitate access Promote access to plantations to improved, maize-pigeon improved, and biofortified, and pea biofortified, establishment resilient varieties intercropping pest- of a robust of cassava (eg. systems to resistant pest control biofortified lower yield and system yellow cassava) variability and resilient enhance soil varieties of Promote fertility maize replanting adapted to practices and off- Comoros season cultivation

Promote linkages with animal feed production systems

Collection/ Professionalization of farmer federations (SNAC, FNAC etc.) to Invest in Immediate ensure organized transport from remote production basins to processing markets infrastructure markets Invest in storage facilities to address seed and output quality and to enable yield losses landing of industrial fishing vessels

Regulatory Joining regional agricultural research and technology transfer Reorientation Invest in partnerships to develop, transfer, and disseminate climate-smart of ONICOR monitoring environment technologies, inputs and management practices that enhance to role of and control / governance productivity, climate resilience market facilities to coordinator prevent Equipping advisory and extension services with physical and and illegal financial resources to ensure rapid dissemination of climate-smart facilitator fishing inputs, technologies and management practices

Setting and Update legal Build capacity to collect, monitor and disseminate real-time application framework agriculture data by establishing an agriculture data unit. of standards for fisheries for grade and quality of imports by private companies

Collection of customs duties and sales taxes

Maintaining a modest buffer stock for an emergency

6 Sources of Supply – II: Agricultural Value Chains

Main Messages

 Cash crop production employs 18 – 28 percent of producers and is expanding in response to surging prices. Strategic investments in quality control, downstream products, and downstream development are required to mitigate against an impending downturn.

 Smallholder growers of cloves and ylang-ylang are receiving the majority shares of export prices in currently favorable market conditions.  Vanilla growers receive much smaller shares of export prices, as their bargaining power is reduced by an exporter oligopoly and a government determined price structure.

 Two livestock value chains are promising – while eggs are the most competitive value chain in Comoros, dairy is developing from a small base in Anjouan.  Local chicken meat production is small and frozen chicken wings, a residual from European demand for lighter chicken meat, are 50 – 100 percent cheaper.

 Market development for domestic poultry meat requires a nutrition awareness campaigns to raise the demand for lighter chicken meat, coupled with supply linkages to maize production for feed, and a robust veterinary system.

 Vegetables are seasonally competitive in segmented markets, but greater domestic competitiveness is constrained by erratic input supply and poor connectivity.

6.1 Introduction

This chapter assesses the current competitiveness and the main opportunities for achieving scale in selected value chains with large economic footprints and high potential for enhanced domestic value-addition. According to the state of development of their respective segments, each of the value chains is classified in the following categories:

 Established: Vanilla, Cloves and Ylang-Ylang,  Growing: Poultry and Dairy  High Potential: Fresh fruits and Vegetables

Rapid reconnaissance interviews of more than 120 representative producers, collectors, processors, traders, industry representatives, and government officials were conducted to collect primary data on the process, geography and margins for each value chain. The structure of production costs and the profitability of a representative production unit is also evaluated, thus providing an insight into the drivers of economic viability and competitiveness. In conclusion, this chapter summarizes production, processing, and marketing patterns, while outlining gaps, structural bottlenecks, policy options and investment opportunities for each of the three categories. Detailed background reports are available for all value chains, segmented by cash crops, livestock, and fresh fruits and vegetables, which correspond to the established, growing and high potential classifications.

Table 6-1: Value Chains at a Glance

Farm Collection/Intermediate Value-added Retail/Export production markets processing Established Ylang 1000 – 3000 Flower collectors; Distillation 2 Buyers/Exporters ylang Smallholders; Essential oils collectors ~ 500 small units (Bernardi, Biolande) 18 Cooperatives; aggregate from distillers 2 Plantations; High share of women Cloves 11,000 – 12,000 Collectors deliver dry Drying and Exporters Associations Smallholders; cloves to exporters removing stems (Anjouan and Moheli) 5 Producer High share of Organizations women Vanilla 5,000 Collectors of green beans; Curing 3 major exporters Smallholders; Collectors of cured beans 6 – 12 curers; (Vaniacom, AGK Sarl, 3 Cooperatives; Controlled by 3 AHH Export); High share of exporters 17 minor exporters women

Growing Milk 4,140 Dairy Mostly direct sales; Pasteurization Consumption centers in farmers and Some collectors are also and packaging Domoni and Mutsamudu 9,450 cows final retailers; Minimal: Informal traders between Niumakélé dairy Anjouan and Grande plant not fully Comore operational Eggs 155 Semi- Non-existent Non-existent Local shop owners, market industrial vendors, some door-to-door intensive units; sales Several thousand backyard units

High Potential Vegetables Subsistence Weakly developed: Non-existent Traditional wet markets; producers Central purchasing market Seasonal market during in Moroni; Ramadan and Grande Local sourcing by one Marriage; supermarket chain Occasional export to Mayotte Fruits Subsistence Non-existent Jam, juice and Roadside sellers and wet producers pickling markets Minimal: Very few small-scale units

6.2 Established Value Chains

59. The three cash crop products have already established segments and links along the value chain. 12,000 – 20,000 smallholder producers (18 – 28 percent of all agricultural producers) are engaged in cash crop production and form the base of each value chain (Table 6.1). The complexity of processing varies across value chains – while cloves are exported in bulk as the simple dried spice, with no value-added processing taking place, vanilla is exported in the form of the cured bean and ylang-ylang oil is exported according to fractions, a proxy for quality. Green bean vanilla producers sell to one three major exporters who control the processing and collecting segments as well. Ylang- ylang has a dynamic processing sector comprising several small enterprises that are struggling to adapt to changing market demands.

6.2.1 Cloves

Production Trends

60. Cloves are the biggest source of export revenue and production has ramped up over the last decade. In an average year, 3,000 – 4,000 MT are produced in Comoros, with output going up to 6,000 MT in peak crop years like 2016 (Fig. 6.1). This level of output is behind the price-setter Madagascar, which typically exports 10,000 – 20,000 MT, but comparable to Tanzania (Zanzibar), Brazil and Sri Lanka, which exporting 2,000 – 6,000 MT, with the broad range reflecting typical annual variability in clove harvests (‘on’ and ‘off’ years).cxxvi As a caveat, these trade data estimates for Comoros are more than 50 percent higher than their mirror in customs data, or in field estimates collected from exporters. However, the trend of cloves maintaining its productive base is supported by all data sources. Field reports of new plantings in Anjouan and Mohéli suggest a slow future expansion of this base as well.cxxvii The downside of new plantings, especially on land traditionally used for food crop cultivation, is that clove trees are an effective monocrop after 3 – 5 years and this pattern is likely to increase exposure to price risk and food insecurity.

Figure 6-1: Clove exports from major origins, 2008-2016, MT

Source: ITC Trade Map; Madagascar Central Bank. See Annex 3 Table A4 for data table

Quality

61. Comoros produces cloves without any unique attributes or markers of high quality due to the absence of domestic standards and weak post-harvest management. Clove quality is determined by the integrity of the post-harvest management – collection, drying and storage processes – that take place before it is shipped.cxxviii At all three stages, there is significant loss of quality due to the absence of domestic specifications, in contrast to Madagascar and Tanzania.cxxix The example of Tanzania is particularly illustrative in that Zanzibar State Trading Company (ZSTC), a parastatal that has a monopoly over purchase and export of cloves, runs a comprehensive network of well-equipped buying centers in clove producing regions, has established grades and transparent uniform pricing for clove purchases and centralized mechanized cleaning, grading and packing facilities. It also operates a large distillery for production of clove oil and fractionation to eugenol, and has been developing retail packaging of selected cloves for distribution in the Middle East markets. Consequently, Zanzibar has a reputation for high quality product meeting all international standards and is able to deliver a high share of the export price to producers.

Structure

62. This value chain is short and comprises of the following segments: producer, collector, exporter. The producer is the owner of the clove trees and sells dry cloves. Harvesting of cloves is typically done on contract, with harvesters paid by weight of cloves harvested. The producer is responsible for drying, and this may also be done on contract. Typical payment rates for the producer are:

 Harvesting: KMF 200/kg fresh cloves with stems; KMF 300/kg fresh cloves with stems removedcxxx  Removal of stems: KMF 100/kg  Drying: KMF 100/kg dry clove outturn

Margins

63. In currently favorable market conditions, cloves growers are receiving more than three- quarters of the export price. The costs associating with harvesting, removal of stems and drying are in the range KMF1,200/kg dry cloves. At a buying price of KMF 2,500/kg, the margin to the producer is KMF 1,300/kg (€5.09; US$6.11). At an estimated current FOB export price of US$7.80/kg, the producer is getting 78% of the export price. This is a reasonably high share, considering exporter weight losses due to collector activities. However, harvesting, drying and cleaning (and collection) costs are largely fixed, and at significantly lower export prices, as there has been in the past, the margins to producers are small and unattractive, resulting in producers leaving much of the crop unharvested. The current strong crop volumes, and new plantings, are driven by the current relatively high market prices that reflect growing international demand (particularly Asian demand) for cloves.

Collectors are paid KMF 50/kg cloves delivered to the exporter, plus additional allowances for transport and handling costs, giving an overall allowance of KMF 70/kg. However, the major complaint of exporters is that the collectors manipulate the product to add weight, damage the product in handling to reduce quality and increase losses, and reduce the stated price to producers, all contributing to increase the collector margin significantly at the expense of both producer and exporter. Inefficiencies in the supply chain, principally collection, offer some scope for improving margins to production, but further improvements require valorization of the waste streams and development of value added products.

Geography

64. Clove production is dominated by Anjouan, followed by Mohéli. Grande Comore has cloves, and could be a substantial producer, but labour costs for harvesting limit production potential.cxxxi Mohéli production has been estimated as typically in the range 600 to 1,000 MT (around 20% of the national crop) with Anjouan providing almost all the rest. Cloves are widely distributed in Anjouan, with centres of production in the west (Sima), the central belt from Mutsamudu across to Bambao, the NE coast, and the south. There are no formal survey data on the number of producers, but it is stated (the basis for this is not known) that three-quarters of Anjouanais producers (185,000) are involved at some level in clove production. Smallholder producer numbers must be in the range 10,000+.

6.2.2 Vanilla

Production Trends

65. Production of cured vanilla is slowly picking up after a long-term decline, falling from 150 – 250 MT in the 1990s to 20 – 30 MT in recent years (Fig. 6.2). Although reliable export data is not available, and different sources give conflicting data, the general picture of a production collapse to such an extent that Comoros is now a minor origin for vanilla, is clear and realisticcxxxii. Driven by extended periods of low prices (US$20 to US$40/kg) since early 2000s, when the global market has expanded, this decline is slowly reversing and export volumes are expected to be 30 – 35 MT in 2017, in response to a big spike in international prices. Since new plantings take 3 years to start bearing beans and yields build slowly, export volumes are projected to grow over the next few years if new plantings are maintained and old are replaced, which in turn is a function of prices. However, the limited availability of existing planting material severely constrains the rate of planting expansion.

Figure 6-2: Vanilla Exports from Major Origins to Major Markets, 1999-2015, MT

Source: USDA FAS; Eurostats; Comtrade TradeMap; WITS; Madagascar Customs

Quality

66. Comoros vanilla has a reputation in the market for producing high quality vanilla, but local producers have not successfully capitalized on this differentiation. Classified as the ‘Bourbon’ label, the vanilla from Comoros has a strong traditional association in the market for quality. Cured bean quality is underpinned by the control of harvesting dates, agreed upon every year by sector representativescxxxiii and Government, to ensure that harvest does not begin before green beans are mature.cxxxiv Given the ongoing boom, schemes aimed at product differentiation and marketing have broken down, as the returns to quality are marginal in current market conditions.cxxxv However, the development of Organic and FairTrade certifications are strategic measures that can help producers access niche markets and capture price premiums, when prices return to their long-term trend after the current uptick. In addition, the long-term development of value-added downstream products like vanilla powder and vanilla extract could be targeted by exporters and vanilla cooperatives, in order to capture greater value domestically.

Structure 67. The traditional value chain is undergoing consolidation, with producer-curers and curer- exporters likely to be the future structure. Traditionally, smallholder producers of green vanilla, collectors of green beans, préparateurs (curers), collectors of cured beans, and exporters comprised the value chain. In this structure, the producers are not contracted and green bean collectors are financed by curers, who in turn are either self-financed or financed by the exporter. In the latter scenario, the exporter provides funds for purchase of green beans at an agreed price, but does not pay an advance fee for the direct costs of curing and the curers’ margin. Collectors of cured beans work for an exporter. The traditional structure of the sector is breaking down under the influence of a number of pressures, the key ones of which are:

 The very low current production levels: 40 to 50 MT cured vanilla can be produced by an individual enterprise in other origins;  There are only 3 exporters, and they are becoming significant curers, developing integrated curing-exporting businesses. Exporters are at a competitive advance to the curers that they

finance, as they set the green bean price for the curers but can themselves offer higher price (making use of the margin to curing);  At current prices, very large sums are required to finance curing operations. For a minimum commercial scale of 1 MT cured/5 MT fresh, with green beans at KMF 25,000/kg, over €250,000 is required for bean purchase. Few traditional curers have the assets and capability to access and manage this level of finance, such that there are almost no self-financing curers operational. The Association of Préparateurs had 300 members in 2004; current membership is 72, but only a handful are active and almost all of these financed by exporters;  Producers have been encourage to organize into groups, and more recently Cooperatives (under the UNDP/ITC Competitiveness Project)1. Whilst there is some advantage to curers to deal with organized producer groups for ease of securing and contracting for larger volumes, the market and price advantage to the group is limited. The principle driver for the formation of producer groupings is for them to take over the curing role and capture the significant added value to the production of cured beans. The result of these changes is likely to be a sector made up for producer-curer groups, and curer- exporters with very few independent curers, with exporters also contracting out curing if and when volumes exceed their installed capacity. Margins 68. Growers of green bean vanilla are benefiting from the price boom for cured vanilla, but receive only 15 – 20 percent of export prices due to export oligopoly and government-determined prices. Prior to the start of each season, the government and sector representatives set a price structure (Structure de Prix) listing prices, costs and margins at all stages from purchase of green beans to cured vanilla, which is exported. In practice, this structure is only used to set a marker for the starting purchase price for green beans and a negotiating position for the costs and margins to curing.cxxxvi In 2017, the Structure de Prix set the green bean price at KMF20,000 but rose to over KMF30,000 and the FOB export price was calculated at €310/kg, but in practice, almost all cured vanilla will be sold at €450 to €550/kg and above. In the current climate of volatile prices, gross farmer incomes have fluctuated between KMF1.875million (€3,800) and KMF3,500/kg only €535 over the space of last two years. It is unclear whether margins to production remain sufficiently attractive to growers to continue with the crop when prices fall back from their current peak. Margins can be maximized by producer cooperatives developing curing operations, and developing Organic and FairTrade certifications to capture price premiums.

Geography

69. Grande Comore dominates the production of vanilla, accounting for an estimated 80% of total production. Mohéli accounts for the bulk of the rest. The Agricultural Census of 2004 recorded 8,584 vanilla producers in Grande Comore, but this number has declined as many growers switched to other crops with the following decline in prices. A recent survey reported 4,300 growers, although it is likely that some of these are not actively growing vanilla but did in the recent past, and still consider themselves potential vanilla growers.cxxxvii In Grande Comore, vanilla cultivation is spatially dispersed and not limited to particular zones.

1 Amélioration de la Compétitivité des Exportations des Filières de Rente (ACEFER)

6.2.3 Ylang-ylang

Production Trends 70. Like vanilla, ylang-ylang essential oil production is starting to rebound after a long decline. In the 1970s and 1980s, Comoros produced 110 MT annually, which dropped to 90 MT in the 1990s, and reached a low of 24 MT in 2014 (Fig. 6.3). Production has since started to increase, with nearly 38 MT being exported in 2016 and an expected 40 – 45 MT by the end of the 2017 calendar year, almost exclusively to France. Although low prices and the supply response – removal of ylang trees and lack of replacement of aging trees – were the primary cause of the long-term production decline, the simultaneous decline in field management practices also contributed to a an inability to rapidly take advantage of an upturn in prices.cxxxviii The recent revival in production has been stimulated by a continuous increase in demand for all oil qualities, supported by the direct involvement of overseas buyers, NGOs and users of the oil to modernize distillation techniques, reduce costs, and stabilize the sector.

Figure 6-3: Ylang-ylang Exports from Comoros & Mayotte, 2008-2014, MT

Source: Eurostats; Comoros Central Bank

Quality

71. Manipulation of ylang-ylang oil quality has contributed to price declines. Ylang-ylang oil is separated out into 5 fractions of increasing density, which is one of the parameters most correlated with perfume aroma and quality.cxxxix In Anjouan, the fraudulent manipulation of quality by heating, which can increase density and result in a doubling of value, has become widespread.cxl Since Comoros’ competitive advantage is built on its ability to supply the highest quality fractions, the inability to detect quality locally has led to reduced prices for Comorian oil. Given the development of field tests which enable spot testing of quality are underwaycxli, and Anjouan’s current moves to develop quality controls at the point of exports, it is likely that the industry will be able to solve this problem in the near future.

Structure

72. Like vanilla, the traditional value chain is consolidating, with distillers grouping to cut out collectors. The traditional segments of the value chain are: grower, distiller, collector, exporter. Many distillers are also growers, and will buy in additional flowers for distillation as required. Distiller – collectors contract pickers for the harvest of flowers, with women supplying a large share of labor and typically picking 10 – 20 kg daily during the harvest period. At the next stage, collectors of essential oil aggregate and transport the product to the exporters. Several distiller groupings which cut out external collectors and supply directly to exporters, have recently been established on both Anjouan and Mohéli. These groupings are expected to increase in number and size, given the benefits associated with organizing, in terms of access credit and economies of scale. In particular, the ability to access modern, stainless steel stills of varying sizes and maximize capacity utilization is a key optimization benefit of grouping. Additionally, access to improved furnaces and sites with good water supplies in another important benefit.

Margins 73. Growers of ylang-ylang flowers are receiving a large share of margins, with distillers facing the most financial pressure. In the 2017 season, flower prices ranged between KMF 200 and KMF 400/kg. Nearly all the benefits of the recent rise in flower prices were retained by the growers, with the large base of growers securing 62 percent of the value and distillers retaining 22 percent. The margins currently accruing to distillers are insufficient to finance investment in improved facilities necessary to improve yields and quality outturn, and address the urgent constraints of reducing wood consumption and securing access to water for distillation.cxlii Two exporters – Bernardi and Biolandes – which are also importer/suppliers for French compounding companies, dominate the ylang-ylang value chain. Consequently, there is no exporter margin in this supply chain, a structural point of difference, when compared to the cloves and vanilla value chains. Additionally, there is a high degree of vertical integration, with both companies being directly involved in distillation (Biolandes – Anjouan, Bernardi – Grande Comore), or buying from distillers.

Figure 6-4: Margins across Ylang-Ylang Value Chain

Source: World Bank Mission estimates (December 2017)

Figure 6-5: Margins across Improved Ylang-ylang Value Chain

Source: World Bank Mission estimates (December 2017)

Geography

74. Production of ylang ylang occurs on all 3 islands, but is dominated by Anjouan. Current estimates for 2017 production of an estimated total of 40 to 45 MT is: Anjouan 25 to 30 MT; Grande Comore 10 MT; Mohéli 5 MT2. Growth in production is being led by Anjouan, followed by Mohéli. Production on Grande Comore is domination by supply from the Fernand/Bernardi plantation at Mitsamiouli (200 ha), followed by CVP at Mbeni. Assuming average smallholder oil yields of around 30 kg/ha3, the total crop area is currently around 1,300 ha. A 2006/2007 survey in Anjouan recorded 449 working stills, 424 of which were owned by individual distillers, with most distillers undertaking 6 to 7 distillations per month. Distillation units were mostly old, and of poor quality (280 made of galvanized iron, 90 of copper, 75 of aluminium, and only 3 of stainless steel) with 50% of the galvanized stills assessed as ‘in bad condition’. The average age of distillers was 47 yrs. An earlier survey (1998) recorded 588 stills of which 427 were functional and 636 distillers. Current estimates are that there are around 250 functional stills in Anjouan, used by a larger number of distillers, up to 400. There are 2 industrial distilleries on Anjouan, one at Domoni (with 10 stills) and one at Bambao. There is no survey data on the number of ylang growers in Anjouan, it is simply stated as ‘several thousand’.

6.3 Growing Value Chains

75. The livestock sector is growing from a small base, especially when compared to food and cash crops. There are three main livestock value chains in the archipelago: the cattle dairy value chain,

2 Anjouan and Grande Comore make direct exports from their ports. Production from Mohéli is mainly transported to Grande Comore for export, but can also be sent to Anjouan. No product insurance cover is available for the crossing from Mohéli – this is at the risk of the product owner at this point. 3 Based on poor average flower yields of 5 kg/tree/yr, and 2% distillation yield.

72 the small ruminants, and the poultry value chain. The bovine meat value chain barely exists and most of the meat produced locally can be considered as a by-product of dairy or mixed systems. The local milk value chain is very rudimentary with no processing of local milk, and very basic marketing channels. The main constraints faced by the local dairy sub-sector are different from one island to the other: in Anjouan, production systems are performing well but the supply now exceeds the market demand; in Grande Comore, this is the opposite, and farmers are unable to produce what the market requires. Intensive poultry production (for eggs) is productive and the potential for substitution to imports is very significant. The potential for job and business creation in inputs and service provision (feed, hatcheries, veterinary) is also important. The most significant productivity bottleneck is the planning and organization of the supply chain for day-old chicks, which are currently imported. Orders often arrive several months after what was initially planned by the farmers, causing laying hens to exceed the normal laying period of one year. A longer laying period decreases their performancecxliii, but farmers keep the hens to avoid losing their customers and market, even if the value of egg production does not cover the feeding cost. If production inputs were reliably available, then the technical performance of poultry farms in Comoros would be comparable with what is observed in other tropical island countries.

6.3.1 Poultry

Production 76. The commercial poultry value chain is young, market oriented and local egg production is competitive with imported eggs. This value chain is already competitive against imported eggs in Grande Comore (price differential of less than 10 percent) and has the potential to be competitive on all islands, without a welfare loss for consumers in terms of prices or quality of eggs. Simultaneously, a traditional poultry value chain also exists on the islands; in fact, free-range backyard chickens are kept by nearly every rural household. Given the widespread prevalence of these two value chains, the potential for job and business creation in input supply and service provision is also significant. 77. At present, locally produced chicken meat represents less than 1 percent of consumption and is not competitive with imports. Most of the imported chicken meat (11,000 MT/year) is in the form of frozen legs and wings from livestock powerhouses Argentina and Brazil. Reflecting the pattern observed across several African countries, local chicken prices are 50 – 100 percent higher than frozen chicken wings, which are often sold at below cost given European preference for light breast meatcxliv – and more than 30 percent higher for whole chicken. cxlv Demand is high and rising, as domestic consumption level – 14 kg per capita annually – is twice the Sub-Saharan Africa average. In comparison, local production is just 16 MT from broilers and 60 MT from layers. Prices do not fluctuate much over the year, despite the seasonal increase in demand for poultry meat in July-August, because of temporary price control mechanisms put in place by the government. Figure 6-6: Estimated Quantities of Poultry Meat by Origin (MT per year)

Locally produced poultry meat Imported poultry meat Number of actors 155 producers 2 to 3 importers (80,000 layers, 16,000 broilers) Quantity of meat 76 MT 11,000 MT

16 MT of broiler, 60 MT of cull (mostly wings and legs) hens (whole chicken) Share of the market 0.7% of market 99 % of market Source: World Bank mission estimates (December 2017) and FAOSTAT (imports)

Structure

Figure 6-7 illustrates the three main segments in the local value chain – input suppliers, producers, and retailers, who are supplied directly by producers.

Figure 6-7 : Local Poultry Value Chain Map – Main Actors per Level

Inputs & services Production Retailing

•Veterinary • Type1 producer • Shop owners technicians (less than 100 • Market vendors •Input traders animals) •Input retailers • Type 2 •Feed producer (100 manufacturers to 1,000) •Hatcheries • Type 3 producer (over •Banks and MFIs 2,000 animals)

Typology of production systems: All intensive poultry producers follow similar production itineraries (same breeds, same inputs), and sell their products on the same market, but their level of technicity varies, as well as the size of the farms and the importance of investment. A simple typology of producers has thus been established, based on these criteria: Small-scale (less than 100 animals): This type of producer is more commonly observed in Anjouan than in Grande Comore. Most of these small units are owned and managed by women. This is an additional economic activity that supplements other sources of livelihoods (usually agriculture is the main source of incomes). The investment is minimal, and the small poultry house is often built with local materials (except for the roofing). In Anjouan, these small units are often built of the family house rooftop, which ensures a good protection against thieves and predators, and avoids spending in a specific concrete slab. Chicks, feed and inputs are bought from specialized input dealers, often collectively (all women from the same village source inputs together). However, because of the limited volume of inputs they require, these units face difficulties in organizing an effective supply chain and thus often have feed and chicks’ shortages. Marketing of eggs (and cull hens) is usually done within the village directly to end customers. Medium scale (100 to 1,000 animals): This is the most prevalent type of production, which supplies the vast majority of local eggs in Comoros. Around 30% of these units (usually the smallest ones, under 500 animals) are owned by women. This type of unit is often the main or even the only economic activity of the owner. The poultry house is more elaborate, investment is significant and has often required a bank loan. Marketing of eggs is done within the village, or through the shops in the area.

Large-scale entrepreneurial farmers (over 1,000 animals): The majority of farmers under this category have 2,000 – 3,000 animals, which is probably the economically optimal size. (the maximum size is 11,000 layers). Setting up these units have required important investments and most of these farmers got loans from MFIs. The workforce usually combines family members and employees. Margins 78. Local eggs occupy 83 percent of the market share in Grande Comore; simplification of import exemptions can lead to total import substitution. The price differential – about 10 percent – in favour of the imported eggs from India is small enough that domestic producers are competitive.cxlvi The lower quality of imported eggs – often arriving in Moroni after three weeks of shipping time – could justify a price premium, but public awareness of Comorian customers on food quality issues and the demand for safe products is not differentiated enough to create two different product markets. The small price differential in favour of imported eggs explains why their market share is 17 percent in Grande Comore. While market shares in Anjouan and Moheli could not be estimated, it is possible that imported eggs comprise a larger share of local consumption in the other islands. Since imported inputs account for 80 percent of the farm gate price of eggs, simplification of exemption procedures can enhance competitiveness on all islands, and possibly lead to total substitution of imports.

79. The net profit and income per worker in a medium-scale egg farm is equivalent to the salary of a mid-level civil servant. Given the short and simple nature of the domestic value chain, a classic analysis of distribution of margins across segments is not meaningful. The poultry farmer is the only value chain actor for which a detailed economic analysis is possible and meaningful. The economic model (see Annex) developed below corresponds to a standard “Type 2” egg producer, with 700 layers (which is very close from the national average of 620 animals). All the data used in this model are information collected by the mission in the field. The model is calculated on a standard cycle of 18 months (6 months of growth, 12 months of production). This model confirms the general sentiment expressed by several interviewed producers, who report egg production being a profitable activity. The net profit over a production cycle of 18 months is 2.7 million KMF which corresponds to 150,000 KMF per month (365 USD). Since such a small unit requires only one person to run, the remuneration is therefore, 150,000 KMF per month, equivalent to the salary of mid-level civil servant. If the size of the unit increases, economies of scale will be generated, especially on the investment. A unit of 1,500 layers will therefore generate a net margin of around 360,000 KMF per month (880 USD).

Box 1 : Case Study of a Medium-Scale Poultry Farmer Mr M. is 40 years old and lives in Anjouan, near Domoni, where most of the poultry farmers of this island are concentrated. He was a high school teacher until recently, but he was tired of recurrent delays in the payment of his salary and he decided to start his own business. He had a family plot with water, not far from the city, which was too small to do agriculture, but ideal to do poultry farming. He was encouraged to start this activity by the fact that many of his neighbours recently started this business and seemed very satisfied by the results. He applied to SANDUK for a first loan of 2 M KMF (5,000 USD) to build his first poultry house for 300 layers and got the loan quite easily. After one year, he took a new loan for another 300 layers building. He buys his feed locally from private traders, and his chicks from Mauritius (he places orders together with other farmers). Getting chicks on time is not always easy. Feed supply is not an issue in Anjouan because there are several suppliers. He has a very good productivity; the mortality is low (5% over one year) and he has never encountered any big problem. He has now resigned from the public service and works full time in his poultry farm, which he says generates better incomes than his previous job.

From this model, it is possible to approach the production cost of an egg produced locally: our model gives a value of 60 KMF for a 700 hens farm. The added value stands at 10 KMF per egg, which gives a total added value generated by the local egg production sector of 200,000,000 KMF (around 500,000 USD). The graph bellows represents the distribution of the production cost (60 KMF per egg) of a locally produced egg by type of expenditure. Only traded inputs are considered in this analysis. The feed represents 80 % of the production cost, and the layer feed only 75 %. In total, imported inputs represent 93 % of this cost.

Figure 6-8: Production Cost of a Locally Produced Egg – Distribution by Type of Expenditure

small equipement (depreciation), 1% day old chicks, 6% investment building (depreciation), 4% feed starter, 2% water, 2% egg trays, 5% feed growth, 3% Vet products and care, 1%

feed layers, 75%

Source: World Bank Mission (December 2017)

80. Total import substitution of poultry meat through trade barriers is unfeasible, given large price differentials, and potentially welfare-reducing impacts of a large price increase. At present, domestic poultry meat is not competitive with imported frozen chicken wings, so widely consumed that they now have a local name -- “mabawa”. Even if import taxes on poultry products (19 percent) were to double, in line with non-strategic food imports, imported wings would remain cheaper than domestic poultry meat.cxlvii Only local cull layers are competitive with imported broilers, but domestic quality is much lower and requires a longer cooking period than a grilled preparation that is now preferred. Paying 50 – 100 percent more for local chicken is only acceptable by customers in very specific conditions, especially for ceremonies such as Grande marriage during July and August, when demand surges by up to 4x and the consumption of fresh whole chicken is preferred.cxlviii

81. However, public investments in nutrition awareness, feed production and animal health can lead market development and partially substitute imports for domestic poultry meat. In the absence of public awareness of the nutrition benefits of light meat, consumer preferences for dark meat drive the demand for frozen chicken wings in Comoros. A public nutrition awareness campaign could shift demand towards more expensive light meat, against which indigenous broiler meat is likely to be more competitive. Simultaneously the development of local feed production, with backward linkages to domestic maize production systems and fisheries, can substantially lower the cost and uncertainty associated with imported feed, thus stimulating supply. Lastly, supporting the provision of veterinary care and inexpensive vaccination against diseases like the deadly Newcastle disease – common in domestic chickens – is critical to prevent shocks that can rapidly deplete a small broiler population. Together, these program elements have the potential to develop a market segment for domestic poultry meat. Even if market-oriented meat production remains competitive against imports, this program can enhance the viability of backyard poultry production systems, which serve as a source of livelihoods and nutrition for remote communities disconnected from urban markets.cxlix

Geography 82. About 155 semi-industrial egg production farms are currently in operation, with Anjouan and Grande Comore accounting for most of them (Table 2).cl Most farms are less than 15 years old and exhibit good productivity. While the production units are younger and smaller in Anjouan, it has the most egg producers. Grande Comore houses more than half of the intensive poultry population and has the highest average size per farm (1,000 animals). Most of the farms are located around Moroni. In Anjouan, the development of the intensive poultry sector is more recent, and this is probably why units are much smaller than in Grande Comore. Anjouan has the highest number of farmers, and most of them are concentrated in Domini area. Mohéli has an average size per farm which is intermediate, but the number of farmers is much lower than in the two other islands.cli Table 6-2 : Number of Farms and Animals Per Island

Grande Anjouan Mohéli National Comore Number of intensive poultry units 60 80 15 155 Number of layers 40 – 50,000 25 – 30,000 7,000 80,000 Number of broilers 12,000 4,000 n.a. 16,000 Average size of farms (animals/unit) 1,000 400 750 620 Source: FAO Comoros 2016 and ASAVIC 2017 6.3.2 Dairy

Production

83. An intensive smallholder dairy Figure 6-9: Animal Ownership, by Island sector exists in the highlands of Anjouan and is emerging in some areas Animal Ownership of Grande Comore. The dairy value By Island 24 chain shows very different characteristics 25 in Anjouan and Grande Comore. 97 20 20 percent of dairy producers are in Anjouan,

15 13 where nearly a quarter of the households 12 Percent

own a cow (Figure 6.9), and the intensive 10 6.1 dairy production is concentrated in the 5.1 5.3 5 3.2 2.2 Niumakélé production basin. Dairy 1.1 .19 .7 0 production is well integrated in the Moroni Reste Ngazidja Ndzouani Mwali agricultural production, intensive and Cow Rabbit productive, with excess supply on the Sheep Source: 1-2-3 Survey (2014) island. However, there is no processing of local milk, marketing channels are rudimentary, and almost no milk reaches Grande Comore’s markets. In Grande Comore and Mohéli, the development of modern dairying is still recent, and farmers face critical capacity deficits that limit the performance of the production system. Cumulatively, local dairy production accounts for a less than 1 percent of the local consumption of milk and milk products at present. However, given the relatively high production in Anjouan and the growing stock of dairy cows in Grande Comore, the potential for local value addition, competitiveness with imports, contribution to livelihoods and the complementarities with other agricultural production is high. Table 6-3: Import of Dairy Products in 2013 (Source FAOSTAT)

Quantity (MT) Quantity % of quantity Value Liquid milk 1,383 1,383(MT Eql) 51(Eql) % 1,184(K USD) Milk powder 121 920 34% 590 Milk condensed 124 310 11% 169 Butter 7 49 2% 37 Cheese 22 66 2% 149 Total 2,728 2,129 Source: FAOSTAT 2013 Structure

84. Dairy value chains are usually complex, involving a wide range of value chain actors especially at the stages of aggregation, collection, processing and distribution. While the value chain is very simple in Comoros, there are significant regional variations between Grande Comore and Anjouan. The figure below gives an outlook of the various categories of actors involved in the local dairy value chain. Figure 6-10: Dairy Value Chain Segments

Inputs & services Production collection and Retailing transport

•Veterinary • Ngazidja • Milk collectors • Milk vendors technicians smallholder and transporters • Shop keepers •Input retailers producer •Banks and MFIs •Anjouan smallholder producer

Anjouan Producers: Most of the milk producers in Anjouan are located in the Niumakélé area. Niumakélé is a high-altitude area (500 to 750 meters) characterized by a very high population density (around 1,000 habitants/km2) and by intensive and highly integrated agricultural production systems. Dairy production has been introduced several decades ago in this area and is now well established (90 % of farmers own dairy cows) and integrated in crop production systems as well. Farmers usually own 1 to 3 cows, plus their progeny.clii Milk collectors: in Anjouan, specialized collectors tour the Niumakélé area every morning to collect milk from farmers from fixed collection points. These collectors then transport the milk to the consumption centers in Domoni and Mutsamudu. Given the absence of a cold chain and quality control, part of the milk is soured and curdled and part is sold raw.

Grande Comore

Producers: Intensive dairy production has been introduced in and around Moroni,cliii with the help of cows imported from Tanzania in a small project implemented by a local NGO (GAD) and MECK Moroni, in 2013-14.cliv Dairy farmers are thus novices, with very limited experience in the business. A second group of farmers in the same area have imported cows from Anjouan, with generally more success because adaptation was easier.clv In contrast to Anjouan, animals are kept in zero grazing systems but under cow sheds, in small enclosures.clvi The inadequacy of production practices and poor animal performance is clearly related to poor capacities and knowledge, and lack of experience of these farmers.clvii This inference is supported by success of a handful of farmers, who are followed by specialized technicians, and apply proper production itinerary: their animals are in excellent conditions, and their production performance are fair. Collectors: Since the dairy basin in Grande Comore is close to the main consumption center (Moroni and urban areas), and some of producers do not require any collection services and sell directly from home. However, most of the farmers sell to collectors, who pass by villages in the afternoon to collect fresh milk. These collectors are also most of the time the final retailers, they sell the milk from their home which is also their sales outlet, at 1,250 KMF per liter. The marketing channel is very short, and the milk is sold raw (not soured), without being refrigerated.

Moheli Producers: In Mohéli, the farming system is very similar to Grande Comore, but animals were all imported from Anjouan and their adaptation to local conditions was easier. In Grande Comore and Mohéli, artificial insemination services are not available, and farmers have to use bulls. There are very few bulls of good breed available and the price of service is therefore high: 20,000 on average. Traders

Given the large price differential in milk prices between Anjouan and Comoros, some collectors informally send part of their milk to Moroni by boat, using small jerrycans of 5 liters. The milk is soured and transported without refrigeration. Due to the unpredictability of sea transport, this marketing channel is not organized and only involves very small quantities. Several past attempts to formalize and organize this circuit (with the support of projects) have all failed, mostly because of the irregularity of sea transport. Irregularity of sea transport is due to a variety of factors including weather and sea conditions which can prevent smaller boats and ships to take the sea, but also condition of boats and maintenance, and erratic availability of fuel. Processors

Niumakélé Dairy Plant The Niumakélé dairy plant has been established in the early 2,000’s by a French NGO, Initiave et Développement (I&D), to tentatively address the marketing problem that was already occurring there. The intention was to collect milk from the Niumakélé basin, to process it and to “export” part of the production to Moroni. This has never worked despite several attempts to revive the project by successive projects. With the support of I&D, the farmers had established an association called Neema, which evolved into a cooperative (CoopEBC) in 2015. In 2016, this Cooperative applied for and obtained a loan of 200 M KMF (approximately 450,000 USD) from SANDUK, in order to rehabilitate the plant and collect, process and market pasteurized milk and yoghurt. Most of the machinery have been received and are being installed. However, the industrial process and type of equipment that was chosen seems to be overly complex for the local capacities, and the key issue of organization of a cold chain and transport circuit from the collecting point to Moroni has unfortunately not been addressed and will not be solved by this project. As of now, there is no other processing of local milk in Comoros – apart from basic clotting done by some traders to extend shelf life – and all the local milk is sold raw or as sour milk. Local milk is very much valued by customers, especially during traditional ceremonies. A small number of units process powdered milk, using it as raw material to produce yoghurt and ice-cream. Most of them have a very limited capacity of a few hundred liters of milk per day. There is only one industrial processor in Ngazidja, in Moroni, but it is not fully operational (after a few years of activity, the first owner closed the factory, which has recently been bought by another investor and should resume operations soon). The management of this factory, met by the mission, confirmed that they would be interested by sourcing local milk but that it is not economically viable given the current market price. Service Providers Veterinary service and technical advice: Intensive dairy production requires a high level of technical expertise, which public services are currently unable to provide. Across islands, a small number of specialized technicians deliver a full package of service: veterinary care,clviii technical advice, and in some cases, artificial insemination (AI). While some technicians are purely private, others are civil servants who deliver these services on a private basis. Most private veterinary technicians also sell inputs, and some even import them, including veterinary medicines and even sometimes semen for AI.clix In Grande Comore, the only importer of feed is CAPAC, which orders small quantities of dairy feed in containers of poultry feed, but supply is erratic. Margins 85. The local milk in Anjouan is competitive with imported Ultra-high temperature processed milk, but not in Grande Comore. There are two to three importers of powder milk and

UHT milk in Grande Comore, as well as in Anjouan. They import 25kg milk powder bags which are used by the food industry but are also reconstituted in smaller quantities by shop keepers and market vendors. Since low-cost production and excess supply characterizes the Anjouan market, the price is 600 KMF at farm gate, and 800 – 1,000 KMF at consumer level. At its peak, part of the milk supply does not even find a market. In contrast, the farm gate price in Grand Comoros is 1,000 KMF (2.5 USD) per liter, possibly one of the highest farmgate prices globally. Consequently, The local milk in Grande Comore costs more than 4 times the price of the reconstituted milk, and almost twice the price of the imported UHT. The table below shows the respective retail price of local raw milk (Grande Comore and Ajouan) and liquid milk (liquid UHT or reconstituted from powder). Figure 6-11: Retail Price of a Litre of Liquid Milk According to its Origin (KMF)

1400 1250 1200

1000 800 800 700

600

400 300

200

0 local fresh Local fresh Liquid UHT Liquid Anjouan Moroni reconstituted

Source: World Bank mission observations (2017) Geography 86. Dairy production is concentrated in Anjouan, but growing rapidly from a small base in Grande Comore. According to veterinary professionals who are the most reliable source of information on number of animals, there are about 9,000 dairy cows in Anjouan, belonging to 4,000 farmers. In Grande Comore, their number is estimated at 350 cows, belonging to 100 farmers. In Mohéli, there are around 100 cows belonging to 40 farmers. The table below summarizes estimated number of farmers, animals, production and productivity, by island: Table 6-4 : Number of farms, animals and production per island

Grande Anjouan Mohéli National Comore Number of dairy farmers 100 4,000 40 4,140 Number of dairy cows 350 9,000 100 9,450 Average size of farms (animals/unit) 3.5 2.3 2.5 2.3 Production per cow per year 1,500 3,000 2,000 % of lactating animals 70% 90% 90% Total production 367,500 24,300,000 180,000 24,847,500 Source: Interviews with Private Veterinarians (December 2017)

6.4 High Potential Value Chains

6.4.1 Horticulture

87. Vegetable and fruit production is integrated into subsistence production systems across the three islands. Fresh vegetables have not been a part of the traditional diet for most Comorians; often, their consumption is reserved for special occasions or restricted to the population segment with higher purchasing power. However, their demand is increasing, and is being met by fresh imports from Madagascar and Tanzania (see Annex). Domestic producers are only loosely connected to market channels. Fruit production is diverse but sparse, which inhibits large-scale collection efforts. Horticulture offers the potential to generate more income than staple crops on small plots, but in the absence of organized collection, structured marketing channels and vertical integration, the scale of production is the principal constraint to the emergence of a processing segment, and the development of a robust value chain.

Vegetables Production

88. Commercial vegetable production is highly seasonal and geographically segmented due to the absence of cold storage and organized maritime transport. Given the growing demand for horticultural products and their suitability to small plots, vegetable production represents a highly promising opportunity for local producers. On all three islands, this production is highly seasonal (July – November) and dominated by small holders. However, the wet markets in Anjouan and Moheli are characterized by excess supply and low farmgate prices during this period, with little storage to smoothen prices inter-temporally. While excess produce reaches Grand Comore via maritime transport on some days, the absence of cold storage at the ports in Anjouan and Moheli leads to significant food loss along the supply chain. Table 6-5: Vegetable production from 2008 to 2011 (MT)

Products 2008 2009 2010 2011 Tomato 6,709 6,910 7,131 7,134 Onion 1,043 1,095 1,156 1,161 Source: Central Bank of the Comoros. Annual reports 2011 and 2013.

89. The low elasticity of off-season vegetable supply is linked to the failures in the input supply and post-harvest market segments. Even as tomato and onion imports from nearby Madagascar and Tanzania are growing during the off-season when prices are 2-3 times higher (see Annex), Comorian producers are unable to respond to market signals by switching to similar, long- season varieties. The wedge between prices and off-season supply is partly driven by failures in the input supply markets. Neglected by both the public and private sector players in the input supply and marketing segments, vegetable producer associations, who successfully supplied the upscale Hotel Galawah when it was operational in the 1990s, no longer have the incentives to modernize their practices and successfully respond to market signals. Fruit Production 90. Fruit production is completely unorganized in isolated plots, and orchards are conspicuous in their absence. While a large variety of fruit species grow wild on the archipelago (citrus, mango, litchi, guava, pineapple, banana, papaya), arboriculture and crop husbandry of the trees

82 is non-existent. Like vegetables, fruit production is highly seasonal and food losses are large due to the complete absence of post-harvest management in the form of aggregation, storage and processing. Given the weak production system, erratic supply and high packing costs, the successful establishment of fruit processing units requires multiple interventions and strong market assessment, which have been largely absent in previous donor-funded projects.clx Developing this sub-sector requires the introduction and multiplication of short-cycle fruit crops like pineapple and sweet bananas, which can be started on a small scale and require low upfront investment. Once these investments bear fruit, linkages with marketing and processing facilities that are shared with the vegetables value chain are likely to be successful. Structure 91. Since vegetable and fruit value chains show significant spatial overlap in structure, this analysis treats them as single value chain. Despite the existence of the CCIA and two large farmer associations (SNAC, FNAC), the key upstream (supply inputs, training and technical advice) and downstream (product collection, transportation, market information) segments are not linked with producers on a regular and effective basis. Figure 6-12: Horticulture Value Chain Actors

Input Supply 92. The erratic supply of seeds and fertilizers and misalignment with the cropping calendar is a binding constraint on commercial production. The supply of fertilizers, plant protection products and vegetable seeds – exempt from import dutiesclxi – is made by the central purchasing the CAPAC (Centrale d'achat des Professionnels Agricoles des Comores)clxii and by private traders. The volume of imported inputs is estimated in 2015 to be: 80 – 120 MT/year of fertilizer; 0.7 to 1 ton/year of plant protection products; 15 to 90 MT per year of potato seeds and 0.5 to 0.7 ton/year of seed market.clxiii Across years, the quantities of imported fertilizer and potato seeds show great volatility that is unrelated to underlying demand, but driven by access to credit for producers and level of working capital for importers. Consequently, the distribution of seeds and fertilizers is erratic and not aligned with the cropping calendar, as cash-strapped importer association (CAPAC) and national producer federations (SNAC, FNAC) place only pre-financed orders for select customers that pick-up inputs from their locations. Other private importers and distributors are also retreating from this market, with their business models no longer placing a premium on origin or quality of inputs.clxiv Producers 93. Vegetable production is dispersed across many small-scale traditional producers, who seek seasonal income than sync with market demand, resulting in seasonal saturation. A small number of producers are seeking to shift to a different cycle, to sync with high demand periods – i.e. big wedding ceremonies and migrants' holidays (August) and the month of Ramadan. However, this

83 has not translated to greater processing (eg. tomato drying), which could reduce post-harvest losses and help smoothen the seasonality of supply. Processors 94. The installation of processing units has been tried by various projects, but none have proven to be sustainable. The processing of horticultural products is so far non-existent, with power supply, potable water and the high cost of raw material proving to be major bottlenecks. For vegetables, the processing of tomato paste is a promising option in theory, given widespread use, but competing with the price of imported past is difficult at current prices.clxv The production of fruit juice and jam, successfully being done by a Moroni entrepreneur at present, is a model that can potentially be scaled up. Sellers 95. The most important distribution channels are traditional roadside markets or covered urban markets, exposing producers to the monopsony power of wet-market sellers.clxvi Aggregation and linkages to modern retail is minimal as producer associations have not invested in the marketing and promotion of their products, which could enable linkages with growing retail sector. At present, only one of the four supermarket chains in Moroni – MAG-Market with 3 stores – sources from 2 producers who regularly deliver vegetables, but not fruit which is more seasonal. A few other points of sale are operated by SNAC/UWEZO and FNAC-FA.clxvii Exporters 96. Island antennas of the SNAC also occasionally organize delivery of products to Grande Comore and Mayotte (partnership with the COOPAC), but these actions are limited due to lack of organization among producers and the shortage of working capital. Further, the centralised management of these operations by the umbrella federation is a significant barrier to more efficient performance. Mayotte is a promising market for horticultural products, given the relatively higher prices and low domestic supply on that island. The major factor which prevents a greater frequency of these exchanges is the inability of the islands of the Union to implement the sanitary and phytosanitary standards required by Mayotte, which has the status of French overseas department and ultraperipheral region of European Union. Margins 97. The profitability of horticultural production was estimated on the basis of the crop budgets established by the mission for some common vegetables (tomato, onion, carrot, cabbage, potato) and a fruit, pineapple.clxviii The crop budgets have been established on the basis of the of cost and price data collected during interviews with producers and economic operators encountered. The table below presents the inputs, outputs and gross margins, calculated for an area of 10 acres (1,000 m2), the details of the calculations is in annexe 4. It should be noted that vegetable crops occupy the land for 3 to 4 months, unlike the pineapple that occupies the land for a year and a half.

Table 6-6: Inputs, Outputs and Gross Margins of Some Horticultural Productions (KMF/10 acres)

Costs Gross Gross Crop Production Transport % Total Product Margin Tomato 295 000 20 250 7% 315 250 445 500 130 250 Onion 218 900 9 600 4% 228 500 336 000 107 500 Carrot 198 650 11 550 6% 210 200 308 000 97 800 Cabbage 221 600 9 400 4% 231 000 352 500 121 500 Potato 382 350 15 200 4% 397 550 547 200 149 650 Pineapple 579 210 49 725 9% 628 935 1 077 375 448 440 Source: World Bank mission (December 2017)

Figure 6-13: Structure of Production Costs

100%

90%

80%

70%

60% Transport

50% Material & irrigation

40% Labor Inputs 30%

20%

10%

0% Tomato Onion Carrot Cabbage Potato Average Pineapple vegetable Source: World Bank Mission Estimates (December 2017)

98. This analysis shows that market-oriented horticultural crops production allows substantial gross margins for the producer. The break-even point is well below the sale price obtained by the producer, and the gap ranges from 138 to 158 percent depending on the product (see table below). This margin allows for the absorption of lower yields or seasonal downswings in prices. On the other hand, these margins become very interesting when prices soar, in times of scarcity of products. The returns for family labor is also very interesting, it is located between 2,850 and 6,840 KMF per working day (WD). The number of working days for a crop cycle (3-4 months) varies between 49 and 67 days/10 acres for vegetable crops. The pineapple crop that occupies the land much longer (18 months) requires about 91 working days for a 10 acres plot (see details in Annex). The shift of production schedules requires heavier investment in nets, tunnels or greenhouses. The products price differential that can be obtained is expected to amortize the cost of these investments without change too much the margin of the producer.

Figure 6-14: Break-even Point and Average Selling Price of Horticultural Products

450 400 350 300

250 200 KMF/kg 150 100 50 0 Tomato Onion Carrot Cabbage Potato Pineapple

Break even point Average selling price

For vegetables, an assessment of the cost structure finds that on average, labor accounts for 38% of costs, followed by inputs (33%), material and irrigation costs (24%) and transport (5%). The break- even point, below which production would be at a loss, is well below the sale price obtained by the producer: the gap ranges from 138 to 158% depending on the product. The returns to family labor are between 2,850 and 6,840 KMF per working day (WD). 99. At the current stage of organization and development of the horticultural sectors, a broad-reaching product strategy would be necessary to ensure a regular and varied supply of fruits and vegetables throughout the year. The market for any individual product is likely too small to justify investment. For the vegetable sector, the priority would be to professionalize producers and build integration from production to commercialization by upgrading associations of vegetables producers, which are already dynamic, to professional organizations. For the fruit sector, the first interventions should focus on the introduction and multiplication, in private nurseries, varieties of fruit and seedlings of quality, well adapted to the eco-climatic context of the Comoros, and linking this production to horticulture value chains. 6.5 Priority Actions for Value Chains

Table 6.8 lists targeted investments and policy reforms, corresponding to value chains segments, that would develop and enhance competitiveness, raise incomes and create jobs for the entire value chain:

Table 6-7 Priority Actions to Develop Value Chains

Intervention Vanilla Cloves Ylang- Milk Eggs Fruits and ylang vegetables On-farm Bulking Advisory Technical Exchange visits, Productive Irrigation to plots for services to selection of support to Alliances match planting improve mother trees for veterinary (PA) with production material technical seed collection service medium- cycles with management providers in scale units market of Nursery Grande Comore and demand. plantations management entrepreneurs to support Training and Technical feed and assistance to selection of producers.

mother trees chick for seed production Multiplication collection of quality fruit required Improve seedlings. water supply Expand size and management of nurseries

Collection/ Development Build collection Cold storage Immediate of field tests to points in detect Niumakélé markets manipulation of (Anjouan) quality

Professionalizat ion of collectors and direct linkages between distillers and buyers

Processing Invest in Upgrade Upgrade Organization of curing drying distillation inter-island facilities facilities equipment transport of fresh raw milk Improved and delivery to furnaces and retailers on alternative Moroni Island. heating technology

Transport Invest in Improve storage transport facilities to conditions for address perishable quality and products: yield losses standardized containers, road rehabilitation Innovation Support Identification Technical developm of higher education and ent of priced entrepreneurshi niche market p support for qualities niches for cosmetics through higher manufacturing Certificati qualities on (Organic, Assess FairTrade commercial ) to viability of minimize clove oil effects of production downturn of

internatio nal prices Finance Development of practical, enforceable loan guarantees Increasing MFI capacity to appraise the viability and risk of agribusiness proposals Support for business plan development

Institutional Producer Long-term ‘Dairy Hubs’ poultry service Development of professional organization groups to support and based on center could associations. capture mentoring of productive imporve access value groupings in alliances. to services and added group inputs. from management, curing. function and operation

Regulatory Registrati Reduce Collect survey Stricter Develop a environment on and government data on current enforcement subsector data role in extent and of food safety strategy policy / governance collection setting status of regulations to of purchase plantings reduce Public growers price, which unhealthy campaigns and delays Register imports promoting processor purchases growers consumption s and reduces Simplificatio of fresh fruits transparency Collect export n of import and Export data exemption vegetables. data Registration procedures collection and data for collection of agricultural growers and inputs (feed) processors

Export data collection

7 Institutional Environment for Agriculture

Main Messages

 At present, the sector lacks a validated strategy that outlines the path toward achieving the government’s commitment to economic transformation.  The achievement of food security, productivity, value chain development and climate adaptation targets identified in SCA2D require substantial investments in agriculture.  Implementation of the National Industrial Strategy necessitates the development of agriculture-derived products and the support to the sector’s enabling environment.  The institutional structure governing agriculture is decentralized and complex, but service delivery is poor due to the absence of sustainable revenue-generating models.

7.1 Strategic Framework

100. Despite agriculture’s large footprint and weak performance, the sector lacks a validated strategy that outlines the path toward achieving the government’s commitment to economic transformation. In 2014, the Government of Comoros developed a multi-sectoral termed Strategy for Accelerated Growth and Sustainable Development (Stratégie de Croissance Accélérée et de Développement Durable, SCA2D 2015 – 2019), which outlines its commitment to structural transformation. As part of the background planning process, the government commissioned the European Union to draft an update of the 1994 National Agricultural Strategy (Stratégie Nationale Agricole, SNA). As in other African countries, the updated draft SNA was also intended to be Comoros’ national chapter for the Comprehensive Africa Agriculture Development Programme (CAADP), the policy framework for agricultural transformation developed under NEPAD.clxix Figure 7-1: Strategic Framework for Agriculture SCA2D Multi-sectoral (national)

CAADP (Pan-African) Agriculture-oriented SNA (National Agricultural Strategy)

However, the draft SNA was prepared in 2014 but has not validated by the Government, leaving a strategic vacuum that has not yet been filled. Consequently, Comoros does not have a National Agricultural Investment Program (NAIP), has not signed the CAADP compact, and its progress towards CAADP goals could not be evaluated for the Africa Agricultural Transformation Scorecard (AATS, February 2018). This Agriculture Sector Review is intended to provide the stocktaking necessary for an evidence-based, participatory and inclusive consultation process that leads to an integrated sector strategy and investment plan that charts the path towards structural transformation. It will serve as an input for the overlap between SCA2D and CAADP processes, as illustrated below:

Strategy for Accelerated Growth and Sustainable Development (SCA2D 2015-2019)

101. The achievement of food security, productivity, value chain development and climate adaptation targets identified in SCA2D require substantial investments in agriculture. The Accelerated Growth and Sustainable Development Strategy (Stratégie de Croissance Accélérée et de Développement Durable - SCA2D, 2015-2019) is a global framework of actions that builds on the Growth and Poverty Reduction Strategy (2010-2014) and covers the 2015-2019 period. It defines

89 priority action programs to achieve the Sustainable Development Goals (SDGs) and proposes a long- clxx term vision of “turning Comoros an emerging country by 2030”.

The SCA2D (2015 – 2019) covers all sectors of the national economy and states four objectives: i) strengthen the foundations of a strong, viable, fair and inclusive economic growth; ii) improve the population’s living conditions and ensure equity in the access to basic social services; iii) promote the Comorian natural and cultural heritage and the optimal use of natural resources; and iv) promote good governance. Each sector has relevant strategies and policies, and these are adopted and operationalized by each island along the relevant sections of the SCA2D. For instance, the agricultural policy developed in 1994, with an updated 2014 draft, was intended to be the strategic reference document for the agriculture sector. Agricultural development targets and climate change adaptation targets are included within the framework of SCA2D. While the draft SNA is focused on the first pillar of the SCA2D, some of the proposed actions are also relevant for the other three pillars.

Comprehensive Africa Agriculture Development Program (CAADP/NEPAD)

102. Comoros has launched the CAADP process, but has not signed the compact. The Comprehensive Africa Agriculture Development Program (CAADP) is the agricultural component of the NEPAD (New Partnership for Africa's Development).clxxi CAADP has served as the framework for action for agricultural transformation across Africa since 2003. CAADP supports member states in increasing investment and productivity in the agricultural sector and its vision is “to maximize the contribution of Agriculture to achieving the ambition of a self-reliant and productive Africa and delivering economic growth and sustainable development for its people as well as playing its full part on the world stage”. Recent panel-data evidence from 46 countries shows that over the 2004 – 2014 period, CAADP implementation has had a positive impact on agricultural value-added and land and labor productivity.clxxii

The pan-African goal for the sector is an agricultural-led development to eliminate hunger, reduce poverty and food insecurity opening the way for an expansion for exports and put the continent on a higher economic growth. CAADP is also organized on four pillars (or thematic priority areas) considered as the best investment sectors for agriculture: (i) Sustainable land and water management; (ii) Market access; (iii) Food Security; and (iv) Agricultural Research and Dissemination of New Technologies.

The development of National Agricultural Investment Plans in every CAADP signatory country is at the heart of CAADP implementation. To comply with this requirement Comores has drafted the SNA, which covers the same period as CAADP and outlines the steps required to operationalize it.

Pillar 1 of the draft SNA is consistent with Pillar 1 of the CAADP, and develops a Sustainable Land and Water Management Program as one of the Comoros priorities. The implementation of this program is based on the experience gained within CAADP.

Pillars 2 of the draft SNA cover Pillars 2 and 3 of the CAADP, which are treated jointly in the organization of the value chains of key products.

On the other hand, Pillar 4 of CAADP is less present in the draft SNA, mainly because of the difficulty of developing Comorian research subjects autonomously form other countries, giving priority to the integration of Comorian research into international networks.

Draft National Industrial Strategy for Domestic Processing by Micro, Small and Medium Enterprises (MSMEs), 2017

103. Implementation of the National Industrial Strategy requires the development of agriculture-derived products and the support to the sector’s enabling environment. Developed by the Ministry of Economy (Ministère de l’Economie, de l’Energie, de l’Industrie, de l’Artisanat, du Tourisme, des Investissements, du Secteur, Privé et des Affaires Foncières), the National Industrial Strategy for Domestic Processing by MSMEsclxxiii is a follow-up document to the Etude Diagnostique sur l’Intégration du Commerce (EDIC), which was the ministry’s strategy document for SCA2D. In particular, the National Industrial Strategy outlines the specific policy measures required to achieve three of goals identified in the former – i) greater food self-sufficiency, ii) better access to international markets, and iii) positioning of Comoros in regional markets.

Based on the COMESA Industrial Policy (2016 – 2015), the National Industrial Strategy identifies nine key priority areas: agribusiness, energy, textiles and clothing, leather and leather products, mining, pharmaceuticals, chemicals and agro-chemicals, light engineering and the blue economy. However, the list of potential products identified for short-term development for the local and export markets are all agriculture-derived products (see Table 4.5 in Chapter 4). The policy measures identified for medium and long-term development of MSMEs and the private sector are also the enabling environment constraints faced by the agriculture sector. Given the degree of overlap between industrial development and structural transformation needs, it is evident that the path to emergence runs through agriculture for Comoros.

7.2 Institutional Structure at the National Level

104. The Union of Comoros has a number of Ministries, with sub-component departments or ‘directorates’. The institutional framework is complex -- the three islands have considerable autonomy as well as their own governing bodies.clxxiv In fact, the governance of the islands is under the leadership of decentralized commissions and technical services responsible for development planning, programming, monitoring and evaluation.

Ministry of Agriculture, Fishing, Industrial Development, Artisan Artifacts & Environment

105. The Ministry in charge of Agriculture, Fishing, Environment, Territorial & Urban Planning (Ministère de l’Agriculture, de la Pêche, de l’Environnement, de l’Aménagement du Territoire et de l’Urbanisme (MAPEAU)) provides sectoral governance at the central-level. Under it are various directorates: the National Directorate of Agricultural Strategies and Livestock, the National Directorate of Fisheries and Fishery Resources; the Directorate of General of Environment and Forests; the National Directorate of Energy and Mines; the National Directorate of Industry and Crafts (DGEF) and the National Research Institute for Agriculture, Fisheries and Environment (INRAPE). These directorates are represented in all the three islands.

106. The MAPEAU is responsible for the implementation of the national environmental framework and environmental policy, as well as being responsible for planning, programming and the monitoring and evaluation of development actions in the fields of agriculture, livestock, fisheries and the environment. Executive functions are delegated to the governorates in each island, Regional Centre for Economic Development (CRDE) and various donor’s projects. The MAPEAU has stated its commitment to achieving the following four goals for the agriculture sector:

 Securing jobs and incomes for those employed by the sector  Enhancement of food security for rural and urban populations  Contribute to the balance of trade between imports and exports  Sustainable management of natural and cultivated environments

National Strategic Directorate of Agriculture and Livestock (DNSAE - Direction Nationale des Stratégies Agricoles et de l’Elevage) 107. The DNSAE is responsible for the coordination of the agricultural and livestock sector in the Comoros and lies under the supervision of MAPEAU. DNSAE coordinates the implementation of the agriculture strategy and agriculture targets under the PRGS. DNSAE is composed of Food Production, Cash Crop and Rural Economy Departments, and a Division of Livestock Management, which has two Departments: Animal Health and Epidemiological Monitoring. Each island also has an agricultural

92 office which comprises six departments: Food Production; Cash Crops; Awareness Raising and Communication; Plant Health; Livestock; and Management and Logistics. Each island office currently has about 40 staff, of which 20 are agricultural technicians (including livestock). The national DNSAE office has around 30 programmatic staff of which 15 are under permanent employment and 15 are contracted, with additional specialized staff, including engineers, nutritional expert, veterinary expert and livestock breeding specialists.

108. Rehabilitation of agricultural support and extension centers on each island is currently a top priority within DNSAE and within island offices, and establishing the long-term technical and financial sustainability of these centers is a pressing concern. Within sub-regions on each island, agricultural support is provided by Rural Economic Development Centres (CRDE). There is, however, a very low level of nationally-led agricultural extension support, with almost all agricultural support initiatives relying on donor-funded projects.

109. Overall capacity is low and unbalanced, as absence of a competitive HR policy and budget resulted in an aged workforce, with the most capable staff seeking opportunities with donors financed projects. Typically, project coordinator functions are taken by the most capable managerial level staff, leading to a progressive weakening of the Ministry and a resulting confusion between activities carried out by projects and those that should be part of the core functions of the Ministry.

110. The generational gap is also fueled by difficulties in providing severance payments that would allow the retirement of several of the senior executives, a known problem of the Small Island Developing States (SIDS). To compound to the problem, qualified human resources are reduced as a result of lack of training and continuing education for executives in place.

7.3 Governorates

111. Each island has its own governorate in which a Commissioner is in charge of the implementation of the Union national strategies and policies and those of the island, in particular promotion of agricultural production, environment, energy, industry and crafts. Agricultural services at the island-level depend on the governorate. Financial and human resources are weak in all three islands. CRDE, newly established following a widespread UNDP model, should be the level of implementation of the policies, the governorate being then in charge of their impetus and coordination at the level of each island. However, given the jurisdictional disputes that persist, commissioners often end up designing their own agricultural strategies. The links between the MAPEAU and the commissioners are not always efficient, which compromises the efforts in implementing agriculture policies

7.4 Agriculture Institutions

National Institute of Agronomic, Fisheries and Environmental Research (INRAPE)

112. The Institut National de Recherche pour l'Agriculture, la Pêche et l'Environnement (INRAPE) is responsible for agricultural and environmental research. Established in 1994, INRAPE is a public scientific and technical institute under the authority of MAPEAU.clxxv INRAPE is the main partner of the DGEF and is responsible for addressing research questions related to the environment. INRAPE has planned to invest most of its research efforts into the conservation of marine ecosystems and to

93 monitor climate parameters. However, funding scarcity constitutes a real obstacle in achieving these goals. The institute has administrative and management autonomy and has the following mandate:

 Prepare and implement research programs (agriculture, water resources &environment)  Consolidate, analyze, and publish findings and experiences, promote exchanges between national and international researchers, develop training programs on rural development, fisheries, and the environment for technical and managerial staff; and  Promote methods and techniques that increase agriculture and fisheries productivity, preserve the environment, and increase conservation.

113. Currently INRAPE’s research projects are heavily reliant on donor-funded projects and there are few long-term research programs and no long-term climate change adaptation research programs. MAPEAU plans to strengthen the institutional capacity of INRAPE, including through proposals to develop a research plan with joint funding from the government and development partners.

Decentralized institutions

114. The Rural Economic Development Centres (CRDE – Centre Rural de Développement Economique) - are key agencies providing agricultural extension and development support to vulnerable farming communities. Rural Economic Development Centres (CRDE) were established in 2013 through an important national decree developed in the Union of Comoros during the design of the CRCCA project (Enhancing adaptive capacity for increased resilience to climate change in the agriculture sector in the Union of the Comoros – 2014-2018).clxxvi These CRDE replace Agricultural Advisory Centres (CCA) as the main rural development and agricultural support institution in the Union of Comoros.

115. The Union of Comoros plans to establish 14 CRDE: 6 in Grande Comore, 5 in Anjouan and 3 in Moheli. CRDE are responsible for providing extension and technical support to farming communities in order to support sustainable rural development. CRDE offer an important platform for supporting climate change resilience at the farm and community level. They work in partnership with NGOs and farmers associations to support sustainable agricultural development, including for climate change adaptation in the agriculture sector. They are endowed with a specific legal personality and administrative and financial autonomy. At the national level, the CRDE are placed under the technical supervision of the Vice President in charge of MAPEAU and will be financially supported by the Vice President in charge of Finance and Budget. The role of CRDE includes:

 Training of farmers and fishermen;  Provision of relevant information to support sustainable rural development;  Technical extension and advice and support to producers;  Supervising professional organizations and community development structures;  Monitoring and evaluation and data management;  The provision of basic services to support improved working conditions for rural communities; and  Support for the development of improved rural economic infrastructure

116. Under the new decree, each local region will have a CRDE, which will be managed by a Director assisted by an accounts manager, administrative assistant and a team of technical extension staff with at least one higher education diploma or equivalent. The centers are financed by the state but

94 are also authorised to use direct contributions from the public and the private sector. CRDE are administered by a Steering Committee, with strong civil society representation. The composition of the CRDE Steering Committee includes: a representative of the prefecture, two representatives of professional organizations (such as fishermen and farmers), two representatives of NGOs working in the agriculture, fisheries and environment, two representatives of service providers (nurseries, seed production, vets etc.), a representative of the local Commissioner, and four representatives of associations and producer groups of all disciplines.

7.5 Non-Governmental Organizations

117. Non-Governmental Organizations (NGOs) and Community Based Organizations (CBOs) are important to the local development context on each island. Most communities have associations, including farmers and women’s associations who play an active role in local management and development. Village Development Associations (VDA) are present in most villages in the Comoros. They are often well organized and implement local development initiatives. These village associations can consist of several 'branches', including for women’s support, youth, community development, etc. VDAs are often important channels through which projects and organizations engage with communities.

Professional Organization and Associations (POA)

118. The statutes of POAs reveal a hierarchical arrangement far removed from the democratic and participative principles that POAs generally follow. These umbrella organizations are managed by notables, especially civil servants or former civil servants, they are not very representative, and are distant from the farmer base. These organizations have lost a lot of credibility, but they have become part of the institutional landscape and have become important stakeholders. The reorganization that is taking place following the passing of the Agricultural Professional Organizations Act, which requires the separation of representation and commercial activities, should enable cooperatives to perform the supply and marketing functions for their members and the representative organizations to play fully their role.

8 Summary of Recommended Actions and Expected Outcomes

119. As this review’s findings demonstrate, agriculture in Comoros has a large economic footprint and is central to the country’s economic growth, food security, export revenues, and employment. Given this agricultural base and the small set of other productive investment opportunities in the economy, the path for economic transformation of Comoros runs through agricultural transformation. In fact, economic transformation in Comoros does not imply an exit from agriculture, but building on it to modernize farming and strengthen efficiency enhancing linkages between agriculture and nascent sectors like manufacturing and tourism. Consequently, progress on the agricultural transformation agenda is necessary and urgent, in order to realize the government’s commitment to becoming an emerging country 2030.

120. The potential for agricultural transformation in Comoros is large and quite evident. Given low productivity and high exposure to environmental shocks, yield-improving investments in climate-smart inputs, modern production and processing technologies, and sustainable practices can deliver large food security gains and build resilience to environment risks. Investments that facilitate internal trade

95 will enhance the competitiveness of crop and livestock perishables, in which Comoros’s remoteness becomes a buffer against shipped imports, bring local supply in sync with market demand, raise farm incomes and lower the food import bill. Export promotion in the form of downstream industrial development of existing products, facing favorable market conditions after a decade long downturn, can allow Comoros to tap into newer markets and enable producers to capture greater value domestically.

The specific recommendations that have emerged from this review are grouped in three strategic pillars below:

 Pillar 1: Public Actions to Boost Agricultural Transformation,  Pillar 2: Boosting Food Production and Smallholder Access to Markets  Pillar 3: Export Promotion of Established and New Products

Under each category, recommendations flow from identified priorities and measures, which aim to address market failures, coordination failures and missing markets in the agriculture sector. The primary role of the public sector in implementing each recommendation along with the time frame associated with implementation are indicated. In addition, the expected direct beneficiaries from each priority are also identified. A few broad patterns can be inferred from the four tables below:

121. First, the provision of traditional public goods for agriculture is qualitatively low, and government involvement is restricted to price-setting in the vanilla value chain, operation of a rice import and distribution parastatal, and identification of high-priority food imports. Consequently, there is an unmet need for strategic public investments in the supply of traditional public goods in agriculture – advisory and extension services, technology transfer and dissemination, land administration services, and provision of agricultural data, as is reflected in Table 8.1.

122. Second, the government needs to play a leading role in market development for food crops in general, and perishables (fruits, vegetables, livestock products etc.) in particular. In Comoros, there is little public investment in transport systems, cold storage, and information systems that enable a better functioning of markets for perishables, and induce market participation by subsistence-oriented producers. While this type of underinvestment is widespread across the developing world, its consequences are far more debilitating for the Comorian archipelago, a small internal market which is further segmented. Table 8.2 outlines a set of direct and enabling public actions that can facilitate internal food trade.

123. Third, the public sector can tangibly accelerate private investment in the sector through regulatory simplification and reduced compliance costs, playing a matching role in the credit market for agriculture, and convening dialogue to enable productive alliances between producer cooperatives and buyers, as recommended in tables 8.1 and 8.3.

124. Last, the private sector can deliver better outcomes in two activities that the government currently undertakes. In terms of rice imports and distribution, private enterprises now import a large share of the rice are now well positioned to take on these activities and free up public resources for more strategic investments of the type discussed above. In the vanilla value chain, the system of establishing a price structure that ostensibly protects growers from price volatility, now prevents them from receiving the gains from favorable market conditions that are benefiting cloves and ylang-ylang

96 growers. An approach which monitors the exporter oligopoly in the vanilla value chain, and provides targeted price or income support during downturns is likely to raise smallholder returns in this sector.

Implementing the agricultural policies and investments outlined below would put the agriculture sector in Comoros on a sustainable growth path that can raise farm productivity and smallholder incomes, reduce the current account deficit, enhance competitiveness and create jobs in the agri-food system. In conclusion, it would jumpstart the process of agricultural transformation and help accelerate the economic transformation of Comoros.

Table 8-1: Public Actions to Boost Agricultural Transformation

PRIORITIES AND MEASURES Direct Role of Public Timeframe Beneficiaries Sector

Long Long Short Short Policy Policy maker maker maker Market Market Medium Medium Financer Financer Convener Convener

PRIORITY 1: Improving access to climate-smart technologies, innovations, management practices (TIMPs), and advisory services (iii) Join regional agricultural research and technology transfer partnerships to develop, transfer, and Crop and Policy maker Medium disseminate TIMPs that enhance productivity, livestock Financer Medium climate resilience, and nutritive value by: producers Convener Long  harmonizing policies, laws, and regulations to Convener Short accelerate transfer of off-the-shelf technologies Export crop  establishing national center of specialization to producers RANSFORMATION

T accelerate development in target crop or livestock product(s) Private seed  enabling private sector involvement in technology and animal development (seed multiplication, animal breeders breeding)  fostering producer-retailer dialogue that enables Retailers GRICULTURAL GRICULTURAL

A farmer organizations to adapt to market demand (iv) Improved access to crop and livestock extension OOST OOST Crop and Financer Medium

B services by:  equipping public advisory and extension services livestock Policy maker Short (CRDE) physical and financial resources producers Convener Medium  promoting technical assistance through productive

CTIONS TO TO CTIONS alliances between cooperatives and anchor Export crop

A investors producers  enabling development of privately managed dairy

UBLIC UBLIC and poultry hubs to disseminate livestock inputs, Extension P animal vaccines and veterinary services agents

 structure the regulatory simplification agenda Agribusiness around the Enabling Business of Agriculture (EBA) startups and indicators SMEs  facilitate supply of agricultural credit by providing MFIs training on appraisal of agribusiness business Microfinance plans institutions  accelerate the development of an incubation ecosystem with linkages between universities and Private sector private sector representatives for preparation of representatives viable business plans, early-stage mentoring, and linkages to MFIs Universities

(v) Strengthening land administration services to promote Food Financer Short land tenure security by: producers Financer Medium  equipping land administration office with Convener Medium geospatial tools and technical assistance Export crop Policy maker Medium  conducting a participatory titling program in producers production basins  forming multi-stakeholder land recognition Producer committees to facilitate demarcation and small- organizations scale cadaster census  developing land-use framework for areas where Small and both food and export crops are cultivated medium agribusiness (vi) Build capacity to collect, monitor and disseminate real- Food Financer Short time agriculture data by establishing an agriculture producers Financer Medium data unit to: Convener Short  conduct an agricultural census to fill knowledge Export crop gaps, better policy making and enable private producers investments  utilize remote sensing tools to monitor and respond Producer to weather, yields, prices, pests and diseases in real organizations time  organize the development of mobile applications to Small and disseminate real-time weather and market medium information agribusiness PRIORITY 3: Reorient role in determination of import tariffs and distribution of food, to enhance domestic competitiveness and free up public resources for investments (iii) Implement a smart targeting approach by:  assessing pass-through effects of tariffs on Crop and Policy maker Short consumer prices and welfare against existing livestock Policy maker Short competitiveness and supply capacity of domestic producers production  enabling annual reassessment of list by developing Consumers criteria-based determination and monitoring of agricultural production and import trends (iv) Reorient ONICOR to the role of a market coordinator and facilitator Food Policy maker Medium  importing and distribution of rice should be importers Market maker Medium handled by established private enterprises in the rice import business Public sector  activities should be restricted to setting and applying standards for imported rice, market Consumers monitoring, collection of duties and sales taxes, and maintaining a modest emergency stock

Table 8-2: Boosting Food Production and Access to Markets for Smallholders

PRIORITIES AND MEASURES Direct Role of Public Timeframe Beneficiaries Sector

Long Long Short Short Policy Policy maker maker maker Market Market Medium Medium Financer Financer Convener Convener

PRIORITY 1: Enforcement of import exemptions on agricultural inputs to reduce input costs and erratic input supply

S association and farmer federations, conditioned on organizations preparation of viable business models  coordinate distribution of time-sensitive inputs at the Private start of planting season importers

ARKETS FOR FOR ARKETS

M PRIORITY 2: Strategic public investments in climate-smart agriculture (CSA) that boost yields and build smallholder resilience to environmental shocks

CCESS TO TO CCESS Target CSA dissemination to food crop producers by: A  piloting improved, biofortified, and resilient Crop and Financer Short varieties of cassava, maize, and pulses in livestock Policy maker Medium production basins producers  providing market incentives to adopt and scale-up tested CSA technologies, innovations, management Consumers practices (TIMPs) for crop and livestock RODUCTION AND AND RODUCTION

P production systems PRIORITY 3: Facilitate internal food trade to raise market OOD OOD

F participation and food access (i) Linking highland production basins with coastal urban markets by: Crop and Policy maker Short

OOSTING OOSTING  enable access to non-motorized (bicycles, carts) and livestock Policy maker Short B motorized transport (motorcycles, tuk-tuks) through producers Financer Medium temporary and targeted import tariff waivers  coordinate organized collection from remote areas Producer and marketing through national-level farmer organizations organizations  upgrade and maintain dilapidated rural roads Traders and through locally managed public works programs logistics SMEs

Consumers

ii) Investing in market access infrastructure to enhance marketability of perishables and reduce price volatility: Crop Market maker Medium  enabling private investments in on-farm warehouses producers Financer Medium and chilling dairy plant (Anjouan) by assisting and Financer Medium cooperatives with preparing viable business cooperatives Policy maker Medium proposals  investing in cold storage at ports to facilitate Dairy maritime transport of perishables (fish, dairy, cooperatives vegetables and fruits) and reduce food loss in surplus markets Fish  operationalizing existing liquid nitrogen facility in cooperatives Moheli to facilitate vaccine storage and artificial insemination for livestock Food  facilitating competition between mobile operators to consumers incentivize wider coverage and increase low-cost access to market information iii) Breaking down barriers to inter-island movement of agricultural goods to reduce transit costs, accelerate volume of Crop Financer Long internal food traded, and increase competitiveness by: producers Policy maker Short  investing in port capacity and infrastructure in and Policy maker Medium Grande Comore and Moheli to reduce congestion cooperatives and tariffs  organization, regulation and rationalization of fares Dairy for maritime transit cooperatives  introduction of a contracting system to establish market competition on high-frequency routes Fish cooperatives

Food consumers

Table 8-3: Export Promotion of Established Value Chains and New Products

PRIORITIES AND MEASURES Direct Role of Public Timeframe Beneficiaries Sector

EW EW N

Long Long Short Short Policy Policy maker maker maker Market Market Medium Medium Financer Financer Convener Convener

PRIORITY 1: Raising smallholder returns in established value

chains by: i) supporting plantation renewal and expansion, ii) STABLISHED AND STABLISHED

E enabling processing yield and quality improvement, iii) promoting productive alliances between cooperatives and RODUCTS

P buyers (v) Supporting plantation renewal and expansion by:  consolidation of bulking plots for supply of high Export crop Convener Short ROMOTION OF ROMOTION

P quality planting material to replace aging vanilla producers Financer Short vines and Financer Short  providing technical assistance for management of cooperatives

XPORT XPORT ylang-ylang plantations E  establish well-managed central nurseries for cloves Exporters and ylang-ylang trees

100

(vi) Enabling processing yield and quality improvements by: Export crop Convener Short  enabling upgrades and investments in vanilla curing producers Policy maker Short facilities by grower-curer cooperatives and Policy maker Short  incentivizing distillery upgrades to water and wood- cooperatives Convener Short saving structures and technologies for sustainable ylang-ylang production Curers,  standardizing low-cost field tests to reduce dryers and fraudulent manipulation of ylang-ylang oil quality distillers  promote investment in drying and storage facilities to minimize on-farm losses for cloves Exporters

(vii) Promoting productive alliances between Export crop Market maker Short cooperatives and buyers by: producers Market maker Short  facilitating formation of producer cooperatives and Convener Short through decentralized extension cooperatives Convener Medium  supporting development of niche qualities through certification (eg. Fair Trade, Organic) to attract new Curers, buyers and capture price premiums dryers and  convening and structuring dialogue with distillers international anchor investors to assess mutual needs and identify scope for collaboration Exporters  ensure long-term support and mentoring of cooperatives through input financing and technical Anchor assistance focused on management and operational investors capacity (viii) Improve regulatory environment by: Export crop Convener Medium  ensuring greater representation of smallholders in producers Policy maker Medium bodies representing each value chain and Policy maker Short  monitoring non-competitive behavior in cooperatives Financer Medium concentrated value-chain segments  removing government involvement in price-setting Curers, in vanilla value chain dryers and  mitigate effects of price volatility on smallholders distillers through targeted income support Exporters PRIORITY 2: Downstream diversification and development of new products by assessing and enhancing commercial viability of new products Assessing and enhancing commercial viability of new products by: Export crop Financer Short  conducting technical and financial assessments of producers Convener Medium commercial viability of derivatives of existing and Financer Short products (clove oil, vanilla powder, vanilla extract, cooperatives Market maker Medium aromatic soap etc.) and new products (ground spices, other essential oils) Agribusiness  establishment of technical university courses aimed startups and at development of derivatives SMEs  multiplication of planting material for crops will small production bases (coffee, pepper etc.)  support establishment of small-scale spice grinding and artisanal extraction facilities Key: (1) Direct beneficiaries are identified based on the expected first-order effects. (2) The role of public sector lists the primary role of the public sector, and it is not restrictive. (3) Timeframe is classified as Short-term for actions that are implementable in 0 – 2 years, Medium-term for 3 – 5 years, and Long-term for more than 5 years.

101

References

African Center for Economic Transformation, ed. Agriculture Powering Africa’s Economic Transformation. African Transformation Report, 2.2017. Accra, 2017. Amoriggi G. (2010) “FIDA: Appui au développement de la transformation des produits agricoles aux Comores. ” Banque centrale des Comores (2014). Rapport annuel 2013. Angelucci, Federica, and Piero Conforti. “Risk Management and Finance along Value Chains of Small Island Developing States. Evidence from the Caribbean and the Pacific.” Food Policy 35, no. 6 (December 2010): 565–75. https://doi.org/10.1016/j.foodpol.2010.07.001. Armstrong, Harvey W., and Robert Read. “The Determinants of Economic Growth in Small States.” The Round Table 92, no. 368 (January 2003): 99–124. https://doi.org/10.1080/750456745. Assoumani, Mohamed Ahmed. “Plan D’actions pour La Mise En Œuvre de la Politique Nationale de Protection Sociale,” n.d., 45. Astudillo, Miguel Fernandez. “Problématiques et priorités d’intervention,” n.d., 21. B., H. “Organisation des Nations Unies pour l’alimentation et l’agriculture.” Population (French Edition) 5, no. 4 (October 1950): 764. https://doi.org/10.2307/1523706. Babu, Suresh Chandra and Shishodia, Mahika. “Agribusiness Competitiveness: Applying Analytics, Typology, and Measurements to Africa. ” (May 2017). Banque africaine de développement – 2010. “Les sources de croissance aux Comores. Rapport d'étude.” Novembre 2010. Banque Centrale des Comores. “Importations 2015 et 2016”. (2017). Banque Centrale des Comores. “Rapport Annuel 2016.” (2016). Banque Centrale des Comores. “Rapport Annuel 2015.” (2015). Banque Centrale des Comores. “Rapport Annuel 2014.” (2014). Banque Centrale des Comores. “Rapport Annuel 2013.” (2013). Banque Centrale des Comores. “Rapport Annuel 2012.” (2012). Banque Centrale des Comores. “Rapport Annuel 2011.” (2011). Banque Centrale des Comores. “Rapport Annuel 2010.” (2010). Banque Centrale des Comores. “Rapport Annuel 2009.” (2009). Banque Centrale des Comores. “Rapport Annuel 2008.” (2008). Banque Centrale des Comores. “Rapport Annuel 2007.” (2007). Banque Centrale des Comores. “Rapport Annuel 2006.” (2006). Banque Mondiale. “Note Sectorielle de Politique Agricole. Orientations Stratégiques.” Version provisoire. FAO,. 18 mai 2013. (2013). Benin, Samuel. “Impacts of CAADP on Africa’s Agricultural-Led Development,” (2014), 56. Bourgoin, Clément & Parker, Louis & Martinez Valle, Armando & Caroline, Mwongera & Laderach, Peter. “A spatially explicit assessment of climate change vulnerability in the agricultural sector of the Union of the Comoros.” (2016). 10.13140/RG.2.2.19744.79366. Centre Douanier de Mutsamudu, Bureau Statistique du Commerce Extérieur. – Importations biens alimentaires année 2016. (2017) Cercle des Économistes et des Experts Comoriens (CEEC). www.ceec-comores.com Comoros Overview – The World Bank Group - December 2017 - http://www.worldbank.org/en/country/comoros/overview Da Silva, Carlos A., Marlo Rankin, and Food and Agriculture Organization of the United Nations, eds. Contract Farming for Inclusive Market Access. Rome: Food and Agriculture Organization of the United Nations, 2013.

102

Dömeland, Dörte, and Frederico Gil Sander. “African Small States,” n.d., 31. Douglas Gollin, Stephen Parente, and Richard Rogerson. “The Role of Agriculture in Development.” The American Economic Review 92, no. 2, (n.d.). http://www.jstor.org/stable/3083394. Doulton, Hugh, Misbahou Mohamed, Gill Shepherd, Siti Mohamed, and Badroudine Ali. “Competing Land- Use in a Small Island Developing State: Using Landscape Approaches to Manage Sustainable Outcomes in the Comoro Islands,” n.d., 9. Easterly, William, and Aart Kraay. “Small States, Small Problems? Income, Growth, and Volatility in Small States.” WORLD DEVELOPMENT, n.d., 15. Elise F Talsma, Inge D Brouwer, Hans Verhoef, Gloria NK Mbera, Alice M Mwangi, Ayşe Y Demir, Busie Maziya-Dixon, Erick Boy, Michael B Zimmermann, Alida Melse-Boonstra; Biofortified yellow cassava and vitamin A status of Kenyan children: a randomized controlled trial, The American Journal of Clinical Nutrition, Volume 103, Issue 1, 1 January 2016, Pages 258– 267, https://doi.org/10.3945/ajcn.114.100164 EU. “Actualisation de la politique agricole et formulation d'une stratégie sur le court à long terme pour le secteur comme vecteur pour lutter durablement contre l'insécurité alimentaire aux Comores.” Landell Mills, (April 2014). FAO. “Linking trade and food and nutrition security in Indian Ocean Commission member states.” Discussion paper. (2017). FAO. “Analyse diagnostic de la filière avicole en Union des Comores.” (2016) FAO. “Rapport d'analyse des résultats du recensement agricole (RGA) Comores 2004.” (24 novembre 2005). Hassani-El-Barwane, Mouhssini. “Le système foncier comorien de 1841 à 1975,” n.d., 340. Hauzer, Melissa, Philip Dearden, and Grant Murray. "The Effectiveness of Community-Based Governance of Small-Scale Fisheries, Ngazidja Island, Comoros." Marine Policy 38 (2013): 346-354. Horus Développement finance. “Programme d'appui au financement su secteur productif. Etude de faisabilité complémentaire.” (April 2017). International Monetary Fund. “Union of Comoros: Selected Issues.” IMF Staff Country Reports 16, no. 394 (2016): 1. https://doi.org/10.5089/9781475562774.002. International Monetary Fund. “Union of the Comoros: Poverty Reduction and Growth Strategy Paper.” IMF Staff Country Reports 10, no. 190 (2010): i. https://doi.org/10.5089/9781455203383.002. International Plant Protection Convention. “Propositions de structure institutionnelle de l’Organisation Nationale de Protection des Végétaux de l’Union des Comores Ministère de l’agriculture”. (2014). Jambor, Attila, and Suresh Babu. Competitiveness of Global Agriculture. Cham: Springer International Publishing, 2016. https://doi.org/10.1007/978-3-319-44876-3. Kida, PRMED Mizuho. “Caribbean Small States—Growth Diagnostics,” n.d., 29. Ministère de l’agriculture. “Mise à jour de la législation vétérinaire en union des Comores”. Version provisoire. (2017). Ministère de l’agriculture. “Politique sectorielle Sanitaire et Phytosanitaire.” Version provisoire. (2017) Ministère de l’agriculture. “Projet de loi portant protection phytosanitaire aux Comores.” Version proviso ire. (2017). Ministère de l’agriculture. “Politique nationale de développement de l’élevage aux Comores.” Version provisoire. (2017) Ministère des finances, de l'économie, du budget de l'investissement et du commerce extérieur, en charge des privatisations. “Etude diagnostique sur l'intégration du commerce en Union des Comores (EDIC 2015-2019).” Rapport exécutif et feuille de route. (April 2015). MNASRI, ISSRAR. “2011-2015 - Comoros - Country Strategy Paper,” n.d., 46. Nimarkoh, Joan, Suffyan Koroma, and Mawuli Sablah. “Linking Trade and Food and Nutrition Security in Indian Ocean Commission Member States,” n.d., 39. OIE. “Rapport d’analyse des écarts PVS”. (2016)

103

Pangaribowo, Evita Hanie, Nicolas Gerber, and Maximo A. Torero. “Food and Nutrition Security Indicators: A Review.” SSRN Electronic Journal, 2013. https://doi.org/10.2139/ssrn.2237992. ———. “Food and Nutrition Security Indicators: A Review.” SSRN Electronic Journal, 2013. https://doi.org/10.2139/ssrn.2237992. Pelling, Mark, and Juha I Uitto. “Small Island Developing States: Natural Disaster Vulnerability and Global Change.” Environmental Hazards, 2001, 14. Pierret, Dorothée. “« Tout d’une grande » : le réseau Sanduk, institution de microfinance de l’île d’Anjouan – Comores,” n.d., 23. Pingali, Prabhu. “Agricultural Policy and Nutrition Outcomes – Getting beyond the Preoccupation with Staple Grains.” Food Security 7, no. 3 (June 2015): 583–91. https://doi.org/10.1007/s12571-015-0461- x. PNDHD. “Rapport Préliminaire à l'introduction d'une Composante "filières et marchés de produits agricoles”. GRET. (January 2011). Roger, Matthieu, Marina Beral, Séverine Licciardi, Miradje Soulé, Abdourahime Faharoudine, Coralie Foray, Marie-Marie Olive, et al. “Evidence for Circulation of the Rift Valley Fever Virus among Livestock in the Union of Comoros.” Edited by Brian Bird. PLoS Neglected Tropical Diseases 8, no. 7 (July 31, 2014): e3045. https://doi.org/10.1371/journal.pntd.0003045. Roy, Étienne Le. “Hybrid Land Regulation between the Commons and the Market Land Tenure in the Comoros,” n.d., 14. Saïd, Mahamoudou. “Diagnostic du secteur agricole comorien à partir des volets transversaux du PDDAA.” (March 2015). Shigeyuki, HAMORI, and Ivohasina F RAZAFIMAHEFA. “Trade and Growth Relationship: Some Evidence from Comoros, Madagascar, Mauritius and Seychelles1),” n.d., 13. Subramanian, Arvind. “The Mauritian Success Story and Its Lessons.” In Achieving Development Success, edited by Augustin K. Fosu, 204–31. Oxford University Press, 2013. https://doi.org/10.1093/acprof:oso/9780199671557.003.0011. UNDP. “Projet d’un laboratoire au comores.” (2017). Vice Présidence en charge du Ministère des finances, de l'économie, du budget, du commerce extérieur et des investissements, chargée des privatisations. “Document de politique commerciale des Comores.” (February 2013). World Bank, David McKee. “Youth Employment Project - Value Chain Analysis and Recommendations. (2016). ” Zafar, Ali. “Mauritius: An Economic Success Story,” n.d., 31.

104

Annex A: Supplementary Figures

Figure 0-1: Benchmarking Trends in Overall CPIA Scores

Overall CPIA Scores 4

0 2008 2009 2010 2011 2012 2013 2014 2015 2016

Comoros Sub-Saharan Africa (IDA total) IDA total

Figure 0-2: Progress in Disaggregated CPIA Scores

Change in CPIA Scores: Comoros 4 3 2 1 CPIA Scores CPIA 0 Economic Structural Policies for Public sector Management Policies social management inclusion/equity and institutions

2008 2016

Figure 0-3: Poverty and Consumption Trends in Comoros

Source: World Bank Macro-Poverty Outlook (October 2017)

105

Figure 0-4: A snapshot of the job market in Comoros

Comoros 788,500

Working-age population 450,157 59.5 percent

Inactive 205,272 Employed Unemployed Unemployed 45.6 percent 184,114 ILO discouraged 40.9 percent 16,206 44,565 3.6 percent 9.9 percent In school: 86,625 (42.2%) Agriculture Non-agriculture 69,963 114,151 38.0 percent 62.0 percent

Wage-employment Self-employment 52,167 61,984 45.7 percent 54.3 percent

Public Private Private XXX sector formal informal Number of workers 30,831 2,869 18,467 Percentage within the category 59.1 percent 5.5 percent 35.4 percent

Source: EESIC 2014. World Bank Staff calculations. Note: The number of working-age people is extracted from the WB Database for the year 2014 and is based on the age-dependency ratio that stood at 76.3 percent in 2014.

Figure 0-6: Overall Trade Trends Figure 0-5: Agricultural Trade Deficits in SWIO Countries

70 3 60 2.5 50 2 40 1.5 30 1 20 0.5

10 Exports/Imports 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Comorros Madagascar Imports as % of GDP Exports as % of GDP Mauritius Seychelles

Source: World Development Indicators (WDI). Note: Agricultural imports and exports are classified as food imports and exports in WDI.

106

Figure 0-8: Benchmarking Yields of Major Crops Figure 0-7: Benchmarking Livestock Production per Capita

Crop Yields Livestock Production per Capita 80 20 60 15 40 10 20

5 2014 Kg/person, 0

Yields, 1000 kg/ha 1000 Yields, 0 Comoros Fiji Mauritius Samoa Comoros Fiji Mauritius Samoa Beef and Buffalo Meat Eggs Primary Bananas Coconuts Maize Rice, paddy Meat, Poultry Milk,Total

Figure 0-9: Comparison of cost to export (US$ per 20-ft container)

1,400

1,200

1,000

800

600

400

200

0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Comoros Mauritius Seychelles

107

Figure 0-10: Comparison of port tariffs (US/TEU)

0 Moroni Anjouan Port-Louis Port-Louis Mombasa (Mutsamudu) (import/export) (Transhipment)

Source : Adapted from Schema Directeur Portuaire de l’Union des Comores (MTBS, 2014) Figure 0-11: Road Density per Land Area (km/1,000 km2)

500 400 300 200 100 0 Seychelles Mauritius Comoros Madagascar Sub-saharan Africa

Source: World Bank staff calculations (2017)

Figure 0-12: Road Density per Capita (km/10,000 Inhabitants)

14 12 10 8 6 4 2 0 Seychelles Mauritius Comoros Madagascar Upper MIC avg.

Source: World Bank staff calculations (2017)

108

Figure 0-13: Rankings on Doing Business Topics (2018)

Figure 0-14: Distance to Frontier (DTF) on Doing Business Topics (2018)

Figure 0-15: Madagascar vanilla prices (nominal), 1997-2017, US$/kg

109

Figure 0-16: Madagascar clove prices (nominal), 1997-2017, US$/MT

Figure 0-17: Seasonal Price Volatility for Fresh Fruits and Vegetables

Seasonal fluctuation of the prices of some vegetable products

2 000

1 600

1 200 KMF/kg 800

400

- Tomato Cabbage Carrot Potato low price 100 200 300 300 high price 1 500 2 000 700 1 000

Source: Interview in focus group with 12 producers of Dibouani (Grande Comore)

Figure 0-18: Export Crop Areas in Moheli

110

Figure 0-19: Export Crop Areas in Anjouan

Figure 0-20: Export Crops Areas in Grande Comore

111

Figure 0-21: Main Dairy Production Areas in Comoros

Figure 0-22: Main Poultry (Layer and Broilers) Production Areas

112

Figure 0-23: Horticulture Production Areas in Comoros

113

Annex B: Supplementary Tables

Table 0-1:Trends for Selected Macroeconomic and Poverty Indicators

Avg. '00-15 2014 2015 2016 E 2017 F Income and Economic Growth GDP (current US$, millions) 457.3 676.6 588.5 615.7 652.2 GDP per capita (US$, 674.8 878.7 746.4 763.0 789.6 nominal) GDP growth (annual %) 2.6 2.0 1.0 2.2 2.5 GDP per capita growth 0.1 -0.5 -1.3 -0.2 0.1 (annual %, real) Private consumption growth 3.6 1.0 2.0 2.8 3.7 (annual %, real) Population, Employment and Poverty Population, million 0.7 0.8 0.8 0.8 0.8 Population growth, 2.5 2.4 2.4 2.3 2.4 (annual %) Unemployment Rate 20.1 19.6 19.9 20.0 International poverty rate 18.6 18.1 19.1 19.1 19.2 ($1.9)a Lower middle-income 37.6 37.1 38.1 38.1 38.1 poverty rate ($3.2)a Upper middle-income 64.7 63.9 65.5 poverty rate ($5.5)a External Accounts Export growth, f.o.b 4.1 8.5 -15.0 34.9 -0.5 (nominal US$, annual %) Import growth, c.i.f (nominal 7.4 1.3 -14.8 21.9 6.5 US$, annual %) Merchandise exports (% of 5.4 3.8 2.9 3.1 3.4 GDP) Merchandise imports (% of 29.5 35.5 31.4 31.8 31.7 GDP) Services, net (% of GDP) -3.5 -2.6 -5.2 -8.6 -10.6 Current account balance (% -7.2 -8.6 0.6 -9.1 -8.9 of GDP) Other Doing Business Rank 152 152 153 158 Human Development Index 156.1 160 159 Ranking Source: World Bank Macro-Poverty Outlook (October 2017). Note: (a) Most recent value (2013), 2011 PPPs. Source: WDI, Macro Poverty Outlook, and official data. Notes: (b) Most recent WDI value (2014). E = estimate, F = forecast

114

Table 0-2: Screening Indicators: Export Crop Value Chains

Ylang ylang Vanilla Cloves Development Footprint Number and type of producers Growers: Growers: Growers: (micro, small, medium)4 Almost all smallholders All smallholders (see All smallholders (see (see below); 2 below) below) plantations (micro) Curers: Distillers: 3 (small); 500 (micro) 3-6 (micro)

Number of smallholders 1000-3000 (no survey 5,000 10,000+ data) Prevalence of producer 18 2 5 associations Share of women in VC Cultivation/harvest: Cultivation/harvest: Cultivation/harvest High High Low Distillation: Curing: Drying: Low/Medium Medium High

Share of youth in VC Low Low Harvesting: High Potential contribution to food Uncontrolled & Neutral/positive Uncontrolled & security excessive expansion provided intercropped excessive expansion (availability/access/utilization) planted area could be with food crops. planted area could be negative negative Environmental Impact (soil, Cultivation: Cultivation: Cultivation: water, forests) Tree crop – positive Medium term perennial Tree crop – positive impact soil, water crop integrated with impact soil, water. Distillation: support & shade trees – Drying: Currently v negative positive impact soil, neutral (deforestation water 10,000+m3/yr; depletion Curing: rivers); potential neutral introduction new technologies heat & water system very positive Distribution of producers Anjouan: G/Comore: Anjouan: across Islands 1000-3000 growers 4,300 growers 10,000+ growers (est; 500 distillers 3-6 curers no survey data) Mohéli: Moheli: Mohéli: 100 growers 400 (est) growers 1-2,000 growers (est; 100 distillers 1-3 (est) curers no survey data) G/Comore: Anjouan: G/Comore: 50-100 growers 300 (est) growers <100 growers (est; no 3 distillers 1-3 (est) curers survey data) Distribution of output across Anjouan: 25/30 t G/Comore: >80% Anjouan: 80% Islands Mohéli: 5 t Moheli: <15% Moheli: 20% G/Comore: 10 t Anjouan: <5% G/Comore: Minimal Economic Importance Average production value 2017: 45 t (est) 2017: 30-35t (est) 2017: 3,600t (est) over last 5 years 2016: 38 t 2016: 20t 2016: 6,607t 2015: 31 t 2015: 31t 2015: 3,797t 2014: 25 t 2014: 68t 2014: 4,836t

4 Micro: <10 employees, <€2m turnover; Small: <50 employees, <€10m turnover; Medium: <250 employees, <€50m turnover

115

2013: 41 t 2013: 77t 2013: 3,889t Avg. production share of GDP n/a. See export share n/a. See export share n/a. See export share over last 5 years below below below Avg. value of exports over 2016: $7.1 2016: $9.7 (6.75) 2016: $43 last 5 years (US$ millions) 2015: $4.3 2015: $6.2 (4.4) 2015: $26 2014: $3.5 2014: $7.0 2014: $34 2013: $4.4 2013: $6.7 2013: $32 2012: $3.4 2012: $2.4 2012: $31 Avg. export share of GDP % GDP: % GDP: % GDP: over last 5 years: 2016: 1.2% 2016: 1.6% (1.1%) 2016: 7.0% GDP 2012-2016: US$m 2015: 0.8% 2015: 1.1% (0.8%) 2015: 4.6% 570; 619; 648; 565; 616 2014: 0.5% 2014: 1.1% 2014: 5.3% 2013: 0.7% 2013: 1.1% 2013: 5.2% 2012: 0.6% 2012: 0.4% 2012: 5.4% Complementarity with other Tourism: Tourism: Tourism: sectors (tourism, services) High (local fragrance; Medium (gift sales, field Low (gift sales; field field tours) tours) tours) Services: Services: Services: High (manufacture & Low Low maintenance stills) Competitiveness Indicators Yield per hectare compared to Low (estimate) Similar Similar world’s top five producers Domestic value added Processing Processing (curing): Low – simple drying (processing) (Distillation): 100% 100% only Downstream value Downstream value added: <1% added: <1% Growth forecast for domestic Domestic mkt: Domestic mkt: Domestic mkt: market Minimal mkt, strong Low Low growth potential Domestic tourist mkt: Domestic tourist mkt: Dom. tourist mkt: Low Low strong growth potential if develops Long-term global demand for High. High High product Access to international High High High markets (current) Export potential 80+ tonnes 150-200 tonnes Within current production range 3,000 to 6,000 tonnes Import substitution potential n/a n/a n/a Price volatility Low V High (extreme) Medium Resilience to production High High Low shocks Enabling environment Quality of the policy Low Low Low environment Level of donor support High High Low received Investment potential Private sector activity (SME High High High share of market/share of investments)

5 figures in brackets exclude potential double counting of Mauritius imports.

116

Recent investment activity by Low Low Low public sector Share of loans/portfolio <5% <5% <5% (banks, MFIs etc.) Lending appetite (banks, Banks: Banks: Banks: MFIs etc.) Low Exporters only Exporters only MFI: MFI: MFI: Currently: Low Currently: Low Currently: Low Potential: High Potential: Medium Potential: Medium

Table 0-3: Screening Indicators: Livestock Value Chains

Small Intensive Backyard Indicator Dairy ruminants poultry poultry Number of producers * *** * ***

Number of other VC actors ** * ** *

% of smallholders ** *** * ***

Prevalence of VC organizations *** * *** *

Share of Women in VC * ** ** ***

Share of youth in VC *** * ** *

Potential contribution to food security * * * **

Environmental impact (*** if none) ** ** ** ***

Value of imports ** * *** *

Import substitution potential *** ** *** *

Productivity ** * *** *

Domestic value added ** * ** *

Domestic demand & growth forecast *** ** *** *

Price volatility ** * ** **

Risks *** * *** *

Resilience to shocks * *** * **

Quality of policy environment * * * *

Level of donor support * * * **

Private sector activity ** * *** *

Public investment * * * *

Access to credit ** * *** * TOTAL SCORE 40 31 43 33

117

Table 0-4: Ylang-ylang prices in Comoros, KMF

2017 2015 2014 2006 Per degree 2,800-3,600 1,700 1,450 750-800 Troisième/kg 60-75,000 17,000 14,500 12,500 Flower price/kg 1,600 500-650 400 250-300

Table 0-5: France & USA import prices, ylang oil, 2011-2017, US$ & €/kg

2011 2012 2013 2014 2015 2016 2017 France/Comoros €/kg 74.81 84.01 80.68 106.27 124.75 172.68 208.07 $/kg 103.99 108.37 107.30 141.34 138.47 191.67 235.12

USA/Madagascar $/kg 176.99 148.75 161.84 104.64 121.55 133.92 140.39

OECD avg. 0.719 0.778 0.753 0.754 0.901 0.904 0.887 annual exchange rates $/€

Table 0-6: Ylang grower's production model

Unit Unit Price Quantity Amount Euro € US$ (KMF) (KMF) Assumptions Trees/ha No. 330 Flower yld/tree kg 10 Exchange rate €/S 1 1.19

Sales: Flowers kg 1,600 3,300 5,280,000 Total sales 5,280,000

Cost of sales Kg 250 3,300 Harvesting kg 50 3,300 825,000 Transport 165,000 Total cost of sales 990,000

Gross Margin/ha/yr 4,290,000 8,737 10,397 Gross margin/100 trees/yr 1,300,000 2,648 3,151

Ylang flower harvester Flower harvest/day kg 250 10 2,500

Table 0-7: Ylang-ylang Oil Distiller’s Production model – wood, unimproved furnace

Unit Unit price Quantity Amount Euro € US$ (KMF) (KMF) Assumptions Oil yield 2% Flower charge (kg) 100 Price/degree/kg 3,500 KMF Price T/kg, KMF 70,000 Exchange rate €/$ 1 1.19

118

Oil fractions: Degrees >900 % total yld ES 70 8 E 60 16 P 50 10 D 40 8 T 58

Unit Unit price Quantity Amount Euro € US$ (KMF) (KMF) Sales of oil: ES kg 245,000 0.16 39,200 E Kg 210,000 0.32 67,200 P Kg 175,000 0.20 35,000 D Kg 140,000 0.16 22,400 T Kg 70,000 1.16 81,200 Total sales 245,000 499 594

Cost of sales: Flowers Kg 1,600 100 160,000 Wood M3 12,500 2 25,000 Labour Unit 5,000 2 10,000 Use of still Forfait 3,500 Total cost of sales 198,500 404 481

Gross margin/distillation 100 kgs flowers 46,500 95 113

Gross margin/year @ 8 distillations/month 4,464,000 9,092 10,819 Gross margin/year @ 16 distillations/month 8,928,000 18,183 21,638

Gross margin per 33 distillations 1,534,000 3,125 3,719

Table 0-8:Ylang-ylang Collector's Production Model

Unit Unit price Quantity Amount Euro € US$ (KMF) (KMF) Assumptions: Price/degree/kg 100 Price/kg T 2,000 Exchange rate €/$ 1 1.19

Revenue: ES Kg 7,000 0.16 1,120 E Kg 6,000 0.32 1,920 P Kg 5,000 0.20 1,000 D Kg 4,000 0.16 640 T Kg 2,000 1.16 2,320 Total revenue 7,000 14 17

Revenue/yr @ 8 distillations/mth 672,000 1,369 1,629 Revenue/yr @ 16 distillations/mth 1,344,000 2,737 3,257

Revenue per 33 distillations 231,000 470 560

119

Table 0-9: Ylang Collector's Production Model (Improved Technology)

Unit Unit price Quantity Amount Euro € US$ (KMF) (KMF) Assumptions: Price/degree/kg, 100 KMF Price/kg T, KMF 2,000 Exchange rate €/$ 1 1.19

Revenue: ES Kg 7,000 0.20 1,400 E Kg 6,000 0.40 2,400 P Kg 5,000 0.25 1,250 D Kg 4,000 0.20 800 T Kg 2,000 1.45 2,900 Total revenue 8,750 18 21

Gross margin/year @ 8 distillations/mth 840,000 1,711 2,036 Gross margin/year @ 16 distillations/mth 1,680,000 3,422 4,072

Gross margin per 33 distillations 288,750 588 700 Table 0-10: Ylang distiller's Production Model (Improved Technology)

Unit Unit price Quantity Amount Euro € US$ (KMF) (KMF) Assumptions: Oil yield 2.5% Flower charge kg 100 Price/degree/kg 3,500 Price T/kg 70,000 Fuel oil liters/hr 5 Fuel oil, No hrs 4 Exchange rate €/$ 1 1.19 Oil fractions Degrees % of total >900 yield ES 70 8 E 60 16 P 50 10 D 40 8 T 58 Unit Unit price Quantity Amount Euro € US$ (KMF) (KMF) Sales of oil: ES Kg 245,000 0.20 49,000 E Kg 210,000 0.40 84,000 P Kg 175,000 0.25 43,750 D Kg 140,000 0.20 28,000 T Kg 70,000 1.45 101,500 Total sales 306,250 624 742

Cost of sales: Flowers Kg 1,600 100 160,000 Fuel oil Litres 450 20 9,000

120

Wood M3 12,500 0.5 6,250 Labour Unit 5,000 2 10,000 Use of still Forfait 3,500 Total cost of sales 188,750 384 457

Gross margin/distillation 100kg 117,500 239 285 fls

Gross margin/year @ 8 distillations/mth 11,280,000 22,974 27,338 Gross margin/year @ 16 distillations/mth 22,560,000 44,947 54,677

Gross margin per 33 distillations 3,877,500 7,897 9,398 Table 0-11: Production Model for a Medium-Sized Poultry Farm

Items Unit price No of unit Price Expenditures day old chicks 850 770 654,500 feed starter 330 750 247,500 feed growth 308 1200 369,600 feed layers 300 28,105 8,431,500 Vet products and care 200 700 140,000 egg trays 6,253 85 529,128 water lump sum 200,000 investment building (depreciation) 5,000,000 10 years 500,000 small equipment (depreciation) 500,000 5 years 100,000 Total expenditures 11,172,228 Incomes eggs 70 187,600 13,132,000 cull hens 1,000 650 650,000 manure 1,500 70 105,000 Total incomes 13,887,000 Financial result 2,714,772

Box 2 : Business Model for Poultry Hub

POULTRY SERVICE CENTER (POULTRY HUB) PROPOSED MODEL The proposed enterprise should serve 30 to 50 farmers (20 to 30,000 animals) to have a sufficient volume of activities. Entrepreneurs could be young graduates but could also be people who are already in the business, youth or not. In order to maximize profits, the services they provide should be diversified: - Sale of feeds: the total volume should be around 1,000 MT per year (required by 25,000 animals in average). This quantity could originate from both local processing of feed (500 MT per year) and importation of ingredients, and importation of already processed feed (500 MT per year) - Sales of chicks: the total annual turnover would be around 30,000 chicks per year. Half of this could be imported, and half could be 1produced locally - Sales of small equipment and other inputs: feeders, drinkers, egg trays

121

- Sales of veterinary products: vaccines, vitamins, etc. - Advisory and technical services (debeaking, vaccination)

The total investment required to set up such a business will amount to 125,000 USD composed as follows: - Feed processing equipment (capacity 2 MT/day): 40,000 USD - Hatchery (capacity 2,000 eggs): 10,000 USD - Building: 30,000 USD - Vehicle: 30,000 USD - Generator: 15,000 USD

The entrepreneurs will require financial support (blending of soft loan and matching grant) as well as capacity support: initial training (theoretical and practical - preferably outside the country in a similar business), and coaching services (technical, but also on management and marketing).

Table 1: Imports of fresh and processed fruits in the Comoros between 2009 and 2016 (MT)

Product 2009 2010 2011 2012 2013 2014 2015 2016 Average Dates, figs, pineapples, avocados, guavas, 223 152 146 75 184 153 12 229 147 mangoes and mangosteens, fresh or dried Coconuts, Brazil nuts and cashew nuts, fresh or 230 95 73 8 2 14 70 dried, whether or not shelled or peeled Apples, pears and quinces, fresh - 11 12 78 36 89 58 153 55 Other nuts, fresh or dried, whether or not shelled 55 2 14 3 45 93 163 54 or peeled (excluding coconuts, Brazil nuts ... Bananas, incl. plantains, fresh or dried 20 26 6 17 Citrus fruit, fresh or dried 7 12 5 26 8 16 12 Grapes, fresh or dried 5 2 4 5 2 5 10 11 6 Fruit and nuts, provisionally preserved, e.g. by 1 2 6 3 sulphur dioxide gas, in brine, in sulphur ... Dried apricots, prunes, apples, peaches, pears, 1 4 1 - 5 5 1 2 papaws "papayas", tamarinds and other edible ... Fresh strawberries, raspberries, blackberries, 3 1 2 back, white or red currants, gooseberries and ... Melons, incl. watermelons, and papaws 1 1 1 (papayas), fresh Total 515 276 266 174 225 343 232 579 326 Source: Comtrade database.

122

Table 2: Imports of fresh and processed vegetables in the Comoros between 2009 and 2016 (MT)

Product 2009 2010 2011 2012 2013 2014 2015 2016 Average Potatoes, fresh or chilled 1 2 14 30 68 115 95 3 41 Tomatoes, fresh or chilled 4 38 86 1 9 39 30 901 139 Onions, shallots, garlic, leeks and other 582 791 821 611 419 2 2 414 455 alliaceous vegetables, fresh or chilled Cabbages, cauliflowers, kohlrabi, kale and 25 15 6 10 14 3 288 52 similar edible brassicas, fresh or chilled Lettuce "Lactuca sativa" and chicory 8 38 13 2 10 6 202 40 "Cichorium spp.", fresh or chilled Carrots, turnips, salad beetroot, salsify, celeriac, radishes and similar edible roots, 1 1 3 3 66 15 fresh . . . Cucumbers and gherkins, fresh or chilled 17 24 49 30 Leguminous vegetables, shelled or 92 61 108 95 36 22 69 unshelled, fresh or chilled Other vegetables, fresh or chilled (excluding potatoes, tomatoes, alliaceous 13 6 6 1 13 8 vegetables, edible . . . Sub-total fresh vegetables 726 969 1 081 753 556 165 127 1 958 792 Vegetables, uncooked or cooked by 51 131 135 73 28 90 156 10 84 steaming or boiling in water, frozen Vegetables provisionally preserved, e.g. by sulphur dioxide gas, in brine, in sulphur 2 5 4 26 18 3 10 water . . . Dried vegetables, whole, cut, sliced, broken or in powder, but not further 42 32 79 34 18 11 189 - 51 prepared Dried leguminous vegetables, shelled, 431 452 387 206 87 720 871 451 whether or not skinned or split Sub-total processed vegetables 526 620 605 339 151 821 1 216 13 536 Total vegetables 1 252 1 589 1 686 1 092 707 986 1 343 1 971 1 328 Source: Comtrade database.

Table 0-12: Screening Indicators: Fresh Fruit and Vegetables

Indicator Tomato Onion Citrus Mango Development Footprint Number and Type of Producers (Micro, Rural Rural Rural Rural Small, Medium With Definitions) Smallholder Smallholder Smallholder Smallholder Number of Smallholders Unavailable Data Unavailable Data Prevalence of Producer Associations Medium Medium Low Low Share of Women In VC Medium Medium Low Low Share of Youth in VC Low Low Low Low Potential Contribution to Food Security High High Medium Medium (Availability/Access/Utilization) Environmental Impact (Soil, Water, Forests) Medium Medium High High Distribution of Producers Across Islands Medium Medium Medium Medium Distribution of Output Across Islands Medium Medium Medium Medium Economic Importance Average Production Value Over Last 5 Low Low Low Low Years Avg. Production Share of GDP Over Last 5 Low Low Low Low Years Avg. Value of Exports/Imports Over Last 5 High High Low Low Years Avg. Export/Import Share of GDP Over Low Low Low Low Last 5 Years

123

Indicator Tomato Onion Citrus Mango Complementarity with other sectors Tourism, Tourism, Tourism Tourism, transport transport transport Competitiveness Indicators Yield Per Hectare Compared to World’s Low Low Low Low Top Five Producers Domestic Value Added (Processing) Low Low Low Low Growth Forecast for Domestic Market Medium Medium Medium Medium Long-Term Global Demand for Product High High High High Access to International Markets (Current) Low Low Low Low Export Potential Low Low Low Low Import Substitution Potential High High Low Low Price Volatility High High High High Enabling environment Quality of The Policy Environment Medium Medium Medium Medium Level of Donor Support Received Medium Medium Low Low Investment Potential Private Sector Activity (SME Share of Low Low Low Low Market/Share of Investments) Recent Investment Activity by Public Sector Low Low Low Low Share of Loans/Portfolio (Banks, Mfis Etc.) Low Low Low Low Lending Appetite (Banks, Mfis Etc.) Low Low Low Low Private Sector Activity (SME Share of Low Low Low Low Market/Share of Investments)

Endnotes i The joint work between ANPI and the IFC Advisory Services has been targeting seven areas of the Doing Business agenda: i) Enterprise creation: simplification of procedures for business registration; ii) Property title registry: revision of the new general tax code and reduction of property registration costs; iii) Access to credit: adoption and promulgation of the law on leasing; information and training courses for professionals at financial institutions; implementation of a Credit Information Bureau; iv) Construction permits: Adoption of the new code of the urban planning which has reduced the procedures, the cost and the duration of granting of permits; v) Execution of contracts: the establishment of the Court of Arbitration and Mediation of the Comoros (CACOM) for facilitating the settlement of commercial disputes; vi) Cross-border trade: work in progress for the simplification of import and export procedures (reduction in steps and of duration); In addition to regulatory simplification and with IFC support, a steering committee and sub-committees for structured Public-Private Dialogue around Doing Business indicators have been created, as have two “one-stop” business registries, a leasing framework and an alternative dispute resolution framework. A study (together with the Bank of France) is underway for the establishment of a credit information bureau. A new Competition Law is being enacted and various new codes have been adopted (tourism, urbanism, etc.) The recent progress in the Doing Business rakings (a 10-point improvement in the ranking for starting a business) is the result of these efforts. However, the advances made are for the most part formal, and full implementation remains a priority going forward. ii ACET (2017) iii The three dimensions of food security are availability, access and utilization. Availability: Food must be available in sufficient quantities and on a consistent basis. It considers stock and production in a given area and the capacity to bring in food from elsewhere, through trade or aid. Access: People must be able to regularly acquire adequate quantities of food, through purchase, home production, barter, gifts, borrowing or food aid. Utilization: Consumed food must have a positive nutritional impact on people. It entails cooking, storage and hygiene practices, individuals ‘health, water and sanitations, feeding and sharing practices within the household (WFP). iv OECD (2011) v CIA World Factbook 2016

124

vi Mohéli is 5 million years old and formed of a low central ridge reaching 790m, while Anjouan is younger (3.9 million years) with steep sided mountains reaching to 1595m at the summit of Mt. Ntringui. vii Dahari viii The rural population of Comoros is growing at 2.3 percent and it is ranked in the 90th percentile globally (Find Friends, World Development Indicators) ix GDP per capita is $770 in 2016 (World Development Indicators). x 42.4 percent of population were below the national poverty line in 2014 (World Bank Poverty Assessment 2017). xi World Bank Poverty Assessment (2017) xii The share of working-age people at school amounts to 42.2 percent of inactive population xiii Nearly 200,000 Comorians are estimated to reside outside the country, with a majority in France. Per World Bank estimates, remittances account for 20 percent of household income, and nearly 25 percent of GDP. xiv World Bank Poverty Assessment (2017) xv World Bank Poverty Assessment (2017) xvi World Bank Macro-Poverty Outlook (2017) xvii Comoros’ international air passenger traffic is marginal, the smallest of the Indian Ocean islands: in 2012, Comoros accounted for less than 3% of the region’s traffic (World Bank 2012) xviii World Development Indicators (2014) xix Remittances from the diaspora accounted for 25% of Comoros’ GDP in 2013, exceeding its exports (15% of GDP) (CIA World Factbook 2016). xx Comoros was in the top three countries of the world in terms of food import share of merchandise imports every year from 2010 to 2012 (CIA World Factbook 2016). xxi CIA World Factbook 2016 xxii Wheat flour, palm oil, and processed sugar are other important food imports. The agricultural trade deficit was $126.9 million (22% of GDP) in 2015. While the food share of imports and exports have declined in the last two years, it is unclear whether this is cyclical or a long-term trend reflective of structural changes in the economy. xxiii Comoros Disaster Risk Profile (World Bank 2017) xxiv Cyclones typically hit during the rainy season (December to April). Karthala (elev. 2,361 m) on Grand Comore Island last erupted in 2007; a 2005 eruption forced thousands of people to be evacuated and produced a large ash cloud (CIA World Factbook). xxv Maplecroft, a UK-based firm that specializes in risk analytics and forecasting, ranked Comoros in 2008 as the most vulnerable country to face the future impacts of global warming, such as increased storms, rising sea levels and agricultural failure. Under certain scenarios, up to 10 percent of the coastal population could be displaced due to erosion and contamination of coastal water tables by sea water, potentially destroying coastal infrastructure - including much of the road network - and raising the incidence of malaria (Country Partnership Strategy 2014). xxvi CIA World Factbook 2016, ReSAKSS Database xxvii World Development Indicators (2014) xxviii The poultry yields/carcass weight in Comoros, Mauritius, Fiji and Samoa is 800, 1114.4, 1804.1 and 800 g/animal respectively (FAOSTAT). xxix Babu and Shishodia (2017) xxx In Anjouan, the annual deforestation rate was 8 percent between 1995 and 2014. Between 2000 and 2005, the annual deforestation rate in Comoros was 7.4 percent, the highest in Africa (UN 2007). https://rainforests.mongabay.com/deforestation/archive/Comoros.htm . xxxi Union of Comoros – Economic Management Guidelines (2009) http://www.lescomores.com/en/economy/agriculture.php xxxii Union of Comoros – Economic Management Guidelines (2009) http://www.lescomores.com/en/economy/agriculture.php xxxiii Agricultural land = 155,000 ha and arable land = 65,000 ha (World Bank). Population employed in Agriculture = 69.963 (World Bank Poverty Assessment 2017) xxxiv Doing Business (2018) xxxv World Bank Climate Change Knowledge Portal xxxvi AQUASTAT (1999) xxxvii UNDP (2015) xxxviii The network covers a total of 834 km, 508 km of which approximately (i.e., 61%) in Grande Comore, 232 km in Anjouan and 94 km in Moheli. xxxix The 2014 update of the road condition survey conducted under the Plan Directeur National des Transports Routiers (PDNTR) revealed that 57.6% of the total road network (67.9% of the priority network) is in poor or very poor condition.

125

xl In Colombia, a reduction in transit time led to a 50 to 200 percent improvement in agricultural productivity (World Bank). xli Agricultural Census (2004) xlii The country's foreign trade is mainly carried out by the port of Mutsamudu (Anjouan), which is the only deep-water port capable of accommodating large-scale vessels and which has a large and efficient loading and unloading equipment enabling the shipment of large vessels which can be handled in less than a day. This port, where handling is conceded to Anjouan Steevdoring Company (ASC, subsidiary of Steevdoring Mauritius), is a redistribution center for the main regional ports. Traffic has grown steadily (about 32,000 containers in 2013 and nearly 40,000 by 2015), of which only about 10% is destined for Grande Comoros. The Port of Moroni, in Grande Comoros, managed by Bolloré Africa Logistics since 2012 under an agreement with the Comorian State and has experienced a significant improvement in its facilities and equipment and in its management. However, with an average annual volume of about 10,000 containers, it remains much less developed than that of Mutsamudu due to the shallow water that does not accommodate large vessels on the one hand and its Close to the city center, which makes it very difficult to access it, especially during working hours, in view of traffic congestion. At Mohéli, a port is being built, by extension of the landing stage of Fomboni; It should make it possible to significantly improve the unloading and storage capacity of the goods, which were very low. xliii Modernisation of Comoros’ main port of Mutsamudu (Anjouan) could help position the country as a regional transshipment hub. Comoros’ Port masterplan study conducted in 2014 indicates that Mutsamudu could reach a transshipment traffic of 150,000 TEUs under the median investment scenario by 2030 (figure 10) provided major infrastructure upgrades are realized to consolidate its fairly competitive position in the sub region. Indeed, Mutsamudu needs to keep up with nearby ports which have comparable draughts (14 meters or more) and larger container terminals. Installation of a trade free zone in the vicinity of Mutsamudu would also facilitate logistics and improve its competitiveness as was done in Port-Louis, Mauritius. xliv World Bank, World Development Report 2013: Jobs for Development (Washington, DC: World Bank, 2012) xlv A large civil service captures a large share of public expenditures – the wage bill consumes three fourths of tax revenues (World Bank Poverty Assessment 2017) xlvi A. Rahmn, “Investment Climate Reforms and Job Creation in Developing Countries. What Do We Know and What Should We Do?” (Policy Research Working Paper 7025, World Bank, Washington, DC, 2014). xlvii Doing Business Indicators (2018) xlviii Early 2014, the banking and financial system consisted of 10 financial institutions approved by the Central Bank of the Comoros: (i) 4 banks (Banque de développement des Comores, Banques pour l'industrie et le commerce, Exim-Banque- Comores et Banque fédérale de commerce); (ii) 3 decentralized financial institutions (Union des Mutuelles d'épargne et de crédit ya Komor (MECK), Union des Sanduk d'Anjouan et Union des Sanduk de Mohéli); (iii) and 3 financial intermediaries (Société nationale des postes et des services financiers, la Maison comorienne des transferts et des valeurs, et Comores assistance internationale). xlix World Development Indicators l World Development Indicators. Difficulties surround the enforcement of financial contracts in Comoros, especially in a few notable cases where the recovery of claims by creditors has been stymied by weak judicial processes. Furthermore, the uncertain legal environment has also led banks to lend against collateral that is linked to fluctuations in gold prices, exposing some financial institutions to systemic market risk. Moreover, tightly-knit social relations in the Comoros also hinder even-handed application of the law. Reinforcement of legal institutions, improving property rights and the quality of collateral, and the rigorous and the systematic application of financial contracts are, therefore, critical for the health of individual financial institutions, overall financial stability and further financial deepening (IMF 2015, ‘Financial Inclusion: Can It Meet Multiple Macroeconomic Goals’, Staff Discussion Note SDN/15/17. li When it underwrote 69% of price support financing due to agreed policies with the government. lii http://prod1-afd-mig.integra.fr/home/pays/afrique/geo-afr/comores/projets-comores/projet-finance-Meck-Moroni liii BCC Annual Report (2016) liv MECK Moroni lv https://www.themix.org/mixmarket/countries-regions/comoros lvi The MECK network offering loans over a maximum period of 5 years for agriculture, for amounts ranging from 100,000 KMF to 5 million KMF, with interest from 9%. The main conditions are collateral to an amount equal to 20% of the amount of the loan. The Sanduk network proposes: (i) ordinary credits whose purpose is the financing of any income generating activity for a maximum capped at 1.5 million KMF, payable in 2 years, a guarantee for a value of 150% of the amount borrowed is requested; and (ii) the credit of solidarity savings to the attention of the poor in addition to financial institutions, for an amount capped at 100,000 KMF, payable in 6 months. The rate of interest for agricultural productive activities is 1%/month, calculated on the amount of principal outstanding. Land represents 90% of the guarantees provided by borrowers.

126

lvii In 2016, the Government of Comoros, made access to credit easier by establishing a new credit registry in the Central Bank (DB16). However, only 9.8 percent of all adults are covered by a credit registry and this number is even lower among agricultural producers (Doing Business Indicators 2018) lviii While the Chamber of Agriculture (CA) split from UCCIA in 2016, ostensibly to cater to specific needs of agribusinesses, it has been unable to secure office premises, financial resources, and its legitimacy as a sector representative is challenged by older organizations. (World Bank field mission interviews 2017) lix The first short list of participants in the contest "Dare to Undertake" was published by the Chamber of Commerce and Industry of Ndzuwani (CCIA) last August. Out of the 167 projects related to the creation of companies that have been submitted to the screening committee, 75 were retained, while 92 have been rejected for not satisfying the required criteria. Among these pre-selected projects, 20 relate to agriculture, 18 to fisheries, 17 to livestock, and 6 to new technologies. Also, tourism and catering have five, and distillery has four, while two are in crafts and other three in service provision. lx Al-Watwan news reports here and here (February 24, 2017) lxi UN Energy Profile, http://www.irena.org/EventDocs/Aboud_Presentation_SIDS_IRENA_workshop.pdf lxii UN Energy Profile lxiii Doing Business Indicators 2018 lxiv World Bank Mission (2017) lxv http://www.thinkgeoenergy.com/comoros-islands-hopeful-for-up-to-30-mw-geothermal-potential/ lxvi https://news.nationalgeographic.com/news/energy/2013/13/131223-earthspark-solar-micro-grid-haiti/ lxvii https://poweringag.org/news-events/video/solar-milk-cooling-siaya-county-kenya lxviii The customary land tenure system is the legacy of the bantu migrants who were reported to be the first settlers on the islands. Land is a common property for a village (community), a clan (lineage), or a family. Each and every family has solely the right of use, and can exploit a piece of land or transmit the occupancy rights to his progeny as long as they cultivate the parcel. Originally, common properties cannot be divided, nor privately appropriated to a member, neither sold (especially to individuals outside the community). The rights to use or to occupy can be claimed by another member if the land is left idling or abandoned. The social legitimacy of use or occupation rights on idling land is at the origin of many land squatting and conflicts. lxix The manyahuli are theoretically the property of the daughters but they are in practice managed by uncles on the maternal side. The sons do not hold the rights to claim the succession regarding the manyahuli. Sharecropping is also a common arrangement between two families. lxx In the tenth century, the Sultans from Iran colonized the islands, and constituted the first Muslim community of the archipelago. They brought in the Islamic socio – politic organization and laws, mainly referring to the Coran and the Minadj-at-Twalibin writings. Village and community properties were rearranged into sultans’ domain called usoyezi. Families and notables who pledge political allegiance to the Sultan were granted land use rights to the usoyezi. The Muslim laws also allow the constitution of private property through the vivification (occupation and valuing) of idling or abandoned land. These private parcels are called milk. The sultan or individuals can donate parcels – called waqf – to the religious authority for the promotion and development of Islam. Apart from the building of Mosque and Koranic school, the waqf can be leased out to private with the rent supporting the expansion of Islam. The Sultan organization, instead of absorbing and replacing the community tenure, went to coexist with it. lxxi In 1886, the colonization (originally the protectorate) of the Comoros by France introduced the colonial, also called French, modern, or State legal system comprising a set of laws issued during the colonial period (specifically Decrees in 1911 and 1926), plus a more recent Civil Code adopted in 1975 that conveys the following principles: (a) all lands belong to State domain except the preexistent recognized community tenure and the private titled properties, (b) the State can transfer private property rights through land registration and titling process, (c) colonial laws will strictly apply to affairs regarding titled lands. The colonial laws also retrieved former Sultan properties into State domain (World Bank Mission 2018). lxxii Most of the rocky soils have poor – quality due to its eruptive volcanic origin which is easily eroded. As a strategy to value small agricultural space, food crops are mixed (maize, banana, cassava) on a small plot – averaging 0.026 ha per household, each farm has an average of 2 plots (Bourgoin et al., 2016). lxxiii Land archives were burned and colonial concession were re-colonized by population on the field. State – own land and forest have over the years been appropriated as community and private land. lxxiv These are new land titles issued after the revolution of 1975 – 1978. The concept of customary system is also present in cities: each urban district is the property of a lineage. Urban households are requesting land title for two

127

main reasons, which have little to do with land securing: 1) it is mandatory to hold a private land title to obtain a building permit, and 2) land title is currently used as a collateral for credit/ loan application with banks/ MFI. lxxv The absence of geodetic benchmarks, the exploitation of old – fashioned topographic material by agents who were mainly trained on-the-job would raise the issue of the location reliability and the security of the mapping of titled parcels. lxxvi Comoros scored 7.0, while the SSA average was 8.6 in the Quality of Land Administration Index (Doing Business 2018). lxxvii With the recent development of decentralization and Communes in the Comoros, ongoing reflections open the possibility of recognizing customary occupation as presumed private property. lxxviii 250 and 600 US$ per square meter (Assessment report GAIC – 2iEC) lxxix Imani (2007) “Survey of Non-Tariff Barriers To Trade: Comoros” lxxx Comoros are members of the COMESA Free Trade Area (FTA), and imports from (and exports to) other COMESA FTA member countries are thus exempted of custom duties (but other import taxes still apply). The Union of the Comoros is also a member of the Southern Africa Development Community (SADC) and the Indian Ocean Community (IOC), but the former has not established a free trade area and the latter is not considered a regional economic community. Tanzania, has become a major trade partner of Comoros over the last decades thanks to its cultural, linguistic and geographic proximity. However, it is not member of COMESA, and its exports to Comoros thus do not benefit from any preferential tariff. A Tripartite FTA, regrouping COMESA, East African Community (EAC) and SADC, is under preparation and should boost further regional trade. The membership to the COMESA FTA will allow business operators to access to cheaper inputs, but it will also expose them further to competition from highly competitive countries (such as Kenya for dairy and poultry). lxxxi World Integrated Trade Solution lxxxii World Bank mission (December 2017) lxxxiii IOC database lxxxiv In theory, exempted inputs only attract a fee of 1% of their value. In practice, obtaining this exemption is a cumbersome process: the importer must be licensed as such by the Ministry of Economy and Finance, then has to apply for an exemption from the same Ministry, for each importation. Exemption is granted by customs only if both documents can be produced. This process is so cumbersome, lengthy and costly, that importers often prefer paying taxes that going through this, especially for small and recurrent importations such as chicks. (World Bank mission, December 2017) lxxxv State of the Seed Sector in Comoros (July 2010) lxxxvi The joint work between ANPI and the IFC Advisory Services has been targeting seven areas of the Doing Business agenda: i) Enterprise creation: simplification of procedures for business registration; ii) Property title registry: revision of the new general tax code and reduction of property registration costs; iii) Access to credit: adoption and promulgation of the law on leasing; information and training courses for professionals at financial institutions; implementation of a Credit Information Bureau; iv) Construction permits: Adoption of the new code of the urban planning which has reduced the procedures, the cost and the duration of granting of permits; v) Execution of contracts: the establishment of the Court of Arbitration and Mediation of the Comoros (CACOM) for facilitating the settlement of commercial disputes; vi) Cross-border trade: work in progress for the simplification of import and export procedures (reduction in steps and of duration); In addition to regulatory simplification and with IFC support, a steering committee and sub-committees for structured Public-Private Dialogue around Doing Business indicators have been created, as have two “one-stop” business registries, a leasing framework and an alternative dispute resolution framework. A study (together with the Bank of France) is underway for the establishment of a credit information bureau. A new Competition Law is being enacted and various new codes have been adopted (tourism, urbanism, etc.) The recent progress in the Doing Business rakings (a 10 point improvement in the ranking for starting a business) is the result of these efforts. However, the advances made are for the most part formal, and full implementation remains a priority going forward. lxxxvii Enabling the Business of Agriculture lxxxviii This data will be available in autumn 2018. lxxxix Tschirley, D., Reardon, T., Dolislager, M., & Snyder, J. (2015). The rise of a middle class in East and Southern Africa: Implications for food system transformation. Journal of International Development, 27(5), 628-646. xc Exports are also made to Mauritius (2013 11t; 2015 26t; 2016 13t), but it is thought that these are then re-exported and counted in the EU and US import figures and so are not included in this data. xci From Central Bank report. The discrepancy with 2014 data from EU & US is extreme. Central Bank data for 2015 is likely to be correct, some minor exports having been made to markets outside the EU/US.

128

xcii Since domestic clove production and exports are also high, Indonesia typically imports 5,000 to 10,000 MT. A significant fall in Indonesian usage of clove in cigarettes would have a major impact on international demand and prices for cloves, as Indonesia would become a substantial exporter into a reduced international market. xciii Total figure for 2016 is probably inflated by double counting in trade statistics, but estimates of exporters in Comoros were also widely variable, ranging from 2,700 to 6,200 MT. xciv The drivers for price increases have been strong demand for all fractions, and prices have risen for all fractions. Prices for the low quality troisième have risen sharply relative to the higher qualities as the margin to production of troisième was so small that some distillers were stopping distillation after the high-quality fractions had been obtained. xcv Trade data for Europe, the major market, is not available for Madagascar, as the HTS code for ylang also includes clove oil and naouli, both of which are also produced by Madagascar. xcvi It is not known whether these plantings have been made with the form genuina (matching Comoros) or macrophylla (cananga), and if genuina whether the environmental conditions will give an oil with the quality characteristics to match those of Comoros or Madagascar. xcvii Cananga oil produced in Indonesia and Philippines from the form macrophylla is not a direct competitor with ylang oil but it can be used as an alternative to the troisième grade when supplies are restricted and/or the price differential between the 2 oils is too large. Currently cananga oil is quoted at US$150/kg. xcviii Clove leaves are not currently used in Comoros, but clove leaf oil is a large volume market. Madagascar produces 1,000 to 2,000 MT annually and current prices are in the range US$15 to 18/kg. Clove stem oil is a standard product with current prices around US$17.50/kg. Clove stems are currently a waste product in Comoros, although occasional purchases are reported at around KMF 450/kg. With an oil content of around 15%, the raw material cost of production of the oil would be in the range US$6.10/kg. Lastly, a niche market exists for clove bud oil. The oil is priced in relation to clove prices, based on an oil content in the 15% range. Current prices are around US$24/kg. xcix Nutrizione Foods, started by entrepreneur Anturia Mihidjai, has led the processing and retailing of sago. http://carrefourentrepreneursoceanindien.org/comores-le-sagou-au-coeur-du-succes-de-nutrizione-food/ c World Bank (2015). ci Covering all three islands, the census was undertaken by sampling 12 agricultural holdings from each of 152 villages/localities (Primary Sampling Units). The enumeration took place in two phases: in the first phase data were collected by enumerators through direct interview, on the demographic and socio-economic characteristics of the agricultural holdings. In the second phase, 912 (i.e. a half of the selected SSUs) agricultural holdings were investigated with objective measurements of area of parcels and crop density and yields. cii Natural meadows and rangelands, a substantial fodder resource for ruminants, are present only on the highlands of La Grille and Karthala Mountains in Ngazidja. However, because of the distance to villages and transport networks, and because of the unavailability of surface water in Ngazidja, they remain underused. In other areas, natural fodder resources such as grassland are not present because of the high population density, but conditions are ideal for cultivation of tropical fodder grasses such as elephant grass. ciii Along-the drier coastal areas, coconuts are a basic staple, with bananas, cassava, breadfruit and some maize is also cultivated. Bananas, cassava and maize are also grown in the wetter, more fertile uplands, together with rice, pigeon peas, yams and vegetables (World Bank 1984). civ Saline groundwater is available near the seashore, but requires treatment to be suitable for agricultural use. Underground acquifers and low-yielding springs are common on the island. cv FAOSTAT data also show a shift towards coconut production and away from other non-cereal crops in terms of harvested area this period, perhaps reflecting changes in relative prices, which are unavailable (Fig. 8). However, there is no corresponding increase in exports of coconut products in the customs data, or evidence for greater local demand for coconuts. cvi Food Balance Sheets, 2014 cvii FAOSTAT 2014 cviii Comoros Embassy; Union of Comoros cix USAID (2016) Food Fortification Initiative cx Under a projected rise of 2 °C by 2050, the area suitable for cassava production is likely to increase significantly in the upland areas of Anjouan, remain suitable in coastal regions, and marginally improve in the hostile, volcanic regions surrounding Mount Kartala on Grande Comore (Bourgoin et al. (2016)). cxi Cassava yields in Comoros are 5.6 tons/ha, the lowest among its comparators Madagascar, Mauritius and Fiji. cxii Talsma et al. (2016) cxiii Seed Sector Baseline Study (July 2010) cxiv Dahari cxv World Bank (1984) cxvi National Investment Promotion Association 2016; Union of Comoros – Economic Management Guidelines 2009

129

cxvii Labor force is estimated to be 245,000 (WDI 2014). The government estimated number of direct fishing jobs to be 8,500 and indirect to be 24,000 in 2009. cxviii Commonly used fishing gear includes mosquito nets and sheets, baskets or cages, iron pipes and spears, and a hook and line (Hauzer et al. 2013). cxix FAO – Comoros Country Partnership Framework (2014) cxx On average, fishing catch is estimated to be 16,000 MT by the National Investment Promotion Association 2016 – significantly less than production potential. However, the UN Environment Statistics estimate for the 2010 catch was 52,000 MT. At the individual level, fishing catch was estimated to range between 75 and 117 kg/month in Grande Comore’s fishing villages in 2010 (Hauzer et al. 2013). cxxi The Comoros is situated on the migration path of tuna and swordfish, the main targets of industrial fishing off the coast of East Africa. The Indian Ocean Tuna Commission has found that tuna species are not overfished, although swordfish show signs of overexploitation. cxxii EU Parliament Press Release: http://www.europarl.europa.eu/news/en/press-room/20180312IPR99504/ending- the-eu-comoros-fisheries-deal cxxiii National Investment Promotion Association http://www.invest-comoros.com/fishing cxxiv UNCTAD (2017) cxxv COFREPECHE, MRAG, NFDS, POSEIDON, 2013. Évaluation rétrospective et prospective du protocole de l’accord de partenariat dans le secteur de la pêche entre l’Union européenne et l’Union des Comores, Contrat cadre MARE/2011/01 - Lot 3, contrat spécifique n° 4, Bruxelles, 111 pp cxxvi Comoros exports to all the major markets but exports have become increasingly focused on India, where it shares a significant duty advantage with Madagascar and other LDC origins against Brazil and Indonesia. cxxvii Since new plantings require 9 years to start producing significant yields, this expansion is likely to be slow. cxxviii Producers sell sun dried cloves, which are not protected from dirt, dust, and animals, thus compromising the hygienic status of the product. When it rains, cloves must be covered or collected inside as any wetting during this stage results in a darker color. If left in heaps for too long before drying, cloves develop a whitish shriveled appearance (‘khoker’ cloves). Exporters also complain that collectors frequently add extraneous matter and water cloves to add weight and packing too many cloves into the sacks (60kg in a 50kg sack), thus crushing the buds and increasing headless cloves and a high percentage of dust. Exporters estimate a minimum of 5% losses from these factors. Storage of cloves is a critical factor, and significant loss of quality can occur if cloves are stored in poor conditions – loss of color, development of molds and pest infestations, fermentation. While exporter warehouse facilities are usually adequate, storage facilities with producers and collectors are frequently minimal. cxxix In practice, exporters often have to meet the International Standard (ISO 2254) or the buyers own standard. With the exception of one exporter (AgriVentures) none of the exporters has any mechanical cleaning sorting or grading equipment. None of the exporters has any re-drying facilities – all drying is by sun drying in the open. cxxx 4 kg of fresh cloves with stems gives 1 kg dry cloves with stems; 3 kg of fresh cloves with stems removed gives 1 kg dry cloves cxxxi The standard contract rates for clove harvesting (KMF200/KMF300 per kg) are not attractive on Grande Comore. These would probably have to more than double to secure wider labour involvement, but no clear response was given on this issue. cxxxii Data for individual years presented in Figure 1 should be treated with caution as the level of double counting included in the figures is not clear and it is assumed that exports to Mauritius are all re-exported to the EU/US. Uncontrolled movement of beans between Madagascar and Comoros is also reported. The overall picture however is clear, from industry reports from inside and outside Comoros and trends in the export data, that production has been in long term decline reaching recent lows of less than 30 MT, but is now increasing on the back of new plantings made since 2014. cxxxiii Assocations of Producers, Préparateurs, Exporters and Office National de la Vanille cxxxiv There is always pressure from vanilla farmers to begin the harvest earlier from their need for funds (a single annual harvest) and fear that beans might be stolen before harvest; and curers want to begin buying before others start and competition pushes up prices. As a result, there is always some low level of ‘early harvesting’. However, when bean prices are very high, as at present, all these pressures are increased significantly, and early harvesting levels increase. However, of all the main origins, Comoros can be said to have the strongest level of control over start of the vanilla harvest. cxxxv This news report is an example of the quality decline taking place due to high prices. cxxxvi Enforcement of the green bean purchase price is variable. In general, as the price is published, all farmers know the price and will not sell beans for less. In the current 2017 harvest season, those buyers who brought early harvested beans before the official price was announced at lower prices than the official price, are being politically pressured through the Office National de la Vanille to make an additional payment to growers to bring the price up to the official price. cxxxvii Survey results reported by Office National de la Vanille; survey data not seen

130

cxxxviii In the short term significant increases in flower production can be achieved through improvements in field management (particularly through imposing a pruning regime to maximize flowering). Well managed ylang trees should yield 10 kg flowers/year, but under poor management (or with old trees) this will fall to an average of 5 kg or less. cxxxix The full physiochemical characterization of the different fractions is given in the International Standard (ISO 3063:2004) and comprises 5 parameters – Relative density, refractive index, optical rotation, acid value, ester value. cxl From Oliver de Bontin, (IFEAT 2006) cxli Additional tests for oil quality can be developed, but none is foolproof and they need to be compatible with use in the field or in simple facilities. Buyers are now also using a lamp test (‘bougie’) to assess oils: when a drop of high quality fraction is placed on a filter paper and heated, no residue should be left; if a residue is left, it indicates that the oil has been manipulated. Optical refraction and refractive index can be assessed simply. If a clear characterization of the fractions could be achieved with Near Infra-Red Spectroscopy, this would allow simple field testing, but this has yet to be tested. Where tests indicate uncertainty about oil quality, then samples could be sent to regional laboratories for detailed Gas Chromatograph (GC) analysis – the standard assessment of the factions in the market. cxlii Under a scenario of higher oil yields (from improved distillation technology and equipment and improved management), returns to distillation can be substantially improved to generate sufficient revenues to finance investment in improved distilleries. Where this is combined with development of sustainable commercial producer groupings based around centralized distilleries, it will provide the basis for technical specialization, employment in the distillery as well as investment in the necessary facilities and equipment. cxliii Modern layer strains reach a maximum laying pace (peak) at 30 weeks of age (around 0.9 egg per hen per day) but this then decreases progressively to reach 0.8 at 70 weeks, 0.65 at 90 weeks and so on. The quantity of feed consumed remains constant over this period. After 1.5 years of age, in the Comoros context, the value of production does not cover the production costs anymore and animals must be renewed. cxliv BBC News https://www.bbc.com/news/world-africa-37617379 cxlv The price differential is significant (+ 100 % for local chicken compared to imported “mabawa”, + 30% for whole chicken). This explains the huge penetration of imported poultry meat in the local market (more than 99% of the market). Only the cull layers (1,000 KMF per head) are somehow competitive with imported broilers, but their quality is much lower, they require to be cooked for a long time and cannot be grilled as it is often the case for chicken in Comoros. Paying 50 to 100 % more for a local chicken is only acceptable by customers in very specific conditions, especially for ceremonies such as grand marriage, where consumption of fresh whole chicken is preferred. The broiler producers therefore target this period for production, because it’s the only moment of the year where customers will accept to pay such a surplus for chicken meat. cxlvi World Bank mission (December 2017) cxlvii Since we do not know the production costs of traders, we cannot evaluate the magnitude of this policy’s effects on market prices. It is conceivable that if importers have a large enough profit margin, they could cut on their margin in order to not pass on price increases to consumers, and maintain their market shares. cxlviii The broiler producers therefore target this period for production, because it is the only time of the year where customers will accept to pay such a surplus for chicken meat. According to a major importer/ met during the mission, the demand can be 4x during July and August. The impact on prices is however limited because the government imposes a price limitation on strategic commodities during these seasons, and this includes poultry meat and “mabawa”. cxlix Gates Foundation letter: https://www.gatesnotes.com/Development/Why-I-Would-Raise- Chickens?WT.mc_id=20160608191716_CoopDreams_BG-TW&WT.tsrc=BGTW&linkId=25350277 cl This estimate was compuled by triangulation of second hand information collected during field visits since there is no recent official data about intensive poultry farms in the country. The 2004 agricultural census did not record cli There is no official data about intensive poultry farms in the country because the large majority of which are less than 15 years old. The 2004 Agricultural Census did not record animal populations, disaggregated by type of production (intensive and free range), because it predated this value chain’s emergence. The number of farmers and animals is thus approached by triangulation of second hand information during the World Bank mission. clii Exotic breeds (Brown Swiss, Holstein, Montbéliarde) have been introduced long time ago in this area and the local cows are mostly crosses with 50 to 87.5 % of exotic blood, which are well adapted to the local conditions. The “grazing”/feeding system is very peculiar: animals are kept tethered permanently on agricultural plots where they stay 2 to 3 months before being moved to another plot. They therefore ensure the fertilization of the plot, and consume the crop residues if any. There are no cow pens and animals are permanently outside, exposed to sun and rain. They are mostly fed with fodder crops such as elephant grass (Pennisetum Purpureum), Guatemala grass (Tripsacum Laxum), associated with legume trees such as glyricidia spp, which are grown between plots and act as anti-erosive devices. Animals are complemented with concentrate feed, purchased locally. Anjouan is free of theileriosis and other major contagious diseases, and prophylaxis is therefore quite simple (regular deworming and bi-annual vaccination against anthrax) and well applied by the majority of farmers.

131

Cows are milked only once a day in the morning because farmers cannot sell the evening milk, because of the absence of chilling and storage facilities; this of course limits the production but benefits to the calves. Male calves are kept up to 3 years and are sold for meat. AI services are available only from time to time (see below services) and farmers mostly use local bulls, at a cost of 3,000 KMF per service. cliii Most of the intensive units are located around Moroni, especially in the area of Mdé, Vouvouni and Iconi, south of the Capital. cliv In the scope of this project, they received imported cows (150 pure Holstein cows and heifers, imported from Tanzania), on credit. GAD provided the technical support. The beneficiaries of this project created a local association, ADEC, which is still active today. A large proportion (50% or more) of these cows did not survive, because of the very different climatic, sanitary and feeding conditions, exacerbated by the very limited capacities of the recipients. Around 50 to 70 of these cows have however survived, but many of them are kept in very poor conditions and are not even lactating because their owners are not able to manage the reproduction aspects (detection of heats in particular). Majority of beneficiaries have thus not repaid the loan to MECK. There are however in this group some farmers who are able to properly manage their animals and generate benefits. clv The highlands of Grande Comore, especially the area of la Grille, are also suited to dairy production, because of the cool climate and abundance of grass, but because of constraints of access and water availability, very few production units are located there. clvi This system is in theory better than the Anjouan tethering system but most of these cow pens are not properly maintained and cleaned and animals live in mud and dung which has an impact on their health and on milk quality. clvii Farmers grow fodder (elephant grass mostly) in dedicated plots, but with limited success, and they therefore also harvest wild grass and weeds to feed animals. Some animals receive complements, but most of them are underfed. The hygiene conditions are also very poor. In addition, most of the farmers do not properly treat their animals, despite the presence of theileriosis in Grand Comoros, which is a major threat for cattle rearing. clviii In Grande Comore, the main sanitary problem is theileriosis that has been introduced from East Africa. It would be difficult to envisage the development of a vibrant dairy sector on this island if this disease is not controlled. There are different ways to control the disease including by controlling its vector, the tick, but the most efficient way to prevent it is through vaccination. The vaccination process is complex and requires availability of liquid nitrogenclviii to store the vaccine, and simultaneous injection of an antibiotic. This has an impact on the cost of vaccination, which is around 40 USD per head (for a lifelong protection). However, considering the impact of the disease, the cost of good dairy animals and the price of the milk, the benefit/cost ratio of this operation is significant. clix Artificial insemination: AI was introduced in Niumakélé by ID around 20 years ago and it has had a very positive influence on the dairy sector. More recently, AI campaigns have been organized between 2010 and 2015 under the PNDHD project (see para 16). Some AI technicians have been trained by these two projects, especially in Anjouan, where around 600 animals have been inseminated since 2010. The main problem that hinders the availability of AI services is the supply of liquid nitrogen, which is needed for the conservation and storage of semen. The PNDHD project established a nitrogen plant, but for unknown reasons, is was decided to locate it in Mohéli, which is by far the island where the demand in AI services is the lowest. In Anjouan, where the demand for AI services is the highest, the AI technician cannot keep AI semen stocks permanently because of the nitrogen issue, and they only organize scheduled campaigns, on synchronized heats. Most of the cows are therefore served by bulls. clx Installation of processing units have been tried by various projects, but none have proven viable. A value chain project, supported by UNDP, financed the establishment of two small processing units – for dried tomato, and juice and jam – with two groups of women in Anjouan. Both units stopped operating after the project due to insufficient market assessment, lack of working capital and inadequate capacity of the units to procure raw material, packaging and other inputs. Unpredictable supplies and prohibitive packaging costs (glass bottles, jars and lids) are the binding constraints to successful operation of any processing units (World Bank field visits). clxi Since 2012, inputs are exempt from all taxes and customs (Arrêtés conjoints n° 012-004/VP-MPEEIA/CAB et 012- 036/VP-MPEEIA/CAB et Arrêtés conjoints n° 17-004/VP-MAPEATU/CAB et 17-007/MFB/CAB portant exonération totale des taxes et droits de douanes sur les intrants agricoles). However importers must have an approval of the MPEEIA, renewable annually and establish an exemption for each import. Despite its provisions, the prices of inputs remains high (see price list in Appendix 3) and the availability is still too uncertain. Fertilizer supply would cover 10% of the needs around and often arrives too late at the producers (especially for vegetable production). clxii After the government withdrew from input supply, CAPAC, an importer association, was created in 2000 by the project PAFIA (Projet d'appui à la filière des intrants agricole), placed under the control of the SNAC and endowed with a working capital by PAFIA was able in the first years to play an important role in the import and the distribution of agricultural inputs. Today the CAPAC no longer has own working capital, but continues to import and distribute inputs independently from SNAC, but only on pre-financing of orders by customers.

132

clxiii Diagnostic du secteur agricole comorien à partir des volets transversaux du PDDAA (NEPAD). Saïd Mahamoudou, mars 2015. clxiv Specialized private companies, such as COMAGRO at Moroni and Réseaux agricoles at Mutsamudu (Anjouan), provide import and distribution of agricultural inputs and also provide technical advice to the farmers, while trying to ensure adequate and continuous availability of inputs. National federations of producers (FNAC and the SNAC) import and distribute also from time to time inputs to their members. Other non-specialized traders sell, from time to time, inputs, but without concern for their origin or their quality. Finally shops, like Doigts Verts at Moroni, stopped distributing inputs recently. clxv Tomato paste in a tube is widely consumed in Comoros households. Tomato paste is imported from the Middle East or Asia and the selling price in shops is 1,150 KMF / kg. At this price, the local transformation of tomato concentrate is difficult to envisage. clxvi The rights to rent a space at Mutsamudu Bazari market location is 3,000 – 6,000 KMF per month, depending on the size of the location and the level of facilities provided. clxvii Modern distribution is still in an embryonic state clxviiand limited to the capital and the main towns of region. 4 supermarket chains of Moroni: Komocash (5 stores); MAG-Market (3 stores); Sara-Market (3 stores) and Sawa-price (6 stores). Single Mag-Market distributes vegetables and fruits premisesclxvii MAG Market sources mainly from 2 producers who regularly deliver vegetables. The delivered products are paid by check, once a month. MAG-Market also sells vegetable (onion, garlic, potato) products imported from the sub-region. The delivery of fruit is not regular, because of the very marked seasonality of production. There is also, in Moroni, a few points of sale, specialized in fruit and vegetables (SNAC/UWEZO, FNAC-FA), private specialty stores. The point of sale UWEZO remains very rudimentary, prices are attractive but it is a very limited supply of product. The specialty store, Rabissi Djema, visited by the mission, is installed in a large well- appointed room, with many refrigerators. But at the time of the visit, the store was virtually empty, the palette of horticultural products was limited to 4 products. clxviii Production cost have been estimated with the use of inputs of good quality, in optimal doses, at the market price. Necessary labor was counted, to ensure good crop husbandry, relying on average the use of 20% of employee at 2,500 KMF/working day.clxviii The yields taken into account are yields expected in relation to the level of crop husbandry and inputs provided. Production level takes into account loss in the field estimated between 10 and 25%, depending on the level of perishability of the products. The products prices are average prices reported by the interviewed producers. The transport costs of the products to the market, currently supported by producers and representing 4 to 7% of production costs, have been included in the calculation of the costs. clxix CAADP provides a set of principles and broadly defined strategies to guide national strategies or “compacts”. Among the key targets of CAADP are a 6% annual growth in agricultural GDP, and an allocation of at least 10% of public expenditures to the agricultural sector. 41 African Union Member States have signed CAADP compacts, 33 of which have developed formal national agriculture and food security investment plans. clxx Stratégie de croissance accélérée et de développement durable 2015-2019 (SCA2D) (Mai 2014) clxxi NEPAD arose from a mandate given by the OAU to the Heads of State of Algeria, Egypt, Nigeria, Senegal and South Africa to develop an integrated socio-economic development framework for Africa. The resultant framework was adopted by the OAU in July 2001. (www.nepad-caadp.org) clxxii Benin (2016) clxxiii Strategie Industrielle Nationale Basee sur la Transformation par les MPME/PME des Ressources Locales aux Comores (December 2017) clxxiv Management of public finances is strongly influenced by the complexity of the political and institutional system. The Union combines a centralized system and a strong autonomy of individual islands. Union law takes precedence over that of islands and is binding on all of the Union. In this institutional configuration, management of public finances involves several entities, including National Assemblies (AN), the executive, and finance ministries and treasuries of the islands and Central government. The budget and treasury management is based on the sharing of responsibilities for tax and budget between the central government and those of the islands. clxxv law n°95-09/AF, decree n°95-106 clxxvi Programme National de Développement Humain Durable (PNDHD) (2013)

133