Report No. 26106-MG Madagascar Rural and Environment Sector Review (In Two Volumes) Volume II: Technical Annexes December 18, 2003 AFTS1 Environmentally and Socially Sustainable Development Africa Region Document of the World Bank ONE NationalEnvironmentOffice (Ofice ~utionalpourI'Environnement) OPCl Public Entity for Inter-CommunalCooperation(Orgunisme Publique de Coopkration ~ntercommunale) PADR RuralDevelopmentAction Plan(Plan d'Actionpour le Dkveloppement RuraT) PAGE EnvironmentalManagementSupport (Project Projet d'Appui Qla Gestion de I'Environnement) PCD Community DevelopmentPlan(Plan Communal deDheloppement) PNAE NationalEnvironmental1Action Plan(PlanNational diictionEnvironnementa~e) PO ProducerOrganization PRSP Poverty Reduction Strategy Paper PSDR RuralDevelopment Support Project(Projet deSoutien au Dheloppement Rural) PTA Annual Work Program(Programme de TravailAnnuel) SAGE Support Servicesfor EnvironmentalManagement(Services deAppui u la Gestion Environnementale) SFI Intermediate LandTenure Security (Skcuritk FonciGreIntermkdiaire) SFO OptimalLandTenure Security (Skcuritk Fonci2re Optimal) SFR RelativeLandTenure Security (Skcuritk Foncikre Relative) SRA ImprovedRiceProduction System (SysdmeRizi2re Amdiorke) SRI IntensiveRice ProductionSystem(SystkmeRiziGreIntensif) TAFA MalagasyNGO for researchon semis direct UNDP UnitedNations DevelopmentProgram USAID UnitedStates Agency for InternationalDevelopment WUA Water Users' Association WWF World Wildlife Fund ZPI Priority InterventionZone (Zone Prioritaire de Intervention) , Vice President: Callisto Madavo Country Director: HafezGhanem Sector Manager: KarenMcConnell Brooks Task Manager: MartienVan Nieuwkoop 3 c Table of Contents LISTOFTABLES............................................................................................................................ 5 LISTOF FIGURES.......................................................................................................................... 7 ANNEX 1 MACROECONOMICSETTING . ............................................................................ 8 1.1 CURRENT SITUATION AND TRENDS....................................................................................... 8 1.2 ECONOMIC AND DEMOGRAPHIC PROJECTIONS ................................................................ 9 ANNEX 2 RURALDCNVIRONMENTSECTOR: SITUATIONASSESSMENT . ................12 2.1 POVERTY PERSPECTIVE .................................................................................................. 12 2.2 RURALPRODUCTIVITY PERSPECTIVE ............................................................................. 16 2.3 ENVIRONMENTAL/ECOLOGICALPERSPECTIVE ............................................................... 38 ANNEX 3 OPTIONSTO IMPROVEFOODSECURITYAND FOODPRODUCTION 49 . .. 3.1 RICE................................................................................................................................ 49 3.2 CASSAVA ........................................................................................................................ 67 4.2 SMALL SCALE FISHERIES& AQUACULTURE ................................................................. 68 ANNEX 4 SUSTAINABLE . INTENSIFICATION AGRICULTURAL PRODUCTION .......................................................................................... & DIVERSIFICATION OF 73 4.1 WATERMANAGEMENT .................................................................................................. 73 4.2 TECHNOLOGY ................................................................................................................. 83 4.3 LAND AND LAND TENURE .............................................................................................. 90 4.4 CREDIT ........................................................................................................................... 92 4.5 RURALINFRASTRUCTURE .............................................................................................. 94 ANNEX 5 . FORESTS& FORESTRYAS PROVIDERSOF EXTRACTIVE PRODUCTSAND ENVIRONMENTAL SERVICES .............................................................. 97 5.1 ECONOMIC IMPORTANCEOF THE SECTOR AND CONTRIBUTION TO RURALECONOMY...97 5.2 ACCESS TO FOREST& BIODIVERSITY RESOURCES ......................................................... 98 5.3 APPROACHES& INITIATIVESTRIED OR CONSIDEREDTO IMPROVE THE ROLE OF FORESTS 99 5.4 WEAKNESSES & STRENGTHSOF FORESTMANAGEMENT ............................................. 106 5.5 CREATING FAVORABLE CONDITIONSFOR INVESTMENT INh4ADAGASCAR'S FORESTS 107 5.6 RECOMMENDATIONS .................................................................................................... 108 ANNEX 6 . SETTINGTHE INSTITUTIONAL STAGE ................................................... 110 6.1 SECTOR PROGRAM FRAMEWORK ................................................................................. 110 6.2 SECTOR MINISTERIAL FUNCTIONS ............................................................................... 118 6.3 FACILITATINGPRIVATESECTOR CONTRIBUTION / PARTNERSHIP............................... PRODUCER/ COMMUNITY E~POWERMENT PARTICIPATION ............................... AND 125 6.4 129 APPENDIX I RICE IMPORTTARIFF SIMULATIONMODEL ................................... 134 APPENDIX I1 POVERTYAND BIODIVERSITY INMADAGASCAR(WDR2003) 136 4 . Listof Tables Table 1.1 Economic Impact ofthe Political Crisis .......................................................................... 8 Table 1.2 Madagascar GDP Growth (annual percentage change in 1984prices)........................... 8 Table 1.3 Madagascar Structure of GDP at current prices (percentage)......................................... 9 Table 1.4 Projection of labor force distribution by sector (2000-2015)........................................ 10 Table 1.5: PRSP projection of GDP by sector (2000-2015).......................................................... 11 Table 2.2: Income Poverty by SelectedHouseholdGroups (2001) ............................................... Table 2.1: lncome Poverty Developments 1993-2001(% of population) .................................... 12 13 15 Table 2.4 Poverty in Rural Madagascarby degree of Remoteness, 1999..................................... Table 2.3: Income Poverty (Rural) by Province (2001)................................................................. 16 Table 2.5 Priorities for Developmentaccordingto Commune Focus Groups.............................. Table 2.5 Average Annual Growth Rates inFood andNon-Food Production.............................. 16 17 Table 2.6: Major Food Crops in Madagascar................................................................................ 18 Table 2.8: Tuber Crops- Strengths afidWeaknesses.................................................................... Table 2.7: Rice - Strengthsand Weaknesses................................................................................. 19 20 Table 2.9: Maize - Strengths and Weaknesses............................................................................. 22 Table 2.10: Major Export Crops in Madagascar........................................................................... 23 Table 2.11: Vanilla - Strengths and Weaknesses .......................................................................... 24 25 Table 2.13: Other Export Crops- Strengthsand Weaknesses....................................................... Table 2.12: Coffee -Strengths and Weaknesses ........................................................................... 26 Table 2.15: Cotton - Strengths and Weaknesses ........................................................................... Table 2.14: Major Industrial Crops in Madagascar.................... 1................................................. 27 28 Table 2.16: Sugar - Strengths and Weaknesses............................................................................. Table 2.17: BeefProduction - Strengthsand Weaknesses............................................................ 29 32 Table 2.18: FattenedDuck Production- Strengths and Weaknesses............................................ 33 Table 2.19: Changes in Shrimp Production over the Past Five Years (official data) .................... 36 36 Table 2.2 1: Traditional Fishing- Strengths and Weaknesses ....................................................... Table 2.20: Shrimp sector - Strengths and Weaknesses............................................................... 37 Table 2.23 Ten Year Deforestation Rate in all 6 Provinces........................................................... Table 2.22 Foresttypes and areas in Madagascar......................................................................... 39 41 Table 3.1: Aggregated account of rice production, processing, and trade..................................... Table 2.23: Causes of Biodiversity Loss ...................................................................................... 46 50 Table 3.2: Regional comparison of rice value chain characteristics(1999) .................................. Table 3.3 : Paddy production (1962-99) ......................................................................................... 51 54 Table 3.5 Regional Comparison of transplanted irrigated rice cropping systems ........................ Table 3.4: Reasonsfor Low Rice Productivity (% ofresponses) ................................................. 55 55 Table 3.6: Productivity ofRice Cropping Systems....................................................................... 56 Table 3.7 Regional comparison of irrigated rice yields ................................................................ Table 3.8 Effect of water managementon paddy yields............................................................... 57 57 Table 3.10: Effect of PerceivedConstraints on RiceProductivity ............................................... Table 3.9: Determinants of Rice Productivity .............................................................................. 57 58 Table 3.11 : Effect of spatial segmentationof markets inthe Northern Region........................... Table 3.12 Regional comparison of downstream operator margins.............................................. 63 63 Table 3.13 :Rice marketing* by poverty quintile........................................................................... 65 66 Table 3.15: Relative caloric share per product and per region...................................................... Table 3.14: Perceivedeffect of effect of rice prices...................................................................... Table 3.16: Distribution of Traditional Coastal Fishermenby Province...................................... 68 69 Table 4.3 :Use of Fertilizers inMadagascar.................................................................................. Table 3.17: Changes in FishProduct Exports 1997-2001............................................................. 70 Table 4.4: Mini-projects categoriedmost significant activities, ................................................... 83 85 Table 4.5: Evolution of soil fertility parameters inAlaotra soils................................................... 86 5 Table 4.6: Rice yields in t/ha infields with varying degrees of sedimentation............................. 87 Table 4.7: Evolution of yields (t/ha) .............................................................................................. 87 Table 4.8 Average labor requirements in man/day/ha................................................................... 88 Table 4.9 Production costs and farm revenues ............................................................................. 89 Table 4.10: Land and Land Tenure Data from EPM-2001............................................................ 91 Table 4.I1: Rice yields under various tenure arrangements......................................................... Table 4.12: Credit extendedto the agricultural sector by BTMBOA (FMG million) .................91 93 Table 4.13: Credit extended by mutual savings and credit institutions in 2000 ........................... 93 Table 5.1 Incomes per product type and beneficiary type oftraded forest products..................... 97 Table 5.2: Approaches tried or considered to improve the role of forests as providers of extractive 100 Table 5.3 Visitors to Madagascar's National Parks and Reserves, 1992-2000......................... products and environmental services .................................................................................. 101 Table 5.4: Community-basedforest management...................................................................... 103 Table 6.1: Sector ProgramFramework....................................................................................... 117 Table 6.2: Recommendedexpenditure realignmentsfor the agricultural sector......................... 118 Table 6.3 Livestock Core Functions........................................................................................... 120 Table 6.4 Functions ofthe fisheries administration (based on law 93-022) ............................... 122 Table 6.5: Institutional Structureof Environment Sector ........................................................... 125 Table A3-1: Effect of rice import tariff on paddy farm gate price............................................... 135 6 L i s t of Figures Figure 2.1 Poverty Evolution in Madagascar ............................................................................ 13 Figure 2.2 Poverty Incidence and Seasonal Food Insecurity ................................................... 15 Figure 2.3 Major Food Crops inMadagascar 2000.................................................................. 18 Figure 2.4 Forest Cover Change M a p 1990-2000...................................................................... 40 Figure 3.2 PaddyYields in Madagascar, Mali and Indonesia(source; FAOSTAT) ............51 Figure 3.1: Map of rice producing regions ............................................................................... 53 54 Figure3.4: Cost-Recovery -Paddy Producer PriceRelationship (2001) ............................... Figure 3.3 :Rice production and availability inMadagascar (1962-99) ................................. Figure 4.2: Relation between fertilizer use and degree of isolation........................................ 64 84 Figure4.3: Distribution of Cultivated Area (EPM-2001 data) ............................................... 90 Figure 6.1 : Map of GTDRs ...................................................................................................... Figure4.4: Correlation between Rice Yields, Fertilizer Use andDegree of Isolation ..........195 12 Figure 6.2: Map of Priority Intervention Zones..................................................................... 114 7 ANNEX 1. Macroeconomic Setting 1.1Current Situation and Trends Madagascaris one of the world's poorest countries interms of GDP. Although significant growth has been seen during the last decade, a political crisis in2002 has largely reversedmany of the gains. The economic fall-out of the political crisis that paralyzed Madagascar untilJuly 2002, following a contestedpresidentialelection held inDecember 2001, is significant. Real GDP is estimatedto decline by 11.9% in 2002, due to output losses inthe transport, export processingzone and tourist sectors. Inrelative terms, the political crisis had a relatively mild adverse effect on the primary sector (Table 1.1). However, farmer's incomes were significantly affected by deteriorating terms oftrade, particularly so in isolated communities, as rising transport costs reducedproducer prices and increasedprices of basic consumer goods (lamp petroleum, salt, sugar, vegetable oil). For example, production inthe cotton sub-sector was hit by the inability ofthe state cotton mill HASYMA to pay about 17,000 small farmers for their output, due to liquidity problems. To the extent that the income reduction has affected farmer's ability to purchase inputsfor the next planting season, agricultural production might be negatively affected in 2003. 2001 (actual) 2002 (estimate) Real GDP 6.0 -11.9 Primary Sector 4.0 -1.4 Secondary Sector 7.6 -25.1 Tertiary Sector 6.1 -12.5 Consumer Price Index 6.9 15.3 The political crisis of2002 follows a period of robust and acceleratingeconomic growth over the last five years (Table 1.2). This trend was driven by: (i)a dynamic secondarysector, with the export processingzones (EPZs) beingby far the most active segment with 22.4% growth in2000; and (ii) the tertiary sector which is particularly pulled by construction, transport, and tourism. Growth of the primary sector has been lagging behind other sectors, thereby staying in line with a long-term trend of falling behind population growth. Growth of the primary sector has been particularly driven by the forestry sub-sector (1999: 17.5%; 2000: 5.7%), while agriculture in fact contracted in 2000 (-2.4%) after growing at 5.3% in 1999. Followingrecuperation from cyclone damages in 2000 and booming vanilla prices, the primary sector grew by a respectable4.0% in 2001. 1997 1998 1999 2000 2001 Real GDP 3.7 3.9 4.7 4.8 6.0 Primary Sector 1.9 2.1 3.4 0.8 4.0 Secondarv Sector 4.7 5.3 4.2 5.6 7.6 Tertiary Sector 4.7 5.1 5.5 6.7 6.1 8 The primary sector represented27.3% of GDP in2001, with agriculture alone accounting for 55% of real value added in the sector (Table 1.3). More than 80% of Madagascar's economically active population works in the primary sector. Although its share in the overall economy is declining due to lower than average growth rates, the primary sector is still twice the size ofthe secondary sector, despitethe spectacular growth of the EPZ in the recent past. Infact, in 2001 agriculture alone was still larger than the entire secondary sector. At the same time, the secondary sector has strongbackward linkageswith the primary sector. For instance, the textile industry, which accounts for 21% of Madagascar's industrial outputs, absorbs 50% of national cotton output (Ravohitrarivo, 2002). Hence, to a large extent, Madagascar'seconomy i s still very much rural-based. This fact is also reflected in the composition ofthe country's exports. Although exports from EPZs accountedfor 35% ofthe country's US$758 million exports in 2001, vanilla and cloves exports stood in 2"dand 3`d place with respectively 17%and lo%, closely followed by the rapidly growing exports of shellfish which amountedto US$72.7 million or slightly less than 10%oftotal exports in 2001. Source: IMF, 2003 Concerns have been voiced about the possibleeffect ofthe strongappreciation ofthe Malagasy franc (FMG) on the country's external competitiveness ingeneraland its ramifications for the primary sector in particular. Over the period mid-1999 to December2001 the nominal and real effective exchange rate increasedby 18% and 33% respectively. This appreciation reflects strong growth in the receiptsfor exports and non-factor services, rising from 20% of GDP to 30% between 1996 and 2001, and rapidly increasingprivate capital inflows, risingfrom 0.2% of GDP to 4% between 1994 and 2000. As a resultMadagascar's current account deficit decreasedfrom 8% of GDP to 1.7% between 1997and 2001. An equilibrium real exchange assessment conductedby the IMF (2003) shows that since 1996the estimated equilibrium and actual real effective exchange rate have been closely aligned, exhibiting differences of less than 5%. Consequently, it is concluded that "the appreciation ofthe real effective exchangerate between 1999 and 2001 would appear to be consistent with an equilibrium change and indicative o f neither misalignment nor a loss of competitiveness" (IMF, 2003). Following the end of the political crisis in July 2002, the Malagasy franc depreciated by 11% in nominal terms, thereby reversing part ofthe appreciationthat was experiencedfrom 1999-2001. Basedon the above, it appears that the exchangerate at this point intime does not imply a bias either in favor o f or against the primary sector. Assuming that the exchangerate climbs back to pre-crisis levels, this could decrease on the costs of imported rice and various primary sector entrants such as fertilizer and pesticides. 1.2 Economic and Demographic Projections Following the high growth rates between 1997-2001,the draft PRSPthat was presented in November 2001, laid out an ambitious target of 6.3% GDP growth starting in2003 and thereafter, so as to position the country to reducepoverty by half in2015. The final draft of the PRSP (May 2003) has set a growth target at 8 to 10%. Based on current demographictrends and assuming 9 that Madagascarwill be able to achieve sustainedoverall economic growth at 8- lo%, the question arises what this would imply for the primary sector so as to obtain a general idea what challenges lie ahead and what forces might start shapingthe sector. Concerning demographictrends, Projectionsfrom the National Institute of Statistics(INSTAT, RecensementGeneral de la Population et de I'habitat, January 97), basedon relatively conservativeassumptionsof 1.5% rural and 4.7% urbanannual population growth, indicate that by 2015 the total population would have grown to about 21.5 million, representingan increase of 42% over the current level of 15.1 million or an average growth rate of 2.4% per year (Table 1.4). This implies that even in the absence of income growth, food production should grow by 2.4%, which representsan accelerationfrom both historic levelsas well the sector's most recent strong performance since 1997-2000when it averaged 2%. As a consequence ofthe projected population growth path, the share ofthe rural population inthe total population would fall from 75% to 64%. However, in absoluteterms, the rural population would still grow from 11.2 million in 2000 to 13.8 million in 2015. At the same time, the share of the urban population would grow from 25% to 36% of the population, representinga massive increase in absolutenumbers from 3,9 million to 7.7 million. Interms of active population, the secondary and tertiary sector would mobilize 40% of all workers (compared with 29% in 2000). This representsalmost a doubling of the number of non-agricultural workers in absolute terms by 2015. Contrary to this massive increase, the number ofworkers in the primary sector would increaseonly by about 18% in absoluteterms. This change inthe demographiccomposition ofMadagascar's population has massive implications.for the primary sector. First of all, the projectedincreaseof Madagascar's urban population would significantly increasemarket demandfor food crops. Although food imports could play a role to meetthis demand, it would simply explode Madagascar's food imports bill. For instance, based on a scenario of stagnantrice production (growing at 0.2%/year), rice imports would increasefrom its current level of aboutUS40 milliodyear to US$235 million in 2015. Second, ifprojected urban demand were to be met by domestic production, it is difficult to envisage that to happenwithout significantly increasingagricultural labor productivity in view of the only modest projected increase in numbers of agricultural workers. Third, given that a large share of agricultural production is currently realized ina subsistencecontext, the structural change in the composition ofthe population and labor force also puts a premium on improving the integration of markets to effectively link food supplies and demand and reflects an important role for agribusinessto properly balancerural production and urbandemand. 2000 2005 2010 2015 RuralPopulation (inhabitants) 11 244 000 12 172 000 13 087 000 13 806 000 % agriculturalpopulation 91% 90% 89% 88% Agricultural population/ primarysector 10270228 10954 800 11647430 12 149280 Number ofagriculturalhouseholds 2 166 000 2 258 722 2 401 532 2 505 006 Number of agriculturalworkers / HH 2,78 2,78 2,78 2,78 Number ofagriculturalworkers (primary) 5898000 6279246 6676259 6963917 No o f ruralnon-agriculturalHouseholds 202 869 250 969 296 819 341 592 UrbanPopulation 3 905 273 4 933 000 6 192 000 7 742 000 Number of urbanhouseholds 976318 1233250 1548000 1935 500 Urbanactive population/HH 2 2 2 2 Urbanactive population 1 952 636 2 466 500 3 096 000 3 871 000 Number of ruralnonagric.Workers 405 738 501 938 593 637 683 184 10 Total Nan Agricultural workers 2 358 375 2 968 438 3 689 637 4 554 184 % labor force in Primary sector 71% 68% 64% 60% % labor force in Industry Services - 29% 32% 36% 40% Total Labor force 8256375 9247684 10365 896 11518 101 Total Posulation 15 149273 17105000 I 19279000 21 5480001 Source: Bockel 2002 As far as the growth projections based on the PRSP are concerned, consolidation of an overall growth rate at 8% would double GDP in 2010, providing a growth of GDPI inhabitant over 40% by 2010 (increase from US$235 to about US$336/inhab in 10 years) and reaching over US$400/inhabitant by 2015 (based on the 6.3% economic growth projection in the earlier PRSP). To ensure such a growth path, the primary sector would have to grow by about 4% per year from 2001 to 2015 and the secondary and tertiary sectors respectively by 8.6% and 7.4% on average until 2015. As far as the primary sector is concerned, a sustained growth o f 4% would represent a significant acceleration from its long term track record. As far as the secondary and tertiary sectors are concerned, the projected sector growth rates also represent a significant increase compared to their average levels o f respectively 5.7% and 5.4% in 1998-2000. Combining the demographic projection with the assumedoverall economic growth rate inthe November 2001 PRSP o f 6.3%, the evolution o f GDP/capita and GDP/worker can be estimated as shown in Table 1.5 which shows that the significant labor productivity differential between the primary, secondary and tertiary sectors for 2000 (1:4.8) will widen. It also demonstrates a continuing shift inthe gravity point o fthe economy away from the primary sector towards the. secondary and tertiary sectors. Table 1.5: PRSPprojectionof GDP by sector (2000-2015) 2000 2005 2010 2015 (billions fmg) (billions fing) (billions fmg) (billions fmg) GDP 22 221 30 160 40 935 55 559 GDP Primary Sector 7 840 9 495 11500 13 929 GDP SecondaryiTertiarySector 14 381 20 664 29 434 41 630 GDP primary sector / GDP 35% 3 1% 28% 25% (000 fmg) (000 fmg) (000 fmg) (000 fmg) Primary Sector GDPiinhabitant 517 555 597 646 Sec.-Tert.Sector GDPIinhabitant 949 1208 1527 1932 Primary GDP / agric. worker 1329 1512 1723 2 000 Sec.-Tert. Sector GDPinon-agric.worker 6 098 6 961 7 978 9 141 Source: Bockel, 2002 11 ANNEX 2. RuraVEnvironmentSector: Situation Assessment Madagascarsuffers three key impediments to meetingthe challenge of sustainable development, including: (i)persisting and extreme poverty, especially in rural areas causingwidespreadfood security problems; (ii)low and declining agricultural productivity, with the result that the rural population remains poor; and (iii)rapid environmental degradationand biodiversity loss due to forest clearance and repeated burning of lands for extensive shifting agriculture. These identified problem areas correspondto three distinct perspectiveson sustainable development inMadagascar and are dealt with in different arenas: (i) the on-going PRSP processfocusing on poverty reduction, covering both ruraland urbanpopulations with the specific aim to cut poverty in half by 2015 through inclusive economic growth; (ii)the Rural Development Action Plan (PADR), which is primarily envisioned as a means to enhance rural productivity and achieve and sustain at least 4% balancedeconomic growth in the rural sector; and (iii)the National Environmental Action Plan (PNAE) which is primarily envisionedas a means to reconcile man with the environment and stop the loss of biodiversity by halting environmental degradation. Each ofthese arenas and their associated programs of action seeks to address a vast array of problems in the rural sector. While the linkages between poverty, low rural productivity and environmental degradation(including biodiversity loss) are becomingbetter understood, these three approaches remain essentially separate at this point. A good understandingof each perspective is a necessary first step to better understandthe big pictureof the sector in an integratedmanner. 2.1 Poverty perspective A general picture of poverty in rural areas can be summedup as follows. First, poverty in Madagascaris largely a rural phenomenon. Eighty-five percentof the poor live in rural areas. Second, poverty rates in rural areas are also significantly higher than in urban areas (Table 2.1). In 2001, the incidenceof rural poverty was 24 percentagepoints higher than inurban areas. Third, extreme poverty, defined as the inabilityto purchasea basket of basic food products to meet subsistence needs, is widespread. Fourth, urbanpoverty has decreased more rapidly than rural poverty since Madagascar startedto enjoy attractive economic growth rates in 1997. A sobering point is that income poverty is thought to have risen sharply by 6 percentagepoints to 75% ofthe population as a result ofthe political crisis, thereby effectively wiping out all the gains that have been achieved since 1997. 1993 1997 1999 2001 Poverty (headcount index) National 70.0 73.3 71.3 69.2 Urban 50.1 63.2 52.1 50.0 Rural 74.5 76.0 76.7 74.0 Extreme Poverty National 59.1 63.1 61.7 61.1 (headcount index) Urban 37.6 54.0 43.2 42.5 12 Figure 2.1 Poverty Evolutionin Madagascar 90 1 " ""1 70 %of 60 Population at 50 or below 40 poverty level 30 20 10 0 1993 1997 1999 2001 Year Source: World Bank,2002 Given the focus ofthe study on the primary sector, arelevantquestion is how farmers would do in terms of povertyvis-a-visother socio-economicgroups. Based on data from INSTAT, Table 2.2 indicatesthat small farmers, fishermen and livestockoperators rank at the bottomof the country's income distribution with poverty incidencerates exceedingthe nationalaverage. Hence, besides beinga rural phenomenon, poverty in Madagascaris also very muchan agricultural phenomenon. In view oftheir sheer numbers and high incidenceof incomepoverty, small farmers are the single- most importantpopulationsegment of poor peopleinMadagascar. There are two alternative cause scenarios for the relationship betweenpovertyand agriculture: one is that individuals and family with no other source of revenue oftenturn to small farming to satisfy basic food needs, the other is that farming does not offer adequateopportunitiesfor familiesto rise out of poverty. HouseholdGroup Yo of Poverty Poverty Share Population Incidence(YO) (YO) Small farmer 62.7 86.1 77.4 Fishermam'Livestock operator 2.3 79.5 2.6 Agricultural entrepreneur 2.0 47.4 1.9 Trader 5.9 32.6 2.8 Salariedemployee 5.0 16.7 1.2 Skilled laborer 13.5 40.2 7.8 Unskilled laborer(rural) 2.4 74.7 2.5 Unskilled laborer (urban) 1.3 44.7 0.8 Other 4.0 48.0 2.7 In view ofthe high incidenceof extreme poverty, other indicatorsthan incomepoverty are also relevant. According to UNDP's HumanDevelopmentReport2002, Madagascarwas ranked 147"' out of a total of 173 countries in 2000, thereby falling into the group of so-called low human developmentcountries, behindBangladeshand Haitiandjust ahead ofNigeria. Its HDIwas 13 0.469, slightly up from 0.441 in 1995 and 0.427 in 1990. The high incidence of extreme poverty not surprisingly is translated in widespread food insecurity in Madagascarwith 40% of the population undernourishedin 1997/99. Consequently, the percentagesof under-weight and under-height children are also relatively high with 33% and 49% respectively (for children under age 5). Sixty seven percent of all Malagasy or 59% of total households, live in a condition of food insecurity. Food security and malnutrition primarily affect the poor but also a fraction ofthe "non poor". Out of those 67% living under food insecurity, 28% are inthe province o fAntananarivo, 9% in Antsiranana, 18% in Fianarantsoa, 13% in Mahajanga, 18% in Toamasinaand 15% in Toliara (SECALINE, 1997). The highest rates of food insecurity occur in the eastern provinces, highlighting a direct relationship betweenpoverty and food insecurity. Following decreasingreal per capita incomes, the level of satisfactionof energy requirements, expressed incalories per personper day, has deteriorated over the last 25 years, falling from 2,490 calories in 1975 to 2,02 1 calories in 1995 and tojust 2,001 calories in 1998 (Bergeron2002). Food intake in Madagascaris also not diversified. It consistsessentially oftwo staples: rice (303g/person/day) and manioc (209gipersonlday). The daily consumption of animal-based products is equally low: 15gfor meat, 15gfor fish and seafood, 21g for poultry and eggs and 8.5g for dairy products. Moreover, 75% of the population lacks access to gooddrinking water. Malnutrition in Madagascarof all forms affects rural areas more than urban ones. The incidence of severe malnutrition in rural areas is 5% comparedto 3% inurban areas (SECALINE, 1997). Nutrition is generally muchbetter in coastal areas where fish and abundant tropical fruits provide valuable dietary supplements. In terms of geographicaldistribution, income poverty rates in rural areas are highest in the provinces of Fianarantsoaand Toamasina, reaching 87.9% (Table 2.3). The incidence of rural poverty is slightly lower in Toliara, Mahajangaand Antsiranana. The province of Antananarivo stands apart from the pack with poverty incidenceat 56.7%. Interms ofnumbers, most rural poor live in Fianarantsoa, followed by Toamisanaand Antananarivo. The relatively low poverty incidence in Antananarivo, coupled with a relatively large share ofthe rural poor living in Antananarivo, implies that poverty targeting mechanisms of social funds and demand-driven rural investments funds would require a higher level of precision as comparedto the other provinces. This observation is in line with Mistiaen et al. (2002) who, while analyzing the geographical dimensionsof poverty in Madagascar, in generalcautions against the use of detailed spatial targeting due to the fact that heterogeneity of poverty across administrative units is much more pronouncedat the provincial than at the commune level. Possibleexceptionsthat are mentioned include urban Antananarivo, rural Antananarivo and Toliara. Even between 1999 and 2001, when poverty was reducedoverall from 71.3% to 69.2%, some regions still became poorer. Thus, while poverty decreased in Antsiranana (-12.1YO),Antananarivo (-8.8%) and Fianarantsoa(-2.8%), poverty increased in Toamasina (+6%), Mahajanga (+4.8%) and Toliara (+2%). 14 Figure 2.2 P o v e w Incidence and Seasonal Food Insecurity 100 -j 90 80 70 60 %50 40 30 20 10 0 Source : INSTATAVorldBanWComellUniversity,2002 Numerousfactors contributeto the conditionof poverty. Perhaps one ofthe bestpredictors of poverty is geographicalisolationas indicatedinTable2.4. The incidenceof povertyis about 17 percentagepoints higher inthe most remotequintile ofhouseholdscomparedto the least remote. Also the povertygap increases significantly with,increasingremoteness. However, even inthe least remote areas, the incidenceof poverty i s still substantial, which indicatesthat other factors also play a role. Inthis context, INSTATflLO findings indicatethat poverty is also strongly correlatedwith: (i)numberof smallchildren: negativeeffect of up to 38% e.g. Antananarivo province); (ii)lack of electricity: +29%; (iii)working inthe rural sector: +15%; (iv) geographical 15 residence:Fianarantsoa+lo% and Toamasina+13%; (v) size ofhousehold: + 6%; (vii) education: primary -6%, secondary & higher:-1 7%); and (viii) rural insecurity. Table 2.4 Poverty in Rural Madagascarby degree of Remoteness, 1999 4 , 77.7 36.5 Source: Paternostro et al, 2001. Iftheprioritiesofcommunesmaybetakenasanindicationoftheunderlyingcausesofpoverty,it would appear that the key factors blockingthe escapefrom povertyare low agricultural productivity, poor road infrastructure, and ruralinsecurity(problems of cattle & croptheft etc.). Poor roads arethe mainconcern inthe east, agricultural productioninthe western highlands& northand insecuritythroughoutthe sparsely populatedwest. Lackof water is a key problem in the extreme south (ILO 2002). It is interestingto notethat the order ofprioritiesis more closely relatedto economic constraints than to socialconstraintswhichwas also an importantfinding of the Bank initiatedVoices ofthe Poor Study in2000 where employmentcame out as the single- most importantpriority aheadof social issues. Table 2.5 Priorities for Develoument according;to Commune Focus Grouus Priority Interventions %of Communities 1. Agricultural development 27 2 Trsnwnrt 26 3. Security 15 I 4. Health 14 1 5. Education 10 6. Water 6 7. Environment 2 Total 100 2.2 Rural productivityperspective Long-termTrends Although beingone ofthe few mega-biodiversitycountriesinthe world andrichly endowed with uniquenaturalresources, Madagascar is on many countsworse offtoday than it was in 1960. Duringthis period, income growthhas notkeptupwith populationgrowthandthe countryhas gradually becomepoorer: GDP/capita'was US$383 in 1960, while it stood at US$246in 2000. Yields ofMadagascar's major food crops like rice, maize, sweetpotatoesandcassavastagnated. Frombeinga rice exporter inthe 1960sand 1970~~ the countrybecameanet importer inthe 1980s. Its roads networkcollapsedfrom 55,000 km in 1960to 33,000 km in 2000. In order to meet the subsistence needsof a populationwhich grew from 5.4 million in 1960to 15.5 million ' Inconstant 1995 US$ 16 in 2002, Madagascar in essence embarkedon a development path characterized by area expansion through slash-and-burn, locally know as t o y . Consequently, Madagascar lost about 12 million ha of forests between 1960and 2000, effectively reducing existing forest cover in 1960by about 50%. This loss of forest, though resulting from agricultural activities did not result in a significant increase in agricultural land. It has been suggested that over the 40-year period the 12 million ha lost was compensated for by an increase in only 100,000 ha of agricultural land. Recent data is more precise and indicates that duringthe decade from 1990to 2000 1,620,000 hectareso f forest was lost (8.6% of forests, Steininger et al. 2003) and the area of rice crop production increasedby only about 81,000 ha (Ministry of Agriculture, this report, total agricultural area increasedby only 28,000 ha). Thus, the forest area lost relative to rice agricultural area gained is a ratio of 20 to 1. At the same time, it is generally recognizedthat the rural sector is potentially competitive. Madagascar has rich and varied agro-ecologicalendowments. While the soils of the highlands are generally sandy and ferralitic, there are extensive fertile volcanic lands inthe north, west, mid-west and around Lac Itasy in the centralhighlands and extensive areas of arable alluvial soils covering the western coastalplain. The various combinations of soil and climate provide favorable conditions for a largevariety of crops -coffee & cacao inthe north, vanilla, spices and oil palm in the north east, rice, fruit & vegetables inthe highlands, tropical fruits inthe east, maize, cotton, tobacco and peanutsinthe mid-west, west and south west and vast rangelands suitable for cattle inthe west and south (AIRD 1993). Added to this are the productive coastal seas of the west coast, south & southeast (Cooke et al. 2000). Besidesfavorable natural conditions, relatively low labor costs offer another competitive advantage. Prior to the 1990s, distorted exchangeratesand poor trade policies could be put on the table as a plausible excuse why the sector could not live up to its widely recognizedpotential. During that period, the transformation and marketing of agricultural productswas inthe handsof national monopolistic parastatals. Poorly designedtaxes and economic barriers were put in placeto extract revenues from export crop producers, leadingto an overall net reduction of export revenues. Domestic food prices were subsidizedand kept low inan artificial manner, which depressed domestic production and led to a surge inrice imports (Randrianarisoa and Minten, 2001). The reforms of the mid-nineties, with a sharp devaluation ofthe exchange rate, reduction of import barriers, liberalization of markets, and privatization of stateenterprises (with the exception of sugar and cotton) have contributed to a relatively balanced incentive structurethat continues to exist through today. In spite ofthese improved incentives, agricultural productivity has remainedalmost stagnant (Table 2.5), thereby suggestingthat the causes of low rural productivity should be considered primarily of a structural nature. Table 2.5 Average Annual Growth Rates inFoodand Non-Food Production 1980-89 1990-2000 Agricultural GDP 2.50 1.40 Food Production 1.63 0.81 Food Production Per Capita -1.04 -2.3 1 Non-Food Production 1.45 -1.77 17 Strengths and Weaknesses of the FoodCrops Sub-sector Food crops dominate the agricultural scene by taking up over 75% of total cultivated area in 2000, which amounted to close to 1.9 million ha (Table 2.6). Rice is the single most important crop, covering on its own approximately 1.2 million hectaresor almost 65% of all agricultural lands. It is the backboneof the rural economy as it is estimatedthat over 10 million people one way or the other are involved in the production of rice in Madagascar. Cassava, provides 14% of all calories consumed in Madagascarand is particularly important to the diet in Antananarivo, Toliara and Fianarantsoa. At the same time, maize, although not extremely important overall, provides 15% o f all caloric requirements in Toliara. Potato is increasingly important as a food crop, but is also said to have export potential. 'Maize Ha 157,391 1 183,840 I 183,580 t/ha 1.o 1.o 0.9 TOTAL Ha 1,796,697 1,809,710 1,877,660 Maize Maize 10% 3% Cassava 41% 9% 5% 18 Rice. With rice takingup such adominant positioninthe sector, its performance effectively determines the overall picture. Unfortunately, average annual growth rates of rice yields have been very disappointing, showing negative growth of-1 .l% from 1975-1984 and only slightly positive growth of 0.2% and 0.1% over the periods covering 1985-1989 and 1990-2001 respectively. Yields of rice are dismal by international standards and even at these already low levels show a declining trend (Table 2.6). As rice production has not kept up with population growth, there has been an apparent deterioration in food security. For example, availability o f rice, measured in kilograms per person, fell from 144 kg in 1981 to 112 kg in 2001(IMF, 2003). In order to make up for the shortfall indomestic production, imports ofrice have increased from US$12.3 million or 2.1% of total imports in 1997 to US$36.8 million or 4.2% of total imports in 2001. According to Bergeron (2001), the prognosis for future availability o f food in Madagascar is not very encouraging. Ifthe current situation o f stagnant productivity persists, the food gap, assuming current per capita consumption i s maintained, is projectedto increase from 92 thousand tons in 2000, to 591 thousand tons in2010. The theoretical potential for increased rice production is high. A comparison o f the results on experimental research stations in Madagascar with other countries that have achieved higher average yields indicates that the yield gap (field level yields as a percentage o f maximum attainable yield under experimental conditions) in Madagascar is much wider than elsewhere. A yield gap of 25% is reported by FOFIFA2,with results under experimental conditions exceeding 8 ton/ha. By contrast, the yield gap in West Java is 80% with similar maximum attainable yield levels. Similarly, experimental research results indicate favorable responseso frice varieties to various nutrients. Responseso f 40% yield increase to 60kg nitrogen, of 36% to 125kg phosphorus and of 43% to 1OOkgo f potassiumare reported on different soils3. Strengths and weaknesses of the rice sub-sector are presented inTable 2.7. Table 2.7: Rice Strengths and - `eaknesses Strengths Weaknesses Production 0 Malagasyrice is competitiveat the 0 Low productivity. Despite efforts farmgate. undertakento improve output, 0 Substantial potentialfor rice productivity remains low. Yields production(1,200,000 hectaresfor barely exceedthree tons per hectare all systems combined). inthe mostproductiveregions. One third of all growers live under 0 Madagascarhas anatural environment(agro-ecological subsistence conditions and barely conditions suitable for different types produce 800 kgha ofpaddy rice on of rice growing) andtraditional an area of less than one hectare. knowledgethat are conduciveto the 0 Isolation of producing areas, due development ofrice growing for both to the poor condition of roads, the local consumptionand export. lack ofbridgesand crossings on 0 There is a high level of most roads, and generally consumption,rankedamongthe inadequate roadmaintenance, world's highest(estimated average thereby increasingcosts of external annual per capita consumptionis 118 inputs. kilograms). 0 Inadequatewater management 0 There is apreference for localrice mentionedas a crucial constraint in Structureand conduct of major agricultural input and output markets andresponseto reforms byrural 'householdsin Madagascar. FOFIFAIIFPRI, 1998. hnp://www.fao.org/ag/agp/agpc/doc/riceinfo/A~ica/MAGBODY.HTM 19 varieties. farmer surveys. Limiteduse ofhigh-performance technologies (improvedvarieties, fertilizer, mechanization,etc.) resultingin low returns on rehabilitationinvestments and irrigationschemes. 0 Rareandexpensivecredit. Fewer than 3% ofrice growers received formal cropyear credit in 1999. The average interestrate is 27% at banks(BTM) and36% at savings and micro-creditinstitutions 0 Lack o f landsecurity: Barely 10% o f agriculturallands are registered with an official propertytitle; this situationmay discourage investment,impedeaccess to credit and hamper leasingof land. ProcessingIMarketing ~ ~~~ 0 Marketliberalizationhas ledto the Hightransportationcosts of developmentofvillage-basedhulling outputs due to poor road facilities which has contributedto a infrastructure. reductioninthe manualpoundingof Poor communicationfacilities producer-processedrice andhas causingstrong geographicalas well stronglyinfluencedadditionalsales as intra-seasonal pricevariation. by farmers. 0 Limited storage capacity at the 0 Competitionfor paddyon the part retail levelalso contributes to of rice processingfactories and significantpricevariation. hulling facilities in major rice producingareas has improved producerprices inrelativeterms. Tuber Crous. Tuber crops, including cassava, sweet potato and potato covered slightly less than 500,000 hectares in 2000 and in terms o fcultivated area come in secondafter rice. Yields of most tubers are stagnant at best. All are to some extent also cashcrops, with a substantial proportion of production being sold onto local markets(e.g. manioc 28%, potatoes 48%, sweet potato 28%). Sweet potato also finds a national market in cattle feed. Ingeneral, yields of tubers are superior to yields of grains (potato & sweet potato 5 t/ha, manioc 7 t/ha). Manioc has shown significant yield improvement since the 1960s from 4 to 7t/ha. For potatoes, high yields are - achieved by specialist farmers. Potatoeshave also played a role in reducing seasonal stress in some areas. Table 2.8: Tuber Crous Strendhs andWeaknesses - I I Strengths I Weaknesses I Production Cassava and sweet potatoesare the 0 Yields are very sensitive to I mainsubstitute food becoming climatic conditions. increasinglyimportantfor the Limiteduse of improvedseeds pooresthouseholds (only 25% of cultivatedarea). Cassava and sweetpotatoes Isolationof productionareas due contributean average of 14% of 0 to the deplorable state ofroads. calories consumedby rural 20 households. Substantialpotato production potential as the crop adapts well to various types of soils (volcanic, ferralitic, alluvial, etc.) and lends itself to three-season farming (rainy season, off-season, inter-season). Dueto strongurbandemand, potato not only representan opportunity for food self-sufficiency, but also income-generation. Potato can be grown off-season from April to August in moderate and low-altitude areas and starts in July inthe highplateaus, thereby increasingcropping intensity of irrigationschemes and offering potential for fertilization carry-over effectsto the subsequentrice crop as well as facilitating soil preparation. Improvedtechnologiesdeveloped byFIFAMANOR are enabling potato production areasto expand on the high plateausand even into coastalareas. Processing/Marketing Previously considereda vegetable Potatoesare usedto produce chips, crop and a substitute crop for rice but the scale ofthis activity is still during the pre-harvestgap, potato is limited. increasingly becoming a cash crop. Sweet potatoesare not processed While just 34% of output was at all, except for a very minimal marketedinthe mid-l980s, the amount of drying inthe south. percentage now exceeds 45%. For cassava, riskof significant The potato supply-chain is post-harvestlosses after drying beginning to integratethe various due to due to weevil damage. stages from seedproduction and Very few tubers are exported due distribution by seedproducer groups and input suppliers to the to the drop inworld prices for intervention of collectors in order to cassava starch. facilitate input supply with payment Low storage capacity: producers facilities as a means to buildtrust use traditional storage systems regarding collection. which have low capacity and a short keeping span (under two Producersare increasingly organizing themselves to masterthe months). marketing channels. Lack of financing for improved Dry cassava keeps well (about eight storage methods months), which facilitates long- distancetransport and trade. Maize. Yields of maize are especially low, at an average of 1.5 tonska, with even lower yields in the dry west. Maize, originally a cash crop for export, is no longer competitive on world markets and is increasingly becoming a food staple and cattle fodder. Studies have shown that maize growing is the biggest single contributor to biodiversity loss inthe western dry forests, and that 21 forestry officials are accomplicesto the processthrough the issue of cutting permitsto migrant farmers (Reau 2001; Casse et a1undated). Existing policies for the promotion of maize do not take any account of these severe environmental impactsand there is a risk that initiatives to promote yet greater production, such as producer-processorpartnerships, will leadto yet further deforestation. The solution appears to lie incontinuing to promotebetter organization of the sector while introducing controls to disqualify maize that is producedthrough deforestation. Such a radical measure would need to be accompanied by support to farmers to enable them to produce maize on used soils and also to introduce new mechanismsto enable prospective maize farmers to obtain access to land. Further support could be provided through a systemto help farmers find land for maize production (such as through a system of land use contracts) and by revision of the laws on land tenureto abolish the principle that `mise en valeur' of land provides preferential title. This anachronistic principle underliesmuch deforestationand environmental destruction in Madagascar (see Section4.3 for details). Table 2.9: Maize -Streneths andWeaknesses Streneths Weaknesses Maize has substantialpotentialfor Low longterm productivityin developmentdue to the existenceof some areas due to unsustainable a variety of soils favorable to this (tavy) productionmethods. crop, including, mostnotably: the Productionis associatedwith very alluvialsoils ofthe west, southwest `highenvironmentalcosts andnorthwest,the volcanic soils of the Vakinankaratraregionandthe The isolationofthe producing ferralitic soils ofthe midwest. areas is a majorimpediment to the developmentofmaizeproduction. Madagascarhas begunto export The bestsoils are the alluvial maize seeds to the countries ofthe deposits, but these alluvial areas IndianOcean. are hardto reach. This poses a handicapfor a productthat does notkeepwell, particularly inthe case o f a deficientstorage system (see below). ProcessingiMarketing Consumptionofprocessedmaize Lack of storage facilities: Maize is (flour, crushed) is increasingfor highlyvulnerable to damage by both humanand animal insects, especially weevils, and consumption(especially for rodents. Damagemay exceed ~ fodder). Many rice hullingfacilities 50% inthe case of inappropriate are increasinglybeingequipped storage and improperlydried with maizegrinders. products. Strengths and Weaknesses of the Export Crops Sub-sector Madagascar's traditional export crops are vanilla, cloves, and coffee. Cocoa, pepper, and more recently litchis are also exported. Export crops account for about 310,000 hectares or 17% of total cultivated area in Madagascar. Most ofthe area under export crops are locatedon the east coast of the country (in the provinces of Fianarantsoa, Toamasina, and Antsiranana) which accounts for just under one-half of the rural population. The previous findingthat the incidence of income poverty is highest in the provincesof Fianrantsoaand Toamasina raises the questionto what extent small farmers are actually able to capture and managetheir fair share o fthe export crop revenues. 22 All export crops are subjectto global demandand price fluctuations as well as increasing competition from other countries. Price fluctuations can be very significant to the extent that crop production almost becomes a speculativeproposition. For instance, vanilla prices exploded by over 600% between 1997 and 2001, thereby boostingvanilla export earnings from 6.9 million SDR to 128.8 million SDR over the same period. Similarly, prices of cloves increasedby over 500% and pushing up export revenuesto 78.5 million SDR between 1997 and 2001. Coffee prices faced an opposite fate by plummeting over 50% causingcoffee export earningsto implode from 23.9 million SDR in 1997 down to just 2.5 million SDR in 2001. Export crop revenueshave been falling steadily over time, but still representa significant share of Madagascarexports, with vanilla and cloves together accounting for 27% of export revenues in 2001. Howlever,just as in food crops, Madagascar's long-term performancein agricultural exports has also been unsatisfactory. Such exports were lower in 1999 (about US$235 million) than they were in 1980 (US$293 million) with a substantial decline in coffee exports dominating this overall trend. Few other African countries have experiencedan actual decline in their agricultural exports over this period and most ofthose that did have experiencedlong periods of civil strife. While several of Madagascar's traditional agricultural exportshave been reduced in recent years by the sharp reduction intheir international prices, a sharp reduction in the volumes of these exports both precededand continuedthrough this period ofprice declines. Madagascar's export crops suffer from generalizedageing. Plantationsare poorly maintained and are not regularly replenished, particularly in the case of coffee. Productionhas become a matter of picking what grows (except in the case of vanilla). Dueto weak output, Madagascar has been gradually losing its position in the world market for all these products exceptvanilla (which has enjoyed a combination o f favorable environmental conditions, high world price and donor support). Quality activities (Rahavohitravo 2002). Worryingly, vanilla quality in2002 was disappointingly low - is also declining for most crops due to poor storageand processingand a lack of extension farmers were forced to harvest early because oftheft problems which represents a new potential threat to Madagascar's primacy for this crop (Randriamanampisoaand van Nieuwkoop, 2002). Not surprisingly, throughout the 1990s,yields have been stagnant for most crops. Table 2.10 Maior Export Crops inMadagascar Source : Ministry of Agriculture Madagascar's exports are destined for a small number of markets. Industrialized countries (the EuropeanUnion, and in particular, France, followed bythe UnitedStates and Japan) are the country's principal trading partners, absorbingtogether more than 80 percent of Madagascar's main agricultural exports. Although the destinationsof Madagascar's principal exports did not 23 change radically during the 1990%there has nonethelessbeena shift away from the United States and Japan to France, the Middle East, NorthAfrica and other Europeancountries. For the most part, Madagascar has failed to take advantage of a large food and agricultural import market locatedvirtually at its doorstep-that of South Africa. Vanilla. Vanilla is an exceptionto the general downward trend for export crops. Madagascar's vanilla production is expected to reach 1,200 - 1,400 tons for the year 2002 comparedwith 800 tons in 2000. There are about 60,000 small producersactive in the sector. Madagascar's vanilla sector has a competitive advantagedue to low production costs and the very high vanillin content of the product. Growing world demand and the economic problems ofMadagascar`s Asian competitors have also helped. Measures taken by the governmenthave further helped the sector including liberalization (through lowering and subsequently suppression of export tax in 1998), improved regulation and quality control, the establishmentof associations (growers, traders and exporters) and involvingthe growers in industry development and international product promotion. The increasedratio of producer prices to borderprices suggest that growers have benefited from the reforms. Additional measures such as certification, quality control and laboratory rehabilitation projects are underway. Ithas beenrealized that Madagascar needs to offer several grades ofvanilla (through the introduction of organic and/or fair trade labeling for instance), notjust focus on high quality. Nevertheless, vanilla production in Madagascaris currently threatenedby a recentfall in international demand(from 2,200 tons in2001 to 1,300- 1,500 tons in 2002), which, combinedwith the increasedsupply and lower quality will translate into lower FOB prices (US$100-120/kg in 2002, comparedto US$2OO/kg in 2001, Randriamanampisoaand van Nieuwkoop, 2002). Substitution of natural vanilla by synthetic vanilla overseas also accounts for the reduction in international demand. Table 2.11: Vanilla Strengths an Weaknesses - Strengths Weaknesses Production 0 Malagasyfarmershavea long Increasingsecurity problems with traditionof growingvanilla. negative impact on quality. 0 Favorableclimatic conditions for Dueto liquidity problems, farmers superior qualityproduct often start harvestingtoo early 0 Availability ofnewlydeveloped thereby compromising quality. plantingmaterial for improved 0 Frequentcyclonespose major yields productionrisk. Low labor costs. Little price-qualitydifferentiation at producer leveldoes not offer sufficientincentivesto improve quality. ProcessingMarketing 0 Marketliberalizationshasenabled 0 Increasedcompetitionfrom Madagascarto regainits dominant synthetic vanillin as new positiononthe internationalmarket, regulations inEuropeallow satisfyingmorethan halfoftotal uniformlabelingof productsthat world demand. are 100%naturalandthat contain 0 Existenceofprofessional vanillin flavors processingfacilities with substantial 0 Increasedcompetitionfrom other capacity (1,200 tons). countries (Mexico, Indonesia, 0 On-goingpromotionefforts to PacificIslands), especially inthe achievegreater visibility both lower quality beanprocessing domestically as well internationally. segmentfor the US market. On-goingeffortsto strengthen 0 Productdifferentiationthrough qualitycontrolthroughbetter labelinghas not (yet) been equippedqualitycontrol successfulto positionMalagasy 24 laboratories andthe issuance of vanilla as a recognizedbrand, plant health certificates which is an importantcondition to On-goingeffortsto improve penetratethe high quality oriented professionalismof growers through Europeanmarkets. certification system. Relatively long domestic marketingchain with multiple intermediarieswhich complicates transfer of price and quality signals from exporters to growers. Lack of vanilla curing capacity at the producer level complicates benefit capture by growers as harvestvanilla has to be processed in48 hours. Coffee. In terms o f cultivated area, coffee is the most widely grown and traditionally was the country's leading export crop. Declining profitabilitydue to low world market prices has led to some exit from production, but it i s believed that nearly 400,000 households continue to cultivate coffee. Coffee exports have virtually disappeared, amounting to less than US$4 million in 2000. Annual production peaked in 1992 at 88,000 tons, falling to 58,080 tons in2000 and to average o f 70,830 tons over the 1990to 2000 period. Poor quality played an important role as well. The share o f national production which was rated as 1St Grade declined from an average o f 57% over the 1992-94 period to only 41% over the 1998-2000 period. Due to quality factors, Madagascar's Robusta coffee i s discounted by some 1520% on the international market, compared with some o f its competitors. According to Jaffee (2002), coffee trees are generally old (between 25 and 40 years) and poorly maintained. Many farmers harvest immaturely which results inan output with bad taste and aroma. The harvest tends to contain large quantities o f broken and flattened beans due to lack o f technical supervision and quality control. In addition, fertilizers are no longer beingused because o f declining coffee pricesand risingprices o fthe inputs. Lack of infrastructure for storing berries duringthe drying process, also causesthe development o f mold and fermentation in products stored in piles. Table 2.12: Coffee Strensrths andWeaknesses - Strengths Weaknesses Production Madagascarhas anaturalpotential Low productivity due to old trees: for producinghigh-quality coffee yields are 200-300 gramsistalk due to the existence of Robusta versus 500-1,000 grams Kouillouand Robustavarietiesthat worldwide. are know for their granulometric Very low use of productivity properties and organolepticappeal. enhancing inputs. Altitudes above 1000meters Immatureharvestingwhich present favorable conditions for compromises quality. Arabica coffee production Productionsubject to frequent Madagascarhas abroadrange of cyclone damage. plantingmaterialwith unique Lack of appropriatequality-price characteristics (e.g. natural differentiation. decaffeinatedcoffee strains) that could be usedto helpboostcoffee productioninterms ofboth quantity and quality. Processing/Marketing Existence of professional Absence or inadequacy of processing facilities with substantial appropriatestorage infrastructure capacity(1,200 tons). and drying equipmentat the 25 Installedprocessingcapacity to grower's and village-based produce coffee at highly collector's despitethe fact that the homogenousquality. harvest takes placeduring the On-goingefforts to revitalizethe rainy season. National and Local Coffee 0 Manual hulling of coffeeresults in Marketing Boardfor quality many brokenand flattenedbeans. improvementacross the value-chain and promotion of Malagasy product. Other Export Crops. Other export crops include cloves, cacao, pepper, lima beans, litchi, palm oil, and specialty packaged spices. Madagascar and Zanzibar are the world's leading producers o f cloves. The Malagasy clove is highlyvalued by Indonesiafor making"kretek" cigarettes. Two o f the plant`s products are marketable: the clove and the oil, which is extracted from the leaves and crowns. Clove oil is very rich in eugenol, which is usedto produce artificial vanillin. Attractive world market prices have sky-rocketed clove export revenues inrecent years, but prices have fallen sharply recently. Pepper production i s stagnant at roughly 2,000 tons per year. Farmers generally grow this crop in association with coffee trees. The pepper plant provides three marketable products: black pepper, white pepper and green pepper. Cocoa is primarily grown in the Sambirano region o f northwesternMadagascar, although it can be produced in other areas as well. Outputhas remainedvirtually unchanged at approximately 4,000 tons per year. Malagasy cocoa is among the world's best. Lima beans can be grown on the wide expanses o f alluvial soils in the southwest and west. Output has increased slightly recently due to an increase in cultivated area. Lima beansare typically characterized by the absence o fMenu Muso (the pigment anthocyanin), which results in a purplishred coloring o f the seeds upon cooking. This coloration depreciates the quality for export. The strengths and weaknesses o f these other export crops include Litchi, palm oil, and the sub- category of fine spices are outlined inthe table below. Tab11 2.13: Other Export Crops Streng is andWeaknesses - Weaknesses Production -Cloves: Strengths 0 largeproductionbase; 0 Clove, Pemer, Cocoa: production trees also provide erosion control subject to cyclone damage 0 Cloves, Pemer: relative low 0 Clove: trees are very sensitive to transportationcosts due to high unit excessive rain, causinglarge values. annual variation inproduct output. 0 Cocoa: high intrinsic quality Cocoa: old trees that have not 0 Limabeans: providesincome beenrenewedinaccordance with opportunities outside traditional the normal regenerativecycle. export regionsof eastern 0 Cocoa: Malagasyproduction is Madagascar; nitrogen fixation not competitive comparedto major qualities reduceneedfor fertilizers. West-African producers. 0 Litchi: increasingproduction 0 Lima beans: low quality for 0 Palm Oil: existing, underutilized export. organic plantations,excellent 0 Litchi:extremelyseasonal climate production, quick sale mentality 0 PackagedOualitv Spices: PalmOil: inadequate scale of production of high quality organic productiontojustify quality spices in various areas processinginvestments 0 PackagedOualitv Spices: small 26 scale bv nature Processing/Marketing Cloves: quality of Malagasy cloves Cloves: large variation in annual has retainedits reputationon the output hampersMadagascarto world market and has enabledthe supplythe market on a regular countryto build customer loyalty. basis, thereby exposing its market Cocoa: Installedprocessing positionto other competitors. capacity to produce marketable Cocoa: competitionfrom productat highly homogenous substitute fats for chocolate quality. production. Clove, Peuuer: Significant Litchi: South Africa holds quality availableprocessingcapacity and after season market, (12,000 tons for cloves and 1,700 govemmentpromotes grouping tons for pepper). operatorsresulting in low quality Litchi: 20% of operators seeking offering, difficult access to higher quality longer seasonal transport. market and differentiatedproduct, PalmOil: Needfor higher quality organic labelfeasible processing equipmentand better Palm Oil: existingknowledgeand collection methodsto realize low bio-label, existingmarket access acidity neededfor organic and connections certification. PackagedOualitv Spices: Vert PackagedOualitv Spices: young Safran started in 1997, upscale high - activity value-addedproduct Strengths and Weaknesseso f the Industrial Crops Sub-sector - Madagascar's principal industrial crops are cotton, sugarcane, sisal and groundnuts. They cover about 160,000 ha or 8% o f Madagascar's total cultivated area (Table 2.14). The industrial crops sector is little diversified, state controlled, and especially for sugarcane, highly inefficient (inefficiencies are perpetuated by current subsidies from the US and the EU). Cotton production and productivity are also declining, mainly due to declining irrigation infrastructure, falling prices and the inability o f the state-owned company HASYMAto pay cotton producers due to liquidity problems. Cotton production is based largely on outgrowing by smallholders, thus offering a considerable source o f rural employment. Most sugar is grown on large plantations -outgrowing schemes have suffered due to'failure o f the state-owned company SIRAMA to pay for cane. Sisal production, in private hands, is efficient but has suffered weak demand. Source :MinistryofAgriculture Cotton. Seed cotton production is based on smallholdings (80%) and commercial farms (20%) mainly in the north west (Ambato-Boeni, Port Berg&)and south west (Toliara). Commercial farms once had a much larger share o f production (75% in 1988) but after cuts in government 27 subsidies many farmers converted cotton fields to more profitable crops. Since 1990, seed cotton production has varied from year-to-year within a rangeof 25,000 to 35,000 tons (Jaffee, 2002). Average seed cotton yields have fallen over the same period from 1.2 tons to 1.O ton/ha but with a sharp increase in the late 1990sas inefficient farms inthe south west fell out of production (due to degraded irrigation infrastructure). Yields vary accordingto soil fertility and irrigation. Best yields (2.5 3.0 t/ha) and best fiber are produced in the north west (Ambato-Boeni, Port Berge). - Yields of commercial farms are only 5- 10% higher than for smallholdings, indicating that small- scale farmers can participate effectively inthe cotton industry. Importantly, the cotton industry cannot produce enoughto supply the tax-free textile industry (Ravohitrarivo 2002), which reflects a major lost opportunity for deriving added value from this sector. Ginning is carried out by three companies, includingthe parastatalHASYMA which produced 74% of ginned fiber in 2000. In 2000, total seed cotton production of around27,400 t was transformed into 11,100 t of fiber (Ministerede 1'Agriculture, Unit6de Politique de Developpement Rural, undated). HASYMA has been losing market share to private sector companies (CFDT, COTONA, DRAMCO) but its privatization has been delayed. HASYMA's closure during the recent political crisis highlights its vulnerability (IMF 2003). Vertical integration of the cotton sector is of crucial importanceto consolidate Madagascar's rapidly expanding textile exports to the USA under AGOA. These exportshavebeen a major driver for Madagascar's dynamic EPZ. However, with domestic content requirementsbecoming effective in sometime 2004, EPZtextile firms canno longer use cheap Asian fabrics, butwill have .to rely on AGOA eligible production, which puts apremium on developing sufficient domestic cotton production capacity. Rapid privatization of HASYMA, also inview ofthe company's dire financial situation, is an important pre-condition to better position the cotton sector to continue to capture benefits provided underAGOA. Given the importance of textile for the success ofEPZs, doing so would helpMadagascarto put its strategy o f rapid and sustainabledevelopment into practice. HASYMA's financial problemsalso have an environmental cost as cotton farmers who have not been paid by HASYMA increasingly turnto tavy-based production of maize as an alternative (George, 2002). Table 2.15: Cotton Strendhs andWeaknesses - Strengths Weaknesses Production Large numberof small growersthat 0 Localpricefixing systemhave are organized in associations. caused producerprices to fall more 0 Existenceof out-grower schemes, than fiber prices, suggestingthat with inputandtransport services effect of low world market prices provided by cottonmills. havebeenparticularlypassed on Growingdemand fromEPZ-based the producers. textile industry. 0 Fallingoutput due to exit caused 0 Roadimprovementand by falling producer prices. maintenanceprovidedby 0 Liquidity problems of Hasymahas HASYMA in cottonproduction affected availability of fertilizers zones hasremovedimportant for smallgrowers with negative obstaclefor on-farmproduction. impact on yields 0 Break-downof irrigationfacilities inthe Southdueto cyclone damage. Lack of storagecapacity at the farm levelwhich lowerslint quality due to sticking 28 ProcessingiMarketing In responseto the dynamic EPZ- Processingfacilities are very old basedtextile industry,private and inefficient. cotton-ginningoperationsare Under-utilization of processing expanding. capacity (30,000 tons for cotton HASYMA has achievedan overall lint). improvementinthe quality of Administeredpricingpolicies and cotton lint inthe recent past. linkage of expert volumesto Strongdemand for finished and domestic sales hamper sector semi-finishedcotton products. expansion. Embroideryis a strong growth market offering downstreamvalue- added opportunities Sugar. Sugar constitutes Madagascar's third largest agricultural product, accounting for around 6% o f the country's agricultural exports in 1999. The state-owned company SIRAMA dominates the sugar sector. As in the case of HASYMA for cotton, SIRAMA i s yet to be privatized. Inthe meantime, output and productivity inthe sector have steadily declined, the country has become a significant importer o f sugar (despite hightariff protection) and SIRAMA's 2,500 out-grower farmers have experienced long-standingdelays in payments for their crop, undermining incentives to produce and maintain production infrastructure. Currently, local production is some 70,000 tons of refined sugar, with SIRAMA operating at about 50% o f its processingcapacity. .Low capacity utilization combined with low sugar cane yields and high inputCosts, render Madagascar's sugar much less competitive than that o f neighboringcountries. Imported sugar (30,000 tons in recent years) is 50 to 60% less expensive than SIRAMA's sugar, hightariffs and other taxes considered (Jaffee, 2002). Despite all the above constraints, Madagascar has been able to export sugar under several preferential access agreements, especially with the United States, which allocates 7,258 t a year to the country, and the European Union (12,950 tons) under two preferential sugar agreements (World Trade Organization, 2001.) The exporting country receives guaranteed minimumprices from the EU(646 Euro/t for white sugar and 524 Euro/t for raw sugar, about two anda halftimes the world price). With the exception o f poor harvest years, Madagascar has generally fulfilled its quotas o f sugar in foreign markets. This means that until now, SIRAMA has survived only because o f the US and EUexport concessions and remains fundamentally uncompetitive. Like HASYMA, SIRAMA should have been privatized a longtime ago. Besides not being competitive, the company's potentially significant environmental liability associated with the untreated discharge o f stillage from its alcohol distilleries might however be an important impediment to come to rapid closure o fthe SIRAMA item on Madagascar's privatization agenda. Strengths Weaknesses Production Existing2,500 producers are Low yields due to low use of inputs relatively well organized. Uncompetitive production; despite 35% import tariff, imported sugar I i s still 50-60% less expensive Strongly distortedworld sugar market due to subsidies in I developedcountries. ProcessingIMarketing Preferential access to the EU with a Processingfacilities are very old 29 basic quota of 10,760 tons and a guaranteedpriceregardlessof Under-utilization of processing fluctuations in world prices, capacity(140,000 tons for sugar). amountingto a CIF price of Euro 64.65kg for white sugar in2000. Govenance issuesassociatedwith sugar imports due to under- Preferentialaccessto the USA with invoicing andnon-transparent a basic quota of 7,280 tons and a provision oftax exemptions. guaranteed FOBprice of `JS$415.90/ton. Market access at the wholesale level restrictedthrough licensing Domesticindustryis protected system which induces rent seeking against imports througha 35% behaviorby existing five tariff. wholesalers with licenses. Potentially significant environmentalliabilities of SIRAMA due to untreatedstillage discharge from its alcohol distilleries. Sisal. Sisal production and marketing are concentrated insouthern Madagascar and the sub sector is controlled by four large companies: Etablissement Galois, SocietC de Heaulme, SociCtC des Plantations du Sisal de Mandrare and SocietC Sisal du Sud. Exports o f sisal have increased recently. The quantity exported duringthe period 1999-2000 increased from 4,600 tons in 1999 to 5,270 in 2000. The sisal industryuses 6,000 employees to produce an average of 18,000 tons per year. The marketed products are fibers, strings, ropes, sacks, packaging fabric, carpets and espadrilles. Sisal is also the basis for different sorts o fhandicraft products that seem to gain popularity due to increasingly innovative designs. Hence, the product is o f interest because o f its capacity to generate value-added and employment indownstream processing activities. Peanuts. Interms o f cultivated area, peanuts constitute the second-largest industrial crop in Madagascar. Peanut production appears to have substantial potential in Madagascar as it can be grown on alluvial soils under rainfed conditions. A strong point is that peanuts are an easy product to store (in the shell) and the risk of losses i s minimal. Peanuts have been the main raw material for producing edible oil in Madagascar. They are also processed by traditional methods to produce local oil. At the same time, the crop, however, faces several weaknesses: (i) there are no producer organizations for peanut growers; (ii)the bulkiness o f the product creates problems in terms of transport and handling o f the product inthe shell; (iii) transport and storage o f the shelled product can result in significant losses due to the fragility o f the nut; and (iv) the development of peanut production is limited by a lack o f high-performance varieties, particularly in terms of yield and resistance to disease. Strengths and Weaknesses o f the Livestock Sub-sector The livestock sector contributes about 15 YOo f agricultural GDP and contributes to the income o f over 60 YOo f rural households. Beef Production. The extensive production o f zebu (humped cattle) dominates the sector. Madagascar's zebu population was estimated at 7.4 million head in 2000, representing a capital asset of about US$200 million. Grazing lands cover 32 millionha or about 55% o f Madagascar's territory. Consequently, the stocking density is very low at 4.3 hahead. Zebu production i s concentrated in the Madagascar's western regions which are home to about three-quarters o f the national population and held in relatively large herds of 30-80 heads. Zebu plays an important 30 role as draught power for land preparation and transport purposesinthe central plateauand middle-west regions. Small herds tend to predominate in these regions. Meat production is low with off-take rates at less than 8% annually. Mortality rates are high due to poor coverage by veterinary services. The vaccination rate was estimatedat 57% in 2000. Diseases are thought to reduceanimal weight by an average 15%. Low productivity is partly due to the fact that for a significant proportion of livestock farmers, cattle farming is traditionally not a commercial undertaking. Indeed,their objective is insteadto accumulate capital or negotiablereserves in order to fulfill social obligations (e.g., for family events) or to prepare for the impact of unfavorableclimatic or biological events (e.g., droughts, locusts, etc.) that regularly cause food shortages. Zebu cattle are used in rituals accompanying various ceremonies, such as births, circumcisions, marriageproposals, funerals, etc. They are also the currency used for resolving conflicts and reachingagreements. Ownership of zebu cattle also determines one's social standing. Incase of adverse weather conditions for crop production, steers are sold in order to purchasefoodstuffs to make up for poor harvests. As such zebu herds function as a personal savings account. Constraints of geographic distance also play a role in explaining low productivity. Livestock farmers have difficulties accessingthe commercial circuit due to the fact that over two-thirds of slaughter facilities are located in urbancenters indeficit regionswhich are far away from the primary zebu raising areas in the West. Consequently, it takes 10to 45 days to arrive at the destination market, resulting in losses, thefts, deaths, increasedexpenses and qualitative deterioration. Longtransportationdistanceshave also resulted inthe proliferation of a wide range of middlemen, who generate costs that reduce livestock farmers' margins. One major problem affecting livestock productivity is widespread cattle theft. The scale ofthis problem is estimated at 8% of the national herd, which would represent about 600,000 heads of zebu per year (Andriantsifereana, 2002). Its effects are numerous and go beyondthe direct effect of cattle theft, including: (i)economic loss of a significant part of livestock production; (ii) reduction of the breedingpopulation, i.e., the productive capital; (iii)hampering of feeding through a reduction in grazing space and the elimination of nighttime grazing; (iv) spreadof tuberculosis due to crowding in nighttime enclosures; (v) slower weight gain, increasedtime spent in the grazing area, delayed sale; (vi) disturbance ofthe animals' physiology in general and reducedfertility rates; and (vii) perturbation of marketing circuits: the greater danger for cattle drivers on foot leads to a great increasein transfer costs due to the greater number of cattle drivers required, an increase in their salaries, a costly increase inthe number ofjourney segments in order to reduce risks during the treks on foot. In 1997, the EU imposeda ban on the importation ofMalagasy meat (Decision No.97-5 17), due to non-compliance with hygiene standards (although market saturation probably played a part in the decision). As a result, existing small beefexports have disappearedaltogether. A costly new abattoir subsequentlyconstructedin Antananarivo with EUfunding to overcome this obstacle has never operated since it is unsuitedto the small scale system of cattle trading (Andriantsiferana, 2002). However, even ifthe ban would be lifted, it is doubtful that Malagasy beef is competitive on the international markets. According to Andriantsiferena (2002), the FOB unit price of Malagasy beef in 1997was FF 12.5/kg as comparedto FF 8.73 inAustralia; FF 10.92/kg in Argentina, and FF 5.31/kg in India. As a result of the above mentionedproblems, supply has decreasedconsiderably, resulting in significant price increases. Andriantsiferana (2002) reports that the'supplyof zebu on the important weekly market at Tsiroanomandidy has dropped from 1,000 head in 1997to only 200 31 head in 2000. Prices have increasedby over 150% over the same period, thereby rendering beef unaffordable to the poor and uncompetitive on international markets. Beefconsumption in Madagascarhas been declining in the 1990s and is relatively low compared to other countries. Table 2.17 Beef Production Strengtl andWeaknesses - I Strengths Weaknesses Production Vast area of pasturesand grazing Lackof commercial orientationas lands. zebus are heldfor social purpose Availability of largecapital asset, Low productivity due to poor i.e. the cattle herd. animalhealthcoverage Zebu beefhas intrinsicquality Low productivitydue to characteristics (e.g. deep redcolor, inappropriateherdmanagement low fat content, low moisture (culling), amongother causedby content). legal restrictionson the slaughter Freeranchingproductionmethods of female domesticated zebucattle is appreciatedincertainniche andyoung animals of localcattle markets. breeds(Decree No. 98-1030). Large scale theft inducesrisk mitigationmeasurethat are counterproductive. . Lack of competitivenessof Malagasybeefonthe international market. Growingpopulationandrising Large scaletheft increase incomesis expectedto increase marketingtransaction costs. demand 0 Largephysicaldistance from consumers increase transportation costs. 0 Large number ofdifferent intermediariesincreaseswedge betweenproducer andconsumer prices (50%). 0 Exportbanto EU. 0 Unsanitary slaughterhousesand disorganizednatureof butchery professioncompromiseproduct quality. Poultrv. Chicken production is currently the most dynamic livestock sector inMadagascar, with 60,000 chickens producedevery week and supplied to small scale out-ranchers, mainly around Antananarivo. About 1.5 million market standard chickens are producedannually, as compared with around 20 million home raised chickens. This may be comparedwith 30 million market standard chickens in Mauritius alone. This type of production is increasing inurban peripheries and suitable for out-ranching to small-scale operators. The production o f fattened ducks, even though modest interms o fvolumes produced, is also important for the rural economy. In2001, duck products included 40 tons offoie gas and 250 tons of duck meat from 4,000 duck farmers in Fianarantsoaand Antananarivo (Andriantsiferana, 2002). As Madagascaris well-known for its fattened duck (foie gras) production, an assessment o f its strengths and weaknesses is presented in Table 2.18. 32 Table 2.18: FattenedDuck Production Strengths andWeaknesses - Strengths Weaknesses Production Low labor costs provide 0 Irregularityof feed supply, competitive advantagein labor- particularlycorn, poses production intensiveproductionprocess. risks. 0 Relativelylow start-up investment Difficulties with supply of costs. 0 vaccines andde-worming Little economies of scale in medicationscan cause high productionprocessleadingto the mortalityrates. emergence ofcottage-type operations. 0 Newcastle disease is endemic due to lack of animal healthprotection. 0 Relativelyshort productioncycle (3 months), especiallyifimproved Genetic problems: low breedsare used, which facilitates productivityof localbreeds. access to credit. ProcessingiMarketing ~ ~ ~~~~~~ 0 Integrationbfvalue-chain: 0 Poor hygienicconditionsinsmall processorsprovidetrainingand duck-producingoperations affect technical assistanceto farmers, so product quality. as to ensure adequate quality. 0 EUexportbanhas closed previouslyattractivemarkets. Animal welfareconcerns might ' reduce internationaldemandfor foie gas. StrenPthsand Weaknesses of the Fisheries & Aquaculture Sub-sector Three key points need to be made before a consideration ofthe role of marine fisheries in enhancingrural productivity, reducing poverty and improving food security: 1 . Fisheriesresourcesare finite. As wild resources,productivity cannot be enhanced by human intervention but depends on the natural limits ofthe surrounding medium. Most fisheries of the world are exploited at or beyond their capacity to regenerateandthe fisheries o fMadagascarare no exception. For most stocks, increasingproductivity will be a matter o freducing and redistributing current levels o f fishing effort, usingmanagementmeasuressuch as closed seasons and no-take zones to facilitate replenishment o f stocks. Madagascar's capture fisheries do have an important role to play in rural productivity, poverty alleviation and food security (in particular as regardshigh quality nutrition) but such contribution cannot exceedthe natural capacity o fthe resource and will certainly not be sustainablein acontinued regime of open-accessand `laissez- faire'. Hence, the greatestopportunities for securing a sustainablecontribution from fisheries lie in improved sector governance, management, developing new forms o f aquaculture and supportive research. 2. Madagascar is surroundedby tropical seas of low productivitv. Exceptions occur ina limited number of small upwellingareas identifiedoff Cap Masoala (NE), Ft Dauphin(SE), Toliara (SW) and Ile Juan da Nova (possession ofFrance). Otherwise, marine productivity is low, especially along the east coast. As a useful comparison, the artisanalfishery o f Senegal, West Africa, with an annual catch of400,000 tons along 750 kmo f coast, is roughly five times the combined traditional and artisanal fisheries catch of Madagascar(estimatedat 75,000 tons along 5000 km of coast). While this large difference is inpart due to the stronger fishingtradition o f Senegal and the smaller range of small scale fishing, the principal reasonis the vastly higher productivity of the waters ofthe West African coast than the WesternIndianOcean. Madagascar's capture 33 fisheries can only ever offer a modestcontribution to improved rural productivity, poverty reduction and food security. 3. The marine resourcesof MadaPascar are poorlv known. Much of fisheries expansion during the 1990shas been based on the supposition that sustainablemarineharvests are around 320,000 tons per year (Andrianaivojaona et a1 1992)while actual harvestshave been only 100-125,000 tons. However, this suppositiondoes not account for the fact that most of the estimatedpotential 320,000 tons relates to stocksthat are little known or unexploited and that most readily accessible resourcesare already exploited at or beyond sustainable limits. Untilbetter information is available on the state and potential of Madagascar's marine resources, a precautionary approach should be applied to any plans for development.At the same time fisheries researchneedsto be greatly scaled up. Madagascarhas approximately 5,000 km of coastlinewith 177,000 km2of continental shelf and an EEZor exclusive economic zone of 1,000,000 km2.This makes it diverse in its rangeof fishery resources. Fishing provides about 62,000 directjobs and 218,000 indirectjobs for the country. Fishing is practicedall alongthe coastline, but is particularly intense along the lengthof the western and south-westerncoasts, as well as in the bays and aroundthe river inlets on the eastern coast. Total production from fishing is officially estimatedto be 125,000 metric tons per year, of which 95,000 metric tons come from marine fishing and 30,000 metric tons from inland fishing (official estimates).The figures for marine fishing include reportednearshorecatchesof shrimp of about 8,000 t and estimated offshore catches oftuna and billfish of 12,500 t. The estimatedtraditional catch of 50,000 t is extrapolatedfrom 1990estimates of daily catch per canoe and is therefore liable to a large margin of error. Addedto that are some 4,000 metric tons of farmed shrimp and unknown quantities of farmed seaweed, carp and tilapia. With receipts of more than MGF 1.06 trillion in 2001, or about USD 160million (official estimates), fish productscontribute an important portion of the country's foreign exchange. Shrimp alone accounts for 73% of receipts at MGF 777 billion. Besides shrimp, the eight leading fisheries export productsthat produce sales figures in the billions are, in decreasingorder of export value: cannedtuna (15.70%) (muchofwhich is actually caught outside the Madagascar EEZ but is canned in Madagascar), fin fish (5.54%), lobster (1.85%), cephalopods(1.01'?40), trepang (0.93%), crab (0.66%), seaweed (0.43%) and shark fin (0.21%). Interms of foreign exchange,therefore, the shrimp sector is by far the most important. The means o f access to resourcesdepend on the type of unitto be operated. For traditional fishing access is open to all citizens at no cost, although statutory powers now exist to impose 1icensing requirements. Compliance with the measures regulating protection of the given resource, such as: closing dates, banon specified fishing gear, etc. is all that is required. In practice, non-compliance with regulations is widespread inthe traditional fishery, and regulations are often ill-adapted to the context. For industrial fishing and artisanal shrimp fishing, access to resources is regulated by DecreeNo. 2000-415 ofJune 16,2000, which defines a systemfor awarding licenses by boat. The following items regarding this system are of interest: (i) the fishingeffort is fixed at 36 artisanal boats and 75 industrial boats, 6 ofthem on the east coast and the remainder on the west. Since this level was reachedduring the year 2000, new entries are no longer possible; (ii)investments inthe segment are secure, since a license is valid for 20 years and is transferable; and (iii)access to resources requires payment. This cost is based on the horsepower of the operating craft andthe 34 c zone of operation. For 2002, the total amount of shrimp license fees has been estimated to be about MGF 30 billion. The Government collects this fee and about 30% ofthe collected funds are transferred to a special treasury account earmarked for promoting the sub-sector: the Fisheries and Aquaculture DevelopmentFund (Fonds de DdveloppementHalieutique et Aquacole, FDHA). With the share of industrial and artisanal shrimp fishing estimatedat about 45% of total shrimp production (see Table 2.19 below), and assumingan equal share in total estimated shrimp exports of FMG 777 billion, fee revenuesto be collected by the Government consequently representedabout 8% of shrimp export value in 2001. This figure would be somewhat lower ifrevenuesof domestically sold production were added to the export values. Fisheriesare regulated under the fisheries law of 93-022 and numerous decrees and ministerial orders, mostly dating back to before 1960.Under the law, the fisheries administration is responsiblefor fisheries policy, managementplans for the conservationof stocks, fisheries researchand fisheries control and surveillance.Apart from a master plan of 1995 which is exclusively production orientated, there are no polices or managementplans (withthe partial exception of the industrial shrimp fishery which is regulated) andthere has been no significant fisheries research. Surveillance is conductedby the National SurveillanceCenter (CSP) which has introduced satellite tracking on shrimpand tuna vessels and carries out a limited amount of air and sea patrolling. The CSP is largely funded by the EUunder the STABEX program, with minor amounts from EUfisheries agreements andthe FDHA.Inthe current situation, the fisheries sector can be consideredas vulnerable to insufficient governance or control which constitutes a significant threat to hardcurrency revenuesas well as to rural productivity, poverty alleviation and food security. An overhaul ofthe sector is required. The collection of fishery products is regulatedby DecreeNo. 10404197ofNovember 13, 1997, which stipulates that practice of this activity is subjectto prior receipt of an approval issued by the Minister of Fishing and Fishery Resourcesbased on an opinion by the head of the province affected by the collection. A collection approval: (i)is valid for one province; (ii)is subject to a fee, the cost dependingon the natureofthe product; (iii)s individual and non-assignable; and i (iv) requirespossessionof a collection permitbefore it may be exercised. A collection permit on its turn is a validation of a collection approval within a specified fivondronana, or district. A maximum of five collection permits may be issuedfor each collection approval. This measure was undertakenfor resourceconservationpurposesand to prevent monopsonysituations. In2000 there were 171 permit holderswith a total of 337 permits. The relevanceof collection permits is however open to question ifthe objective is to expandthe market for traditional fishery products. From that perspective,the actual number of 37 permit holders with a total of 106permits seems rather limited in view of the fact that the province is home to over 20,000 traditional fishermen. Shrimp. According to official data, total shrimp production has risen by more than 30% from 1997-2001 as demonstratedin Table 2.19. Different producer segments contributed to the overall production of 17,305 tons in 2001, including industrial fishing, deep water fishing, artisanal fishing, traditional fishing and aquaculture. The figure for aquaculture for 2001 (5,399 t) is an official estimate, the true figure consideredto be closer to 4,000 t for 2001 (GAPCM, pers. comm.). Industrial shrimp fishing has stabilized at about 8,000 metric tons per year, reflecting the natural limit to this resource. Fluctuations betweenyears appear to be due to natural factors such as rainfall. Traditional fishing has grown strongly by over 70% accompanied by the spread o f new and more efficient and often destructive fishing techniques, particularly the use of gill nets recommendedby the fisheries administration and NGOs (GTZ), as well as to the expanding onsite support provided to traditional fishermen bythe middlemenand exporters, both technically and financially. 35 Table 2.19: Changes inShrimp Production over the Past Five Years (official data) Unit: metric tons "This figure is consideredan over-estimate by industry representatives-4,000 t is more likely Aquaculture is the spearhead of the shrimp segment. It is still inthe beginning stages in Madagascarand cultivation potential is substantial. The 2001 production of about 4,000 metric tons from 2001 represents the production of four companies, tyo of which are still not infull production mode. Total surface area under cultivation is about 4,000 ha while the country may have a further 20,000 ha available for farming. At the current yield of 4.43 metric tons/ha/year, shrimp production from aquaculture could grow to be 80,000 metric tondyear. Madagascar's success on the international shrimp markets can be explained by various factors, including low production costs and high quality. Success has also beenassisted by awell- organizedand self-regulating production sector. Interms of sanitary quality, the shrimp industry overall meets the level ofthe Europeanstandard. The official structuresfor health and sanitation controls are in place and operational. Superior quality ofthe Malagasy product is reflected inthe increasing share of whole shrimp exports as well as the fact that the highly demanding markets of France, Japan and Spain are the major export destinations. Interms ofcosts, the lowest operating costs found in the African shrimp industry are those ofMadagascar. In addition to being extremely important for overall economic growth ofthe primary sector as well as associatedforeign exchangeearnings andtax revenues, the sub-sector generates significant employment estimatedat 10,000 directjobs and 40,000 inducedjobs (GAPCM pers. comm.) de 2.20: Shrimp sector -Strengths mdWeaknesses Strengths Weaknesses Production(capture Inherentlyresilientnatural 0 Fishingeffort at or beyondoptimal fisheries) productivityalongwest coast level (mangroves) 0 Traditional sector disorganized Industrialsector relativelywell andunmanaged structuredand managed 0 Competition/conflict between Substantialemployment creation industrialand traditionalfisheries (traditionaland industrial) Traditionalfisheryunableto capture full benefits (poor hygiene) Industrialfishery causes ecologicaldamage (by-catch& bottomtrawling) Production 0 Low productioncosts 0 Environmentalfactors (winds) (aquaculture) Substantialpotentialfor expansion Environmentalimpacts induced by (current 5000 t, estimatedpotential immigration 20,000 t) Well managedsector, free of 36 I ' disease ProcessingIMarketing Highlycompetitive product Decline in world shrimp prices (capturefishery and aquaculture) Integrationof industrialproduction Artisanal and traditionalfishing: (capture fishery and aquaculture) banon exports to EUreduce Potentialfor eco-certification ofthe market opportunities and depress shrimp fishery and aquaculture prices Poor air freight frequency and capacity hampersexports of live and fresh products Traditional Fishing. The number oftraditional coastal fishermen is estimatedat about 40,000 who exploit a total of about 20,000 dugout canoes. The highest concentrations are found in Toliara, where over 50% of all traditional fishermen operate. Mahajangaand Antsiranana account for respectively 7,000 and 6,000 traditional operators. In terms of production, traditional fishing is estimated to supply about 70,500 metric tons or 73.6% of total production from marine fisheries. However, it must be pointed out that the estimated annual finfish catch of 50,000 t is an extrapolation from an average estimatedyield per canoe measuredin 1990, allowing room for a very large margin of error. About 25% of artisanal production is sold directly to consumers in local markets, while 15% of production finds its way through professionalcollectors. About 60% oftraditional fishing production is channeled through so-called"mareyeurs", a fishtrader, generally a female who buysthe product at the landingand resells them at the village market or in the surrounding villages. An estimatedtotal of 3,000-4,000 "mareyeurs" are active inthe country. Artisanal processing(drying, salting, smoking) is necessary inremote areas, since it is the only preservation method that allows productsto be moved to large centers of consumption. Traditional fishing is also practiced aroundthe inland bodies of fresh water that extend over an area of 52,138 ha. Freshwaterfisheries in rivers and lakes, important as a source of local protein, are not subject to significant managementand have suffered overfishing, reducing yields (the Lac Aloatra fishery yield declined from 3000 tons in 1990to 700 tons in 2000). Yields have further declined as a result of sedimentation and species introductions. Sincethe 1960's the practice of family-based fish farming of tilapia (of which there were some 85,000 ponds in operation in 1963) has declined to almost zero due to a lack of know-how and supplieso falevins for pond stocking. Schemes during the 1980sto develop the production of alevins for re-establishing family fish farms had limited impact, and most ofthe state-ownedalevin ponds have been abandoned. The practice of fish farming in rice paddies, which has major potential, had scarcely developed (Adrianaivojaona et a1 1992) untilthe late 1990's. Recent expansionon parts of the high plateau is substantial. Table 2.21: Traditional Fishing Streneths andWeaknesses - Strengths Weaknesses Production Fishingcan generate higher than Intrinsicallylow productivity of average rural incomes western IndianOcean marine Highdistributivepotentialof ecosystems,especially coralreef benefits areas Vital source ofhighquality Highly inefficientcaptureof nutrition benefits (poor conservation Potentiallyless damagingto techniques, lack of access to ecosystemsprovidedthere is good markets) management Limitedrange andseaworthiness offishingcanoes I I 0 Widespreadabuse offisheries 37 regulations Nearshorefishingmay damage recruitmentofoffshore stocks Absence ofrights-basedfisheries management (open access regime) 0 ProcessingiMarketing Emergingsigns ofvertical Lack of awareness o f or access to integration underwhich preservationtechniquesthat adhere middlemenand exporters provide to sanitaryrules or add value to technical and financial support to product traditionalfishermen. Lack ofmeans of storage for live Signs ofimprovedbenefits capture and fresh produce through bettercollection systems 2.3 EnvironmentaVecologicalperspective Biodiversity Assets The island of Madagascar's long-lasting geographical isolation together with its highly varied geomorphology and micro-climates has resultedin the presenceof extremelyvaried and archaic life-forms that have evolved into some of the most unique biodiversity inthe world. Madagascar has been called the single highest major biodiversity conservationpriority inthe world, owing to its combination of high diversity, endemism, and degree ofthreat. Madagascar is one of the 17 recognizedmegadiversity countriesthat together represent 80% of the world`s biological diversity. Madagascar alone representsabout 2.6% ofthe world's non-fish vertebrate biological diversity on only 0.4% of the world's land area. About 10,000 plant species have been identified with 940 species oftrees and large shrubs alone of which 96 per cent are endemic. Although Madagascaroccupiesonly about 1.9% ofthe land area ofthe African region, it has more orchids than the entire African mainland, and is hometo about 25% of all African plants. In addition, Madagascar is home to a large diversity of medicinal plants which are of potentially critical importanceto the pharmaceuticalindustry. The complete knowledge about the biodiversity o f Madagascar is still to be explored and more facts on the exceptional biodiversity remain to be described. For example, over the last 15 years, there have been no fewer than 10 new species of primatesdiscovered and future studies may push this number even higher. Overall, about 80 per cent of Madagascar's plant species are endemic, and for most animals, the proportion is even higher, the best example being the lemurs, of which 100per cent of occur naturally only in Madagascar. O fthe 280 bird species recordedwith 204 species breeding in Madagascar, 110 species are endemic. O fthe 346 reptile species recorded, 314 are endemic. Only two other eco-regions in the world, Le., Caribbeanand Meso-America, can match Madagascar's diversity in reptiles. Raxworthy (in press) suggests a high risko f extinction for 18 reptile species. Most remarkable is that Madagascarharbors endemism at the higher taxonomic level (genus and family level). It contains 5 endemic botanic families and 5 endemic primate families. In comparison, Brazil, the largest tropical country of the world, does not have one single endemic family. Madagascarhas over 5000kilometers of coast, with the second largest barrier reef inthe world found in the southwest.These coral reefs have recently beenfound to have some of the highest 38 levels of diversity of fauna and flora of any of the coral reefs ofthe region. Additionally, Mangrove forests are common in the west. Recognition ofthe importanceofthe marine biodiversity associatedwith its 5000 km of coastline has only grown recently. A rapid assessment conductedby C1in 2002 year showedthat Madagascar's coral reefs have some ofthe highest levels of species diversity ofthe region combinedwith a significant level of endemism. Terrestrial ecosystems & biodiversity The land surface area of Madagascar is 590,000 km2of which an estimated 15% (88,800 km2)is covered by forest or spiny desert (as ofthe year 2000, Steininger et al, 2003.) There are various analyses of forest type and historical trends and a few recent analyses are presentedbelow. The main trends include the following: 0 Widespread deforestation that has occurredincreasingly over the last half century 0 The dry western forests are highly fragmented and underrepresentedin protectedareas 0 Only protectedarea status and enforcement appears to slow deforestation The main conclusion is that ifdeforestationcontinues at its current rate of approximately 1% a year, at some point within the next 100years the only remaining forests will be those in highly threatened, completely isolatedprotectedareas. The CI forest change study (Steininger et al. 2003) found that from a total of 105,000 square kilometers of forests and spiny desert that existed in 1990, only 88,800 km2remained in 2000. The analysis enabled an estimation of deforestationratesamongthe six different provinces (see Table 2.23.) Table 2.23 Ten Year DeforestationRate inall 6 Provinces Province Forestsand YOdeforestation spiny desert '00 1990-2000 (km2) Antananarivo 791 16 Antsiranana 13258 4.9 Fianarantsoa 1I7595 II11.5 Mahajanga 16813 7.8 Toamasina 15300 8.7 Toliara 34875 9.5 1Madanascar I 88.800 I 8.6 I Source: Steininger et al. 2003 Of special note in the above table is the fact that even provinceswith little remaining forests (Antananarivo and Fianarantsoa)still experiencehigh ratesof deforestation eventhough the absolute number of h2forest lost is low. Another main conclusion is that the dry and spiny of forests are being cut at a very rapid rate. 39 40 Dufils (in press) has also conducted a forest cover change analysis and forest type map. The categories of forest maintained by Dufils include evergreen forest, dry forest, complex secondary forest and mangroveswhich together are estimatedto cover approximately 29% (17 million ha) of the land area (Table 2.22). The remainder comprises transformed grasslands, savannahs and agricultural lands containing a tiny fraction of the original biodiversity and a high proportion of introduced species. The extremely large area of "forest" in this analysis comparedto the CI study i s due to the inclusion of secondary forest. Secondaryforest is often called either "savoka" or "tanety" and is the fallow stage of the slash and burn (tavy) agricultural system. With the removal o f Dufils' "complex secondary forest" the two studies result in relatively similar areas of forest cover. Madagascar`s unique terrestrial biodiversity is essentiallyrestrictedto the terrestrial forests. Protectedareas cover only 12.8 % of total remaining forest (19.14 % accordingto CI study) or about 3% of national territory (with dry forests underrepresentedinthe protectedareas system). Thus, expansion ofthe protectedareas and conservation of forests outside protected areas remains an essential element of Madagascar's biodiversity conservationstrategy. Table 2.22 Forest tvues and areas inMadagascar Formation H a Evergreenforest 5,532,800 Drv forest 4.1 18.300 Complex secondary forest 7,199,100 Mangrove (98% in West coast) 453,000 TOTAL 17,303,200 The original amount of forest cover is still muchdebated, butthe predominance of forest- dependent species amongthe endemic fauna suggests that the island was originally mostly forested. Despiteconsiderable efforts over the past century, the rates of deforestationfor the entire island cannot be evaluatedobjectively, largely associatedwith the lack of agreement between results obtained from various estimatesand inventories. Evergreenforests have,been better studiedthan other forest formations, and in 1999represented approximately 5.5 million ha with an average loss of 102,000 ha per year during the period 1950-1993 as well as the recent period 1993-2000(Dufils in press). Another estimate is that Madagascarlost about 12 million ha of forest between 1960and 2000, effectively reducing forest cover by 50% injust 40 years. The severest lossestook place during the 1970s and early 1980s during the height ofthe socialist revolution when the practice of slash-and- burn, locally known as tuvy, was actively encouraged in order to produce more rice to feed the growing urban population without need for investment in the declining irrigation infrastructure. -Deforestation rates have since declined from over 400,000 hdyear in 1975-1985to around 100,000 200,000 ha/year during the 1990s. The national averagerate of deforestation is currently estimated to be around 1.5% generalizedacross all forest types (e.g. Casse et a1undated). Exampleso f estimatedannual rates of deforestation include: -- Evergreen(dense humid forest, all altitudes) -0.76% -- Low altitude evergreenforests - 1.53% Dry south western spiny forest - 3.44% (the most threatenedtype) Dense savannahs - 0.67% (Messerli, 1999) 41 The vast majority of forest loss is due to agricultural encroachment. Inthe East, most forest loss is through slash & burn cultivation (tavy)by expandingor migrating resident humanpopulations. In the west and south-west most forest is lost through clearanceby migrants for growing maize or producing charcoal. Forestsare also degradedby extractive activities such as selective logging, cattle grazing and by fire the extent of which i s too difficult to evaluate without on-site monitoring. Regenerationtimes for humid eastern forests are slow (100 years) and for dry forests they are even longer or infinite (Messerli, 1999). The highest deforestation rates are for dry forest in the west and south west (around4%) where clearanceby fire demands less labor and where an estimated 80% of deforestationhas recently been due to clearancefor maize production (Reau 2001;Casse et al undated). Rates of deforestation are higher for fragments (generally surroundedby agricultural land) than for blocks. Additionally, small fragments are more subjectto internal depletion by grazing or hunting. As stated above, during the time frame 1960-2000when Madagascarlost 12million ha o fforests, the total area ofagricultural land increasedby a mere 100,000 ha, emphasizingthat conversion of forest to tavy has made no significant contribution to agricultural capital (Pavy 2002) as tavy lands are typically abandoned after 2-3 years of production. These figures demonstratethat food consumption needs of a growing population havebeen(barely) met through the simple conversion (nutrient mining) of natural capital. In addition to forest loss and degradation, Madagascar's grasslands and savannahs are subject to annual bushfiresthat effectively ensure elimination of original biodiversity and prevent any ecological regeneration along forest margins. An estimated 50% (800,000 ha) of all of Madagascar's grasslands and a further 1000-3000km2of clearedforest areas are burned every year effectively preventing any ecological regeneration(Kull2002, Fennet a1 1999; Reau, 2001). This burning results in increasederosionand large scale C02 emissions. Following transformation of forest into agricultural land, conventional land use practices continue to degrade agricultural lands. Under forest cover, soil erosionratesare negligible. As soon as forest cover is removed, annual erosion ratesrise dramatically to as much as 14.6tons/ha (Brand et al 2001, with some estimates as high as 200-400 tonsha e.g. Pavy 2002). Sediment run-off also - chokes agricultural land downstream. With the regrowth of vegetation on fallow land, erosion rates are much reduced but soil fertility is already low. Severalyears of fallow are required before fertility is sufficiently restoredto support a further rice crop. Erosion rates are also influenced by farmers' choice ofcrop -thus, ifland is usedto grow ginger (a popular cash crop), soil erosion may reach 144tons/ha (Brand undated). With continuing use of low-input agricultural practices and the absence of any mitigating measures to maintain soil condition and fertility, soils progressively erode away and the land becomes barren, as is happeningin areas of intensive cultivation such as Lac Alaotra. By contrast, anti-erosion techniques such as reforestation, contour live-fencing and direct sowing can reduce erosion by 80%, from 8 t/ha tojust 1.6 t/ha (Fara 2002). Understanding Tavv It has often been saidthat poverty isthe fundamental cause ofdeforestation (e.g. Pavy 2002). The reality appears to be more complex than this. Numerous factors influence deforestation rates including population pressure, migration, shortageo f labor or capital required for more intensive farming methods, remoteness, lack of access to markets, conflict avoidance, cultural tradition and politics. Appendix I11provides details of the relationship between poverty and biodiversity in Madagascarthat was preparedfor and included inthe World Development Report 2003. 42 The largest areas of historical deforestation occur along the eastern escarpment. This is also the most densely populated region of Madagascar. Consideration o f population maps indicates that people have spread inland from the coastal plain and up into the highlands moving away from the sea and towards the mountains as the land was depleted. Assumingno major changes incurrent practices, with a population growth rate of 2.9%, deforestation rates can be expected to double, like the population, every 30 years or so. The fastest deforestation rates in Madagascar are achieved by Antandroy migrants in south-western and westernMadagascar. Unhindered by the local taboos that protect the forest and driven by the need to earn cash to buy zebu, they clear the forest as fast as they can (Reau 2000). Shortage o f labor is a key factor. Inthe east, most farmers carry out tavy as a complement to intensive farming on the valley floors. Tavy enables them to produce rice for the family with the lowest labor input, freeing them up for other productive tasks (Brand / BEMA 2000). Remoteness is also a factor. Lack o f access to markets reduces the incentives for cash cropping and enhances the need for social security, since food may well prove unavailable at certain times o f year, especially duringthe cyclone season. There is a strong correlation between remoteness and poverty and tavy may be the only agricultural option for the poorest farmers. Conflict avoidance may also play a part. Crowding o f farmers onto the valley floors often leads to conflicts over rightsto land which are avoided on the higher slopes where tmy is carried out. Crowding creates challenges interms o f social organization. Some o f the more remote and less cohesive societies may be unableto cope with such crowding, thus encouraging tavy (Messerli 1999). Cultural factors also contribute to deforestation. For example, the drivingforce for deforestation in the west and south-west is the desire o f Antandroy migrants to acquire cattle. One hectare o f forest converted to maize will generate enough revenue over 3 or 4 seasons to acquire one zebu (Reau 200 1). Assuming 750,000 cattle exist in the southern province o f Toliara, many tens o f thousands of hectares o f forest would be required to maintain the herd. Conversely, the Mahafaly tradition which regards forests as sacred and as a preferred location for tombs has contributed in some areas to forest conservation. The practice of tavy in the east i s also undoubtedly partly driven by cultural imperatives -as a potent expression o f ownership o f land and o f the prestige and independence associated with self sufficiency in rice. There is also evidence that cultural attitudes are hardened by poverty, reinforcing the tavy response (Messerli &Kirstler undated). There is a clear link between policy and loss o f forest. Thus, while repressive policies for addressing tavy and illegal bush fires have often beencriticized as ineffective and unsustainable (e.g Casse et a]., undated; Kull2002; Reau 2001) it is equally clear that the pro-tavy policies o f the socialist revolutionary period o f 1975-1985 caused a very substantial increase in deforestation rates, estimated at 400,000 hdyear over 10years, representing 4 millionhectares, destroying over 10% o f Madagascar's forest cover. Based on a doubling o f population between 1975 and 2000, the tavy rate per capita in 1975-1985 was ei~& times today's estimated rate. Lack o f access to land and insecure tenure are also key catalysts o f deforestation. While most cleared land is subject to the prior customary right o f existing or former occupants, farmers can be sure that forest has not been farmed in recent times and is therefore subject to no prior claim. When further supported by an official cutting permit, the farmer gains effective title to the land. 43 Bush fires are primarily a means of preparinggrasslandfor cattle grazing before rains inorder to promote a soft `green bite' for cattle, sometimes coupled with a sense of injustice (Kull2001). The empirical evidence presentedabove implies that system intensification does not necessarily substitute for subsistenceproduction: farmers continueto grow food cr.ops under slash and burn systems that require very little labor, ratherthan buying food stuffs with revenuesfrom cash crops generated under intensified systems. Hence, it appears that system intensification is a necessary, but not sufficient condition to reduce slash-and-burn. This finding goes againstthe basic premise of PNAE which is more in line with the notion that system intensification initselfwould be a sufficient condition. It also confirms similar emergingevidence from other projects (e.g. LDI), and raisesthe question of how to focus follow-up investments. More emphasison land tenure security arrangements, better and enforced slash-and-burnregulations, and better market integration to ensureadequate supplies offood stuffs at reasonableand predictable prices are some of the directions that are suggested to increaseadoption oftechnologiesfor sustainable production and environmental preservation(Fara 2002). Freudenbergerand Freudenberger (2002) working in Fianarantsoahave made the interestingobservationthat householdsthat have become richer through intensification, have usedsome of their extra capital to pay labor to conduct tavy for them -thus greater intensification may also leadto greater extensification. This would suggest that although tavy is non-sustainableand has numerousnegativeexternalities, it is still a viable economic activity given the present economic and legal conditions. Enforcement of existing legislation that severely restrictscutting and burning could impose increasedeconomic costs and would decrease its financial attractiveness. Aquatic ecosystems Aquatic systems compriseMadagascar's network of rivers, lakes and wetlands (freshwater), brackish coastal swamps (mangroves) and near-shoreand offshore seas including extensive coral reefs and numerous small islands. Aquatic systems have not received nearly as much attention as terrestrial systems so remain poorly known and are highly threateneddue to a combination of factors. Freshwater systems (rivers, lakes. wetlands). Madagascar's freshwater ecosystems are considered important from a biodiversity perspective,with high rates of endemism (70% in the case of fishes). Freshwater biodiversity is threatened by introduction of exotic species, habitat degradation (sedimentation) and conversion ofwetlands to agriculture. Intact freshwater systems essentially remain only in remoteforested areas. Lakes and wetlands can be highly productive for fishing, based mainly on introducedspecies (tilapia & snakeheads.) Some freshwater systems are (or will soon be) part ofthe international RAMSARConvention sites. Brackish water systems (mangroves). Madagascar possesses over 425,000 ha of tidal marshes, of which an estimated 327,000 ha are coveredby mangrove. Mangroves are overwhelmingly (98%) concentrated along the west coast, mostly (70%) in large stands of over 500 ha. Mangroves are important for maintenanceof marine ecosystemprocesses, coastal defense, as birdareas and as sources of wood and food for local populations. Marine ecosystems (coral reefs. islands, shelf seas, ocean). With a coastline of over 5000 km, Madagascarpossesses an estimated3000 km of coral reefs, mainly along the west and north east coasts, and a large area of continental shelf (117,000 km2). The total EEZcovers 1,023,000 km2. Madagascarhas some 251 small islands that are important as nesting sanctuariesfor seaturtles and sea birds or as seasonal bases for traditional fishermen. Certain islands already make an 44 important contribution to tourism and many others have tourism potential. Islands and submerged coral reefs help to keep shippingaway from fragile coastal habitats. While rates of species endemism are very low in marine ecosystems, Madagascar's marine ecosystems are of regional importance for biodiversity because of their size and variety - Madagascar possesses an average 75% o f the shallow water species known for the entire western Indian Ocean (Cooke et al. in press). Trends. Marine ecosystems are threatened by over-fishing (especially o f coral reefs species and o f high value inshore species such as shark, sea cucumber and ornamental gastropods). Shrimp resources appear to be healthy and relatively well managed, thanks to the cooperation o f shrimp fishing companies through an association (GAPCM). The state of offshore tuna stocks is imperfectly known but recent studies consider the Western Indian Ocean tuna stocks to be significantly over-exploited by the licensed EUfleet, exacerbated bythe presence o f many illegal unlicensed vessels from Taiwan, Korea & China. Pressures on coral reefs cause demonstrable loss o f biodiversity and ecosystem productivity, with a 50% decline inthe number o f fish species reported for the Grand Recifbarrier reef of Toliara. However, creation o f set-aside o f no-take zones to ensure maintenance o f fish stocks and ecosystem processes can have a positive impact on fisheries and introduces some resilienceto the system. Factors & linkages. Degradationo f near-shore marine ecosystems is clearly linked to poverty and low rural productivity. In the arid Toliara region where the opportunities for agriculture are limited, coral reefs and mangroves come under intense pressure and are highly degraded. As resources become scarce, fishers employ increasingly destructive methods untilthe reef is razed flat and devoid o f life. Similar linkages exist in the case o f mangroves. Studies commissioned by the marine & coastal component of EP2 have shown that degradation of the coastal environment is strongly linkedto land use practices inthe interior. Erosion o f soils causes vast quantities of sediment to be transported to the coast, damaging sensitive coral reef ecosystems and altering marine ecosystem function. Localized degradation of the coastal zone through removal of stabilizing vegetation frees up sands that develop into shifting dunes which smother mangroves and entire villages. Problems, Threats, Effects and Root Causes o f Biodiversity Loss Biodiversity conservation in Madagascar is made more difficult by the exceptionally fragile terrestrial ecology -the fauna and flora have a very poor ability to regenerate following forest clearing for agriculture or man's frequent use o f fire. There are almost no native pioneer species that can re-colonize severely disturbed sites where forests have been converted to grasslands. Although man only arrived about 1500 years ago, about three fourths o f the island i s now covered by grasslands that are dominated by a low diversity o f pan-tropical, exotic grass species. Many anthropogenic impacts in Madagascar seem to be largely irreversible. Loss and degradation of natural ecosystems have complex, interconnected relationships with other ecosystems and with production systems. Loss o f forests can result in sedimentation o f freshwater and marine ecosystems, changes to the whole hydrologic cycle and, loss or reduced productivity o f lowland rice fields and reduced productivity o f freshwater and marine fisheries. The development o f productive, sustainable land use systems inMadagascar is also made more challenging by the fact that Madagascar has some of the least fertile and most highly erodible soils in the world. The soils are so infertile that a large part o f the grasslands leave much of the 45 soil surface still exposed even at the peak of vegetative development at the end ofthe rainy season. Some principal direct threats on biodiversity include the following: Deforestation -- Half of the remaining naturalforests of Madagascarhave been lost over the past 40 years. Land clearing for agriculture in the dry forest almost always results inthe conversion to grasslandsafter only one cycle of cultivation. Slash-and-burnagriculture in the humid forest can support one to half a dozen cycles of field and fallow before sites are abandoned. Forestsare being "mined" for the short-term, unsustainable use of their nutrients. Unsustainableharvest of biological resources-This is a widespread problem for marine and coastal resources, for natural forests and for freshwater ecosystems. Uncontrolled burning. Dry forests are often convertedto unproductive tone@or grasslandsby fire alone. Fire in the humid forests are an unquantified, highly destructiveandperhapsgrowingthreat. Widespread, poorly quantified, degradationof rangelands. Sedimentation/siltation of marine/coastaland freshwater ecosystems. The effects of these threats include: 0 Conversion of highly diverse forest habitats into extremely low diversity, low productivity grasslands, often of very low economic value; ` 0 Degradation and loss of coral reefs, mangrovesand sea grass beds from sedimentation; 0 Soil and hydrological degradation; 0 Fragmentationof forest ecosystems and loss of ecological corridors; 0 Degradation and loss of productivity of grasslands; Root causes ofthe identified problemsare summarized inTable 2.23. Table 2.23: Causes of Biodiversity Loss 1. Paucity of At best, the permitting systems for commercial use of natural proven natural resources is a systemfor exploiting natural resources-not for resource managing natural resources. management Not a single natural forest inMadagascar is being harvestedfor systems that have wood productsbased on a forest managementplan. Many resources been shown to be investedin developing managementplans but no plansare ecologically, implemented. economically or Emphasishas been on transfer of access rights and of non- socially viable commercial usage rights with the obligation to protect. Very little emphasis on the developmentof sustainable use/management systems that ensurethe regeneration ofthe resourcesharvestedand of the productivity of the ecosystem. Government agents confuse laws, regulations and permits with sustainableNRM. Very few institutions or peoplehaveexperience innatural resource management. 46 2. Lack o f clear 0 Permitting systems generate no economic incentives for sustainable financial incentives use. for sustainable Transfer of management rightsto local communities have rarely natural resource been associatedwith commercially oriented natural resources use management 3. Lack o f Costs of management o f the national PA network is runningat about sustainable US$250,000/month; about 20% o f this is covered by sustainable financing for sources, e.g., park entrance fees and by government contributions. protected areas and 0 Very few naturalresource management systems have systems for for natural resource reinvesting a portion o f revenues to cover management costs management 0 Modest budget allocations for the sector 0 Slow development of market mechanisms for the valuation o f environmental services. 4. Inequitable, 0 Nearly all non-agricultural lands and natural resources owned by the insecure land State -a condition established during colonial times. tenure and adverse 0 Traditional land tenure systemsthat accord rights to those who clear incentives the forest for agriculture 5. Unsustainable or Some o f the most infertile, easily erodible soils in the world low productivity Large number people living on unsustainable slash and burn agricultural agricultural (toy)-a land-use system that converts rainforest into systems un-productive grasslands over a 3 to 40+ year period. Near-subsistence agriculture that does not permit the purchase o f essential nutrients like phosphate that are depleted over time. 6. Poor Lack o f political commitment for enforcement o f laws and governance, regulations for the past three decades. especially in the Lack o f transparency and accountability. forestry sector Very poorly paid government agents are placed in charge o f high value natural resources. Lack o f internal incentives that would reward government agents for sound NRMand respect o f laws and regulations. Government agents and political parties use their control over natural resourcesto enrich themselves. People o f wealth and power use their influence over government to access resourcesto become more wealthy and powerful. 7. Poverty 0 Desperately impoverished people more and more willing to go combined with anywhere and do anything to meet their basic needs. mobility and de 0 Rural poverty is at 76.7% and has not been decreasing -- GDP per facto open access capita is US$246 to common resources 8. Rapid The combination o f an expanding population and stagnant demographic productivity have generated pressures for agricultural expansion growth o f 2.8 %- through forest conversion population doubling time o f 28 years. 47 9. Policy Policies for transfer of management rightsto communities stress constraints procedures (which are relatively cumbersome) and de-emphasize sustainable management I O . Institutional e Roles and responsibilities between MEEF and ONE remain unclear. constraints Paucity of institutions qualified to support the capacity development of community-based institutional structures for natural resources management. Paucity of institutions with natural resource management expertise/capacity. 40 ANNEX 3. Options to Improve Food Security and Food Production The previous chapter has shown that as a result ofthe high incidenceof extreme poverty in Madagascar, food security is a problem. It has also shown that in order to satisfy basic needs, the rural population has invested its assets, that is land and labor, massively inthe production of food crops that occupy 80% oftotal cultivated area. And, last but definitely not least, given Madagascar's mega-biodiversity assets, the low productivity of food crop cultivation in combination with a large and growing ruralpopulation is one ofthe principal drivers for environmental degradationthrough tavy and other forms of unsustainableagriculture and resource use. This chapter explores options on how best to increaseproductivity of food crops inMadagascar as, following the Situation Assessment presentedinthe precedingchapter, it is believed that increased food crop productivity holds an important key to simultaneously: (i)improve food security; (ii) free up assets to explore diversification opportunities; and (iii) reducepressureon Madagascar's precious environmental assets. Clearly much ofthis improved productivity should be centered on rice production. This is because rice: (i)is Madagascar's staple crop, providing 50% of all calories consumed, with the poor even relying disproportionately on rice (Hirsch 2000); (ii) occupies65% of all cultivated agricultural land, more than any other crop; and (iii) provides food and income to about 10million people (70% ofthe population) that are actively involved in its cultivation. Improved rice productivity would go a long way to improve overall food crop productivity and is therefore the primary focus ofthis chapter. After having explored in detail possible optionsto improve rice productivity (reflecting the results ofthe commissioned backgroundstudy on the rice subsector), the chapter will look briefly at options to improve productivity of cassava and small-scale fisheries. 3.1 Rice Overview Rice cultivation is the single most important economic activity in Madagascarwith direct value- added in 1999contributing to 12% of GDP. Consequently, accelerationof economic growth in the rural sector as envisaged under the PRSPwould be difficult to envisage without setting the stage for substantial productivity gains in the rice sub-sector. Both the large number o frice producing householdsin Madagascaras well as the fact that the vast majority (85%) of the value- added in the rice value chain goes to rice producers in the form of net income and to salariesof agricultural workers suggest that the benefits of any rice productivity increasemost likely would be widely distributed and would directly support the primary revenue source of the poorestrural households(Table 3.1). Total value-added ofthe rice value chain is composedof 74% of net income (including value of family labor) of rice producers, processingagents and traders. Another 19.6%oftotal value added is paid out in the form of salaries for agricultural workers. Only 6% oftotal value added reflects compensation for renting land, 1% is usedto compensate financial (interest) costs and 0.3% is collected as taxes. Family labor costs of rice producersare estimated at 33% oftheir net income (1,627 billion FMG in 1999). 49 Unit: million Fmg Rice Processing Trade Whole producers Agents operators and chain detailers Of operators Production Gross Value 2.517.504 3 12.599 3.785.643 Intermediary Consumed inpu 249.302 200.363 3.505.178 692.758 Added Value 2.268.202 112.236 280.465 2.660.90 1 Salaries and gifts 490.193 8.630 24.407 Land renting cost 150.790 1 Financial interest costs 1 I 4.589 I 22.081 I/ 26.669 I Taxes 2.256 4.616 6.872 Net Income I,627.219 96.761 229.361 1.953.341 Therefore, in a strategy aimed at settingthe stage for more inclusive economic growth as a means to reduce poverty, such as the PRSP, it appears that efforts aimed at increasingrice productivity would potentially have a high pay-off in strengthening the link between poverty reduction and economic growth inMadagascar. The highdistributive potential o fthe benefits associated with productivity growth inthe rice sub-sector also suggestthe likelihood o f significant multiplier effects as most o f the increase innet incomes o f rice producing households would be translated into increased consumption o f products and services generated by the expanding secondary and tertiary sectors in urban areas. However, the lack o f scope for added value inthe rice sector limits its potential for generating rural prosperity. And it further needsto be pointed out that rice provides only the calorific part o f nutrition. Over-dependence on rice would lead to poorer nutritional balance and reduced rural productivity. Sector Structure Given the heterogeneityand geographical and seasonal variability o f the rice sub-sector in Madagascar, numerous factors need to be considered when seekingto identify the enabling conditions for increased productivity, ensuring food security and improved market integration. From a geographical perspective, rice cultivation takes place in six different regions: North, North-West, Center-West, Hauts Plateaux, East Coast, and Lake Alaotra (Figure 3.1). 50 Figure 3.1: Map of rice producing regions I A Antsiranana 1Nord Ouest i ICentre Ouert I itsoa Toliara w ' 71 v-1 Non cowerte 1 Taolanaro Two regions are consideredto produce a surplus relativeto regional consumption (Lac Alaotra and Center-West) while four runa rice deficit (North, North-West, Hauts Plateaux and East). In terms of number of farmers and cultivated area, the Hauts-Plateaux and East Coast regions are the most important (Table 3.2). 'YOof total rice farms Average Yield (kgka) 1285 2429 1918 RiceArea croppedper farm (Ha) Productionper farm (T) Quantity Kg de paddy/working day Rice Volume tradedby producers %of national volume of local rice trade Rice Volume sold by retailers Degreeof coverageofmarketdemand Number o fother operators 'YOoperatorsnonproducers From a production perspective, rice is cultivated through six distinct production systems. Four of these are irrigation-based production (lowland transplanting, lowland direct sowing, improved and intensive) and two are rainfed basedproduction systems. Irrigated systems cover about 80% 51 of rice-cultivated area. Slash-and-burnsystems (tavy) cover almost 150,000 ha or 10% of total cultivated area. Low-land transplantatiing systems are the most common under irrigated conditions and cover about 820.000 ha or almost 60% of total cultivated area. From a farm householdperspective, atotal of 18 rice-basedfarming systems can be distinguished. Based on production managementand householdstrategies, rice producerscan be grouped into three categories: (i)highly vulnerable micro-producers; (ii)rice-selling specialized producers; and (iii)multi-activity oriented self-sufficient producers. Highly vulnerable micro-producersgrow rice under unfavorableconditions with limited access to land, water and capital. They representabout 34% of all rice-producing householdsand do not even produce sufficient rice to meet subsistence needs. Highly vulnerable micro-producers comprise mainly rainfed and slash-andburn farmers inthe Northern and Easternregions. For them, the average cost of paddyproduction is FMG 802/kg, which is relatively high. Rice-selling specialist producersgrow rice underrelatively favorable conditions that allow for double cropping. Access to land and capital is also relatively favorable. This category of farmers represents23% of all rice-producing householdswho are found around Lake Alaotra, and in certain areas in the North-Western, Central-West, and Hauts-Plateauxregions. They are highly competitive, with an average cost of paddy production amountingto FMG 477/kg. Multi-activity self-sufficient producersgenerate a small rice surplus, indicating that rice is grown primarily for home-consumptionwithin the context of an otherwise diversified farming system. They represent43% of all rice-producing householdsand can particularly be found inthe Hauts- Plateauxand Center-West regions. Multi-activity self-sufficient producers are moderately competitive with average cost of paddy production estimatedat FMG68l/kg. From a poverty perspective,the vast majority (>SO%) of rice-producing households are poor - basedon a poverty threshold of US$1of purchasingpower per personper day. Rice producing householdsthat show signs of a partial escape from poverty are particularly found within the category of multi-activity self-sufficient producers. This confirms the importance of agricultural diversification and non-farm activities for reducingpoverty in rural areas once food security in rice is assured The positive impact of diversification inreducing poverty is further supported by national household survey data which show that off-farm income accounts for 65% of the income of the richest rural quintile and only 38% ofthe income ofthe poorest rural quintile.. Rice Productivity From an historical perspective, the dismal record ofthe rice sub-sector impliesthat achieving the desired improvementswould be a major departure from the establishedtrend. Average paddy yield per ha in Madagascarhas been stagnating for the last forty years (Figure 3.2). In 1960, before the GreenRevolution, when yields where largely determined by natural factors, the average paddy yield in Madagascar was slightly higher than in Indonesia and significantly higher than in Mali, thereby indicating that conditions for rice cultivation are inherentlyfavorable in Madagascar.However, yield-enhancing technologies associatedwith the Green Revolution have never become establishedin Madagascar.As a result, yields inIndonesia are now more than double those of Madagascar, while yields inMali, formerly significantly lower than in Madagascar, have recently overtaken yields in Madagascar. 52 Figure 3.2 PaddyYields in Madagascar,Mali and Indonesia(source; FAOSTAT) 5.00 ~, . . . -... ............, - -- I ,.. .^ ...... .... .,......" ..,,. ..,,, 4.50 4.00 3.50 3.00 -~ Indonesia f 2.50 m I- ---ICMadagascar --t-Mali 2.00 1.50 1.oo 0.50 0.00 Duringthe last twenty years, domestic rice production has not kept pace with the growing demand for rice. Consequently, Madagascar, a country that used to export rice, has come to rely on imports to compensatefor its deficit. At the same time, rice consumption per capita has steadily declined, reflecting on the one handpossible changes infood preferencesassociated with on-going urbanization and on the other hand growing poverty levels which may have induced substitution of rice with cheaper altematives such as cassava (Figure 3.3). 53 Figure 3.3 :Rice productionand availability in Madagascar(1962-99) 1 500 - C c D p 1000 - 1202 x- 100 - 80 .g b, 500 - - c ~ - 60 2 250 - - 4 0 Zu,i D 0 +Production nz +-Importations +Disponrbrlrte I hab -. Source :Instat, Minagri In absolute terms, domestic paddy production has morethan doubled between 1962 and 1999. However, this increasehas been driven more by area expansionthan by improved productivity, with the former causing about 80% ofthe incrementalproduction over this period and the latter about 20% (Table 3.3j. There are many limits to increasingproduction through area expansionas has been pointed out in the previous chapter. Furthermore, area expansion has not preventedthe rural population from falling deeper into poverty and has led to widespread deforestation, with the consequentialloss of economic opportunities and ecosystemfunctions associatedwith standing forests. Extensive rice production has also caused massivesoil loss and sedimentation effects that severely reduce the productivity of irrigated lands in the Lac Alaotra, Easternand Northern Regions. The high environmental cost of area expansion impliesthat a shift towards agricultural intensification would potentially generate indirect benefits adding to thejustification for investment in intensified production.. I Area (Ha) Yield (Tons/ha) Total Production (tons) 1962 830,000 1.63 1,350,000 1999 1,448,945 1.92 2,778,39 1 Increase 75% 18% 106% Reasons for Low Rice Productivitv When askedto identify the factors that limit the production of rice, farmers cite a host of reasons, as indicated in Table 3.4. Closer analysis of these responses, that come from 1,470 rural households included inthe EPM-2001, revealsthat farmers particularly perceive constraints to the productivity of labor (e.g. lack of access to equipment, lack o f hired labor, lack of animal draught power). Access to land is also perceivedas important. Landtenure and sedimentation problems are perceivedas less important. 54 Lack of access to land 26.6 20.3 53.1 1Land tenure insecurity 46.5 28.7 I 24.8 I Lack of access to manure 37.4 26.1 36.5 Lack of access to animal draught 22.9 19.5 57.6 power Lack of access to hired labor 21.0 24.9 54.1 Lack of access to agricultural 19.3 14.5 66.2 equipment Lack of access to agricultural 35.6 22.3 42.1 inputs Lack of access to credit 38.4 21.6 40.0 Degradation of irrigation 29.8 27.3 42.9 infrastructure Sedimentationproblems 41.4 31.4 27.2 Labor. As far as labor inputs are concerned, regional comparison indicatesthat labor intensity, input use, and labor productivity vary significantly (Table 3.5). Inthe Lake Alaotra region, the strategy appears to be to maximize returns to labor, with returns that are actually 8-9 times higher than in the HautsPlateaux(HP) region. This strategy is reflected in the greater use of mechanizedagricultural services (e.g. plowing by tractors) and wider use of chemical weeding (41YOo f areas) which reduces labor input and increases returns per family working day. At the other end ofthe spectrumis the Hauts Plateaux(Hp)where the smaller averagefarm size orients producerstowards maximizing yieldsha. The Center West region demonstratesan intermediate strategy. The large difference in labor input per habetweenLake Alaotra and HP (ratio of 1 to 3.5) for a very similar yield (7% difference) is very striking. As far as access to land is concerned, rice productivity varies according to existing cropping systems, including: (i)irrigated rice; (ii)rainfed rice; and (iii)tavy rice or upland slash and burn rice. Yields are highest under irrigated conditions, followed by rainfed cropping systems. Yields under slash-and-burnconditions are very poor (Table 4.6). Another source of variation is cropping practice, including: (i)direct sowing; (ii)manualtransplanting; (iii) intensive external input use (SRA); and (iv) intensive labor use (SRI). As can be noted from Table 3.6 each type of irrigated system has its advantages and disadvantageswith regional preferencesfor specific systems. Direct sowing systems are attractive because ofthe high return on labor, which is 2-4 times higher than for traditional transplanting. However, yields/ha tend to be lower, makingthese systems particularly attractive where access to land is less problematic. SRA systems tend to give higher yields per ha, but require access to external inputs. SRI systems also aim at maximizing yields per ha, but require massive labor inputs. Transplanted rice cultivation is the most common practice and is preferred because water managementrequirements under this systemare less demandingthan under direct sowing, SRA or SRI. In addition, yields 55 c vary according to geographical region reflecting differences in agro-climatic conditions, irrigation, use o f inputs etc. Average yield (kg) 2 010 1691 2 738 1653 678 Labor requirementdha (days) 348 253 459 865 232 Vol. Paddy kg/ working day 6 6 6 3 Gross Income/ha 1 304 996 1412 031 2 202 791 956 377 342 425 Margin /day of family labor 11 348 8 209 8 189 1576 2 395 LAKE ALAOTRA AverQgepricepaddy: 922 of total rice cultivated 22% 54% 6% 1Yo 10% 7% Average yield (kg) 1863 2 818 3 656 4 274 1958 619 Labor requirementdha (days) 44 88 95 237 90 224 Vol. Paddy kg/ working day 39 31 38 18 20 3 Gross Income/ha 1 478 648 1813 036 2 650 443 2 859 761 1 127734 368 892 Margin/ dry of family labor 82 147 78 828 66 261 23 250 30 479 2 236 Inputs. Farmers widely claim that they do not use fertilizers becausethey are: (i)too expensive; (ii)nocredit isavailabletoassistintheirpurchase;or(iii)simplynotavailable. Long-established habits may also be part o f the cause, together with a reluctance to invest in inputs for crops mainly intended for home consumption (Hirsch 2000). Impositiono f VAT indeed caused a sharp drop in fertilizer use. At the national level, 69% of the area under irrigated rice is cropped 'Giventhat tavy is generally consideredto be asystemrequiring relatively low labor inputs, this figure is anomalous and may indicate '''problemswith the dataon labor(BMinten, pers. comm.) The volume of paddyiworking day is basedonthe net production supplied after subtractingthe volume of local seeds Gross Income per ha= Net production per ha (after subtractingseeds x average price/ kg of paddy monetarycosts including - inputs:phyto., fert, purchasedseeds, salaries, rice workers' food, equipmentrenting: GIMu=P.h,,xprice -production costs 56 * without any mineral or organic fertilizer application. An exception is the HautsPlateauxregion, where fertilizers are applied on 76% of all areas under irrigated rice, mainly inthe form of manure. With an average level of fertilizer application amounting to 33 kg/ha in the Hauts Plateauxregion as comparedto 15 kg/ha in Lake Alaotra, one observes a significant yield differential of 568 kg/ha or 22% betweenthe two regions (Table 3.7). The difference with the Central-West, where 78% of all irrigated rice is grown without fertilizers is even larger. This differential suggests that at the currently low levels of fertilizer application in Madagascar, the yield improvement associated with a modest increasein fertilizer use would be significant. Average yield irrigatedrice (kg/ha) Fertilizer use(YOofcultivated area) HautsPlateaux 3,200 76% Lake Alaotra 2,632 40% Center West 1,966 22% Water. Basedon farmers' responses, the impact ofthe quality ofwater managementon paddy yields under irrigated conditions appearsto be highly significant. As indicated inTable 3.8, improvement of water managementincreases paddy yields by 30-60%. Water management Lac Alaotra Center-West Hauts Plateaux North West (kg/ha) (kg/ha) (kg/ha) (kglha) Good 3282 2087 3535 1779 Medium 2490 1748 3424 1519 Bad 2139 1788 2740 971 Determinants of Rice Productivity To obtain amore quantitative idea of factors that determine rice productivity, a logarithmic production function was estimatedbasedon recently available data from the EPM-2001. The dependent variable in the model is the logarithm of paddy yield (kg/are). The possiblecausative factors include use of inputs, type of household, type of production system and the degree of isolation (measured as the distance ofthe plot to the nearestmain road). The results from 767 rice plots are presented in Table 3.9. They show that paddy yields are positively correlated with input use, sowing density, applied labodha andthe use of animal draught power per ha. Isolation of a rice plot has a strong negative effect on paddyyields. The results also show that paddy yields are associatedwith diversification ofthe agricultural production system, measuredas rice cultivated area as a percentageof total cultivated areaper rural household.This could either be because diversification provides cash income year round thus permitting higher rice productivity or because more productive farmers are able to invest in diversification (Mintenpers. comm.) 57 l s e m h a 1 .164415 .0406888 4.04 0.000 .0845392 .2442909 l a n i m 1 .0550683 .0186921 2 . 9 5 0.003 .018374 .0917625 l m a i n h a I .1369547 .0213813 6 . 4 1 0.000 .0949813 ,1789282 lduree I - . l o 7 2 8 5 .0251839 -4.26 0.000 -.1567233 -.0578468 l v a l - i n t r a n t h a .1899186 .0361996 5.25 0.000 ,1188555 ,2609817 l p n o r i z 1 .1158097 .032785 3.53 0.000 ,0514496 ________________________________________----------------------------------- -cons 1 -1.611279 .3416707 -4.72 0.000 -2.28201 -..1801697 9405488 Imainha- log o f labor inputiha: lduree- log of distance from nearest main road; Iva-intrantha- logof the value of Key to Table 3.9 : lrendt-logproductioniha; Isemha- log of sowingdensity/he; lanim - log of animal power inputiha; inputsiha: lpnoriz - type ofproductionsystem; -cons -constant. In order to obtain a better ideato what extent the constraints as perceivedby farmers, and those mentioned in Table 3.3, affect paddy productivity, an ordinary least square regressionanalysis was carried out basedon 2,183 plot data from the EPM-2001. The resultsofthis analysis are presented in Table3.10 and suggest that sedimentation problemsand problems with irrigation infrastructure and lack of access to inputs appear to be significant productivity shifters. Table 3.10: Effect of PerceivedConstraints on Rice Productivity S o u r c e I ss df MS Number o f obs = 2183 -------------+------------------------------ F ( 11, 2172) = 230.12 M o d e l 1 405.250545 11 36.8409586 P r o b > F = 0.0000 R e s i d u a l 1 347.723595 2172 .160093736 R-squared = 0.5382 -------------+------------------------------ A d j R-squared = 0.5359 T o t a l 1 752.97414 2183 -344926312 R o o t MSE = .40012 r e n d t q I C o e f . S t d . E r r . [ 9 5 % Conf. n t e r v a l ] -------------+----------------------- .--------------I -------- , SDS I -4.922708 2.594637 -1.90 0.058 -10.01094 .1655233 p l u v i a l i -5.81069 1.104094 -5.26 0.000 - 7 . 9 7 5 8 8 1 -3.645498 t a v y I -6.520035 .4357028 -14.96 0.000 -7.374472 -5.665597 semence/ha I .121126 .0100164 12.09 0.000 . l o 1 4 8 3 3 .1407688 d i v e r s i f i c a t i o n 1.618043 .5946369 2.72 0.007 .4519263 2.78416 a c q u i s d e f r i c l -3.444011 .6993015 -4.92 0.000 -4.815381 -2.072641 v a r i e t e a m e l i 1 -3.56401 .7309741 -4.88 0.000 -4.997492 -2.130528 c o n t r a i n t e l i e e a : a c c e s i n t r a n t l -1.271173 .3365432 -3.78 0.000 -1.931154 -.6111932 i n f r a s h y d r a u l -2.765436 .3733523 - 7 . 4 1 0.000 -3.497601 -2.03327 e n s a b l e m e n t 1 -5.689576 .3119951 -18.24 0.000 -6.301417 -5.077736 c o n s t a n t e 1 8.831777 .6593994 1 3 . 3 9 0.000 7.538657 1 0 . 1 2 4 9 diversification - diversification; acquis defric - credit availability; variCte ameIiorC- improved varieties; accts intrant Key to Table 3.I O - SDS -direct sowing; pluvial -rainfedrice; tavy - slash& bum; semence/ha- sowing density; -accesstofertilisers; infrashydraul-irrigationinfrastructure;ensablement-sedimentation;constante constant. - Options to Improve On-Farm Rice Productivity Based on the presenteddata and analysis, the challengeto improve on-farm rice productivity can be translated into the question of how to make rice production system intensification work. In order to achieve intensification, the following options are proposed: (i)the development o f locally adapted productivity enhancingtechnologies, (ii) the creation ofan enabling environment to inducetechnological innovations at the farm level; and (iii)the establishment of productive on- farm infrastructure. Given a fixed farm-size, successfulrice production systems intensification would subsequently lower the threshold for production diversification as rice producing households would need less land and labor to meettheir food security needs. 58 Productivitv Enhancing Technologies. As far as irrigated rice is concerned, results indicate the potential yield enhancing effects of Green Revolution type technologies based on improved seed and fertilizer application. Given the currently very low use o f modern inputs, particularly chemical fertilizers (95%) do not have accessto credit to finance yield enhancing inputs. At the same time, access to credit i s reportedly 2.3 times higher among farmers who do apply fertilizers compared to those who do not. The fah-gate price o f fertilizers is relatively high due to hightransportation costs and a poor distribution network; input suppliers also have to deal with distortions inthe fertilizer market due to availability o f fertilizer donations. The considerable variation o f the paddy farm-gate price is perhaps the key limiting factor because farmers runthe risk o f not being able to recover fertilizer costs, despite significantly improved production. For instance, with the marginal product o f fertilizer application on paddy production amounting to 5.0 and given a farm-gate price of fertilizers o f FMG2,5OO/kg, the farm-gate price o f paddy should be at least FMG 500/kg to recover fertilizer costs. In2001, the farm-gate price o f paddy varied around FMG 500/kg, renderingthe use o f fertilizers a break-even or loss-making proposition. Ifit is assumedthat the ratio o f the marginal product o f fertilizer application to the ratio o f the farm-gate price o f fertilizer over the farm-gate o f price o f paddy should at least amount to 2.0, the farm-gate price of paddy in 2001 should have been FMG 1,000 per kg'. Alternatively, to generate the same result, the price o f fertilizer should drop from FMG 2,500 to FMG 1,250 in case the farm-gate price ofpaddy stays at FMG 500kg. This example shows that the establishment o f an acceptable ratio o fthe farm-gate price o f fertilizers to the farm-gate price o f paddy relative to the marginal product of fertilizer application on paddy production would be a necessary condition to significantly increase fertilizer use in irrigated rice production in Madagascar. Such a ratio might be achieved either by ensuring an acceptable minimumfarm- gate price o f paddy through import tariffs or by ensuring an acceptable maximum farm-gate price o f fertilizer through eliminating tariffs on fertilizer (or applying a fertilizer subsidy) or through a combination o f both measures. Since farm-gate prices o f paddy should not be raised beyond the reach o f the urban poor and rice-deficient rural poor) a tax exemption o f imported fertilizers would be the minimumrequired "subsidy" to ensure that farmers adopted the use o f fertilizers. However, further study here is recommended. Ensuringaccessto means of financing the purchase o fagricultural inputsbyrice producing households is another critical element of an innovation-inducing environment. Taking into 7Let MPfp be the marginalproductof fertilizer applicationon paddyproduction; Pf=farm-gate price of fertilizer per kg; and Pp= farm-gate price of paddy. IfMPfP/(PWp) =2; MPQ=5; and PfiFMG2,500/kg; then Pp is FMG1,OOO. 60 account existingcreditconstraints for the vast majorityof asset-deprivedrice producing households, there is a needto scale-upthe positiveexperience of communityvillage granaries ("Greniers Communautaires Villageois" or GCVs)under which micro-financeinstitutions provide credit secured on stored paddy. Giventhe very limitedexposureofthe vast majorityof rice producinghouseholdsto the application of yield enhancing inputs, a case couldalso be madeto providethem with a limited amount of inputsthat would allowthem to experiment with these inputs intheir own fields. In order to do so one couldthink of setting-up an inputvoucher schemefor a limited periodoftime in the context a researchoutreach program. Sucha schemewould havethe additionaladvantage o f puttingeffectivedemandfor yieldenhancinginputs inthe hands of rice producinghouseholds, thereby contributingto improvement ofthe inputdistributionnetwork. Inorder to deal directly with the poor inputdistributionnetwork, additionalactionsto improveaccess by new entrants couldbe considered. These could includethe makingcredit availableand providingstart-up subsidiesto new privateinput suppliers inthe contextof initiativesto promoteprivatesector development in rural areas. On-FarmProductiveInfrastructure. An importantconditionfor successful systems intensification is adequatewater management. The resultsofthe study indicatesignificanteffects of good water managementon paddy yields (Table3.6). Inthis context, there is a needto stop the deterioration of irrigation infrastructurecausedby eroded soil and sandsilting'in canals. At the nationallevel, there are about 700,000 haof lowlandsaffectedby poorwater managementconditions. This situationexistsagainst the background of ineffectivepublic investmentsthat havebeenmade in the past, which focused particularlyon hardwarerequirements of large-scale irrigationschemes and neglectedthe institutionalmechanismsrequiredfor long-termmaintenance. Resolvingthis problemwould enable rice farmers to adopt either labor-savingdirect seedingtechniques or yield enhancingtechnologiesbasedon either externalinputsor labor-intensivecultivationpractices. Consequently, there is a needfor irrigationrelatedinvestmentsto improveand rehabilitate Madagascar's most importantproductiveasset, which is its irrigationinfrastructure(second- largest in Africa). Since 80% ofall irrigationinfrastructureis concentratedin eight (out of 20) existingagro-ecologicalregions, these investments should beparticularlyconcentrated inthe Marovoay,Melaky, Diana, Alaotra, Atsimo-Andrefana, Menabe, Vakinankaratraand Imerina Centralregions. The BetsibokaRizicultureProjectinthe Marovoayis a leadingexampleof an approach which combines infrastructuralworks with institutionaldevelopment for their maintenance(FTFABE/AHT2002). In light ofthe fact that past irrigationinvestments have not generatedexpectedresults, irrigationrelatedinvestments shouldbe incorporatedintoa watershed management approach under which investments in irrigationinfrastructure, environmental protectionmeasures, localcapacitybuilding, andtechnologytransfer should beadequately balanced. Additionally, attentionand fundingmust be considered for ongoingmaintenance of irrigation infrastructuresince past investments havebeen lost due to poor maintenance. MarketIntegration The needto increaseon-farmproductivitycannot be seen in isolationfrom the needto improve market integration. Marketintegrationis requiredto set the stage for productivity increases(e.g. through lower pricesofexternalinputs) and can be considered a preconditionfor capturingthe benefitsfrom increasedon-farmproductivityor diversification. The needto improvemarket integrationrequiresdealingwith a set of inter-relatedissues, including: (i)the organizationof the domestic marketfollowing liberalizationefforts inthe 1990s;(ii) the extreme segmentationofthe domestic market dueto the deficientruralroads networkandthe lackof informationflow; and (iii)theintegrationofthedomesticricemarketintotheinternationalmarket. 61 The impact o fthe market liberalization of the late 1980sand early 1990s by dismantling state- owned rice marketing and processingcompanies has been positive in reducingthe trade gross margins of downstreamoperators. Between 1996 and 1999, gross margin for paddy collectors decreased from 19%to 13%, while wholesalers' gross margins decreased from 14% to 10% and retailers' gross margins from 10-16% down to 8% with large disparities remaining between regionsdue to access constraints. This reduction of gross margins is an indication of increased competition in local rice markets. However, the benefitsofthe reducedmargins appear to have been passedon mainly to rice consumers. To the extent that rice producing households also purchase rice, this is not entirely negativefor rice producers. In terms of traded volumes, domestic paddy production was estimatedat around 2.8 million tons in 1999. Taking into account harvestassociatedlosses and seed provisions for the next crop, this provides an available net supply of 2.6 million tons, which is equivalentto 1.7 million tons of rice. A total of about 1.6 million tons ofpaddy or 62% oftotal production i s usedfor home consumption. Out ofthe remaining 976,000 tons, atotal of about 200,000 tons (or 6-7% of total paddy production) is used for payments in kind for land, labor, services and inputsas well as gifts. Hence, a total of 776,000 tons of paddy, amountingto 28% ofthe total domestic paddy production is effectively traded. This translates into atotal of 523,000 tons of domestically traded rice. Taking into account annual rice imports of about 186,000 tons in 1999, the annual amount of rice traded in Madagascaris 709,000 tons. Importedrice accountedfor 27% of all traded rice in Madagascar in 1999. Traded rice transits through a series of operatorsbeforebeing sold through retailers to the consumers. It is estimatedthat out of the total of 709,000 tons of rice traded in 1999, 294,000 tons (41%) were consumedby rural householdsand 415,000 tons (59%) by urban households. If it is assumedthat imported rice is overwhelmingly consumedin urban areas, it means that imported rice covers 45% of urban demand, while domestically producedrice covers 55%. In terms of market structure, it is estimatedthat there are over 4,700 collectors that operate in the country. They trade annually about 640,000 tons of paddyand 80,000 of other rice. The major clients of the collectors are the millers who take 63% ofthe collected quantity o f paddy, wholesalers take 35%, while 2% is directly delivered to final consumers. There are about 2,100 small rice millers. They are very well organizedwith a network of collectors inthe North Western (82% of paddy delivered by collectors), Hautsplateaux(79%) and Lake Alaotra (82%) regions. in other regions such as Center West andNorthwholesalers buyingpaddy from collectors (77% in Center-West; 73% inthe North) play a more important role in providing paddy to rice millers, which is most probably due to the lower concentration of small millers in these regions. Dependingon the region, the average retailer trades between 8 and 58 tons ofrice per year, with daily sales of between 27 and 193 kg(on the basis of 300 trading days per year). It is estimatedthat there are about 23,800 rice retailers. Despiterecent improvements, market competitivenessremains affected by the limited number of downstreamoperators. Almost 50% o f all villages remain faced with a situation of permanentor occasionalmonopsonies. As a result ofthe deficient rural roads network, rural markets are segmentedwith both strong spatial as well as seasonal price variations. For instanceinthe Northern region the retail price of rice was 55% higher inthe most expensive market (Diego) comparedto the cheapestmarket (Andapa) in the period immediately following the harvest of rice. This price difference decreased to 35% betweenthese two markets during the deficit period (Table 3.11). Seasonalvariations of rice prices are quite strong as well. At the collector level, the rice price typically shows a seasonal variation ranging from 18% to 41% inthe same region. 62 r Diego Sambava Andapa After harvest 2.600 1.750 1.675 During deficit period 2.975 2.300 2.200 Downstream operator margins can also be traced back to access constraints caused by the absence or poor state o f rural roads. Margins tend to be higher inthe North and Center West regions where the quality o f the road infrastructure is far below national average. For instance, in the Morondava region there are only 0.8 kin o f paved roads per 100 sq-km, in the Northern region this figure is 2.3 km per 100 sq-km, while the national figure is 6 km/ 100 sq-km. Under these conditions, collectors manage to keep margins in the 17-25% range, while they are down to 9% in the regions that are better served (Hauts Plateaux, North-West, Lac Alaotra). The same holds true for wholesaler margins (Table 3.12). Table 3.12: Regional comparison of downstream operator margins YO(Sales- Collectors Wholesalers PurchasesYSales -NORTH 25.3% 10.6% 6.4% NORT-WEST 9.6% 7.7% 4.5% CENTER- WEST 17.1% 14.5% 6.7% HAUTS 8.8% 12.1% 6.1% PLATEAUX EAST 8.8% 6.3% L A C ALAOTRA 8.9% 8.1% 16.7% Average 13.1% 9.9% 7.6% i n terms o f accessto the domestic market by international competition, the existing 15% import tariff and 20% VAT on imports are o f particular relevance. With an effective protection of about 38%, the objective o f this policy is to ensure a reasonable paddy price to domestic producers and a reasonable market price to rice consumers. Inview o f fluctuations indomestic production caused by external shocks (e.g. weather, locusts), the import tariff is adjusted accordingly. For instance, following extensive cyclone damage in2000, it was reducedto 5%. Enforcement of the existing tariffs has however been problematic, causing a major decline in paddy producer prices in 2001 to levels as low as halfo f 1999-2000price levels. Hence, without enforcement o f the existing tariff, integration o f the domestic market into the international market has de-facto progressed much more rapidly than the integration o f domestic markets which is hamperedby problems o f rural infrastructure and market organization as indicated above. Options to Improve Market Integration Based on the presented data and analysis, the challenge as far as the market integration objective is concerned will be to synchronize the pace o f liberalization o f external trade with the pace o f integration o f domestic markets so as to provide an incentive to rice producing households to capitalize on the intrinsic competitiveness o f rice production at the farm level. This would entail: (i)puttinginplaceanappropriatericeimportpolicy; (ii)generatingconditionsfor increased 63 competitionamongdownstream operators of the rice chain, includingthe improvementof rural roadinfrastructureand (iv) increasingthe capacity of rice storage at the local levelthrough GCVs. RiceImport Policy. There exists a very sensitive relationshipbetweenthe paddy farm gate price and the ability ofrice producinghouseholdsto obtaina labor returnper day that is higherthanthe localwage rate over a rangeof betweenFMG 600-900/kg:at a paddy priceofaround FMG 600/kg, only 46% ofall rice producers are able to do so; at a paddypriceof over FMG 900/kg over 82% are able to do so (Figure 3.4). At the same time, the producer priceof paddy depends on the effective importtariff for importedrice. Figure3.4: Cost-Recovery-Paddy ProducerPriceRelationship(2001) P a r t o f v i a b l e f a r m e r s a n d p a d d y f a r m g a t e , p r i c e , 1 0 0 % 9 0 % 8 0 % 7 0 % al E 6 0 % b le E 5 0 % L 4 0 % v w- m 3 0 % 2 0 % 1 0 Yo 0 % 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0 1 0 0 0 1 1 0 0 1 2 0 0 1 3 0 0 P a d d y p r i c e f m g / k g I .~ __ - ~ Giventhat the vast majorityof rice producinghouseholds are poor and giventhat abouttwo- thirds ofthem produce a surplus inquantitativeterms, there is aneedto ensure that the farm-gate price of paddy does not fall belowthe FMG 700-9OOkgrange, so as to avoidpushingalarge number of rural householdsfurther intopoverty. Inorder to establisha minimumfarm-gate price for paddy, the challenge is to set an importtariffthat keepsthe farm gate priceat the upper level o f the FMG 700-900 range. Ratherthan increasingexistingtariffs, thiswould requireeffective complianceof actual importtariffs andtaxes by rice importersas demonstratedin a simulation model presented in Appendix 11. Tariff enforcement should be ensuredthroughamore aggressiveapproach in resolvingthe actual governance issuesthat are currentlysurroundingthe collectionof importtariffs andtaxes. Since lack of compliancewith the current importtariff and tax structure is inter alia caused bythe manipulationof invoices, there is a needto at least temporarilyreplacethe currenttransactionbasedcollectionsystem with onethat applies tariffs and taxes basedon a referenceprice linkedto the internationalmarket. This requiresclose monitoringofthe world market priceinorder to adjustthe importtariff accordingly incaseofa significantdecrease so as to avoid an abrupt negative impactonthe farm gate price ofpaddy in the sensitiverange. Ensuringa farm-gatepriceofpaddythat would fall somewhere inthe FMG 700-900rangewould translateinto a consumer priceofrice of about FMG2,000 per kg, which given historic price levelswould not be consideredexcessive. The effect ofbetterenforcementoftariffs andtaxes on the priceof importedrice might be particularlyfelt inthe Easternand Hauts Plateauxregionsas 64 these are the major destinationsof importedrice, accountingfor over 80% of all imports in 1999. Rural households that would suffer from a stricter enforcement of rice import tariffs due to an increase in rice pricesabove the low price levelof 2001, would bethose that do not produce enough rice to meet home-consumptionneeds. These households representthe categoryof so- called vulnerable micro-producerswho particularly depend on slash-and-burn farming systems in the Eastern and Northernregion. Sincethese farming systems produce rice at high environmental cost, the effect of a higherprice of importedrice on the incomes of rice producinghouseholds engaged in tavy could be seen as part ofthe `stick' dimensionofthe carrot-and stick approach that is beingproposed to arrest further expansion of toy. At the same time, the second-order effects of a guaranteed minimum farm-gatepaddy pricein the FMG 700-900 range might offer opportunities for rural householdswith rice deficitsthrough increased agricultural wages and new employmentopportunities in the rural off-farm sector (trade and services). An alternative viewpoint concerningthe influence ofrice import tariffs as a means to improve rural agriculturehas been providedby Minten (pers. comm.) One question is whether higherrice prices might encourage extensificationas opposedto intensificationof production. For example, the increased profitability of maize ledto widespread deforestationinthe southwest, and the greater wealth of farmers in Fianarnatsoahas encouragedthemto hirelaborto carry out slash & burnfarming on their behalf(Freudenberger & Freudenberger2002). A second question concerns the influenceof price seasonality on productionlivelihoods. The following table is an assessment ofthe rice markets basedon INSTAT-DSM, andEPM2001data. And finally it must be pointed out that the extreme fragmentationand priceelasticityofthe rice marketwill act as a buffer to both positiveand negativeeffects ofthe rice tariff (B.Minten, pers. comm.) Table 3.13: Rice marketing* by povertyquintile Percentage of householdsthat Total Q1 4 2 43 p4 Q 5 ... sold rice 52 47 54 56 53 49 ... boughtrice 72 76 77 72 67 64 ... boughtimportedrice 13 11 11 12 15 19 kgs of importedrice bought 28 18 20 31 40 32 ... sold and boughtrice 33 35 37 36 29 23 ... are net rice buyers(quant.) 53 58 55 54 47 49 ... are net rice buyers(value) 55 61 58 56 49 48 ...... are net rice sellers (quant.) 38 29 38 37 43 41 are net rice sellers (value) 36 27 35 34 42 42 Source: Mitten and Colleagues calculations based on INSTAT-DSM,EPM2001 *: Q1: poorest quintile; Q5: richest quintile The most important point ofthis table is that the poorest segments ofthe rural populationare net buyersof rice and would likely not benefitfrom increasedrice pricesunless they were only experiencedduringthe main harvest(and sale) period. The impact of seasonality is very strong in some regionsandthere is a large difference betweenpoor and non-poor focus groups intheir perceptionofthe benefits of alteredrice prices. 65 Table 3.14: Perceived effect of effect of rice prices (qualitative evaluation by communal focus groups) Proportionof the populationthat... Poor Non-poor ... prefers a higher rice price during the 63% 61% harvest period ... prefers a higher rice price during the lean 16% 32% period ... uses rice as the major staple during the 48% 68% lean period ... uses cassava as the major staple during the 56% 3 1% lean period ... depends on imported rice during the 24% 21% harvest period ... depends on imported rice during the lean 56% 43% period Source: Post-crisissurvey, ILOprogram, Come11University,NovembedDecember2002 Increasing Competition among DownstreamOperators. A total o f 523,000 tons o f domestically produced rice or 28% o f total domestic production is effectively traded. Taking into account annual rice imports o f about 186,000 tons (1999), imported rice accounts for about 27% o f all traded rice in Madagascar. To collect paddy and rice from rural households, there are over 4,700 collectors that operate inthe country. The major clients o fthe collectors are the millers who take 63% o f the collected quantity, wholesalers purchase 35%, while 2% is directly delivered to final consumers. There are about 2,100 small rice millers who handle between 212 tons (East) and 1,104 tons of paddy (Lake Alaotra) per year. Wholesalers, who are usually located in urban areas, have an important role in linkingregional supplyand demand, facilitating interregional transfers o f domestically produced rice and supplyingthe domestic market with imported rice. It is estimated that there are about 100 wholesalers who are involved in rice trading with an average of 5,000 - 10,000 tons o f rice sales per wholesaler. Wholesalers fall into two categories. Those from rice surplus regions are more involved innetworking with upstream collectors and operators, while those from rice deficit regionstend to be heavily involved in imported rice trading. However, recent trends show a growing involvement o f wholesalers from surplusareas getting involved in trading of imported rice as these supplies are easier to manage. Retailers in Madagascar sell 710,000 tons o f rice yearly. With average rice sales per retailer varying from 8 to 58 tons per year, it is estimated that there are about 23,800 retailers in the whole country. The aggregate storage capacity at the retail level is estimated at around 56,000 tons, which appears low compared to the total turnover of 710,000 tons. On-going liberalization o f the rice market i s effectively reducingtrade margins o f collectors and wholesalers, reflecting increased competitiveness o f local rice markets. However, lack of market integration continues to cause wide geographical and seasonal price fluctuations due to a variety of reasons. First, the process of market liberalization is only progressing slowly as almost 50% o f rice producing households still remain faced with a situation o f permanent or occasional monopsonies. Second, the dismal status of the rural roads network implies excessive transportation costs, thereby significantly increasingtransaction costs o f trading. Third, market information and intelligence is poor, thereby hamperingthe identification o ftrading opportunities. To increase competition inthe market place requires a number o f actions on different fronts. First, there is an urgent needto accelerate improvement and expansion o f the roads infrastructure 66 particularly benefit from improving the following connections: (i)RN44 Moramanga - Lac in rural areas. Feedback from downstream operators suggests that the rice subsector would Alaotra: (ii)Toliara - Morombe, (iii)Sofia - Bealanana access road, (iv) Mainiarivo access. Second. there is a need to facilitate entrance o f new downstream operators inthe market. With lack o f credit to finance working capital being a major constraint for potential new entrants, the establishment o f a Guarantee Fund should be considered in order to facilitate access to bank credit. Third, to improve availability of market information and intelligence, there i s a need to significantly strengthen and expand existinginformation collection and dissemination capacity. Fourth, enforcement o f tariffs and taxes on imported rice, with its corresponding impact on relative prices o f domestic and imported rice, will provide incentives for wholesalers to refocus their operations on domestic rice, thereby contributing to increased competitiveness o fthe domestic market. Local Storape Capacity. As market integration is progressing only slowly and will take considerable time to be fully achieved, especially considering the significant challenges o f rural road network improvement and expansion ,there is a need to equip farmers to better deal with the current reality o f imperfect markets and position them to take advantage o f the observed significant rice price fluctuations. Inthis context, the experience of the GCVs is highly relevant. Under these communal storage schemes, farmers can extend the paddy selling period to , collectors, thereby benefiting from higher prices that normally occur inthe lean period. By doing so, they could typically boost sales revenues by about 30%. At the same time, credit provided by rural micro-finance institutions secured by stored paddy, allows farmers to meet immediate . consumptionneeds as well as finance working capital for the next production campaign, thereby contributing to addressing the existing credit constraint in rural areas. Currently, GCVs cover about 24,000 tons o f paddy or lessthan 5% of the existing paddy trade flow in the country. Taking into account GCV's capacity to simultaneously contribute in dealingwith two critical constraints in rice production (ease impact of imperfect markets, enable access to credit), there is a need to expand the application of the concept in the country. This would require investments to establish paddy storage capacity at the community level; something which perhaps could be pursued more aggressively under the implementation o fthe Rural Development Action Plan (PADR) through the Bank-financed Rural Development Support Project (PSDR). 3.2 Cassava From a food security perspective, cassava i s a special case that deserves special attention. It is the second most important source o f calories o f the Malagasy population after rice. At the national level it represents 14% of calories consumed; it is particularly important inthe South where it provides 27% o f all consumed calories. It is attractive for poor people as costkalorie from cassava i s only half of the costkalorie from rice. Consumption o f cassava increases during the lean season as the increase inthe price o f rice forces poor households to switch to relatively cheaper cassava (Table 3.I 5). The latter i s facilitated by the fact that the seasonalprice fluctuations o f cassava are not synchronizedwith those o f rice (Dostie et al., 1999). Dryingo f cassavatakes place inthe dry period before the lean seasonwhich allow collectors to stock-up and start selling when demand increases duringthe lean season. At the same time, fresh cassava can be kept in the ground for long period o f times, which reduces price fluctuations. 67 Table 3.15: Relative caloric share per product and per region Region 1I Total consumed calories 1 Change in consumptionfrom harvestto leanperiod Harvest 1 Lean Deriod I Rice 1 Other cereals 1 Tubers Fianrantsoalhighlands 2738 I2357 -14% 1Yo 20% Ranomafana 2613 12353 -1 1% 1Yo II9% MaGngahighlands 3240 I3172 -7% 4% 2% Maiunaarest I1 2906 I2782 1I-10% II10% 1I-6% Source : Mintenand Zeller (1998) In terms of food security, these characteristicsmake cassavaan interesting crop as some sort of buffer against increasingfood prices or falling income. At the same time, it has been shown that cassava is an economic inferior good, with an income elasticity of-0.88, implyingthat overall consumption decreases with rising incomes. Consequently, from a commercial/marketing perspective, investment in cassava production is not a very interestingproposition inthe long run, thereby potentially undermining its benefits as abuffercrop in the short runbecause of its limited prospects. In effect, national cassava production has been stagnantin 1990sand production in 2000 at 2.2 million tons was in fact the lowest since 1991. Yields have been stagnant at most. Previously attractive export outlets have been effectively wiped out due to falling world market prices. The loss of export markets is to some extent compensatedby increasingdomestic demand for cassava as animal feed. The central question then is how cassava could continue to play its role as a calorie provider of last resort, despite its limited financial prospects inthe long run. This question is even more relevant following the collapse of the export market, which implies that demand for cassava even more than before depends on unpredictableevents such as loss of production due to cyclones or locust problems, delays ingetting supplies of rice or an extraordinary long leanperiod due to late harvest of domestic rice following adverseweather conditions. Possibleresponses to this question include the following (adapted from Dostie et al. 1999): (i) Cassava ifproperly dried keeps well for about eight months, which facilitates long- distance transport and inter-regional trade. However, dryingtechniques are not optimal causing serious mouldproblems. Improvement indryingtechniques could allow better conservationof cassava, thereby reducing currently significant post- harvest lossesthat are estimated inthe order of 1520%. (ii) Cassava yields are low and there is a needto explore crop intensification opportunities. Given its importance for food security reasons, cassava should be an integral element of any agricultural researchagendathat seeks to be pro-poor. (iii) Changing eating habits ofa growing urban population could offer opportunities for cassavaprocessingand the establishment of linkagesbetweenthe sector and agribusiness, thereby providing a more stable outlet for cassava production. (iv) Iftheobjective would beto integrateafood security element intothe country's road construction and road repair program priority setting, including isolated cassava production and consumption regions would go a long way inachieving this. 4.2 Small Scale Fisheries& Aquaculture Two key points need to be made before a consideration ofthe role of marine fisheries in rural developmentand poverty alleviation: 1. Madagascar is surroundedby tropical seas of relatively low productivity. As a useful comparison, the artisanalfishery of Senegal, West Africa, with an annual catch of 400,000 tons 68 along 750 kin of coast, is roughly eight times the combinedtraditional and artisanal fisheries catch of Madagascar(estimated at 50,000 tons along 5000 km of coast). While this large difference is in part due to the stronger fishing tradition of Senegal and the smaller range of small scale fishing in Madagascar, it serves to illustrate that Madagascar's marine fisheries are relatively unproductive (although high in species diversity) and can only ever offer a modest contribution to improved rural productivity and food security. 2. The marine resourcesof Madagascarare also relatively poorly known. Much ofthe optimism that has driven fisheries expansion is based on the assertionthat sustainable harvestsare estimated at around 300,000 tons per year (Andrianaivojaona et al 1992) while actual harvests are only around 125,000 tons. However, this position ignoresthe facts that most of the estimated potential 300,000 tons includes stocks that are little known and not yet exploited and that most readily accessible resourcesare already exploited at or beyondsustainable limits. Untilbetter information is available on the state and potential yields ofMadagascar's marine resources, a precautionaryapproach should be appliedto any plans for development. I Marine small scale fisheries, known as `traditional', are based on dugout canoes and fishing on foot. The number of canoes expandedfrom about 5,000 in 1980to about 25,000 in 1996, and has since declined to around20,000 (Table 3.16). Assuming an average householdsize of 6 or 7, small scale fisheries provide a livelihood for around250,000 people. Traditional fisheries target all edible or saleableresourcesin shallow coastalwaters but are unevenly distributed along the coast as a function of population density and navigable sea conditions. The major small-scale fishing centres are Toliara, Morondava, MahajangaandNosy Be, all on the west coast. Inthese areas, small scale fisheries are vital sources of employment, food security and good nutrition, especially for Toliara which is poor in most agricultural resources. Small scale fisheries supply a growing national market for marine produce as well as several products for export (sharks' fins, sea cucumbers, octopus and some ofthe higher quality finfish. The robust price of fish on local markets combined with the exportablecomponentoftheir catch ensures that fishermen escape from being within the poorestsegment ofthe population but yieldsper fisher (average no more than 5 kg per day's outing), and therefore fishing incomesare still low (in the order of $1000 annually allowing for low seasons and inevitable lost days). Table 3.16: Distributionof Traditional Coastal Fishermen by Province I Province I NumberofFishing I Numberof 1 Numberof 1 Villages Fishermen Dugout Canoes Antsiranana 370 6,000 5,000 Fianarantsoa 80 2,400 1,500 Mahaianca 350 7.000 3.500 1 Toamasina 1 200 I 4.000 I 3.400 I Toliara 300 21,000 6,500 Total 1,300 40,400 19,900 Production ofthe traditional fishery has been partially monitored since the early 1990s. Export data for high value export products such as lobster, sea cucumber and shark-fin may provide a better indication of production trends than low value catch since so much of the latter is locally consumed. The traditional finfish catchhas beenofficially estimatedat 50,000 t for most years since 1990but this is an extrapolation basedon daily yields last estimated in about 1990, with a large potential margin of error. Export data for the period 1997-2001are presented inTable 3.17. Overall, the traditional fishing sector appears to be in stagnation or even decline ifone considers the number of canoes and the trends in key export products (trepang, lobster & shark fin). The 69 increasing catch of cephalopods (squid & octopus) is an indication o f a degraded fishery from which high grade fish have been extracted and the maintained catches o f trepang are due to the advent o f industrial-scale exploitation illegally using scuba gear, local stocks having been exhausted. It may be surmised that the traditional fishery has reached or exceeded sustainable yields within its current range o f operations. This meansthat it is not a strong candidate for reducing poverty or improving food security. Table 3.17: ChangesinFishProductExuorts 1997-2001 Source: Rakotozanany 2002 Editor's note: official fisheries statistics are not subject to independentevaluationand mustbe treatedwith caution. For example, the productionfigure for shrimp exceeds industryestimates. Potential for expanded production therefore exists only inareas o f low existing fishing activity, for certain less exploited products such as mangrove crab and for offshore and demersal species not currently accessible by canoe (e.g. Adrianaivojaona et a1 1992). However, it should be noted that the demersal offshore fisheries (Le. bottom fishing on the continental shelf for sharks and `noble' fish species) are the most vulnerable to over-exploitation and need to be strictly managed. In degraded coral reef areas, there is an urgent need for restorative management based on no-take zones and gear controls. Small-scale marine aquaculture (`mariculture') that involves local people is virtually non-existent in Madagascar. A project to develop seaweedfarming north o fToliara which began well inthe early 90s has in recent years regressedto the collection o f wild seaweed, due to social and organizational factors. Shrimpaquaculture, while havingan excellent environmental record, is exclusively industrial and situated in less populated regions and does not specifically employ ex- fishermen. Experiments with the ranching o f sea-cucumber, urchins and mollusks are technically promising but no projects have yet been developed to scale these up commercially. 70 Scaling-up challenges As noted above, the principal challenge to scaling up arethe facts that: 1) marine fisheries resources in Madagascar have limited productivity - most nearshore fisheries are fished to or beyond sustainable levels; 2) marine resources and their potential are poorly known and 3) the traditional fishery has a very limited geographical range with the result that nearshore resources are intensively exploited, damaging the recruitment o f certain resources further offshore. For several over-exploited stocks, such as shark, sea cucumber and reeffisheries, the challenge is more to restore depleted stocks to former levels and to ensure their sustainable exploitation, rather than scale them up. The traditional marine fisheries sector is remarkable for the lack o f development beyond sail and oar powered canoes (capable o f annual catches o f 1-2 t per canoe) to larger motorized vessels (`artisanal fishing') based on small teams o f fishermen (capable o f annual catches o f upto 10t per vessel). There are less than 100 motorized small fishing boats in all Madagascar. Projectsto promote motorized artisanal fishing have virtually all failed at the technical and socio- organizational level. The reasons are various, but includethe highly individualist nature o f fishermen`s strategies, general illiteracy o f fishing communities, lack o f technical capacity and inability to obtain credit. The transition to motorized commercially-orientated artisanal fishing fleets tapping more offshore resources requires many socio-economic changes and may require a generation to accomplish. Inthe medium term, small scale fisheries face a crisis due to the lack of seaworthy canoes, due to the reduced availability o ftrees o f sufficient size. This is leadingto the construction o f smaller and less seaworthy canoes with smaller crews and fishing range resulting inever increasing pressures on near-shore resources. There is an urgent need for affordable, seaworthy canoes to support a transitional phase o f the small-scale fishery and available credit to purchase boats and gear. Improved canoes would also serve to improve fishermen's access to offshore resources. The current opportunities for scaling-up small scale marine fisheries are either geographical i.e. range extension or expanding into unfished sectors o f coast or based on improved management aimed at getting improved productivity out o fthe fishery while conserving the resource base. Substantial stretches o f the west coast are subject to low fishingpressure and could support the development of new fishing centres. However, such development i s constrained by the lack o f roads and storage facilities or the local demand and social context generally required to support a small-scale fishery. Fisheries on the east coast are generally very little exploited but are also less productive than fisheries o f the west coast. For the east coast, rough sea conditions are the principal limiting factor where expansion o fthe fishery would depend on the adoption o f larger motorized vessels. Lack o f credit and perhaps the general lack o f fishingtradition are likely constraints on the east coast. Most o f the high value marine resources are already over-exploited within the accessible range o f fishing canoes, including lobster, sharks, mollusks, finfishand sea cucumber. Some o fthese fisheries are suitable for community-based management and here a GELOSE initiative has been proposed for the principal lobster fishery at Fort Dauphin (although this now needs funding and technical support). The shark fishery requires a combination o f assessments, policies and trade measures at the national level and the implementation o f participatory regional assessmentsand plans at the local level (following internationally established guidelines for shark fisheries & trade). Near-shore sea-cucumber resources should fall under local community based management whereas the offshore industrial sea-cucumber fishery requires firm management and surveillance 71 at national and regional levels. Policy and trade measures are also required for sea-cucumber at the national level. Shrimp fisheries are a special case, since here there is an issue of direct conflict with industrial fisheries. The share of the catch by small scale fisheries has been increasing, but represents an inefficient use of the resource since it targetsjuvenile forms and the poor sanitary conditions at village landing sites prevent export o f the products to the high value markets (e.g. Europe). A specific project has been proposed by GAPCM/AFD to address these problems through the establishment o f special shrimp fishery management zones (`Zones d'Amenagement Concerte' or ZAC), one o f the objectives of which would be to help small scale fishermen derive a higher value from their catch through better sanitary conditions and better accessto markets while limitingthe use o f unsustainable practices such as the harvestingo fjuveniles. Opportunities for innovative approaches The major opportunity for innovation is to promote locally-basedfisheries management based on a system o f use rights(potentially based on the GELOSE law) within a context o f integrated coastal zone management (ICZM) or participatory management. A useful management tool to accompany this approach for coral reefareas is the establishment o f `no-take' or `set-aside' zones which help to ensure maintenance o f ecosystem processesand the reproduction o f target species (particularly in coral reefareas). Integratedmanagement approaches have been piloted for Toliara and Nosy Be (Ocean Consultant 2001a & b) and the Baie d'Antongil but should be optimized and extended to many other areas. Important opportunities also exist for mariculture o f usefulor valuable species, especially for species for which there is a local or national demand as food such as mollusks that cannot readily be met by wild-caught produce. There are also opportunities for the culture or ranching o f specialist foods for export such as sea-cucumber. Farming sea fish in the coastal zone is not widely recommended since it is more technically demanding and environmentally risky than freshwater fish farming, with which it could not readily compete. Artisanal shrimp farming should also not be encouraged since it would certainly fail to meet stringent hygiene standards and would probably lead to severe loss o f mangrove and add to the risk o f disease inthe industrial aquaculture sector. Finally, opportunities may exist for farming specialist products such as sponges or the ranching o f post-larval fish collected from coral reeffronts, butthe feasibility for these has yet to be tested. 72 ANNEX 4. Sustainable Intensification & Diversificationof Agricultural Production After having laid out anumberof optionsto improvefood productionand food security in Madagascar as an importantfirst step to improverural productivity, this chapter takes acloser look at the building blocks for sustainable intensificationand diversification of agricultural productionsystems as an engine for sustainedrural productivity growth. Inthis context, the following factors are reviewed in more detail: (i) water management; (ii)technology; (iii)land and landtenure; (iv) credit; and (v) road infrastructure. 4.1 Water Management With about 1 million hectareso fMadagascar's cultivatedarea under some sort of irrigation, water management is akey factor affectingsector performanceand it is difficult to imaginethat more inclusive rural growth can be achieved without improvingthe performanceof irrigated agriculture.Despitethis tremendous potential, past performanceof irrigatedagriculturein Madagascar has beendisappointing. The overall image is one of stagnation where irrigated I productivity levels are not very different from rainfedpaddyproductivity levels, and where investmentsin irrigation have not providedan incentivefor agricultural intensification. Yet, researchdata indicatethat the story of irrigatedagriculture inMadagascar is not one of low potential. A comparison ofthe resultson experimentalresearchstationsinMadagascar with other countries that have achieved higher average yields indicatesthat the yield gap (field level yields as a percentage of maximumattainableyield under experimentalconditions) in Madagascar is much wider than elsewhere. A yield gap of 25% is reportedby FOFIFA8,with resultsunder experimentalconditions exceeding 8 ton/ha. By contrast, the yield gap in West Java is 80% with similar maximumattainableyield levels.Thus, the differencebetweenMadagascar and other countries is not relatedto potentialpaddy yield levels, but to actual field productivity levels. Improvedwater management embodies the promiseofachievingthese higher yield levelsthrough the use of Green Revolutiontechnologies.Madagascar's 1million haunder irrigation, the second largest in Africa, providesan excellent startingpoint for agricultural intensification.This section will address the questionwhat has to be done to finally unleashthe existing potential of irrigated agriculture in Madagascar.Itwill look at experience from past investmentprograms, will identify constraintsto the sustainability of irrigatedagriculture, and will propose specific actions to boost the productivity of irrigatedrice. Outcomes of Past Interventions In light of past interventionsand investments, the key questionthen is why water management continues to be a problemand why improvedwater managementhas not resultedin agricultural intensification, higher productivity levelsand sustainable O&M. An analysis of past interventions show very mixed resultsand point towards a history of overemphasizingconstruction and a simultaneousneglectof institutional issues. Ratesofreturn have generally been low, and sustainabilityunlikely. The SAP ofthe PPI-2 project (1995)provides a summary: `Experiencewith lendingto the irrigation sub-sector inMadagascar has been disappointing. The first IDA-financedoperationwas designedto improve irrigation and drainageof some 8Structureand conduct ofmajor agriculturalinput and output marketsandresponseto reforms byrural householdsinMadagascar.FOFIFA/IFPRI, 1998. 73 12,000 ha in the Lac Alaotra area. The projectwas carriedout between 1970and 1976. The post evaluationimpact ofthe projectwas initially consideredsatisfactory,basedon satisfactorycompletionof physicalworks, but a later review concludedthat the projecthad failedto meet its fundamental goals, giventhe lack of sustainable impacton productionor farmer incomes.Poor maintenance, ineffectivenessofthe extension services and environmentalproblems resultingfrom excessivedrainage were cited as reasonsfor poor performance. The projecthada negative rate ofreturn. This projectwas followed by the Morondavairrigationand rural development projectwhichwas implementedbetween 1974 and 1982. Benefitswere low andthe project'srate of returnwas zero or negative. A later phaseof irrigationdevelopmenton Lac Alaotra followed between 1984and 1991, with co- financingfrom CFD and covering35,000 hamanagedbythe former parastatal SOMALAC. This projectwas rated satisfactory, havingsuccessfully achieved its physicalobjectives, with paddy yields and productionincreasingsignificantly. However,sustainabilitywas not assured because of uncertaintyregardingmaintenancearrangements. A fourthproject in support of irrigationrehabilitation,the Mangokyagriculturaldevelopment project, was implementedbetween 1979and 1986. Cropyields andprojectbenefitswere considerably lowerthan anticipatedand once again the projectterminatedwith a negativerateof return. The ex-postimpactevaluationrecommended, among other things, the needfor private sector involvementand closecollaborationwith water users andwater userassociations (WAS).' Seven years of projectimplementationlater, the ICRofthe PPI-2(2002) strikesasimilar chord: `The extentto which W A Swere able to organize and conduct O&M onthe rehabilitated infrastructureis of key importanceto achievingsustainableresults. This sustainabilityis at risk if,duringprojectimplementation,paymentof irrigationwater fees is not firmly and lastinglyestablishedat the 100%level. It is difficult to see how recoveryof O&M fees would regaina sustainable levelafter project completion, ifrecoveryis significantlybelow that levelduring implementation.' Recoveryof O&M fees scoredgenerally low throughout project implementation,with only few positiveexceptions for a smallnumber of schemes. The total financialcontributionis on average inthe order ofUS$6per hectare, excluding manuallyexecutedmaintenance. The nominallevelofthe fees (as comparedto similar projectselsewhere(e.g., Mali: US$85 per hectare), and as comparedto the benefitsthat beneficiariesderivefrom the investments) reinforcesthis argument. Inaddition, proper O&M was estimatedto cost betweenUS$23 andUS$38 per hectare(althoughsome questions have been raisedwhether this has beenestimated correctly). With an average fee of US$6 per hectare, it is evidentthat continuedgovernment support of maintenancewill be unavoidable,even iffarmers pay 100%. The ICR furthermorerecommendsthat `irrigation projectsshould be part of a multi-sectoralapproach, integratedinto watersheds management programs and includinglandtenure, credit, transport, accessto marketing, inputsupply and extension'. Several key issues affectingsector performancecan be identified. They are presented on a national.watershed and scheme level. National Level Promote coherentstrategicsupport to the sector. Support to irrigatedagriculture in Madagascar has oscillatedbetweendirect involvementofthe State inoperationand maintenanceduringthe 70's and 80's and lower levels of State involvementduringthe 90's. Agricultural PIP investment levels(as percentageof overallPIP) are presentedinFigure4.1, as well as the contributionof agricultureto overall GDP. The data show declininginvestmentlevelsthat are notcommensurate with the nationalimportanceof agriculture. 74 Recent GoM policy has been characterized by Irrigation Management Transfer(IMT), an approach inspired by the objective to reduce high levelsof public investmentsin irrigation schemesthus releasingscarce funds to support other services. GoM expendituresfor O&M decreased from 50% of the budget of the Ministry of Agriculture in 1985 to 285M Fmg (US$42,500) and an undeterminedallocation for salaries of staff still employed on `perimetres classes` at present. At the same time, 36 schemes have beentransferred, or 8,607 ha out of 270,000 ha of public irrigation, slightly more than 3%. IMT in Madagascarthus had a major impact on the GoM`s O&M budget, which is however hardly paralleledby a reductionofthe number of `perimetresclasses'. This situation is prone to lead to serious confusionon who is responsiblefor scheme O&M, and a formal confirmationofthe defacto situation is urgently required.Overnightor `big bang' IMT has some successful internationalprecedence in Mali and Andhra Pradesh. What is more, state withdrawal appears to be adriving force behindIMT, ratherthanthe need to redefine rolesand responsibilitiesof stakeholders inthe sector, to create a moredynamic rice sub- sector and to facilitate involvement and investmentofthe private sector. While defacto IMT can be done overnight, support to the sector needsto be thoughtthroughmore carefully. A comprehensive sector strategy with aclear identificationof mutualroles and responsibilitiesof stakeholders, includingthe responsibilityfor O&M is urgently required. Absence of such strategy is putting Madagascar's irrigationpatrimonyat risk, as evidencedby Andapa, where some 8,000 ha appear to have been lost, amongothers due to failure to undertakemaintenance works. Figure4.1: Agriculture's contributionto GDP,and investmentsin agricultureas percentage of overallinvestments. Ag GDP and Ag PIP 45 40 35 30 25 20 15 10 5 0 year 75 Watershed Level One ofthe reasonsofpast poor performanceof irrigatedagriculture inMadagascar is relatedto the systematic failure to address irrigationfrom a holistic,multi-sectoral(`watershed') perspective.Improvedwater managementis one amongmany constraintsthat havean impacton productivity.When addressed in isolation, noneofthe mono-sectoralprograms, includingthose that target irrigation infrastructureimprovement, will achievetheir objectives. Watersheds impacton irrigationschemes(and vice-versa) in many differentways. One ofthe most visible ways is the impactof environmentallyunsustainablepracticesonthe slopes ofthe watershed, the development oflavakas andthe subsequent loss of productivelanddownstream due to sedimentation. Another example is the fact that upstreamwater abstractions for and releases from irrigationhavean impact on water quantityand quality further downstream inthe watershed. Inadoptinga watershed approach, the size ofthe schemeshas beenshownto bean important criterionfor determiningits relationshipwith the watershed and for identifyingthe kind of activitiesthat are neededtofacilitate sustainable agriculturalintensification. Insmall-scale irrigation, located in smaller sizedwatersheds, a same farmer typically owns landinboth lowlandsand slopes. Effortsto improve landmanagementpractices(e.g. erosioncontrolby terracingor plantingof vetiver) cantherefore focus on onetarget group, as some ofthe benefits from investmentinerosion controlwill beharvestedinthe downstream irrigation scheme. In larger sized watersheds, however, such linkagesoftendo notexist, and farmers inthe upper parts ofthe catchment needto derive direct benefitsfrom investments insoil protection.In larger-scale catchments,.thechallenge is thus to identifyactivitiesthat, amongotherbenefits, increase productionandprotectthe soils. Alternatively, river basinmanagementprojectshaveundertaken efforts to invest in upstreamcatchment protectionbytaxingdownstreamwater users. Poverty is at the center oftop- andtail-endlinkages. There is evidencethat inparticularpoor people are forced intomarginalizedfragileupper catchment where unsustainable land management practices( e g slash-and-burn or tavy)cause erosionand landdegradation, andthat the resultingsedimentationdegradesand causesfloodinginthe lower catchments, where in particular poor and marginalizedpeople are forcedto make a living. An analysisofEPM2001 data showed a significanteffect of sedimentation on rice yields. With poverty of such crucial importance, holistic approachesto sustainable agriculturalintensificationwithin awatershed context needto address povertyreductionas a priority issue. These include improving performanceof irrigatedagricultureso as to sustain a largernumber ofpeople. Italso includes promotionof sustainable landmanagementon hill slopes, which is the focusofthe technology section ofthis chapter. Scheme Level Increasingthe low levelsofproductivityin irrigatedagricultureis keyto transformingthe sector into an engine for moreinclusiverural growth. Inaddition, higherproductivity levelsare keyto ensuringthe environmentalsustainabilitywithin a watershed perspective. Inview ofthe multi- dimensionalcharacter ofthe problems, andcontraryto what was done inthe past, a holistic agenda needsto be adopted, in whichthe central challenge is to providesufficient incentivesfor farmers to adopt use of fertilizersand highyielding varieties, andto invest inimprovingwater management and maintainingirrigationinfrastructure. The following outlinesanagendafor increasingproductivity levels. 76 Increase cropping intensitv. Irrigation infrastructure is still very much used exclusively for wet season rice production. Only few farmers have ventured into second season or `contre-saison' production o f alternative non-rice crops. To the extent that water availability i s a constraining factor, the introduction and promotion of affordable irrigation technologies should be considered. Inall of the sites that were visited, groundwater tables appear to be favorable (i.e., less than 7m below field level). Experience (e.g. Niger, Kenya, Bangladesh) demonstrates that the impact o f such programs can be significant. Promotion o f affordable irrigation technologies involves the establishment o f sustainable manufacturing, supply and marketing chains, including training o f local blacksmiths for manufacturing and repair o f pumps, village stockists for sale o f pumps and spare parts, and identification o f market opportunities for vegetables and niche crops. Additional advantages of increasingproductivity by increasingthe cropping intensity are the more common use of fertilizer in non-rice crops, the spillover effects to wet seasonrice production, and the fact that product chains for some crops (e.g. potatoes) show a better vertical integration with traders proposingproduction support in terms o f extension, credit and inputs. Improve access and securihr. Since hightransportation costs due to the poor rural roads network imply a heavy tax on the use o f productivity enhancing inputsas well as the marketingo f output, public sector investments in the rural roads networks are considered a highpriority. Evidence has been presented in the previous chapter that suggest that access to roads is a crucial factor inthe intensification o f agriculture. Efforts to increase production therefore needto target those areas where road projects (PST, FID)have completed road works. Improvingaccess, however, is not only related to roads. It also involves making sure that at the end of the road sufficient quantities o f produce are available tojustify hightransport expenses, e.g. by grouping produce in storage rooms. Relatedto this is improved communication between local traders and urban markets. In addition, and to the extent that remoteness and insecurity are directly correlated, improving security may have the same effect as road rehabilitation. Support vertical integration o f supply and marketing chains. Before liberalization, downstream operators (transporters, collectors, millers, wholesalers and retailers) were often regarded with suspicion for taking advantage o f their monopolies, and often selling rice duringoff-season to the same farmers whom they bought from at harvest time, for abusive profits. With liberalization, reduced trade margins of collectors and wholesalers reflect increased competitivenessof local rice markets. Gross margins for paddy collectors amount to 13%, wholesalers to 10%and retailers to 8%9. These figures suggest that, on average, there does not appear to be much scope for further reducing operators' margins by promoting competition. Yet, market imperfections and large disparities between regions due to access constraints remain, as evidenced by large price fluctuations between regions and seasons. Where this is the case, there seems to be scope for targeted interventions to improve access, increase competition, and empower farmers. In all other places, collaboration and integration between farmers and downstream operators should become key elements o f any approach to increase production and establish a more dynamic sector. An example o fthis is the USAIDfunded LDIproject that has supported the signingo f contracts between farmer groups and rice traders inthe Lac Alaotra region. Interestingly, the support responds to increased competition that followed liberalization o f rice`trade, the emergency o f many small agro-processors and traders, and the deliberate strategies that some o f the larger traders now employ to ensure their market share, includingthe signingo f production contracts with farmer groups, more focus on production from own land, and importingrice. This model should be further explored, e.g., by construction o f storage facilities for farmer groups, provision 9 Bockel, ibid. 77 of guarantee funds for private banksto facilitate traders signing production contractswith farmer groups for the provision of production credit, inputs, tools/equipment and/or extension in exchangefor exclusive sale of produce", and supportingfarmer groups in selling their produce by public tender to the highest bidder, in exchange for production credit. Renting out of small- scale affordable irrigation technologies by agro-processorsto producer organizations may be another interesting option. Other possibilities include supportto agro-processors (sugarcane, cotton) to establishcontractualrelationshipswith outgrowers, and exploring opportunities for expansionof this model into other crops. Tomatoes and potatoesare examplesof crops where some form of vertical integration is taking place. Specifically, promotion of high value niche export crops (e.g., organic rice) should be considered, and targetedsupport could be provided to bring producer organizations, exporters and international demandtogether. In general, the aim would be to tap the synergiesin the sector from all those interestedin production increases, to analyze and identify their constraintsto support more directly intensification, and to provide targeted support to remove some ofthese constraints, preferably through existing channels such as private commercial banks. A more aggressiveapproach would be to attract and promote new investmentswith the above support as billboard. Among the advantagesare most significantly: (i)the provision of credit to farmer groupswho do not have access to formal credit; (ii)improved access to productive technologies, including fertilizer and extension; (iii)promotion of crop diversification; and (iv) creation of off-farm employment in agro-processingindustries. Improve the managementof irrigation schemes. Establishmentand strengtheningof WUAs is the foundation of any approachthat aims to improve the performanceand sustainability of irrigated agriculture. There is an increasingly relevant body of literature relatedto the questionof establishing and training of WUAs. Much of this has, however, been written for an audience of smaller scale irrigation schemes and is not a priori suitablefor larger scale schemes. New approaches are required that build on the muchhigher level of complexity of large-scalewater managementand O&M, find credible responsesto the challengesrelatedto the involvement ofthousands, ifnot tens ofthousands of farmers, and use the opportunity ofthe potential larger financial clout o f larger scale schemes. It is unrealistic to expect farmers to be able to operate and maintain close to 20,000ha (case of Marovoay). These new approachesshouldtake their inspiration from business- like operations, as opposedto the often family-style managementapproachesthat have shown to work on smaller schemes. Key elements of such approach are: 0 Rather than farmer representativesbeingresponsiblefor daily operations, professional executive managementis required to implement the annual work plan, as agreed by the general assembly. Farmers are represented in a Boardthat would meet only once or twice a year. The Executive Director is recruited by and accountableto the board, and prepares detailed implementation plans, prepares backgrounddocumentation for the general assembly, allocates water, launchestenders for O&M work, and ensures quality. 0 In an alternative model, a private company replacesthe Executive Office. The WUA (or a Federationof WUAs) puts out a bid for O&M service provision, including a bonus for good performanceas evidenced by clear indicators. The State could initially contribute, but would reduce its contribution over time. In addition, and ifthis is deemednecessaryto attract loPrivate commercial bankshavenot beenwilling to providecreditevento the largestplayers, andhave insistedon a 100%guarantee.It appears that last year's rice importanarchy is relatedto this, as confirmed by the fact that banks do providecredit for importingrice. 78 private service providers, the State could guarantee the contract up to a certain percentage. This and other options" could be further exploredwith IFC. Feedback links need to be establishedbetweenthe ones who pay O&M fees andthe ones who provide O&M services, so that quality of service provides an incentive for payment, and payment o f fees providesan incentive for quality. These linkages should be strengthened through performance-basedcontracts. Cost efficiency considerationsneed to permeate management.Board and Executive Director need to be responsiveto the fact that the budget consists of farmers' contributions. At the same time, fee collection needsto be professionalized,with a clear incentive structure for those who are responsiblefor collection. Key managementperformance indicatorsneed to be identified, and a monitoring and evaluation system needsto be set up. An Administrative Manual needsto be prepared, outlining rights and obligations of staff. Improve O&M fee recovery rates. Realistic assessment of O&M fees, and their full recovery is key to sustainable irrigated agriculture. The analysis in this section intendsto provide some guidance on realistic levels of O&M fees, as well as on ways to improve recovery. Many schemes have only recently startedto collect O&M fees and little is known about the required level. One ofthe exceptions is the PClSMarianina scheme in Lac Alaotra, where fees have been collected for over 10years, and where fees have beenamended eachyear, suggesting that more realistic levels have been reached. The actual average fee (Fmg 64,00O/ha) is five times higher than the original fee in 1993, and twice as high as many small-scale schemes. The following analysis will therefore be based on PC1Smarianina. A comparison with other countries indicatesthat the O&M fees paid inPC1SMarianina are still at the bottom end (see table 4.112). Annual Total Output O&M irrigation Output per O&M per unit costs as water supply unit cost per irrigated % of per unit Cost irrigation unit area area output per irrigated area recovery supply (US$/ha) (US$ka) ha (m3/ha) ratio (us$/m3) Australia Jemalong, 16.69 NA NA 2892.89 1.17 NA Coleambally 19.13 655 2.92 5877.78 1.14 0.11142 Murrumbidgee 17.33 142 12.20 3519.79 1.17 0.40343 I West Corugan 38.84 280 13.88 3427.67 1.14 0.08164 Western I `See e.g. http:llwww.inpim.orglLibrary/NewslettersRl IInl1.html 12 From: Emergingexperiencesof benchmarkingin the irrigationand drainage sector (draft). Fernando Gonzalez and Arum Kandiah, The World Bank, 2002. 79 Morna 10.26 587 1.75 6560 0.3 0.08956 Waghad 40.96 1619 2.53 7270 0.09 0.22587 Rio Yaqui ID 1Mexico 041 45 888 5.07 8342 1 0.183 Module 06 46.6 1090 4.28 6509 1 0.19 Module 10A 54.7 I090 5.02 6937 1 1.09 Module 10B 55.4 1088 5.09 8089 1 0.22 Module K79 Module P4 - 63.9 1554 4.11 7632 1 0.25 CPB 56.7 1868 3.04 8988 1 0.2 Morocco Loukkos 435.2 2598 16.75 55 11.23 0.42 0.6 Actual O&M fee payment depends on the capacity and willingness to pay. Capacity to pay depends on the O&M fee as percentageof the output per hectare. This is supportedby the data in table 4.1 that show a significant correlation (0.75) between fee rate and output per ha. With Lac Alaotra among the lowest globally (at 1.7% of output per ha), there is ample reasonto support the claim that capacity to pay should not be a problem. InMali, e.g., where rice prices are similar to those in Madagascar, closeto 15%of output/ha is collected as O&M fee. Assuming an average productivity of 4.5 tons/ha, and maintaining 1.67% ofthe output per ha, an O&M fee based on capacity to pay of over US$] 1 should be a realistic option. Willingness to pay is relatedto the marginal production that can be attributed to irrigation. The data in table 4.1 show a significant negative correlation (-0.75) betweenthe annual irrigation water supply per ha, andthe O&M costs as percentageof the output per ha, suggestingthat payment of O&M fees is lower in situations of relative water abundance (Le., low marginal production of irrigation). Inthe Office du Niger, Mali, the water fee as percentage of marginal production amountsto 15%, assumingthat, with average annual rainfallof 200mm, all production is attributable to irrigation. With a marginal production of irrigation of around 0.6 tonha, the O&M fee in PC1YMarianina amountsto 10.6% of marginal production, and there appears to be scope for upward revision. An estimation of a realistic O&M budget has been done by AFDI3.The estimation indicatesthat the average annual water fee needs to be increasedfrom US$9.48 to over US$37 per ha. Seventy five YOof this budget is allocated to investments, and 25% to salaries. This is a much higher percentage allocated directly to investmentsthan other countries that typically show a 50-50% partition (e.g., Office du Niger, Mali). Assuming that US$37 would represent the percentage of output per ha o f irrigated land, equalto the cases referred to intable 5.1, an estimation has been made ofthe productivity that is requiredto increasethe water fee from the current US$9.48 to the new level of US$37.31. This analysis indicatesthat the required productivity levels would be unrealistically high. An increase in the O&M costs as percentageof output from 1.67% is therefore necessary. I L'Organisation-Gestion des PerimktresIrrigues A Madagascar, AFD. 80 . I n summary, both the capacity and the willingnessto pay higherO&M fees have not yet been fully exploited in Lac Alaotra. Increases in production levelsare a necessary condition to increasing the capacity to pay. Increases in production levels are also necessaryto increase the stakes of irrigated agriculture, increase the returns on investments in irrigation, and increase the willingness to pay. What applies to PC1S/Marianina equally applies to other schemes: current O&M fees levels and recovery rates are insufficient to sustain irrigated agriculture. Improve recovery rates of O&M fees. O&M fee recovery rates of 100% have been achieved in Mexico and Mali, among others. In Mexico, this may be related to the upfront payment on a volumetric basis. In Mali, high recovery rates may be related to the fact that farmers can be (and have been) evicted from their land in case o f non-payment of O&M fees. The examples show that credible incentives and deterrents need to be inplace to increase water fee recovery rates to sustainable ,levels. Incentives. Irrigation management transfer requires adoption o ftransparent, accountable and farmer-controlled collection and allocation o f funds so that feedback linkages are established between those who pay O&M fees and those who provide services. Quality o f service delivery, good governance, efficiency and participation are key to achieving higher recovery rates. WUAs therefore need to establish appropriate procedures for identifying, processingand evaluating investments, expenditures and fee collection. This includes: 0 Criteria and procedures for allocating funds for annual maintenance 0 Criteria and procedures for identification and prioritization o f maintenance works 0 Procedures for water distribution and allocation 0 Procedures for and control over implementation o f maintenance works, work and progress supervision, and quality control 0 Identification o f performance indicators (O&M fee recovery, equitability and timeliness o f water supply, etc), and their monitoring and evaluation 0 Identification o f corrective measures on the basis o f M&E 0 O&Mmanual Other incentives include discounts in case o f payment before the due date or in case o f full payment. Where a Federation o f WUAs exist, O&M fees need to be collected on a WUA basis, who are responsible for timely and full collection. Finally, State financed investments can act as a carrot by linkingthem to performance on O&M fees, recovery and actual implementation o f maintenance works, as is already the case in PC1YMarianina. Deterrents. International experience indicates that inmany cases recovery rates are unsustainably low, and that sanctions are critically absent or insufficiently credible to act as deterrent. Cutting offthe water supply is often promoted as sanction, butmany times implementation is technically not feasible if done on an individual level. Closure at a group level offers better opportunities, as it brings into playthe social collateral. Water fee collection should be organized at the same level. Although many WUAs refer to local authorities as the ultimate person for implementing sanctions, few o fthem have been involvedother than by collecting "ristournes" from traders. In some cases, these taxes make up a significant part o f the commune's budget, making it a stakeholder with an interest in ensuring the sustainabiiity o f irrigated agriculture. There is therefore an opportunity to mobilize local authorities for enforcement o f WUAs rules and regulations and measures against defaulters. 81 Incases with a high percentageof absentee landlords, low recovery rates often result from poor clarity about who is responsible for payment.In Marovoay, where indeed absentee landlordism is high, ownership titles include a clause authorizing the WUA to lease out the land concerned in case of defaulting. Another option would be for the WUA to find an arrangementwith the office for collecting land tax, authorizing it to collect and transfer the water fee. Other deterrents include penaltiesfor late or incompletepayment and, in case WUAs decide to develop income-generatingactivities, the denial of benefitsof these activities to those who default. Provide incentives for cost-sharingto upgradeirrigation infrastructure. Whereas full recovery of O&M costs by farmers is the international norm, there is very little international precedencefor full recovery of for replacementcosts of irrigation infrastructure by farmers. The example from PC1YMarianina in Lac Alaotra provides an interestingcase. Farmers appear to be capableof paying for routine O&M costs, even though per hectare fees and recovery needto be increasedto fully cover costs. Inaddition to routine O&M costs, farmers pay 20% of replacement costs in cash, with the balancebeingpaid by AFD on the condition of minimumrecovery and spending of O&M fees. The 20% payment by the Federationof WUAs is allocated on the basis of priorities set by the General Assembly. The processappears to be thoroughly participatory, with a high level of responsibility and ownershiptaken by the Federationof WUAs (FWUAs). The approach provides an interestingmodel for financing of irrigation expansionthat merit replication in other parts of the country. Indoing so, the conditionality acts as an important `carrot' for increasingO&M fees and recovery levels. Further improvement can be achieved over time with a more important role for the FWUA in recruitment of contractors and supervision of works, now the responsibility of AFD. Inaddition, a clear distinction needs to be made between routine maintenanceand replacement, which is not always straightforward. Strict adherenceto the 20% and to the conditions relatedto recovery and implementation of works will pose a clear upper limit to the overall 80% envelope. The approach distinctly differs from the approachproposed inthe `Guide d'Interventi~n'~',in particular with respectto the upstreammobilization offarmers and evaluation of proposals. The Guide distinguishes six different `many questionsasked' steps to be accomplishedbefore State investment'5. Investments in expansion will dependon the retums on such investments. On the basisof econometric modeling, Minten and Randrianarisoa(2001) argue that agricultural production shows a significant but low magnitude of direct irrigation effects inMadagascar. While the study is among others complicated by the low quality data, it i s clear that, after decenniaof donor investments and with 2 ton/ha as result, Madagascarhas not found the golden bullet for turning investments in irrigation into a profitable proposition. It is equally evident that credible models for linking investments in irrigation infrastructure to rapid agricultural intensification will serve to provide an immediate solution to the challenge of attracting private investments inthe sector. 14Guide d'interventionpour la mise en Oeuvre des projets, January 2002. 15Inthis respect,the `Guide d'htervention' appearsto outline an intervention approachthat is more suitable for small-scale schemes. 82 FutureAgenda The above analysis outlineskeyissuesrelatedto the poorperformance of irrigatedagriculture in Madagascar. It respondsto the objectiveto propose an alternativeapproachthat would enable Madagascarto bringthe performanceofthe irrigationsector moreclosely in line with its potential, athree levelagenda is proposedthat includesthe following elements: At the nationallevel, confirmthe defacto transfer ofresponsibilityfor scheme O&M to W A S ,while at the same time establishinga sector strategythat clearly outlinesthe rolesof all stakeholders, includingthe state, privatesector and farmers. 0 At the watershed level, upscalethe experienceswith sustainablesoil and water management practicesof ANAE. Consider puttingin place an appropriateframework for water resources management at a basin levelthat would allow for financingof investments in the upstream parts of a river basin. At the scheme level, promoteincreasedcroppingintensity,facilitate linkagesbetweenagro- processorsand farmers' groups, and support WUAs inmanagingtheir schemes. Promote alternativeO&M models, includingestablishment of executiveofficesand promotionof privatesector involvementin O&M throughperformance-basedcontracts betweenW A S and privatesector. Institutionalissues should beaddressedwith priority, while investments in infrastructureshould beconditionedbythe performance of O&M fee collectionand implementationof O&M activities. 4.2 Technology Green RevolutionAssociated ProductivityEnhancingTechnologies. Productivityenhancingtechnologies associatedwith the GreenRevolutionare barely used in Madagascar. Resultsofthe EPMsurvey conducted in2001, indicatethat less than 10%of respondenthouseholdsuse improvedcrop varieties. Also the use of fertilizers is dismally low, estimated at about 10kg/ha, which is very low comparedto major rice producingcountries such as Indonesia,Vietnam andBangladesh, with applicationrates of respectively290 kg/ha, 160 kg/haand 70 kg/ha. It is also low comparedto applicationrates in other African countries such e.g. Tanzania and Zimbabwewith applicationrates of 65 kg/haand 25 kg/ha. Total volume of fertilizershas hardly increasedsincethe mid 1980s(Table 4.3). A largepartof fertilizers used in Madagascar originatesfrom donations. Imports Gifts Total Volume Kgfertilizer (tons) per ha 1975-84 14,600 4,800 19,400 991 1985-90 10,300 11,700 22,000 998 1991-95 13,750 11,250 25,000 10,5 1996-99 11,150 8,363 19,513 8 2 2000 17,277 10,000 27,277 10.9 In light ofthe experiencewith GreenRevolutionassociatedproductivityenhancingtechnologies, it is difficult to see how ruralproductivityinMadagascarcouldbe increasedwithout embracing these technologieson a significantly larger scalethancurrentlyisthe case. This observationis supported bythe fact that, as reportedinthe previouschapter basedon EPM-2001data, rice 83 . yields are positively correlated with expenditures on modern inputs. Similarly, based on an analysis of EPM-93 data, Randrianarisoa and Minten (2001) also suggest that the poor would benefit enormously from increased use of modern inputs. It should be noted that for export crops which are currently grown organically (by default), the increased introduction and application o f chemical fertilizers and insecticides could result inthe loss o fthis potential competitive market advantage iffarmers were to become too dependent on non-organic inputs. The disappointing use of Green Revolution associatedproductivity enhancing technologies can be attributed to a number o f factors. First, as pointed out inthe previous section, their success very much depend on the availability o f water and the quality o f water management. Second, the poor state o f the rural roads network poses a high implicit tax on the use o f lumpy inputs such as fertilizers. Recent evidence based on EPM-2001 data shows a strong correlation between fertilizer use and degree of isolation as shown in Figure 4.2 (Minten, 2002). Third, the availability o f donated fertilizers generates price distortions inthe fertilizer market, thereby hampering the development of proper distribution channels. Fourth, the changing and unpredictable application o f rice import tariffs makes adoption o f fertilizer use inrice cultivation a risky proposition as indicated in the previous chapter. Fifth, financing o f fertilizers is a problem since credit is not readily available. Figure 4.2: Relation between fertilizer use and degree of isolation %menages utilisant des engrais 30 25 20 15 10 5 0 least remote 2 3 4 most remote Inview o fthe fact that 75% o fthe farmers inthe least isolatedareas do not use fertilizers, it is clear that the improvement o f rural roads i s necessary, but not sufficient to generate enabling conditions for the adoption o f Green Revolution associated productivity enhancing technologies. Since proper water management is a crucial factor for the adoption o f these technologies, it implies that they are particularly relevant to farmers that have access to irrigation infrastructure. Since productivity o f irrigated systems, albeit low in absolute terms, is typically higher compared to rainfed-based systems, the implication of a strategy that promotesthe application o f Green Revolution type technologies, is that it will most probably benefit those farmers that are already somewhat better off in relative terms. Ifit is accepted that Green Revolution type technologies in light o fthe experience in other countries will have to play a crucial role inunleashingthe potential o f Madagascar's currently underutilized and non-performing irrigated agriculture, there is a case to be made that for equity reasons efforts inthe field should be accompanied by parallel efforts to increase productivity o f livelihoods that depend on non-irrigated systems. Inthis context, the emergence ofNGOs promotingmore sustainable agriculture, such as no-or low till techniques, agro-forestry and use o f organic fertilizer (manure and compost) is an interesting development as these so-called low-external inputtechnologies might be particularly relevant for livelihoods in marginal areas. 84 Low External Input Technologies The potential o f low external input technologies is being explored based on ANAE's experience in this field. Points o f particular interest include the effects of these technologies on soil fertility, soil erosion, yields, and farmer's costs and revenues, as well as factors that could possibly drive scaling-up o f these experiences. ANAE was created in 1990 within the context of Madagascar's NationalEnvironmental Action Plan (PNAE), with the mandate to implementthe "Soil conservation and improved rural life" component. ANAE has the overarching objective of contributing to a sustainable management of the productive natural resources and a better rural life by means of: (i)promoting farm production technologies not harmful to the environment, and a sustainable use o f soil resources; (ii) encouraginga rational land use; (iii)promoting the transfer o f management o f natural resources to the local communities; and (iv) seeking to improve farm production and farmer revenues in a sustainable manner. To achieve this objective, about 4,800 mini-projects had been financed, with participation o f approximately 370,000 families covering a total surface area o f 75,839 ha by the end o f 2001. The total number o f distinct activities within the mini-projects amounted to 15,169. Out o fthese projects, sustainable water and soil management activities accounted for 55%, farm production 38%, infrastructures 1YO,and social activities 6%. Reforestationaccounted for a large part o f the sustainable water and soil management activities, and one fifth o f the total distinct activities (see Table 4.4). The semis direct (horticulture, cropping on tanety, culture contre-saison) accounted for 22% o f total activities. Horticulture seems to be the most favored activity in farm production. Horticulture, along with fruit tree production, cropping on tanety, small livestock, apiculture, pisciculture, culture "contre-saison," and forage production, which account for 50% o f total activities, helpeddiversify production systems and contribute to a better and more balanced diet. Production intensification(SRI, small livestock) did not prove to be well received nor practiced, and accounted for only 8% o f the total activities (Table 4.4). Table 4.4: Mini-projects categories/most significant activities, Number Share Category/Activity # of Total # for % within % oftotal Cost ($) activities category category activities Sustainable water and soil management 8,305 55% Reforestation 3,233 39% 21% 58ha Improvementof tanety 2,718 33% 18% 70ha Fruit tree production on tanety 2,270 27% 15% 0.2iplant Farm production (diversificationhntensification) 5,798 38010 Horticulture 1,803 31% 12% 200ha Croppingon tanety 1,210 21% 8% 70ka Small livestock 5 13 9% 3yo 3head SRI 465 8% 3yo 103lha Apiculture 390 7% 3yo 12heehive Culturecontre-saison 357 6% 2Yo 109ha Piscicultureand rizipisciculture 309 5% 2% 0.04lalevin Forageproduction 79 1Yo 0.5% 54ha Infrastructures 152 1Yo 85 Small irrigationand dams 59 39% 79ha irrig. (constr./rehab.) 470ha irrig. Bridges (constr./rehab.) 55 36% 6,453hridge Ruralroads(constrhehab.) 9 19% 676km Riverbank protection 29 6% 35 h a recovered Social programs 914 6Yo We1Is construction/springs 777 85% 5yo 2,79 1/unit protection TOTAL 15,169 Source: ANAE, 2002 '/ Costlunitiactivityvaries from one locationto the other. Costs are specific to ANAE projects and do not include ANAE's administrative costs Soil Fertility. No-tillage technologies and mulching practices, which were particularly popular in ANAE's portfolio of mini-projects, ledto an improvement of both the chemical and physical properties of soils as can be appreciated from Table 4.5. Based on soil analysis intwo pioneer regions of no-tillage practices (Vankinankaratra, and Alaotra) at the end of each cropping season over a 3-year period, it turns out that: (i)activity and content in soil micro and macro-fauna increased resulting in a higher content of soil humus; (ii) soil structure evolved into a fine and medium polyedric structure; (iii) soil compactionwas reducedover time; and (iv) top soils increased in thickness from 10-15 cms to 20-25 cms over the 3-year period. Soil analysis showed that organic matter content increasedby 45%, Nby 440%, P by 600%, and K by 218% over a 3- year period. All these changes contributed to a higher quality/productivity soils. Table 4.5: Evolution of soil fertility parameters inAlaotra soils anolving; direct Dlantine "semis direct" over a 3-vear Deriod (1997-2000) Soil parameters 1997198 1998199 199912000 YOincrease Soil nutrients N(Yo) 0.5 1.8 2.7 440% P (0100) 0.02 0.1 0.14 600% K (meqil00) 0.11 0.15 0.35 218% Organic Matter content (YO) 3.75% 5.18% 5.45% 45% Topsoil thickness(cm) 12.5 15 22.5 80% Source:ANAE 2002 Soil Erosion. Soil erosion measuredon an ANAE site applying anti-erosion technology showed a highly significant decrease over a 5-year period. In 1996, soil erosion amounted to 8 t/ha in fields in Ambatomanga inthe Antananarivo area on an average slope of 12% (the maximum slope reaching 40%). After 5 years of applying anti-erosion technologies (reforestation, amenagement tanety, semis direct), soil loss amountedto only 1.6 tlha, in2000, an 80% decrease. Another study on erosion was conductedduring the 1999/2000cropping season inthe watershed o f Antsaloana, the BealananaRegion, Mahajanga, where average soil erosion could reach 150tlhdyear. Soil erosionwas measuredon 3 types of parcels with 20% slope: P1(traditional 86 farming), P2 (live fences + mulching), and P3 (fallow). Results showedthat applying anti- erosiontechniques reduced soil erosion by 3 and 5 times as compared to traditional farming and land left fallow, respectively(39.8 t/ha for P2, 132.0 tha for P1, and 192.8 t/ha for P3). Erosionof tanety soils notonly affects productionand revenueson tanety farming, but also causes siltation of, and reducedproductionon, downstreamrice parcels. A study publishedby PAGE in 2001, quantified productionand revenue losses in silted-uprice fields in the Region of Tamatave, mid-coaston the Eastern side of Madagascar based on a survey conductedon 300 fields with varying degreesof siltation. Itturns out that households with highly-silted rice fields loose 1,488,500 FMG ($229) a year per householddue to siltation, while those who have moderately-siltedfields loose an average of 741,000 FMG ($1 14) ayear per household(Table 4.6). Table 4.6: Rice yields int/ha infields with varying degrees of sedimentation Highly silted field Moderately-silted Not affected Average 1.2 1.4 2.4 1.8 Source: PAGE 2001 Yields. The increase in soil fertility had apositiveeffect on yields observable from the first year of no-tillage and mulchingtechniques (semis direct) as shown in Table 4.7 indicatingthe evolution in yields of selected crops in the Alaotra region on farmers' fields applyingthe technology in ANAE's mini-projects, in comparisonwith traditional farming, and yields on TAFA demonstrationsites on poor soils. Table 4.7: Evolution of yields (*a) Crop/ cropping method Year 1 Year 2 Year 3 Year 4 Year 5 Green Beans Traditional 0.60 0.60 0.55 0.45 0.40 99% Semis direct 0.90 0.95 1.10 1.10 1.13 Soybean Traditional 0.32 0.36 0.36 0.26 0.25 l,oy, Semis direct 0.5 1 0.64 0.73 1.10 1.20 Maize Traditional 0.60 0.40 0.40 0.35 0.40 201% Semis direct 0.70 0.80 0.90 1.50 2.70 Rainfed rice Traditional 0.60 0.60 0.30 0.35 0.30 188yo Semis direct 0.80 0.80 1.00 1.30 2.30 Source: ANAE 2002, and TAFA 2001 87 * Costs and Revenues. Semis direct significantly reduces overall labor time in the field by about 40% from Year 2 and on, as shown in Table 4.8. As far as the distribution o f work is concerned, one observes that semis direct reduces labor for soil preparation (no plowing) and maintenance (no manual weeding), but increases it for sowing (it is more difficult to plant a field with a vegetative or dead cover), and harvest and post-harvest, due to the ensuing higher production. The average savings in labor was 73 man/day/ha which translates into an average savings o f 730,000 FMG/ha ($1 12) ifthe average wages for hired labor is 10,000 FMG/man/day. Year 1 of semis direct would require higher labor than inthe traditional farming for: (i)crop establishment (planting), since the soil is harder to work and benefits o f crop residues on soil structure have not yet kicked in; and (ii)crop maintenance (weeding), since benefits o f mechanical weeds control through tillage are lost, while new balance between weed control through crop residue cover and weeds population has not yet been established. Table 4.8 Average labor requirements inman/day/ha Crop/ Soil Sowing Maintenance Harvest & Total O hsavings (semis direct/ croppingmethod preparation post-harvest traditional) GreenBeans Traditional 50 22.5 30 42 144.5 Semis direct 4 32.5 6 53.5 96.0 34% Soybean Traditional 50 22.5 64 68.5 205.0 Semis direct 4 32.5 12 78 122.5 40% Maize Traditional 50 20 61 62.5 193.5 Semis direct 4 22.5 5 75 106.5 45% Rainfed rice Traditional 50 22.5 62 39 173.5 Semis direct 10 41 5 44.5 100.5 42 Source: TAFA 2001 A study on the economic evaluation of naturalresources, publishedby PAGE in 2001, looked at the impact of soil conservation on farm revenues in the highplateaux o f Madagascar. The technologies evaluated were soil mulching and contour planting o f live fences (Tephrosia, Crotalaria). A survey of 78 households and 358 parcels (where ANAE financed soil conservation mini-projects) planted with the principal lanew crops (manioc, sweet potato, maize, green beans, taro, and soybean), showed that profits on one ha could increase by 137%. Table 4.9 shows that mulching alone increased production by 60%, decreased labor requirement by 40%, and removed the need for fertilization. Contour fencing increased revenues by 50%. 88 Table 4.9 Productioncosts andfarm revenues Coststrevenues Mulching Contour live fencing `000FMGIha control mulching Increased Increased revenuelsavings control fencing revenuelsavings Production eq. 4,045 6,645 + 2,600 4,500 5,600 Laborcost 1,530 918 - 612 1,518 1,242 +-1,100 Fertilizationcost 100 0 - 100 100 60 -276 40 Netrevenues 2,4 15 5,727 + 3,312 2,882 4,298 + 1,416 Source: PAGE2001 Scaling-upFactors. Although the results of low external inputtechnologiesare encouragingin terms oftheir effects on soil ferti1ity;soil erosion, yields, andfarmers' costs andrevenues, adoption is at the stage of farmer innovatorswith the mainstreamingofthese technologies still havinga longway to go. However, an encouragingsign is that spontaneousadoptiondoestake place. A series of surveys, conducted by ANAE, involving farmers who were not direct beneficiariesofANAE's mini-projectsbut livingclose to the ANAE project sites, showedthat 26% of the respondentshad started applyingthe demonstratedtechniques and practicesintheir own fields. The potentialfor scalingup is highly influencedby the existenceof severalroadblockswhich might frustrate any major attempts at expanding adoptionofANAE technologies. These roadblockscouldbe summarizedas follows: ( i )Agricultural services: the trainingand extension services offeredby the operatorswhich implementthe ANAE mini-projectshavebeen citedto be less thanadequate interms of quantityand quality. Low externalinputtechnologiesare knowledge-intensiveand farmers havetremendousdifficulties in masteringtotally newtechnologiesduring2 or 3 visits by the operators duringthe plantingseason. Moreover, availableNGO capacity is limitedto introducenew technologieson a large scale. (ii)Landtenureandsecurity:totheextentthatlandsecurityisanissue, tenantfarmersmight have less interestto investin soil improvementtechnologiesas dueto their status they are not necessarilysure to reapthe mediumto long-termbenefitsfrom these investments. (iii)BushfiresarearealprobleminMadagascarforalltypesoffarmproduction. The legislationthat regulates bushfiresis outdated. Farmers could loosetheir crops and be discouraged from adoptingimprovedfarmingtechnologiesifbushfireskeepthreateningtheir effortsto improvesoil fertility through no-tillageandmulchingtechniques. Besidesthe above-mentionedconstraints, which are technology-specific,otherconstraintslike lack of credit and lackof marketintegrationdo play a role. However, inview ofthe resultsso far, low external inputtechnologiesclearlyshould form part of a comprehensive strategy to both boost productivityof rainfedagricultureand arrest soildegradation. Doingso would benefitthe marginalsegmentsof the farmers' populationand generate positiveoff-site effects that favor productionon irrigatedlands. One addedbenefitor low externalinputtechnologiesis the low entry costs. The technology is accessibleby all populations-this is notthe casewith more capital-dependentGreen Revolutiontechnologies. 89 4.3 Land and Land Tenure Agriculture in Madagascaris practicedon very small holdings; the medianand averagecultivated area of households coveredunder the EPM-2001 survey was 1.O ha and 1.71 ha respectively. Not surprisingly, problems of access to land figure high on the list of agricultural production constraints; as mentionedby over 50% of EPM-2001 survey respondents. The distribution of land is neither extremely unequal nor equal. Based on EPM-2001 data, the gini-coefficient reflecting the distribution of cultivated land is estimated at 0.566. This figure put Madagascar in the league with Mexico (0.607), Turkey (0.595), Philippines (0.560), Pakistan (0.556), Indonesia (0.555), Uganda(0.549) and Egypt (0.549) and way behind extremely unequal cases that are typically found in South-America, including Venezuela(0.917), Argentina (0.856), Brazil (0.841), Ecuador (0.840) and Colombia (0.829). Consequently,the low median cultivated area per householdappears to be less driven by an unequaldistribution of land than by other factors. Also the largest landholdings included in the EPM-2001 survey are not large in absolute terms as they do not exceed20 hectares. Figure4.3: Distributionof CultivatedArea (EPM-2001data) Households Distributionper Cumulationof CultivatedArea 100 - 90 80 70 - 60 - 50 ~ 40 30 - 20 - 10 - 0 ._~____ 0 10 20 30 Percent of50 40 60 70 80 90 100 households As far as land tenure arrangements are concerned, the following picture emerges from the data presented in Table 4.10. First, the vast majority of farmers are owner cultivators. Second, share cropping is the most common form of tenancy arrangement. Third, the vast majority of farmers have obtained their landsthrough heritage; which might reflect the reported excessive subdivision of land through customary inheritancerules (DCDSF 2001). At the same time a small land market is active as 13%of surveyrespondentsmentioned to have bought land. 90 Table 4.10: Land and LandTenure Data from EPM-2001 I Characteristic I Freouencv 11Land tenure @De: (% o f all dots) i Owner-cultivated 191% Ii Share cropping Fixed lease contracts II 47% 36% I "For Free" rental arrangements 17% Acquisition method: (YOof all sample plots) Purchase 13% Heritage 78% Gift 5% Slash-and-Burn 4% The effects of land tenure on productivity are not straightforward. Randrianarisoa& Minten (2001) report a slightly higher productivity on lands under rental agreementsbased on EPM-1993 data. A comparisonof rice yields on owner-cultivated plots with leasedplots based on EPM-2001 data, however shows higher yields on the former (Table 4.1 1). As far as rice yields under various rental arrangementsare concerned, it has been observedthat they are lowest under sharecropping agreements. Although this might reflectthe effects of the alignment ofthe incentive structure underthese agreements, it could also be that the illegal status of sharecroppingunder the current law preventstenants from investing in productivity enhancingtechnologies. However, both EPM- 1993 and EPM-200Idata suggest that land titling hasno significant effect on productivity at the national level (Randrianarisoa & Minten,2001; Stifel et al. 2002). There are indications of the existenceof an inverse landsize-productivity relationship as reported by Randrianarisoa& Minten (2001) and Stifel et al. (2002). Basedon observations from 2,495 plots included in the EPM-2001 database, rice yields and rice cultivated area were also found to be negatively correlated(-0.3 1). These findings point inthe direction that mechanismsthat would promotethe transfer of relatively large holdings to smaller ones would have a positive effect on productivity. A land tax would be an example of such mechanism. Madagascar currently has a landtax which is progressively applied according to the size ofthe holding. A total of six categories are distinguished with nominal tax amounts for the first five categories and a 1% rate appliedto landsthat fall in the sixth category. The current system has three weaknesses: (i)collection is almost non-existent; (ii)inflation adjustment of nominal amounts is long overdue; and (iii)serious under-reporting of the value of properties. In2001, the collected amount of the landtax (IFT, Imp& Foncier sur le Terrain) inMadagascarwas slightly over FMG 484 million or less than US$80,000 equivalent. The weak functioning o fthe landtax system prohibits a transfer of currently under-utilized landsto more efficient uses. Besidesthe fact that this reflects a serious governanceproblem, it also implies the loss of potential fiscal revenuesthat could usedto improve the land administration system inthe country. 91 Under customary systems of landtenure two kindsof land are considereda commons -forest and grassland. Farmed land is considered the property ofthose who have traditionally cultivated the land, reinforced by the principle of modern law that requires prior proofof use (mise en valeur) before a user can seek title to the land. Where use is not maintained, the land revertsto a commons. However, any signs of former use will deter newcomers, particularly where they are not inhabitants of the region. This can leadto inefficient land use and is an important factor forcing migrant southernersto cut forest to grow maize rather than cultivate unoccupiedsuitable land (e.g. Reau 2001). The principle of mise en valeur operates to discouragenon-consumptive use of land and this is one of the important values of GELOSE, enabling communities to secure a prior right over forest without necessarilyusing it. The lessonof GELOSE could be applied to the modification of landtenure law to enable titles to be obtained over land (especially forests) for conservation and potentially for consumptiveuses. As far as recent experiencewith landtitling programs is concerned, it is worth noting that ANAE has collaborated with CEPRO, at the Direction Generale des Domaineset des Services Fonciers, to implement landtenure "securization" activities in 4 areas, covering a total of 5,763 ha and 1,666 beneficiary families. The average cost per haof land was about 91,000 FMG (US$15). Dueto cumbersome proceduresand centralizeddecision-making, processingof the titles turned out to be problematic. In light of the above, it appears that the rationale for an aggressive, supply-driven individual land titling program is questionable. At the sametime, the case for communal titling using mechanismsfor establishingrelative security of land tenure (e.g. GELOSE) so as to create incentives for proper managementof currently open access resourcesappears more compelling. More specifically, there is a needto: (i)extendthe definition of `mise en valeur' to include non- consumptive uses of land and specifically citing use for conservationor watershedmaintenance; (ii) develop regulations to facilitate access to agricultural land that has fallen into disuse; (iii)to focus collective titling on disusedagricultural lands under communaljurisdiction so as to make them more readily available for renting out to users of any origin; (iv) to reexaminecurrent legal restrictions on the use agricultural rental agreements, also from the point of view to ensure payment o fwater fees for the coverage of O&M costs of irrigation systems; and (v) to explore the application of the land tax as a mechanismto achieve a more efficient allocation of land resources. In any case, in a resource-constrainedenvironment, it appears fair to say that any proposed improvement in land tenure andtitling arrangements should be linked with improvements inthe functioning ofthe land tax system. 4.4 Credit Production credit from the commercial banking sector is insignificant and in decline as demonstrated by lending figures from BTMBOA, the leading commercial bank for the agricultural sector (Table 4.12). In effect, total lending from 1997-2000amountedto FMG 21,115 million, which is less than US$4 million at the current exchangerate. Commercial banks appear to concentrateon trade credit as indicated by lendingvolumes that increasedby over 50% from 1997-2000, as comparedto production credit that decreasedby almost 50% over the same period. The driving force behindthese figures has beenthe shrimp industry. 92 Table 4.12: Credit extended to the agricultural sector by BTM/BOA (FMGmillion) Activities Production credit Trade credit 1997 1998 1999 2000 II 1I I Total 1997 1998 1999 1 2000 I Total I 1I 1I II I I Crops 5,528 6,182 3,738 3,283 83,150 108,114 117,426 112,132 Livestock 281 854 327 31 764 566 1,036 55 Fish 584 307 16,555 29,101 15,920 37,576 Forests I 2,2351 TOTAL 6,393 7,343 4,065 3,314 21,115 99,7051 137,781 134,3821151,9891 523,857 Duringthe 1990s, major efforts have been investedto improve access to formal credit through the establishmentof mutual savings and credit institutions. The largestofthese had about 90,000 members in 2000. At that point, the total annual lendingvolume ofthese institutions had reached FMG 43 billion (4.13). Hence, mutual savingand credit institutions are certainly a much more important source of formal credit in rural areas than commercial banks. Interms of coverage, the mutual savings and credit institutions had a presence in about 20% ofall 1,392 communes in Madagascar(Cornell/FOFIFA commune survey). The financing of collective storage schemes (GCVs) is one ofthe micro-finance institution success stories. Under these schemes, farmers escape low output prices immediately after the harvestperiod by obtaining a credit that is secured against stocked output which is eventually sold at a higher price during the lean period. loans 2,591 934 12,69 1 4,276 4,049 20,492 Credit FMG 1,270,647 1,063,796 14,957,354 14,836,210 10,936,900 43,086,907 Despitethese efforts, access to formal credit remainsrare. Basedon data from micro-finance institutions and expert opinions, it is generally acceptedthat the overall penetrationrate is about 3%. EPM-2001 data confirms the low penetrationrate as only 18 out of 5,080 households (0.35%) reportedto have obtained credit from formal sources. Access to credit also depends on available assets. Assets of householdsthat reportedto have obtainedcredit exceeded the average assets of poor households.bya factor 10(EPM-2001 data). Hence, access to credit i s skewed towards the wealthy. This observation is line with Randrianarisoaand Minten(2001) who report, based on EPM-1993 data, that only 1% ofthe poor gets credit through the formal system as compared to 6% for the richest quintile. The latter figure also implies that, despite being rich, 94% of the richest quintile either does not have access to credit or chose not to seek it, thereby suggestingthat there are a lot of factors other than credit accessthat set the richest quintile apart from the poorest. 93 The effects of lack of credit and liquidity on input use, rice yields, lengthof on-farm storage are well-known and it is difficult to argue against the notion that with access to credit farmers would be much better positioned to adopt improved varieties, avoid distress sales and purchase productivity enhancing inputs to improve productivity and incomes. However, at the same time, most farmers are poor, do not have the requiredasset levels and operate under conditions that in most cases do not makethem attractive clients to rural financial institutions, including mutual savings and credit institutions, which operate on a commercial basisto ensure their financial viability in the longrun. In light of this, it is difficult to envisagethat rural credit would provide the key to ignite the primary sector as an engine for inclusive economic growth. Rather it is believed that the provision of productive assets to the rural poor on a matching grant basiswould bring assets levels, income-generatingcapacity and eventually saving-generatingcapacity up to a standardthat enable mutual saving and credit institutions to work with them on a commercial basis for subsequent investmentrounds. Doing so would putthe expectations regarding the role of credit in setting the stage for rural productivity growth on a more realistic footing, although it does not deny that rural credit can play a very useful role in certain niche markets, such as GCVs. 4.5 RuralInfrastructure With only 9 kmof pavedroads and 60 km of secondary roads per 100sq-km, Madagascar's roads infrastructure is extremely poor. Dueto delayedmaintenance 80% of the network is now ina very precarious situation. The situation has been aggravatedby a series of recent cyclones. As a result, the access of producersto local marketspresents a significant constraint. Depending on the region, the walking time to the nearest market equals 1to 5 hourswith distances ranging from 5 to 21 kin, with an average distanceof 10km. Each additional km of distanceto a main road decreases the producer price of rice by FMG 14. Villages locatedfar from the roads are negotiating paddy prices that are 15-22%lower than villages with access to the road (Bockel, 2002). The Northern region contains areas where the transport of paddy is managed by human carrying with transactions being done in kind. Under these conditions, the price of paddy price may be as low as FMG 350/kg. The isolatednature of rice production areas is one ofthe major challengesto improve the competitivenessofthe rice sub-sector. With most of production areas constrainedby difficult access, many producers are faced with very poor marketing conditions. On the contrary, the very good state of the main commercial road which joins Antananarivo with the country's largest port in Tamatave, provides easy conditions to supply the major rice consumption urban areas with imported rice, thereby effectively disqualifying part of the existing rice growing areas as potential suppliers due to transport cost constraints. The benefitsof improved rural roads might be considerable. An example in Mahajanga shows that road rehabilitation induced an increaseofthe farm gate prices of food crops by 50%. In Bangladesh, a comparison betweenvillages servedand not served by roads showed that productivity was 33% higher in villages having good access (Ahmes and Hossain 1990). An increaseof 10% in rural road coverage in one region inthe Philippines increasedproductivity by particularly benefit from improving the following connections: (i)RN44 Moramanga - Lac 3% (Everson 1986). Feedbackfrom downstreamoperators learns that the rice sub-sector would Alaotra, (ii) Tulear - Morombe, (iii)Sofia - Bealananaaccess road, and (iv) Mainiarivo access. Recent evidence from the INSTAT 2002 householdsurvey shows a strong correlation between rice yields, fertilizer use and degree of isolation as shown inFigure 4.4 (Minten, 2002), suggestingthe possiblepositive impact ofrural road investments. Rice sub-sector surveys in 1999confirmed that farmers explained as the main reasons for not using fertilizers, their own 94 . ignorance.aswell as the excessive price of fertilizers. Both these factors are conditioned by the absence o f deficient rural deficient rural roads network, which isolates the population by limiting the flow o f goods and information. At the same time, the fact that even in the less isolated areas more than 70% o f farmers do not use fertilizers also suggest that other factors than market integration might play a role as well and that market integration is most probably a necessary but not sufficient condition to improve rice productivity. __ - - - Figure 4.4: Correlation between Rice Yields, Fertilizer Use and Degree of Isolation . - - - nce yield as hnction of distancefrom roads %&ages utiiisant des engrais 30 25 20 15 10 5 0 least remote 2 3 4 most remote least remote 2 3 4 most remote ____ -- - The effects o f rural isolation have been quantified recently based on EPM-200.1data (Stifel et al., . 2002). Followingthis research, the following picture regardingthe relationship betweenrural, infrastructure, rural productivity and rural poverty emerges: (i) Inrural areas the cost o ftransportationaverages 15 percent o fthe price o frice. (ii) Market integration decreases with isolation as indicated by the fact that autoconsumption intotal consumption rises from 20% inthe least isolatedto 40% in the most isolated areas. (iii) Rice, maize, and cassavamedianyields inthe most remote areas are only about halfthe yields inthe least isolated areas. (iv) Although overall use is low with only 12%o f households use chemical fertilizers, this figure drops to 5% in the most isolated areas as compared to 25% in the least isolated areas. (v) Isolated areas are poorer: the incidence o f poverty is 85% inthe most isolated areas vis-a-vis 55% inthe least isolated areas. (vi) Rural insecurity is associated with isolation and has a negative effect on rice productivity. Based on these findings, simulation results suggest that: (i)investment inmajor highways would reduce total travel time by 30% on average and increase rice production by 1.3%, with most o f the gains in the least remote quintile; (ii)investment infeeder roads would increase overall production by 1.O%, with the most remote areas benefiting most as they would see rice production increase by 1.6%; (iii)investments in trails accessible only to foot traffic and zebu- drawn carriages would increase overall rice production by 0.2% and 0.5% in the most isolated areas. Inconclusion, it can be stated that rural roads investments are essential to reduce currently significant transaction costs with subsequent positive impacts on rural productivity, market integration and poverty. The results ofthe simulations suggest a trade-off between focusing investments on major highways with largest overall benefits or on feeder roads with benefits 95 particularlygoingto the mostremote(and poorest) areas at the expense ofthe overall benefits. As far as import substitutioncrops are concerned, there is a needto ensure that roads investment planningand priority settingmaintaina reasonablelevelplayingfield betweenthese crops and importsby avoiding linkingup urbancenters with the country's mainports, while ignoring linking-up major domestic productionareas with the country's major urban centers. 96 ANNEX 5. Forests & forestry as providers of extractive products and environmentaI services After having explored optionsto increaserural productivity, the focus of the current chapter is to identify options that would enable Madagascarto better capitalize on its forests as a source of rural growth through their provision of extractive products and environmental services. By doing so, it builds on the environmental perspectiveof the situation assessment as presented in Annex 2. 5.1 Economic importance of the sector and contribution to rural economy According to statistics from the Ministry of Financeand from the IMF,the forest sector represented5% of GNP and 17%of the primary sector in 2000. Its value in current prices amounted to FMG 1,208 billion and its growth rhythm between 1996and 2000 was the highest in the primary sector. The value offorests and forest-based products that are traded annually in Madagascarhas been estimatedat FMG 497 billion (Topa, 2002). Wood and charcoal are most important and account for respectively 50% and 35% of total trade revenues. Forestsprovide all of the domestic energy needs inrural areas and over 70% ofthe total energy consumption ofthe country. Non- timber forest products and eco-tourismmake upthe remaining 15%, with the observationthat direct revenuesfrom the tourist sector interms of park entrance fees are still negligible in relative terms (Table 5.1). Table 5.1Incomes per product type and beneficiarytype of traded forest products (in FMGMillion) Beneficiaries Products ~ Wood Charcoal Medicinal Eco-tourism TOTAL Plants, animals State 19,500 32,254 893 52,647 Local communities 165 90 207 894 1,356 35,629 61,094 3,085 99,808 Woodcuttersand workers Economic operators 161,609 89,318 26,924 1,265 279,116 TOTAL 216,903 150,502 62,470 3,052 432,927 As far as how the benefits of this trade accrueto different groups from different forest products, the following picture emerges. First, most of the forest revenuesaccruing to the state come from export taxes of a relatively small amount of products, especially of non-wood products. Second, forest revenuesfrom fees paid to the local population are negligible except for ecotourism. Third, the largest benefit to rural peoplecomes in the form of employment. Fourth, a significant minority of the sector's economy is driven by the growing markets for animal, vegetal and pharmaceuticalproducts, tourism, and other non-wood products (excluding the charcoal business). Fifth, more than 60% of the total sector's value is capturedby fragmented chains of 97 transport, processing, retail and value added. Profit margins are often shared among a large number of actors in the value chain. There is no significant concentration of businesswith large dominant companies and groups. Sixth, the vast majority of wood products are for the internal market. However, the value of standingtimber for the internal market is closeto zero and export prices are at least 40% below international levels. Interms of value, exports are dominated by pharmaceutical products and CITES and non-CITES plant and animal species. The export of natural and processedwood are dominated interms ofvolume by sawn wood from industrial plantations, and in general, remain low. The low or negligible amount the governmentreceives from charcoal production can be seen as a subsidy that encouragesthe practice at the cost of more environmentally and healthy sources of energy such as hydro and propane gas. As far as forest exploitation is concerned, whether for wood, or for other products, the exploitation is typically an artisanal gatheringactivity. Most forest enterprises are small scale. No very large logging company or forestry industry operatesinMadagascar. Producergroups and companies are ill-equipped and undercapitalized, and many workers are part-time farmers using tree harvestingto supplement subsistence agriculture. Tree-cutting and squaredtimber sawing are typically done with the handtools inthe forest. Hauling is done on man's back and transportation towards industries and markets is carried out by low volume individual transporters. As a result ofthis inadequateharvestingand transportation system a good part of sawing is also done inthe forest and is extremely wasteful. The quantities producedper man and per month are 3.6 cubic meters (see FOFIFA/ CIRAD studies of industries,2000 and 2001) in the case of timber and 1I.5 kilograms of charcoalper month. It is estimatedthat the sector generates over 16.5 million remuneratedwork days per year representingan approximately total value o f MGF 100 billion (US $ 15.5 million) which are distributed amongthe poorest layers of the population. Gathering of non-wood products, as well as timber production and skidding, are done manually. 5.2 Access to forest & biodiversity resources Ifforest and biodiversity resourcesareto contributeto sustainablerural developmentandpoverty reduction, then appropriateregimes for access to those resourcesmust be developed. According to the existing law, all natural forest is the property of the state and even extraction of products from private propertiesrequiresa permit, whether for commercial or subsistencepurposes. In respect o f protectedareas, ANGAP exerts direct control over access to biodiversity resources within protected areas. In practice, the capacity ofthe forests administration to regulate access is limited and is essentially restrictedto the issue of permits at the regional level to commercial loggers and at the local level to agricultural or subsistenceusers inthe case of forests where the administration is present nearby. Permitsare typically issuedwithout any systematic field verification or monitoring for compliance. In remote areas where forest is plentiful and state authorities absent (such as the such the zones where tavy is heavily practiced) access may be essentially free and open. In more densely populatedareas where resourcesare in shorter supply, there is typically some form of local regulation that affects access linkedto customarytenure rights or locally declared taboos. Where the state is present, it may operate againstthe local interest, such as awarding permits to commercial loggers or migrants to clear forest for agriculture, as in the case ofthe growing of maize inwestern dry forests (Reau 2001). Ninety- three percent of logging (from natural and secondary forests) inMadagascar is for firewood, charcoal or poles for local markets. Harvesting for quality timber is highly selective and wasteful, rejecting all but a few known species such as palissandre, rosewood, ebony and katrafay. Timber production from natural forests in 1998 accountedfor 2.3% of deforestation, mostly in the province of Toamasina. 98 Much timber production is unaccountedfor, with an estimated60% of logging being illegal (Pavy 2002). Permits are typically of short duration e.g. 3-5 years. This encourages official logging companies to `mine out' resources as rapidly as possible rather than managethe plot. Logging companies are often individuals that acquire an exploitation or transportationpermitand employ local peopleto provide products to pick up points are referredto as "exploitants." Exploitants on average harvest200% morethan the permitted area. Exploitation also encouragescopy-cat operations (without permits) that increasestill further the area affected. The operations are also too sinalIto enable rationalization of operations. The fragmentation of forest exploitation and the weak structure of existing enterprises is a serious constraint to re-organization and better control of the sector. Permitsare not granted accordingto any rational plan and are inadequately documented or evaluated. Benefit capture is weak, with paymentsfor concessionstypically low and much of the wood directly exported as raw timber without added value. A ban has now been imposed on the export of unprocessed rosewoodsand ebonies but other high value woods remain unaffected. On paper, sector policies are generally sound. However, while size and complexity of forest and environmental programs have grown for the past decade, the national capacity for law enforcement has not been adequately strengthened. With atotal budget equivalent to US$450,000 per year, the Water and ForestDepartmentis called on to represent the state and capitalize on the actions ofdonor assistedprogramsfifty times bigger (over US$20million/year), which are independentlyexecuted by international NGOsand new parastatalorganizations. Effectively the Water and Forestadministration has been neglectedinterms of capacity building during the first two Environmental Programs(EPl and EP2.) Poor governanceand law enforcement is increasingly recognizedto undermine costly ongoing environmental management programs, and has the potential to discouragemost qualified long-term investorsfrom doing businesswith Madagascar inthe field of biodiversity and environmental services. Under these circumstances, the disconnect betweenwritten policies and ground implementation has become so wide that further focus on improving policies and strategies would distract from real capacity building, accountability and implementation issues. Systems of community based management, whether basedon GELOSE, GCF or dina, have been introduced recently and now cover an estimated200,000 ha in Madagascar. These systems seek to engage the local resourceusers (base communities) to regulateaccess to forest & biodiversity resources, typically by providingpreferential access to these local community users and granting the communities the authority to bar outsiders who do not form part ofthe communaute`de base. Although this approachhas the potential to move an open access resourcetowards a controlled access resource, there is enormous support and monitoring requirementsthat have beencreated but not addressed by this process. 5.3 Approaches & initiatives tried or considered to improve the role of forests The desire to conserve forests as reservoirs of biodiversity has driven a search for ecologically sustainable forms of extractive use and for a better valorization of forests interms ofthe environmental services they can provide. The very high conservationvalue ascribedto Malagasy forests results from their unique biodiversity inrelation to other forests and habitatso fthe world. While this high conservationvalue is clearly felt internationally, the challenge is to internalize those values or to develop other values such that the forests contribute to sustainablerural development and poverty alleviation without compromising their biodiversity value. A total o f 12.8% of Madagascar's forests (1,717,379 ha) are within the national protected areas network. Current protectedareas legislation generally bans extractive uses whereas non-extractive uses like ecotourism may be encouraged. Extractive uses are thus currently available for the remaining Approach considered / tried Status Establishmentof forest protectedareas as limited providersof A numberof protectedareas (especiallynational extractiveproducts.revenues(from ecotourism)and parks) are operatedto these ends environmentalservices (watershedsetc.) Sustainabletimber extraction(selectivelogging, rotational Tried for dry forest- found unsustainable; widely logging etc.) and ecocertificationschemes considered but not tried for eastemforest (long cyclesstill required> 50 yrs), although national 1 standardsfor eco-certification are being established Transferof state forests into community-basedmanagement Numeroustransfers effectedduring EP2 but no (GELOSE, GCF) coupledwith sustainable extractionof forest managementschemesyet operational, no or little productsor non-extractiveuses monitoring Controlled extractionof speciesvaluable for the wildlife trade Considered; no credible scenario identified; Madagascarrecently suspendedtrade in CITES listed species Use of forests for the prospectingo f bio-active compounds (bio- Considered, has occurred on a limited scale through prospecting) I I direct contracts, other legal mechanismsnot yet in place; potential demand not filly evaluated Controlled extractionof essential oils. medicinal ~lants& other I Considered. limited imdementation to date non-timberforest products (Manongarivo,WWF). highly fragmented market: riskof beingcounterproductive (Messerli 1999). Establishingforest plantationsfor wood production, agro- IIIAdopted for wood production & agro-forestry; forestry. erosion protectionor carbonsequestration carbonsequestration considered(requires ratification of Kyoto protocol and elaborationof carbonmarketmechanisms). Settingaside forests under carbon storage schemes for the Considered, requiresMadagascarto ratify Kyoto potential sale of carbon credits on internationalcarbon markets Protocol and extensionof Kyoto to carbon storage (currently limited to carbon sequestration Le. reforestation) Protectedareas 2 eco-tourism. The principal alternative to extractive use is the creation o f protected areas, sometimesallowing limited extraction o ftimber and non-timber products in the buffer zone (this extraction often classed as droits d'usuge). Protectedareas can be promoted as providers of environmental services (such as watersheds for agriculture) and as sources o f revenue from non-extractive activities (eco-tourism, research, media & education & bio- prospecting). The network o f Madagascar's protectedarea system is composed of 18 National Parks, five "Integral" Nature Reserves and 23 Special Reserves. Inaddition, two marine areas have been brought under protection with four more identified that are currently beingcreated. Less than 3 per cent of Madagascar's total land surface area i s currently protected by statutory 100 conservation areas comparedto a world average of 8-12 per cent of a country's land surface. Besides its importancefor conservingunique biodiversity, Madagascar'sprotectedareas system has become a critical asset for the developmentof its tourist industry. Madagascar is primarily an eco-tourism destination. A recent survey showedthat 55% of all tourists to Madagascarinclude eco-tourism activities intheir itinerary, indicating that the country's ecological image is a strong part of the draw to visitors (Crompton 2001). The statistics for the national parks indicate that 86,964 visitors entered 22 sites in 2000, basedon ticket stubs. The numbers of visitors have grown steadily in all categories from a total of 5,898 in 1992(Table 5.3). Given the range of assets and the relatively small number oftourists currently visiting Madagascar,tourism still has considerablepotential to boost economic growth in a number of regions and to benefit more communities throughout the island. Preliminary results ofthe on-going economic analysis of the upcoming third phase of the PNAE estimatethe discounted on-site net revenuesfrom the tourism sector that can be attributed to the existing protectedareas network in Madagascar at US$57 million over 25 years. This amountwould exceed ANGAP's operating costs over the same period (Carret, 2003.) Table 5.3 Visitors to Madagascar's NationalParks and Reserves, 1992 2000 - 1992 1993 1994 1995 1996 1997 1998 1999 2000 Foreigners 4619 10985 12702 20747 25366 31072 41573 51720 54440 Nationals 1260 3930 4615 15408 16484 19209 28297 29138 32306 Researchers 19 47 101 81 84 52 1 1 1 200 218 Total 5898 14962 17418 36236 41934 50333 69981 81058 86963 Under regulations introduced by the protectedareas service, ANGAP, 50% of park entry fees are to be used to benefit local communities living adjacent to the park. Popular protected areas such as Andasibe (26,000 visitors annually) and Isalo (17,000) generate considerable sums for community projects. The ability of local populationsto benefit inother ways, such as through guiding, is limited by their lack of educationand training. Apart from park entry fees, popular protectedareas attract tourism investmentand createemployment, contributing to local and national economies. All biodiversity-basedtourism necessarily focuses on the countryside and therefore potentially brings benefits of tourism closeto rural populations such as revenues from food supply, lodging, handicrafts and the like. Benefit capture neverthelessremains a challenge since local residents are rarely equippedto take part intourism operations. There is thus a need for increasedoutreach of tourism to rural people, such as through training schemes andthe establishment of community-basedstructures. Conversely, there is also a need to `sell' Madagascar as an international eco-tourism destination. This requires not only increasedand improved infrastructure but also a system of standards for eco-tourism (as is the case for EastAfrica) such that operators can be guaranteedthat tourism will be run along ecological and, where appropriate, appropriate socio-economic guidelines with a balancedshare of benefits (`fair tourism'). Additionally, the quality ofthe ecotourism experience is limited by the current poor quality of infrastructure available to satisfy higher numbers of tourists. This infrastructure constraint includes roads and air travel opportunities as well as hotels and restaurants. At present, revenuegenerationfrom protectedareas in general is limited bythe low number of visitors to Madagascarand only applies to protected areas accessibleto tourists - most protectedareas or classified forests are simply too remote to capturethe benefits of eco- tourism. The recognition that revenuesthat can mostly be capturedfrom eco-tourism probably will never be enough to finance the national parks network was one of the factors leading to current efforts to establish a national trust fund to finance Madagascar's protected areas network. 101 Based on historic willingness to pay for the environment in Madagascar, preliminary results of the EP3 economic analysis estimate the discounted value o f the existence value o f the protected areas system at between US$12 and 25 million over the next 25 years. However, the successful establishment o f a trust fund to capture some o f these existence values will strongly depend on Madagascar's ability to establish a track record of good governance for the management o f its natural resources. In view o f the location of a significant number o f the country's protected areas, estimated hydrological benefits are likely substantial. About 50% o f Madagascar's 401protected areas are situated in locations where the protection o f natural resources provide significant benefits to downstream irrigation systems, dams, ports and water collection systems. For instance, preliminary results of the EP3 economic analysis indicate that about 340,000 ha o f irrigated lands, or about one-third o f the country's total, benefit from environmental services provided by existing protected areas in terms o f foregone lossesthat would be caused by sedimentation in the absence o f the protected areas' hydrological services. Given the negative impact o f sedimentation problems on rice yields as suggested in the previous chapter, these foregone losses imply real benefits to farmers. Preliminary estimates putthe total discounted value o f hydrological services provided by the protected areas system at well over US$ 80 million based on a 25 year time- horizon. The fact that hydrological benefits are larger than estimated benefits from tourism implies that national residents (primarily farmers) are amongthe largest beneficiaries from protection efforts and demonstrates one area where environment-poverty linkages are real and strong. This is further supported by an analysis o fthe EPM2001 that shows that the poorest quintile experiences the highest levels o f sand and mud sedimentation inthe their rice fields (Mitten and Ralison, 2003.) Sustainable timber extraction and eco-certification. It is believed that considerable potential exists in the eastern rainforests for long cycle sustainable exploitation o f specialist hardwoods and wood products, ideally coupled to added-value woodworking industriessupported by origin- labeling and eco-certification. Under these scenarios, rotational or selective logging would permit the extraction o f valuable timbers without significantly reducing a forest's ecological functions or the maintenance o f biodiversity. However, untilnow no credible attempts have been launched to systematically explore the existing potential along these lines. Scattered examples o f partial regeneration of formerly logged forest offer some prospect o f sustainable logging (e.g. Ranomafana). The only scheme so far attempted has been for the dry western forests north o f Morondava in the Kiriiidy forest (now a national park). The experiment comprised logging in lines, leaving intervening forest intact. Duringthe timeframe o f the experiment (10 years) there was no measurable regeneration and monitored growth rates were so low that the required rotation lengths were too longto be economically viable. The experiment was abandoned. Requirements for sustainable harvestingwould include the development o f national and regional forest zoning plans identifying the areas to be set aside for production and making each exploitation the subject of a long term permit and exploitation plan (at least 50 years). However, sustainable timber extraction activities should be adequately balancedwith the needto set forest aside for conservation purposes in order to ensure adequate biodiversity and watershed protection. Ifthis need is taken into consideration, remaining areas o f natural forests for sustainable timber extraction might be relatively limited. Hence, longterm potential for sustainable timber production thus seems unlikelyto lie in a significant increase o f exploitation o f natural forests and it is therefore expected that the future o fthe Malagasy forest products industry will inevitably shift from natural forests infavor o f plannedand managed forests (ARD 2002). In this context, the large, currently untapped potential for sustainable plantations is relevant. An important exception of this observation could be mangroves which are well suited to sustainable 102 harvesting over quite short cycle times. Mangrove forestry could prove o f decisive importance as a source of wood fuel on the west coast and scattered locations on the east coast. However since mangroves provide numerous ecological services for the marine and coastal ecosystems, care should be taken to avoid exploitation by one group at another group's economic expense. Eco-certification of sustainably harvested natural timber has been considered as a potential instrument for promoting conservation of forests while extracting the maximum value from rare and specialist timbers. Madagascar, with some mostly endemic 1200 tree species and its `exotic' international image would be expected to offer considerable potential for this form o f sustainable timber use and marketing. In 2001, the country initiated the establishment o f an eco-certification process with the private sector. However, eco-certification requires high standards o f governance and full traceability o f products. This transparency and professionalism is very difficult to imagine given the current state o f the forestry sector. Its success would also depend on a highly motivated and professional private timber sector, currently lacking in Madagascar. Accordingly, apart from a few potential plantations and ranchable animal species, the introduction o f wide- spread timber certification would require a considerable effort (Topa 2002). Communitv-based forest management (GELOSE, GCF). Law 96-025 ("GELOSE = Gestion Locale Securisee) provides for the transfer o f management responsibilities for renewable natural resources to local communities. This is optionally coupled with limited security o f tenure over the resource management area via mechanisms such as SFR (Relative Land-tenure Securing) SFI ( Intermediate Landtenure Securing) and SFO ( Optimal Landtenure Securing ). Normal operation o f GELOSErequires the intervention o f an official mediator to supervise the transfer and resolve conflicts that may arise. InFebruary2000, the government approved the GCF ("Gestion Contractualisee des For6t:ts" or "Forest Management Contracts") decree, which transfers the management o f forests to local communities. The GCF falls under the GELOSE law butputs a specific emphasis on community-based forest management according to a simplified procedure. To date, the number and scale o f community-based forest management arrangements include 104 contracts, covering a total o f about 140,000 ha (Table 5.4). Table 5.4: Communitv-based forest management Management transfer scheme Signed Contracts Area covered in hectares GCF 42 97,4 17 GELOSE 62 44,228 TOTAL 104 141.645 Source: Razafindrakoto,2002 The highest number of GCF and GELOSEforest contracts are found inthe provinces o f Toliara (22 and 16 respectively, covering a total area o f 94,192 hectares) and Mahajanga (2 and 24 respectively, covering a total area o f 7,494 hectares). Attention so far is particularly focused on the number of GELOSEcontracts issued with the intention to put as many contracts in place rather than assuring or perfecting management systems. A key motivatingfactor for communities has been to secure natural resources for their option value, as a means o f preventing official concessions to outsiders, and as a meansto keep outsiders from squatting. GCF experience is similar, in that while numerous contracts are in place, few, ifany, have become operational management systems, and even fewer have been monitored for effectiveness. Inaddition, there are examples that indicate that for communities engaging in other forest-based activities such as forest products exploitation in Fort-Dauphin or mushroom farming in Moramanga, revenues from this forest management transfer may be considerable (Razafindrakoto 2002). 103 Linked to community-based management is the principle of equitable sharingofthe benefits from forest use. A draft law on access to genetic resources proposesa framework for access to genetic resources, based on the principles of the biodiversity convention (CBD) and including provisions on equitable sharing ofthe benefits from biodiversity. An amendment has been proposedto the law on intellectual property to take account ofthis. However, none of these laws are effective and it thus remains the case that the forests administration hasthe prerogative to control access to and use of biodiversity and fprest resources. For marineand freshwater fisheries resources, the prerogative lies with the fisheriesadministration. Extraction of non-timber forest products. Forestscan provide a rangeof non-timber products useful to local residentsor of potential value in national and international trade including raphia, resins, fungi, essential oils, wild silk, fruits, leaves (for essentialoils), honey, tree fern boles and strangler fig cases. The commercial value of such products is estimatedat US$3 million, or 40% of all exported forested products including timber (Pavy 2002). A study ofthe exploitation of non-woody forest plants found the sector to be highly fragmented and disorganized, in which individual collectors secured only a small proportion (3-1 1%) ofthe export value of the products (MEF 2001). Particular scope exists for increasingthe sustainableproduction of essentialoils, mainly basedon extraction from harvestedleaves and medicinal plantssuch as Prunus africanus, although the highly disorganized natureof these sectors renderstheir managementdifficult. At the same time caution is required because encouraging use of forest productswill continue to attract peopleto the forest fringe where they are likely to engage in destructiveagricultural practices. Thus, improved valorization of forests, howeverdesirable from a development standpoint, may simply serve to intensify occupationof forested regionsand leadto an increasedrate o f slash & burn (Messerli 1999). This points in the direction that the extraction of non-timber forest products should be embedded in on-going efforts to transfer the managementof forest resourcesto local communities. Medicinal Plantsand Bio-prospecting. Madagascar's status as a mega-diversity country and the presence of highly varied habitats makes it relatively attractive as a provider of medicinal plants and as a bio-prospecting hunting ground. Approximately 4,000 out of Madagascar's total o f 13,000 species have been reportedto have medicinal values, and nearly 50 have been exported commercially (Rasoanaivo and Bodeker, 2000). The most well-known medicinal plant includes Catharanthusroseus (rose-periwinkle) that forms the basisfor achild leukemia cancer treatment drug and a anti-hypertension drug and tranquillizer. Other plants include CenteZZaasiatica (Gotus kola), Drosera ramentacea.and Voacangothouarsii. Despite the importance of these plants and the potential of others, Madagascar currently capturesvery little value-added from the export of medicinal plants. On the one hand, this has to do with the fact that sub-sector is not well- organized which has resulted in overall poor quality of productsthat are brought to the market (caused by poor collection as well as inadequatedrying methods.) On the other hand, this also results from the fact that the permit allocation processas managedbythe Ministryof Water and Forests has not been designed to capture some of the monopolistic rentthat Madagascar should potentially enjoy in view ofthe highly endemic nature of its plant species. To address the first problem area, there is a need to explore mechanismsthat ensure better quality control, possibly linked with the establishment of more controlled production environments such as ex-situ forestry boundary cultivation, village/community garden cultivation or community plantations. As far as the second area is concerned, Madagascarshould consider putting in place an auctioning system ofthe collection permits for those medicinal plants that are endemic to the country. Inparallel, Madagascarshould also participate more actively inthe international trade arena conceming 104 intellectual biodiversity and bio-prospectingrights. The recently formed group of mega-diversity countries would be a natural platform for Madagascarto enter this arena. Harvesting hiah value species for the wildlife trade. There has been some debate on the potential for developing forests as sustainable sources for species of high value inthe wildlife trade, mostly reptiles, many of which are highly prized by collectors. To date, no credible system for the sustainablecommunity-basedharvestingof reptiles has been developedand the possibility remainstheoretical. Such ideas have also been thrown into doubt by the failure of Madagascarto comply with quotas on CITES listed species leadingto a recent blanket ban on export of all CITES listed species. Given the lack of institutional capacity to effectively enforce a system of wildlife trade permits and the considerablereputational risk associatedwith cases of non- compliance, especially so in view of Madagascar'sefforts to capture international willingness to pay through the setting-up of atrust fund for the environment, the harvestingof high value species for wildlife trade should be phased out by putting CITES quota at zero. Forestrv plantations. Plantation projects in Madagascarhave been relatively successful. There exist some 320,000 ha of plantations including 80,000 ha of industrial plantation of eucalyptus and pine. Most ofthese are close or ready to be exploited (> 20 years old). Commercial plantations for timber provide labor and are the basis of a developing furniture making industry. A large area of 40,000 ha at Matiatra is being consideredfor transfer to surrounding communities. The national reforestation initiative (reboisement)proposedthe planting of 2 ha for each commune in Madagascar for the 2001-2002 campaign. In addition, there are numerous diffuse family-sized plantationsthat are quite important for meeting local needs (Topa 2002). The ANAE program of EP2 has proven the potential of forest plantationsto reducepoverty and enhance rural productivity as sources of wood and fuel, for the maintenanceof watersheds and control of erosion (Fara 2002). The ANAE program catalyzed the planting of 20,000 ha of plantation (Topa 2002). In the coastal zone, mangroveshave an important role to play. Inaddition to providing an ecological support function to fisheries (especially shrimp fisheries), mangroveshave potential to produce wood fuel for coastal towns such as Mahajangaand Toliara. There is also potential to develop and valorize plantations and mangrovesfor the purposes of carbon sequestration, permitting the sale of carbon credits in international markets (e.g. Rarivoarivelomanana 2002; Keck 2001). Carbon storage & sequestration schemes. The Kyoto Protocol to the United Nations Framework Convention on Climate Change(FCCC) will soon be renderedactive through ratification by an adequate number of countries. Evenwithout the participation of the UnitedStates, the Kyoto Protocol will create opportunities for Madagascar to benefit from the evolving global market for carbon sequestrationand emission reductions. Additional marketsfor carbon beyond Kyoto are also developing. With regard to forests, the primary opportunities currently lie with the sale of carbon credits for forests that are actively sequesteringcarbon. Under the Clean Development Mechanism (CDM) -the part of the Kyoto Protocol with the most relevancefor financing carbon projects in developing countries-only afforestation and reforestation are considered acceptable sequestrationactivities. Thus, the greatest potential for the sale ofcarbon credits by Madagascar will be with plantation forestry and with energy projects. Nonetheless, there is a strong lobby that wishes to promote carbon credit markets basedon pure conservation, given that forest conservationwill: (i)prevent releaseof carbon that might 105 otherwise be released; (ii)conserve biodiversity, and iii)provide additional ecological services including hydrological services. In order to translatethis notion into practice, the Bio Carbon Fund was launched recently. This Fundwould consider financing carbon sequestrationprojects that go beyond plantationsand include natural forest managementand agro-forestry. However, given the position taken by the Kyoto protocol, and the lack of a clear plan by the U S for international carbon trading, carbon schemes basedpurely on conservation(avoided deforestation) are the least likely to find financing for the time being. Given the general interest in "conservation carbon" it remains a strong possibility that private deals will emerge where corporations pay for the conservationof forests for their carbonvalue. Studies have shown that Madagascar's forests store largeamounts of carbon. The averagecarbon storage capacity of naturalMalagasy forests has been estimatedat about 160t/ha, with values as high as 400 t/ha in wet eastern forests (Rarivoarivelomanana, 2001). The opportunity cost of conserving one hectareof forest in Madagascarhas been estimatedatjust $2.58 / ha (DGEF / CI 2000), whereas the potential revenue generatedfrom the commercialization of fixed carbon has been calculated to be as high as $356 per ha (Pavy 2002 basedon Rarivoarivelomanana, 2001). A particular interestofmangrovesis that sequestrationrates are particularly high andthat carbon storage, when soils are taken into account, is also large. Madagascar's large areas suitable for mangrovegrowth (>3 50,000 ha) representa considerablefuture opportunity for carbon sequestration. Pre-conditions for participation in the carbonmarkets includeratification by Madagascar's national assembly ofthe Kyoto protocol and establishmentof an institutional framework to facilitate investmentflows and approval proceduresfor running a carbon market. The recent fusion of the Ministries for Forests& Environment favors such an institutional mechanism. Madagascarshould pursue aggressively the opportunities presentedby the Kyoto protocol and of both the Bio Carbon Fund and the Community Development Carbon Fund announcedby the World Bank at the Johannesburgsummit in2002. Madagascarwould also benefit from a technical south-southpartnershipwith other developing countries embarking on a similar path. 5.4 Weaknesses & strengths of forest management The fundamental problem for forest management is the poor standard of governance within the sprawling forestry sector. Evidence for poor governance includes: (i) the frequent issueo f illegal permits for timber extraction; (ii)failure to adhereto CITES quotas for CITES-listed species; (iii) awarding forest clearance permits for maize farming (Reau 2001); (iv) the discovery of blank permit books used by ministry officials; and (v) the very poor recovery of official fees (running at about 25% in August 2002, Topa 2002). For decades, illicit appropriations o f forest goods and incomes by political and administrative authorities at the local and central levels have integrated into the normal administrative managementsystem. A high level of tolerance for petty corruption and offenses ofthe forest code exists at the level of forest and non-forest national institutions. A detailed critical analysis ofthe forest taxation and ofthe problems relatedto recovery is being prepared by the new team of the Ministryo fEnvironment, Water and Forests. The general conclusion that may be drawn from these sources andfield observationsis that structural problems exist and political willingness is a necessarycondition, but not sufficient, to come up with generaland sustainable improvements of forest governance. Among these problems we may cite: (i)gaps in the law, a disparity of regulatory texts, and a lack of integrated procedures for the technical and administrative monitoring of recovery; (ii) the inexistenceofprocedures/ systems which are legally defined to prosecuteoffensesto forest regulations, transactions before 106 judgment, the fines on seized or confiscatedproducts, and legal actions; (iii)the separation between the functions of grantingpermits, of monitoringthe exploitation, the transport authorization and the recoveryo f taxes, andthe lack of crosscheckingof relatedinformation; (iv) the lack of organizationalstructure which is speciallytooled and benefits from sufficient technical and financial support to bring sustainable solutionsto the problemof recovery. The persistingproblemsof governance and corruption inthe forestry sector havedriven donorsto circumvent the forests administration in supportingnew structuressuch as ANGAP, ONE and NGOs in forest and biodiversityconservation. However,the powersfor awarding permits and policing infractionshave remained with the forests administration. Given that most of Madagascar's remainingbiodiversity occurs outside protectedareas but within forests managed by the administration,there is a critical needto correctthe governance problems in order to be able to transformthe forests administrationinto a positive force for rural developmen`t and environmentalconservation. Among the strengths of forest management in Madagascarare the extensivecoverage and infrastructures(albeit degraded) ofthe forests administration, its information base(albeit outdated and incomplete)and its comprehensive suite ofpowers regulatingaccess to forests and biodiversity. The human resources ofthe administrationare very considerable, particularly in areas such as plantationmanagementand production. A reformedforests sector couldhave a major positive impacton rural development and biodiversity conservation. As notedabove, the forest sector makes an importantcontributionto rural livelihoods, amply justifying efforts to improve its management and development. The prospectof substantial additional benefits from innovativeuses such as carbon storage, eco-tourismand erosion control providefurther technicaljustification for development of the sector as a contributorto rural development. Constraintson the contributionofthe forests sector to the rural/environmental sector includethe persistentabsence of any forests zoningplan for Madagascar andthe vast numbers of legaltexts, some obsolete, that act as an impedimentto managementofthe sector. 5.5 Creating favorable conditionsfor investmentinMadagascar's forests The most urgent necessity for the transformationandthe growth ofthe forestry sector relevantfor all potentialforest users is to restore favorable sector governance conditions, andto regain local populationsand internationalpartners' confidence. The best way to do this would be through a transparent and participatoryprocessto radically realignthe objectives, strategies and structures of the sector towards biodiversity conservationandthe provision of new set of forest-based environmentalservices. The processwould lead to an entirely new nationalforests policy and structure, supported by a regionally differentiated zoningplan. The revisedpolicies, plansand corpus of laws would include regulationsfor accessto, management and use of forest resources and to the equitable sharingof benefitsderivedfrom biodiversity (inaccordancewith the conventionon biological diversity). A major focus shift will likely follow from the exercises notedabove that will move the forestry administrationaway from its current focus on exploitation towards a more inclusivefocus on multiple use management including a significant development of a conservationdepartment. At the level of local communities, measures shouldbe introducedto encourage community-based structures or the decentralizedcollectivesto initiate andtake part in forest management, if necessaryat the inter-communallevel through structuressuch as OPCIs (OrganismePublique de Cooperation Intercommunale). The practiceof GCF shouldbe extendedand refined, with the use of GELOSEwhere circumstancesrequire it. However, any expansionofthe GCF and GELOSE 107 activities should be accompaniedby clear plans to monitor and supportthe managementof the transferred forests. Tax reform is also fLindamenta1to re-launch sector performance and improve governance. Infact, taxation may be usedto select the partners, modify the behavior ofthe players, facilitate and motivate governance, redistribute rent, set and have the observanceofexploitation quotas, and structurethe control operations. The currenttaxation system does not lend itselfto be effectively enforced and monitored. Recastingthe forest taxation can be anchoredto the decentralization process and can be specially designedto build the capacitiesof the managementand monitoring bodies at the local level and that ofthe territorial authorities. The issue oftaxation, of exploitation rules, of a system to monitor, control, and enforce sanctions, the size ofthe permits, and the bidder selectionrules and attribution of forest exploitations, andthe annual exploitation quotas per region must be studied in keepingwith a coherent perspective and taking account of the institutional capacityat the local and central levels. In general, there is a needto considerably simplify the framework of exploitation and oftaxation by carefully avoiding to introduce rules and prescriptions which cannot be easily enforcedby communities and the forest authority would not be systematically ensured. There is also a need to buildpartnershipsbetween local interest and government institutions to control illegal exploitation and bushfires. Responsibility for monitoring and recovering forest revenues should no longer be the sole domain of DGEF but be subjectto ajoint program betweenthe Ministryof Environment, Water and Forestsandthe Ministry of Finance tlirough the Directorate ofTaxes. The latter Ministry is currently little informed about the functioning ofthe forest sector and of the potential it presents with regard to revenues at the central and local levels, but has a direct interest to ensure that the current short- falls in tax revenue collection tax losses recordedannually interms of forest recovery be reduced. Measures should be developedto enablethe forests administration to contract out the management of large forest blocks for carbon storagethrough national or international tenders, usually requiring involvement of community-based structuresthat would obtain a share of the benefits in return for their contribution to managementand protection. Such options should be extendedto national NGOswith proven capacity inthe types of managementrequired. International entities should likewise be encouragedto invest in forest managementactions based on partnershipswith national and local institutions. In order to effectively capturebenefits from international carbonmarkets, it would be desirable to establish a national carbon marketing board, coordinatedby the Ministry ofthe Environment, Water and Forest and based on international partnerships providing technical assistanceand other necessary support. Such an entity would aggressivelymarket carbon credits for Madagascar and would support Madagascar's representationto the climate change convention. This could follow the model establishedin Costa Rica for the marketing of ecosystem services. 5.6 Recommendations Fundamentalrevision of forestry sector mandatetowards biodiversity conservation and ecosystem services. The forestry sector mandate, policies and institution should be fundamentally revisedto reflect a general orientation away from exploitation towards conservation of natural forest, expansion of plantations and the provision of environmental services. There should be a specific emphasison the efficient capture and equitable distribution of benefits generatedfrom biodiversity, in conformity with the Convention on Biological Diversity and the national biodiversity strategy. Inview of the importance of the protected areas system for the tourist sector, there is particularly a needto develop mechanismsthat would enable ANGAP to capture more revenuesfrom the tourist sector to cover its operating costs. At the 108 same time. the importanceof hydrological services providedby protectedareas shouldbe recognized and reflectedaccordingly in the public budget. Wildlife trade shouldbe discontinued until a system of control can be put in placethat is effective and transparent. Reforestationefforts should be preferably embedded in carbon storage and sequestrationinitiatives with existing opportunities beingpursued aggressivelyby the Ministry. While extractive forestry would remain a significant activity, a process of technical assessment and planningwill be required before extractive forestry could be recommenced accordingto a revisedapproach, based on long- cycle permitsfor management and exploitation, possibly supported by eco-certificationof extractedtimber. A11 non-specializedforest exploitation activitiesshould be gradually phased out and replaced with plantationforestry over the next 10to 20 years. The revisedmandate and policies shouldbe promulgatedas rapidly as possibleso as not to hold up the developmentof a national forestry zoningplan, beginningwith the elaborationof regional forest plans prepared in line with the new policies. Finalization and Extensiono fthe zoning plan amongthe localcommunities. Completion, discussionand disseminationof the nationalzoningplanwith local governingbodies and local communities, accompanied by demarcationand effectivepreservationof areas to be managed permanently under forests. Based on the contributionof decentralizedauthorities via the agro- ecological regional development plans(PRDRs), communaldevelopmentplans (PDCs) and plans relatingto PNAE priority interventionzones (ZPIs) the sector should producea GIs-basedmaster zoning plan for forests, capitalizingso far as possibleon existingwork towards such a plan. The planningprocess should link into national spatial and economic planning beingundertakenby the new `super ministry' for economic programs, landuse planning, transport& public works. The practiceof linking the availability of public financing for community-basedprojectsto environmentalperformanceindicatorsembedded inthe PCD shouldbe continuedand expanded. Ensuringgood governance in tlie forest sub-sectorvia suitableprocesses. Given the proposed radical transformationof the forest sector, measures for promotinggood governance would be built into the new institutional design, including: (i)the design of transparent, competitive systems for selectingcompanies interestedin commercialforest operations; (ii)simplification and consistencycheck of regulations, including modalitiesto actually enforcerights and obligationsand apply penaltieseffectively; (iii)the developmentof atransparentplanto discontinuecurrent loggingcontractsand other forest use concessions over a periodof three years, and to allocate newly-designedpermitsto qualified candidates; (iv) a forest taxation reform and creation of ajoint forest revenue securityprogramby the Ministries incharge of Forestsand Finance; and (v) a human resources management planthat reflectsthe direction ofthe proposed reform cited above. Imuroving forest managementthrough community-basedapproaches. The reformedforest sector should make a much stronger connectionwith local users and managementthrough acceleration and quality improvement o f community-basedmanagement systems and participation of decentralizedcollectives (Le. Communes) in forest management. Again, this would be fully integrated intothe new sector design. PADR to avoid agriculturaldevelopmentbasedon area expansionthat threatens natural forest. The PADR shouldadopt an explicit working principle of `zero tolerance' to agricultural developmentbased on area expansionwhich threatens naturalforest or agricultural soils. This would necessitate, for example, careful examinationof supportto maizeproduction inwestern areas and productionof damagingcrops such as ginger inthe east (Brand undated). 109 ANNEX 6. Setting the InstitutionalStage After having focusedon the questionofwhat could or should be done to improve rural productivity in a pro-poor and environmentallysustainablemanner, this chapter centers on how to go about this from an institutional perspective. Its purpose is to provide guidance on possible realignmentsthat are requiredfor better targeted and moreefficient public functions & services in the rural / environmentalsector. Account is taken ofthe current situation and trends as well as the opportunities & constraintsthat have been identified. Specific attention is givento the ongoingtrend towards withdrawal ofthe state from economic activities, decentralizationofthe public sector and the recent restructuringofthe government. The analysisand recommendations are grouped around four themes: (i)sector programframework (ii)ministerial functions; (iii) participationand empowerment, and (iv) partnerships. 6.1 SectorProgramFramework Recent governmentrestructuring. The government of Madagascar has undergone a substantial restructuringin January 2003 that is highly significant for the rural / environmental sector. Key changes have been: (i)integrationof economic programs, landuse planning, transportand public works into a single `super-ministry' under the vice PrimeMinister; (ii)combinationof Agriculture with Livestock and Fisheriesinto asingle ministry; and (iii)combinationof Waters & Forests with Environmentinto a single ministry. These changes are importantto the rural/environmentsector for various reasons: - The fusion of economic programs, landuse planning, transportand public works will facilitate an integratedapproachto nationalspatialdevelopmentplanningand represent an importantopportunity for the sector to incorporaterural and environmental dimensions into nationalspatial planning; - The combination of agriculture, livestock and fisheries regroupsthe `.food producing' sectors and should facilitate a greater emphasis within the fisheries sector on food security, rural developmentand poverty reduction as a complementto the establishedorientationon generatingrevenues from fisheries exports and licenses; - Fusionof forestswith environmentmay be seen as aradical moveto create atransformed forests sector orientedtowards conservationand biodiversity as opposedto extractive production. This should greatly facilitate the developmentof conservationprogramsoutside protectedareas, improvedsector governanceandthe efficient captureand distribution of benefitsfrom biodiversity. National RuralDevelopmentPlan. Madagascaradoptedthe National Rural DevelopmentPlan (PADR) in 2000 as the principal framework instrumentfor promoting rural development in Madagascar. The PADR is a general framework comprisingfive objectives, five thematic orientations,and specifies 13 strategic axes and 28 programsof action. The five objectives are: (i)ensurefoodsecurity; (ii)contributetoeconomicgrowth; (iii)reducepovertyandimprove livingconditionsin rural areas; (iv) promote sustainable managementofnaturalresources; (v) promotetraining and information for improving rural production. Following these objectives, the Plan consists of five thematic orientations: (i)bettermanagementof rural sector through legal & institutional reform; (ii) encouragethe emergenceof economic operatorsas partnersto rural development;(iii)expand & promote agricultural productionwith optimal use of resources & 110 . . infrastructure; (iv) ensure food sufficiency in all regions; and (v) developsocial infrastructureto ensure access to social services. Action under the PADR is based around 20 agro-ecological regions. For each regionthere is a steering committee(GTDR)comprisingfive colleges (local authorities, decentralizedtechnical services, private sector operators, producer associationsand NGOsiprojects)with the mandateto coordinateimplementationofthe PADR in the region. All 20 regions have been subject to agro-ecologicalassessments and actionplanshave been developed for most regions (Figure6.1) 111 . Figure 6.1: Map of GTDRs 112 * It is important to underlinethat the PADR is a policy for rural development in acomprehensive manner. The importance ofthe PADR is that it has put rural development firmly on the development agenda as a key engine for endingpoverty in Madagascar. At the same time, when the institutional framework for rural developmentwas much more fragmentedthan it is today, the PADR provided a useful platform for institutional coordination. Donors have recognizedthe importance of the integratedapproachtowards rural development as laid out by the PADR and have lined up specific projectsthat support implementation of various aspects of the Plan. However, since the PADR is relatively new, it is somewhat prematureto assess impacts of the efforts at this point in time. Key weaknesses ofthe PADR are: (i) lack ofspecificity in particular on the priorities and modalities of implementation; (ii)ambitious scope by also including the social dimension (health, education) of rural development; (iii) weak linkage betweenthe strategic orientation ofthe Plan and the budget allocation process ofthe participating ministries; and (iv) absence of systematic evaluation efforts to link progresson the ground ofthe various sector ministrieswith the goals and targets ofthe Plan. In general, PADR exists somewhatas a passive (`demand driven') framework with `wish lists' and lacks an active institutional champion. National Environment Action Plan. The GOM, with support ofthe international donor community, initiated in the !ate 1980s what has been perhapsthe most ambitious and .comprehensive environmental programto date in Africa. The National Environmental Action Plan (PNAE) was given legal power by adopting the National Environment Charter and the National Environmental Policy in 1990 (Law 90-033, December 21, 1990). The PNAE was adopted by the Government of Madagascar in 1989, while implementation started in 1991with the support of a broadcoalition of bilateral donors(Germany, France, Switzerland, USA), international agencies (GEF, IDA, UNDP) andNGOs (WWF, ConservationInternational). The Plan, which was put into operation in 1991, recognizesthe link between environmental protection and economic development and includes six elements: (i) protecting and managingthe national heritage of biodiversity, with a specialemphasison parks, reservesand gazettednatural forests, in conjunction with the sustainable development oftheir surrounding areas; (ii)improving the living conditions ofthe population through the protection and management of natural resources in rural areas with emphasis on watershed protection, reforestation and agro-forestry; in urban areas this would involve improving water supply and sanitation, waste management and pollution control in general; (iii)promoting environmental education, training and communication; (iv) developing mapping and remote sensingtools to meet the demandfor natural resourcesand land management: (v) developing environmental researchon terrestrial, coastal and marine ecosystems; and (vi) establishing mechanismsfor managingand monitoring the environment. The PNAE was designed from its inception as a fifteen year investment program divided into three five-year phases. The first five year phase aimed at creating a proper policy, regulatory and institutional framework so as to generate the conditions for genuine country ownership ofthe environmental agenda which prior to the PAE usedto be set and driven bythe donor community. The second phase of the PNAE aimed at consolidating the programs initiated underthe first phase by putting the establishednational institutions firmly inthe driver's seat. The third phase, which is currently under preparation, is focused on consolidating the varied past efforts and establishing sustainablefinancing mechanismsfor the environment. Unlikethe nation-wide approach adopted by the PADR, the strategic approachof EP3 is to focus actions on priority intervention zones (ZPIs), selected on the basis oftheir importancefor biodiversity (Figure 6.2). 113 114 Major achievements of the PNAE include: (i) the enactment of enabling legislationfor the protection of country's natural resourcesandthe promotion of proper environmental management; (ii)the set-up of environmental institutions(such as the park serviceANGAP) for the implementationo f environmental activities and programs; (iii)the developmentand implementationof community-basedapproaches for naturalresources management; (iv) the emergingevidenceof positivefield-level impacts; and (v) the establishmentof a platform for sustained donor support and coordinationfor the environment in Madagascar. However.there are numerous areas where the PNAEcould improve its track record. As mentionedearlier, the application of policies and regulationsremains achallengedue to weak institutional capacity and serious governance problems, particularly in the forestry sector. Resourcesunder the PNAE have been disproportionally investedin parallel structuresat the central level, while too little has been investedto strengthen institutional capacity on the ground. Lack of rigorous priority setting has also ledto a situation in which PNAEtends to drift somewhere betweenconservationand rural development, sometimes seekingto fill gaps that PADR now seeks to fill. Consequently, there is the notionthat the operational programsof the PNAE have spreadthemselvestoo thinly, thereby contributingto the widespreadfeeling that more could have been achieved than actually has been. The challengefor mainstreamingofthe environmentalagenda is reflectedin: (i)the relatively modestbudgetallocations for the sector; (ii) the existing limited knowledgeand awarenessofthe Malagasypopulationconcerning environmental issues; and (iii)the slow development of market mechanisms for the valuation of environmentalservices. Need for more Focus, Coordination and Synergv. In view ofthe large rural dimensionof Madagascar's poverty agendaand the country's mega-biodiversityassets, the existence of separate strategic programframeworks for rural development and the environmentprobably makes sense. However, as bothframeworks stand now it appearsthat they are bothtoo ambitious, while at the same time the division of responsibilitiesbetweenthe two has not been specified. These are reasons for concern because, ifnot addressed, they will complicate achieving resultsofthe ground, while runningthe risk of duplicating scarce resources. Previous experience under PNAE has clearly shown the needfor focus. Perhaps the best example of the needfor s.vnergy is that thefocus of conservation has been historically onprotected areas, the forestry administration hasfocused on exploitation, developmenthasfocused on irrigated rice andyet the most serious threat to all three -tavy -requires a massive coordinated effort and has not nearly been adequately addressed. The case for a more focused PNAE could be adoptedby redefiningthe core objectiveofthe Plan towardsthe conservationand valorization ofnaturalresourcesthrough participatory modelsin zones of high biodiversity threat and importance(principally eastern escarpment and dry western forests). By positioningthe Planas such, biodiversity conservation as well as poverty reduction in selected rural areas could be achieved simultaneously. Issuesof environmentaldegradation that are not critical for biodiversity, such as soil loss on established and transformed agricultural lands should be regarded primarily as a concernof agriculturaldevelopment, althoughcertain specialist activitieswith a specific environmentalobjective such as tree plantationsfor carbon storage would merit coordinationthrough the PNAE. However,the PNAE shouldaccept that it alone cannot hopeto achieve sustainable naturalresourcemanagementand conservationin all of Madagascar andthat coordinationwith other programsis necessaryto achievethis. As part of such coordinationefforts, PNAE could play a usefulrole to integratethe environmental 115 dimension of development into public sector programs, by mobilizingstrong capacity in biodiversity information management and conservationplanning. As far as the PADR is concerned, the need for focus equally applies. Given the importance ofthe primary sector for livelihoods in all of rural Madagascar,the actual national coverage of the Plan seems appropriate. As said, the inclusion of the social dimension of rural development appears overly ambitious, especially so as the correspondingsocial sector ministries have never taken up an active role in the PADR or showedany special interest. Consequently, a thematic focus would most probably position the PADR best as a key identifiable element of Madagascar's poverty alleviation strategy. In this context, it should be made clear that theprimary role of PADR is the delivery of enhancedruralproductivity andfood security through intensijication & diversiJication while ensuring maintenance of natural capital and the natural environmental services that support agriculturalsystems (e.g.watersheds). Application ofthe PADR at the regional level will highlight the priorities andtrade-offs to be considered for each region. However, the GTDRs, as voluntary committees without access to specialist assistance or advice from the central level, willjudge on the options alone and there is no mechanismwhereby regional experience will be recordedand fed back into refining the national vision. Sharpening PADR thus requiresestablishingprocesses whereby it builds upon its experiences, interactswith other actors and develops its own sense ofvision. Itneedsto develop more of its own dynamic, similar to that which has developedwithin PNAE. Increasedassistance to the GTDRs will be necessary to achieve this aim. Given that both the PADR and PNAE operate inrural space, the needfor close coordination betweenthe two is evident. As a first step, this would require that both Plansadopt the same vocabulary or at least agree on a dictionary that clearly indicateshow GTDRs and ZPIs relate to each other. At the same time, the PADR shouldadopt a principle of no tolerance to increased agricultural production through expansion into primary forest and coordinate with the PNAE on the maintenanceof biodiversity and the captureof benefits from biodiversity. Althougli the need for focus in both PADR and PNAE would position both programsto deliver results more effectively, it is most probably not sufficient to establish a dynamic rural / environmental sector as an engine for inclusive and sustainableeconomic growth that benefits the rural poor. As pointed out on numerousoccasions, the needfor improving rural roads infrastructure is a critical factor to achieve poverty reduction, improve agricultural productivity, increasetechnology adoption and pave the way for improved environmental managementand benefit capture from biodiversity. It is therefore important that the principal actors of PADR and PNAE shouldjointly and actively seek a role in national land use and infrastructure planning in cooperation with the new `super-ministry' of economic programs, land use planning, transport and public works. Such new national land use plan should fully integrate considerationsof poverty reduction, food security, rural development, environment and biodiversity and constitute the spatial reference instrument for the actions of all national programs, thereby setting the stage for a fully integratedapproachthat seeks to maximize synergy betweenthe various sectoral interventions and programs. The national land use plan should be developed using bottom-up and top-down approachesworking at the levels ofnation, regions and communes. The agro- ecological regionswill play a pivotal role through the GTDRs in integrating communal priorities into regional development plans and making the link to the national land use planning process. In line with these observations, the division of responsibilities betweenPADR and PNAE as well as areas for joint action and coordination with the new Ministryfor Economic Programs, Land Use Planning, Transport and Public Works could presented along the lines as indicated in Table 6.1. 116 Table 6.1: Sector ProgramFramework Program Rural developmentI PADR ~~ ~ ~ Environment & forests I PNAE Focus on increasingproductivity of the rice sub- Undertakeand commit to rigorous biodiversity sector and non-rice staplesthrough priority settingand risk assessment measuresappropriateto each region Identify priority zones for intervention precisely Generatefood surpluses of key staples (rice, Maintain and developthe national protectedareas cassava. maize. potatoes) network Building blocks Promote wide applicationof measuresto reverse Developenhancednational biodiversity strategy for to mobilize degradation of agriculturalresources(soil capturingbenefitsfrom biodiversity conservation. anti-erosionmeasuresetc.) Developcarbon storage and carbon sequestration Ensure reasonableproducer pricesof food crops strategies Provideincentivesfor intensiveproduction Developsustainable financing mechanisms(trust methods funds, eco-taxes) Developcash food crops for domestic markets Integratebiodiversity conservationinto PRDRsand through producer-operatorpartnerships PCDsof priorityregions Progressivelydevelopcash'crops for export Assess and publicize the values of biodiversity to Adopt an approchefiliere for all cash crops Madagascar Implementand promotethe development of Includehydrological benefits in conservationpriority nichemarkets through aNatural Product of setting Madagascar label and associated sustainabilitystandards(ecology, fair trade etc.) PADR & PNAE Jointlv & moactively: jointly Harmonize implementation(ZPIs, agro-ecologicalregions,GTDRs, commmunes etc.) Contributeto land useplanning, economic planning, transport infrastructureand public works in partnershipwith the new `super-ministry' for these sectors Jointly seek to tackle the problem of tuvy Economic Furtherdevelop andrehabilitateroadnetwork programs, Emphasizerural roadsbasedon HIM0 land use Launch landtenure reform and support program(abolishrule on "mise en valeur"; create new planning, agricultural landrentalagreements, developcollective titles (SFO, SFI) transport and public works In the event of linking PADR and PNAE to create a combined rural/environment sector program which is governed by the three perspectives-poverty (& food security), rural productivity and environment/ecology and further supposingthat an eco-regional approach is adopted, the priority should be on those regionswith rural poor which display the lowest agricultural productivity and who present the greatest threat to biodiversity. The allocation of priority setting, assuming poverty reduction as the priority, can be illustrated as follows: The classification could be refined by quantifying parametersor adding others such as food security. The `trade o f f would be the possible lost opportunities for poverty reduction, productivity enhancementor biodiversity conservation in lower priority areas. Inpractice, each perspectivemay be governed more by its internal priorities than those external to it. Thus, the PNAE might still prioritize threatened areas of important biodiversity, even ifsurrounded by rich, highly productive farmers (or even an especially valuable wilderness area empty o f people), while 117 PADR might prioritize poverty-stricken areas of low ruralproductivity even where they were of low biodiversity importance. 6.2 Sector Ministerial Functions Ministry of Agriculture, Livestock and Fisheries With the integration of all food producing sectors into a single Ministry,there is a unique opportunity to truly realign the budget allocation and functions ofthe new Ministry with the rural productivity orientation o fthe PADR. Budget Allocation. An analysis of budget allocation and expendituresthat was carried out as part of the study beforethe recent reorganizationwith the objective to assess their linkage with the strategic orientation of the PADR, provides a useful point of departurefor such realignment process (Rarivoarivelomanana 2002). In general terms, this analysis confirmed that budget allocation should be better brought in line with the PADR and that better provisions should be made to cover the recurrentcosts of initiatives launchedby projects and programs (Table 6.2) Table 6.2: Reco imended expenditure realimments for the amicultural sector Problems Recommendations Budget processis - Define a more precisepolicy concerning budgetary allocation not linked with aimed at achieving PADR targets PADR orientations - Specify operationalbudget in accordancewith each strategic axis I - of the PADR AIlocate more operationalbudgetto decentralized services (especially secondaryregions), the amounts varying according to each region's specific needs. Compositionof GOM should provide more financial assistance to activities that agricultural sector relateto objective # 3 of PADR (increase and promote budget not conform agricultural production with optimal use and sustainable with PADR managementof resourcesand infrastructures) and back these activities upwith conservationactivities (watershed protection, vetiver plantation, agroforestry More effortdfunds should be put into strengtheninglencouraging the participation of and collaboration (orientation 2 of PADR) with private entities in the prospect of sustainable financing of PADR activities and withdrawal of the State from production activities Promote/encourageactivities that develop a supply chain approach and those which promote the professionalism of producers, in the context of the MinAg's transitionary functions Improve rural finance performance Compositionof - Allocate more funds to support ongoing activities inrelation to livestocksector the PADR's objectives, (especially institutional capacity budget not conform building and diversification o f production) with PADR - Dedicated support to dairy and forage productions at the national level, as well as to animal feeding and small livestock production in selected regions. 118 Core Functions: Ayriculture. Numerous studies (Berger, 1998; Fusilier, 2001) have been carried out in the past which, in line witli the overall principle ofthe withdrawal ofthe state from direct economic activities, emphasize the need to rationalize functions and to greatly reducethe size of the Ministry's agricultural operations. In line with these studies and in view ofthe strategic orientation of the PADR, a number of agricultural core functions could be identified. First amongthese would be agricultural and food security policy analysis and formulation so as to obtain an adequate handle on the incentive framework for private sector investmentsas well as benefit capture by the various stakeholders in each ofthe relevant production chains. The establishment of the Policy Unit for Rural Development(UPDR) inthe Ministry following the launch of the PADR has been a step in the right direction to establish capacity in this field. However. there is a need to strengthenthis Unit, while assuringthat its agenda better reflects the sector implications of the poverty reduction strategy adopted under the PRSP. Closely associated with this core function is the need for the establishmentof adequate agricultural statistics to allow for monitoring and assessment of sector performance. The last agricultural census has been carried out in the mid-1980s. The lack ofreliable statistics has affected the analytical capability and credibility ofthe Ministry and thereby its effectivenessto take a lead inthe setting the sector policy agenda. Lack of intellectual leadershipcoupledwith diffuse institutional responsibilities are among the factors that have hamperedprogressin this field. There is a need for external assistance with a solid track recordto address this problem on an urgent basis. A second field would be the provision of adequateresearchand extension services. Overall support to agricultural research has been falling and total spendingper scientists has more than halved (Neinkema, forthcoming). Agricultural R&D also remains heavily dependenton donor contributions, totaling about 60 percent in 2000 for FOFIFA. The latter raisesthe question of how priorities for researchare being established and highlights the needto explore the feasibility and relevance of demand-drivenagricultural researchfunding schemes. As far as extensionactivities are concerned, there is a need to ensure progressivetransfer of the SNVA (Systeme National de Vulgarisation Agricole) to communes. Since the allocated budgetbythe GOM to the Ministry will never be sufficient, a rationalization at the central and decentralizedlevel is required. Priority should be given to communes that are willing to cooperate with the Ministryand which have high agricultural potential. Additionally, the administration should elaborate an incentive strategy to attract private sector investments in agricultural extension. GTDRs (20 nationwide) play a vital support role to SNVA and should be given support to perform this role. The pivotal role of GTDRs as intermediaries betweenthe Agriculture administration, decentralizedservices, extension workers, private operators and producer organizationswould helpto ensure the transition to private sector participation. In light of on-going demographic trends depicted inchapter 1, which increasingly leadto a further separation of production and consumption function, the importance of establishing and ensuring compliance witli phytosanitary norms and regulations will grow. Ongoing liberalization and integration of the domestic market witli international markets, resulting in increasedagricultural imports and exports, will reinforce this trend. Establishing institutional capacity in this field should be considered a core function of the Ministry. Besidesthese core functions, a number oftransitional functions fall within the agricultural mandate of the Ministry. Within this realm, there is a need to finalize the unfinishedtransfer of irrigation infrastructure to water use associations ( W A S )within a watershed managementframework. Along the lines suggested in chapter 4, the Ministryhas an important role to play and must collaborate with the land use managementauthority. Among the services it should deliver are 119 extensionservices and securingof landtitles for the irrigatedareas (with the assistance of the Service des Domaines). After the formal transfer o f irrigation infrastructureto W A Sas a first step, required public support for the rehabilitationof irrigation systems shouldbe managed entirely by W A S,whoenter into contracts with private sector suppliersfor this purpose. Provision of technical and management training of farmers' organizationsandwater users' associations is also a function the Agriculture administrationshould support as partofthe transfer process. Another transitionalfunction would be the provisionof productiveassetsto the poorest segments of rural livelihoods as part of the country's poverty reductionstrategy to generate enabling conditionsfor inclusiveeconomic growth. The overall aim of efforts in this field would be to increase rural productivity and incomes through the financingof demand-drivenactivities on a matchinggrant basis which will eventually improvecreditworthinessof beneficiariesand facilitate their accessto the formal financial system. Activities aimedat improving synergywith rural financial institutionsand agribusiness should be internalizedunder this effort. In line with structuraladjustment, the government haswithdrawn from seed production, removing its subsidies to seed productioncenters (Centres Multiplicateurs des Semences) and privatizing all public seed producingunits. However, the developmentof a private-sector seed supply chain is hinderedby the lack of interactionbetweendifferent stakeholders (farmers, researchers, extension workers and operators). Support to a process aimed at improvingthe performanceofthis critical supply chain might be warranted for a limited time. A number of other functions should be removedfrom under the Ministry's umbrella. First, acknowledgingthat the public sector does not have a direct roleto play in economic activities, imply that HASYMA and SIRAMA shouldbe privatizedwithout delay. Second, in line with the proposed approachto the transfer of irrigation infrastructureandthe empowermentof WUAs inthe subsequent rehabilitationprocessconform amarket-based mechanism, the need for in-house engineeringcapacity will effectively disappear which meansthat the rural engineeringsection (Genie Rural) ofthe Ministry should be set on its own, privatefeet. In line with the withdrawal of the Ministryfrom the irrigation infrastructurescene, water management structuresthat should stay in the public domain becauseof public safety considerationsshouldbe put underthe umbrella of Public Works. Core Functions:Livestock. Beforethe recent reorganizationofministries, it was recommended that the Ministry of Livestock focus on the following functions (Fusilier 2001) which are primarily regulatory and therefore still applicable: Functions Strategies Policy orientation Redefinethe main orientations -ActivitiesCoordinationof central and decentralized of the livestock policy entities' activities Define legislative framework - Elaboratelawddecrees Gain the trust of international - Promoteactivities with potential for improving agencies food securityand poverty reduction Plan investments Produceand publish national statisticson Liaise with other ministries in livestock charge of rural development Set up monitoring and evaluation system Produceand publish livestock -- Economicanalysisof the sector (material and statistics non-materialinvestments,recurrent costs...) Monitor national strategies for livestock developmentand act accordingly Support to Perform studies on dynamics I - Incollaboration with other stakeholders, 120 supply chain of supply chains elaborateand launch a national policy on genetic (filieres) Support activities in favor of improvement. species selection. ... conservation of livestock EIA o f animal production genetic assets Elaborate a labeling system Support to natural resource management Sanitary standards Define quality standards Set up national epidemiological surveillance Improvement of animal health system Safeguard hygiene standards Monitor impacts of biological products on of animal products animals Monitor widespread prophylaxy Monitor privatization of veterinary professions Application of quality norms Source: Fusilier, 2001 In addition,the following specific recommendations were made regardingtechnical supportto the sector: 1. The Livestock administration should withdraw from all research activities and delegate them to FOFIFA, which, due to its close established links with producers, would be better placeddefine appropriateresearchthemes. The GTDR at the regionallevel, shouldalso be made responsiblefor identifyingregion-specificresearch needs. 2. Making producers more professionalshouldbe a priority inthe livestock sector. Unfortunately,due to lack of funds, the "Direction de 1'Appui a la Professionalisation des Eleveurs" (DAPE) cannot, at the moment, performthis activity. 3. Establishmentof a Livestock producers' federation (Federationdes Groupementsde Eleveurs)would considerably facilitatethe monitoring and evaluationof regionalprojects and programs.The Ministry sliould put particular effort into helpingthe creationof such a Federation. Core Functions:Fisheries. Essentially, the missionofthe fisheries departmentis to administer, manage and developthe sub-sector (Table 6.4). In its administrative role, it is incharge of issuinglicenses and other operatingpermits, whether for fishing, aquaculture, collection or marketing. In its managementrole, it is responsiblefor monitoring operations and resources, assessing stocks and stock potential, and draftingworking plansfor the various segments in each subsector. Management plans exist so far only for the shrimp sector, driven by private sector and donor demand. There is no overall fisheries management& developmentplan. No major revisionofthese functions appears necessary.However,as a relatively cash-richadministration because of the Fonds de DdveloppementHalieutique(FDHA), a special accountwith the treasury receivingpart ofthe licenserevenues, there is considerablescope for aligning expenditures accordingto the needs of the sector. Fisheriessurveillanceis conductedby a quasi-autonomous surveillancecenter funded by the EU and the FDHA which is actively seekinggreater administrative & financial autonomy and an expanded role in endangeredspecies protection. The center represents Madagascar concerningregionalcooperationon offshore fisheries surveillance and management. 121 Function Activities Management of fisheries ------ Participate in the Interministerial Commission on fisheries Support provincial consultative fisheries' councils Prepare & maintain fisheries managementplans Regulating fishing activities Define regulations for fishing within management framework Propose parks and reserves where fisheries are strictly controlled -- Regulatethe protection o f marine mammals and endangered species Control the use of illegal fishing methods Regulatethe protection of species in mangroves & intertidal zones Providing the legal regime for - Control fishing activities fisheries & aquaculture Control licensing of vessels Control aquaculture activities Sanitary & quality control of Regulate creation & function of fisheries treatment plants fisheries & aquaculture Applq control measures in collaboration with other ministries products ------ Control import o f eggs, larvae, alevins (fry) and livingaquatic plants Control export of fisheries & aquaculture products Policing fisheries & Investigate and document infractions aquaculture / enforcement of -- Ensure application ofpenalties, confiscations etc. rpviilatinni There exist a number of opportunitiesto betteralign expenditureareas with the strategic orientationof the PADR. Need for a national fisheries managementplan basedon regional& thematic plans. The absence of a global plan for fisheries is an obstacle to transparence and sustainabilityofthe sector and is a major omission in sector functions. Such a plan is requiredto bring fisheries into line with national policies including PRSP, PADR and PNAE. Frameworkmanagement plans developed for Toliara and Nosy Be would be amongpossibleregionalcontributionsto such aglobal plan. The plan would include specificthematic plans for highly exploitedresources such as sea cucumber, sharks & lobster. An initiative ledby GAPCM and fundedby AFD to develop regional shrimp fisheries managementzones (ZAC) would contribute further piecesto the `jig- saw`. Need for a suuport fund for small and medium-sizedfisheries entemrises. It is of primary importancethat all small and medium-sizedenterprises-which handle collection of the products supplied by traditional fishing throughoutthe islandnation-will be able to overcome the difficulties presented by upgradingto meet quality standards, ifthis branchof fishing is to grow and become sustainablymanaged. Currently, EDFETABEX financing is inthe processof being implemented. The Fundwill providetotal financial aid ofFMG 5,191 million for 26 eligible small and medium-sizedfirms. The amount of individual aid for each business varies from FMG 46 million to FMG 483 million. These 26 small and medium-sizedcompanies together account for annual sales of FMG 75,193 million, FMG 57,982 million of that from exports. The fund could be established with resources from FDHA and aid from donors. This would makethese small and medium-size businesses operational at the export level. The upshotwould be a reactivationof traditional fishing through an increaseddemand for its products. Taking the amount of the STABEX fund as abasis, about FMG 10 billion will be neededto bring the remaining50 small and medium-sizedbusinessesinto compliancewith the standards, whereas anticipated sales would reach FMG IO0 billion per year. Creation of regional support centers for traditional fishing. Such centers are neededwithin all existing marine districts. They will needto handlethe following tasks: (i) provisionof fishing 122 gear (transactiontype still to be defined); (ii)training in product preservationand storage; (iii) setting up a system of loans aimed at fishermenand mareyeurs; (iv) potentialprovision (with donor funding) of ice p1antsl6.The organizationalnature of these centers is of little importance, providedthat they have access to all needed resources(human, financial and material). This could be achieved by extendingthe actions currentlyunderwayas apilot programby the fishery componentofthe UNDPIRPPMED'' project in the provinceof Toliara, by the GTZ (German Agency for TechnicalCooperation) in the Nosy Be regionor, inthe case ofthe southern and western coasts, by the African DevelopmentBank `integrated littoral developmentproject' (PADIL) due to come on stream in 2003 or 2004 (supported by funds ofUS$11 million). Openingup of coastal zones. Since it is necessary to open up isolatedzones nationwide, this can only be done over the longterm. Inthe immediatefuture, openingup these areas through sea routes would seem to be a good alternative. The fishing community alreadydoes this on a daily basis, but under highly precarious conditions(dugout canoeswith paddles). The role of the Government will therefore be to provide incentivesfor less risky and faster means oftransport (e.g. a coastal marinetransportsystem based on private motorizedlaunches). Ministry of Environment, Water and Forests The recent merger ofthe Ministry of the Environmentandthe Ministry of Water and Forests into a single Ministry providesa uniqueopportunityto put the managementof naturalresources on a truly conservation-orientedfooting yet creates an enormous institutional challengeto successfully integratetwo,basicallydifferent mind-setsand approaches. The aim ofthis section is to lay-out the agenda of issuesthat needto be addressedto effectively seize the opportunity and meetthe institutional challenge. It is envisagedthat an upcominginstitutional assessment ofthe sector as part of the on-goingpreparationefforts of the third phaseofthe PNAE will provide moredetailed answers and directions. Origins Environment Institutional Framework. The Charter for the Environment(law 90-033) providedfor the creation of a national structureresponsiblefor developmentofnational environmentalpolicy. Investmentsmade under the PAE from 1991 untilto date have ledto the establishmentof a numberof environmentalinstitutionsall aimedat conservationof the country's natural resources. New institutionsthat were created includedthe National Office ofthe Environment(ONE) charged with coordinatingthe implementationofthe PAE, supportto the development o fenvironmentalpoliciesas well as the evaluationof EnvironmentalImpact Assessments (EIAs). A hybrid NGO-parastatal,the National Association for the Managementof ProtectedAreas (ANGAP) was created and chargedwith the management of protectedareas system. The National Associationfor EnvironmentalActions (ANAE) was createdto undertake soil conservationand household-leveldevelopmentactivities. SAGE was pulled out of ONE more recentlyto provide supportservices for environmentalmanagement, particularly in coastal zone regions. The National EnvironmentResearch Center (CNRE) was created for the purpose of carrying-outenvironmentalscientific research. The Ministry of Environment, responsiblefor the developmentof national environmentalpolicy was somewhat of a late arrival on the environmental scene, but has progressivelymovedto the center ofthe environment sector institutional stage and is now mergedwith the Ministryof Water and Forests. l6 Ice plantsprovide an excellent means for monitoringfishing effort andencouragingcompliance by fishers with licensingregulations etc. -access to low cost ice is conditional uponan up-to-datelicenseand provisionof informationabout catch, effort etc. "RPPMED: PovertyReductionandPromotionofSustainable Livelihoods(Reductionde laPauvretCet promotion d'une Moded'Existence Durable), MAGi98007. The fishery componentofthis programwill be completedat the end of 2002. 123 Origins Water and Forests Institutional Framework. The forestry administration was originally created in colonial times as a production sector and this orientation persists in the procedures and operations o f the sector despite recent laws and policies that orient the sector more towards conservation. Under decree 2002-821 the Ministry for Water and Forests is responsiblefor the conception. implementation and coordination of the national forests policy. The administration is based on a central ministry, general direction, 6 provincial directions and 19 district offices, staffed by 1 102 personnel of which 28% are centrally based. A new administrative structure was established in 2003, that creates two General Directions: one for Water and Forest and one for Environment. Under the general direction of water and forests are three directions (valorization of forest resources. biodiversity conservation, and technical assistance). The 6 provincial directors will take responsibility for all actions of the combined ministryat the regional levels. Law 97-0I7 and related decrees have revised and reorganized the forestry regime. While the policy and laws are in line with the recent laws and policies on environment and natural resources. the forestry policy lays down only vague orientations which lack strategic direction. The sector thus remains vulnerable to the contiiiuation of former production orientated practices, as permissible within the policy's four principal axes: (i) end degradation o f forests; (ii)improve management o f forestry resources; (iii)increasethe area and potential o f forestry; and (iv) improve the economic performance o f the forestry sector. Agenda to be Addressed. The principal challenge o fthe merger would be to mainstream conservation as the guiding principle for natural resources management throughout the new Ministry. Doing so effectively would require a newhumanresources skill mix inthe Ministry; something which appears difficult to achieve with the current pool o f Water and Forest Department employees in view o f their educational background and working experience. Besides bringing the human resource skill mix in line with the new strategic orientation, there i s a need for systematic thinking on how to effectively change an institutional culture based on exploitation to one o f conservation. An essential element o f the integration effort would also be to define the optimal balance of resource allocation between the central level and the frontline. In both Ministriesvery little investment has in fact been effectuated at the field level duringthe course ofthe PNAE. With increased emphasis on measuring impacts on the ground, it will be necessary to increase investment on the ground as a means to achieve concrete results. Inaddition to the challenges that are immediately related to the merger, there is also a need to address internal institutional issues at the same time. As far as the Water and Forests Department is concerned, these institutional issues have been presented in detail inthe previous chapter. As far as the Environment Department is concerned these issues have been around for some time, but have never been addressed systematically and decisively. First, there is a need to define once and for all the exact division o f functional responsibilities between the Ministryand ONE. Second, there is a need to specify a results-based relationship between the Ministry and the various AGEX institutions that operate in the sector. These agencies have been set on their own feet recently, but require continued direct donor support. Third, there is a need to define the division o f responsibilities between the Ministry o f the Environment and the environmental units in the various sector ministries as far as compliance with environmental legislation is concerned. In order to present some sort o f a mindset to approach the institutional challenges that lay ahead and in anticipation of the results of the upcoming sector institutional analysis, a possible specification o f institutional functions and services inthe environment sector is presented in Table 6.5. This specification leaves open the question whether there is a need for a separate 124 Environment Department and a separate ONE, or whether the two should be merged to mirror the institutional set-up o f the Water and Forests Department. Table 6.5: Institutional Structure of Environment Sector Institution Function Ministry o f Environment, Waters, Forest policy & regulations Forests MinENV.EF) Forestsinventory & statistical service Direction for Water and Forests Managementof forests & plantations Issue of forest exploitation permits Policing of forest & wildlife laws Control of wildlife trade (CITES etc.) MinENV.EF Environmental policies & regulations Direction for Environment Env. Permits for investmentprojects Environmental pollution regulation & control CIME Interministerial committee on Inter-ministerial coordination on environmental environment issues, chaired by Ministerresponsible for environment HCNN HighNational Council on Ultimate body for approving protectedareas and Nature on wildlife issues CNE National Environment Council 1I deciding Consultative collegiate body on environment (attached to presidential office) ONE Office National de Coordination & programming of PNAE I'Environnement II Addlication Supportto environmental policy making of EIA uroceduresunder MECIE law CORE Committee on Orientation of Orientatesenvironmental research Environmental Research Environmental researchpermits ANGAP National Parksmanagement Managementof national protectedareas network service (NGO/ Parastatalstatus) CNRE National environmental Environmental scientific research researchcenter SAGE Support services to Provision of support services for environmental environmental management management(NGO status, derived from ONE) 6.3 Producer/ CommunityEmpowermentand Participation The empowermentand participation ofthe rural poor in the development processis central to PRSP, PADR and PNAE. Success in poverty reduction, enhancingrural productivity, improved environmental managementand conservationof biodiversity all depend on identifying and applying the right mechanismsfor enabling participation and empowermentof local actors. The general consensus and strategyof governments, donors and projects is that effective participation of rural people requires that they organizethemselves into structures generally referred to as community-based organizations(CBOs) or, inthe case of agriculture, producer organizations (POs) or cooperatives. Federationso f organizationsare important when dealing with entire production sectors or across large geographical areas. Rural development projects and the PNAE have therefore encouragedthe creation of a range of CBOs, POs and committees as program partners yet funds or resourceshave rarely been available to enablethe structures to become functional and autonomous. Many hopes are pinned on these organizations but most lack the maturity to operate without outside support and it would be unrealistic to expect them to become strong in less than 4-5 years (Razafintsalama2003). 125 Historical overview Participation and collective action figure strongly in Malagasytradition, as manifested inthe constitutional principle of,fihavanana (`togetherness') and persisting institutions such as the , fokololoiiu (social group) and dim (social pact), guided by respectfor the ways of the ancestors (razana)and for elders or patriarchs(rqvanzan-dreny). Colonization introduced European forms o f property (including land tenure) and legal personality (e.g. companies). Transition to independenceadded the non-profit association(law 60-133), the law 011 cooperation(60-136) and water-user collectives (law 61-026). The socialist revolution of the 1970`s and 1980's promotedMarxist conceptsofcommon ownership and collective labor and introduced socialist cooperatives (laws 77-038, 77-039 and 77-040). Structures introduced since the 1980`s include mutual associations (94-020), foundations(95-028) andNGOs (96-030). There has also been a movement, under the aegis of the Ministry of Economy and Planning, to relaunchthe cooperative resulting in the modernized law on cooperatives (99-004, Soh0 & Razafimaharo2001). Inthe field of natural resourcemanagement, law 96-025 ("GELOSE") provides for transfer of managementof renewablenatural resourcesto `base communities', the community being manifested in the form of 2 dina whose implementation is typically entrustedto an association under law 60-133. Optionally coupled to resourcemanagement, there exist various forms of collective landtenure known as SFR (se`curisationfoncibre relative), SFO and SFI although these do not constitute definitive title. A variant of GELOSE is GCF (contractual forest management) being promoted by the forests administration. Mention should also be made of the decentralizedadministrative authorities - commune, region, and province. Decree 99-952 introducedthe public inter-communal organization (OPCI) providing an institutional mechanismfor communesto cooperate over the managementof inter- communal assets or activities. Experiencewith Cooperatives and Producer Organizations In the 1970`s there were some 1,300 peasant organizations(or producer organizations, POs) that were mostly general purpose structures for coordinating rural development, often highly politicized. Many of these are now defunct while many more havebeen created. Today there are some 4,000 PO`S,including 600 concernedwith dairy cattle raising, established as a result of the Livestock Sector Project. Complete data are lacking on cooperatives butthere are some 237 cooperatives in Antananarivo province, of which 144are for transport, 44 in crafts, 33 in agriculture, 12 in commerce and 4 in fisheries. It appears that cooperatives have been most successful in commercial transport (Soh0 & Razafimaharo2001). Experienceswith cooperatives were certainly not all favorable. Inthe case of rice, producers are reportedly skeptical about collective storage schemes for rice because of bad experiences with socialist cooperatives(Bockel 2002). Socialist sea fishing cooperatives establishedinthe 1980's were unsuccessful, blamedon lack of training o f fishermen and managers' poor understanding of the realities of fishing. A project in the 1980sto provide groups and associationsof fishermen with vessels also failed becausethe groups and associationsdid not function - insteadthe vessels were leased at sub-commercial ratesto companies, most of whom also failed due to lack of experience in fishing or poor management (Andranaivojaona et a1 1992). 126 Inthe field of POs, the Producers` Associations FederationFIFATA stands out. FIFATA specializes in lobbying and information provision. consisting of 200 POs with a total of 30,000 individual members. FIFATA receivedtechnical assistance from the FrenchNGO, FERT. Member associations of FIFATA have become increasingly specialized in supply of equipment & materials, technical assistance, credit provision and marketing. Their supply activities operate through a network of cooperativesand buying centers. FIFATA has developed its own extension service. The inaiii lesson to be learnedfrom FIFATA's success are the use ofthe participatory approach. involving all producerssince its creation, and the needfor goodtechnical assistance in the establishmentphase. Another successful PO is ROVA, the federation of dairy producers for the region of Vakinankaratra. ROVA receivedtechnical, organizationaland managementtraining funded by NORAD in collection services which enable it to exert an influence over producerprices, large scale purchase of fertilizers and obtaining training for members. While funding is now over, partnerships with technical and financial institutions have allowed ROVA to stay in operation. The above success stories relate to federations providing centralized services to producersthat they themselves cannot easily provide. At the production level it i s important to be realistic about the extent to which producerswill cooperate. In a proposalfor a producer-operator partnership in the maize sector, it is recognized that individual producerswill havetheir own plot of land and that the maizethey produce remainstheir individual property untilpayment is made, eventhough maize is stored collectively. Fisheries cooperativesmay have failed because ofthe highly individualist strategies adopted by fishermen. Legal constraints for POs and cooperatives. A number of legal constraints have been identified. The status of non-profit association(law 60-133), still the most used from, is unsuitedto commercial operations. The NGO status, which must be non-profit, requiresdemonstration ofthe professional aptitudes ofthe personnel, is in practice difficult to comply with andmost so-called NGO's are "Associations" under law 60-133. For cooperatives,there remains an ambiguity inthe finance law which requires `societes coopkratives' to paytax inthe same way as companies whereas the law on cooperativesexcuses them oftax. Distinction needsto be drawn between non-profit rural cooperatives and commercial operations such as the transport cooperatives. Recommendations. In order to re-dynamize the use of cooperatives and make laws known, there is a need to: 1. Reform the law of associations(cooperatives, associationsandNGOs) and laws on "mutuelles d'epargneet de credit" to allow for commercial operations. 2. Introduce favorable fiscal status for ,cooperatives,such as tax exoneration on benefits for the first operating five years. 3. -Develop capacity buildingprogramson: organizational aspects (legal status, transparent compensationand salary system, sanction system for non complianceto rules); - managementaspects (annualwork plans, a monitoringand evaluation system, reporting system, staff managementandtraining policy); and - financial aspects (clear financial managementsystem, cost saving practices, audits) 4. Promoteparticipatory approaches which give responsibilities to producers incontrol and managementof resourcesand activities 127 ! Experienceswith community-basedmanagement Phase2 of PNAEhas providedconsiderable experience in establishment community-based management. Most initiativeshave been under one oftwo variants under GELOSE (law 96-025): (i) `mainstream` GELOSE incorporatingpublic mediation(to deal with potentialconflicts) and with the option of providingsecurity of tenure over the managed area (SFR); and (ii)contractual forest management(GCF =gestion contractualiske desfor&) which is specifically adapted to the managementof state-ownedforests without mediationor the tenure option. There has also been limited use of the customary dina or social pact, in most cases as aprecursorto aGELOSE transfer but which serves to legitimize immediateactionby the community members. Experienceto date is essentiallylimited to the establishment phaseof GELOSE, with few managementsystems yet operational. To date there are 62 GELOSE and 42 GCF contractsin place covering 141,665 haof forest (Razafindrakoto2002) or about 1% of Madagascar's remainingforest. While GCF covers only forest, GELOSEcontractstypically includeawider management zone surroundingthe resource. There exist additionalGELOSEcontractscovering lakes, marineareas and non-forestecosystems althougha full inventoryof GELOSEcontracts is not yet available. The general coiisensus is that GELOSEproceduresare too longand complex, a factor which has encouraged recourse to dina and development of the simplified GCFfor state-ownedforests. A draft directive has now been issuedfor the harmonizationof resource managementtransfers. There should be a move towards more streamlinedsystems during EP3. EP3 shouldalso provide experience of actual operationof community-basedmanagement systems supported by capacity- building measures. Recommendations. In order to scale-up the initial successof community-basednaturalresource management schemes, there is a needto: - Proceedwith harmonizationof transfer procedures - - Provide for simplified GELOSEwithout public mediationwhere conflicts are not present Invest in building capacity for managementand establishingpermanent management support services at the regionallevel. Decentralizationprocess and agriculture/ rural development In line with the decentralizationprocess, core public sector functions needto be devolved in a subsidiary manner to regional and local authorities. However, actual laws on decentralizationare unclear and sometimes contradictory on the precisepowersand duties ofthe decentralized collectivesand `deconcentrated' governmentservices. Inprinciple, the regionalexerciseof most public sector functions should be devolvedto the provinces, and it will be the provincial governmentsthat decide on the mechanisms for delegationand exercise ofthese powers. However, the scope and mechanisms for such devolutionto provincesare not at all clear and the process is likely to take many years. In the meantime, decentralizationexistsdefacto and for specific purposes and offers important opportunities for programs such as PADR and PNAE, thus: - The GTDRshave legitimate if not legal guiding authority on rural developmentwithin their respectiveregions and can exert considerableinfluence; 128 - The law on inter-communality provides for the establishment o f inter-communal bodies (OPCIs) to manage common assets or activities (e.g. irrigation schemes, forests, coastal zones): - Through the elaboration of comtnunal development plans (PDCs), communes are increasingly defining their role in rural development and environmental management; - The GELOSE law is now a well established mechanism for decentralized management o f natural resources, and the customary dina has also gained considerable ground; - Freedom of association into legal structures (associations, cooperatives) adds a further dimension to decentralization. As far as the PADR is concerned, a start has been made with the establishment o f GTDRsas a platform for participation at the regional level. Each GTDR comprises o f five colleges (local authorities, decentralized technical services, private sector operators, producer associations and NGOs/projects) with the mandate to coordinate implementation o f the PADR in the region. All 20 regions have been the subject of agro-ecological assessments and action plans have been developed for most regions. However, they are a recent feature on the institutional scene and it not clear what their role would be in the context o f an on-going decentralizationo f the state. At the same time, the role o f farmers' organizations and rural municipalities has not yet been clearly defined vis-a-vis the role and functions o f the provincial governments (Bockel 2002). Owing to the size o f the regions and the small scale of interventions, most GTDR have established sub regional antennae or branches. The GTDRs have proved to be dynamic but subject to the inevitable limitations imposed by the voluntary status o f GTDR members. A recent GTDR evaluation identified needs and gaps inthe following areas -capacity reinforcement, system of monitoring, lack o f participationfrom POs and private sector colleges: external consultants' support, platform for exchange, validation o f regional development plans, material resources, compensatory payments for GTDR personnel, auditing o f accounts and revised standardized statutory status for GTDRs (EPP 2002). GTDRs have been criticized for acting too much as an extension of the Ministry o f Agriculture Le. as an arm o f government - rather than in their role as facilitating and guidingstructures (Fusilier 2001). To redress that situation, it was suggested that GTDRs be encouraged in their attempts to spread the Systeme National de Vulgarisation Agricole (national system for agricultural promotion). As such, their specific role of intermediariesifacilitators between the Ministryo f Agriculture, the Services Deconcentres, the agricultural promoters, the decentralized units, private operators and farmers' organizations, would be to ensure that the State withdrawal from production activities is progressive and that the taking over by the private sector is achieved in a subtle manner. In view o f the uncertainties surrounding the decentralization process, while taking into account what has happened defacto, it is recommended that PADR and PNAE should promote all mechanisms and tools for decentralized management in order to help establish a vibrant decentralized culture and thus favor an eventual devolution o f powers consistent with established practice. 6.4 FacilitatingPrivateSector Contribution/ Partnership Inview ofthe errors ofthe collectivist past (still felt in some sectors), it is now generally accepted that the private sector has an essential role to play in rural development based on the realities of the free market. In certain cases, the private sector and state may collaborate in areas 129 where both state and private sector have a contribution to make (e.g. agricultural extension services. research, social services deliver!, etc.). In general, however, the private sector will be exclusively responsiblefor investment and production, preferably in partnershipwith rural populations. In the past, the private sector has often sought to influence the government through discretionary relationships with decision-makers. Typically, this has ledto market distortions, monopolies, restricted information flow and other obstacles to efficiency, poverty reduction and sustainability. In the modern context of good governanceand transparency, the private sector can continue to play an influential role subject to certain safeguards that promote transparencyand improved governance. Structural adjustment measureshave facilitated a rapid upturn in private sector investment in Madagascarsince the mid-1990s. Facilitating private sector involvement at this stage is therefore a matter of removing remaining obstacles and using incentives and regulations (`carrots & sticks') to guide private sector contributions so as to promotethe policy objectives of accelerated national growth, poverty alleviation, inclusive rural development and environmental conservation. As has been noted, poverty and low rural productivity are mainly the result of isolation andthe fragmentation of rural sectors and markets. Partnershipsbetween producers and operators are one way of overcoming such isolation and fragmentation as they offer the possibility to go around certain constraints simultaneously (e.g. Randrianjohary 2003). Partnershipsmay extendto include millers, brokersand exporters, ultimately encompassingthe entire product market chain (approche,filiBre). Producer / operator partnershipscan deliver many benefits to producers including assured business, improved incomes, better information on market prices, alternative sources of credit and self-development or empowerment. For the private operators, benefits include economies on collection costs, stable prices and guaranteedsupplies. Producer/ operator partnershipsare rare in Madagascar, with most producersand operators still operating individuallyandoften in competition with one another. Becauseoftheir lack of organization and information, producers generally suffer most as a result of poor market integration. Successfulpartnerships depend on the common interest of the parties and freedom from externalities that might disturb the relationship, such as price fluctuations or excessive competition. Possibledrivers for the identification of opportunities for partnershipsbetweenPOs and agribusiness include: (i) economies of scale indownstreamprocessingactivities, e.g. palm oil, essential oils; or (ii) need for mutual trust for effective operation of eco-certification and labeling schemes, e.g. organic vegetables. At the same time, while some ofthe best economic potential may lie in specialist rural products, their inherentinstability and novelty makethem also quite risky for the development of stable partnerships. Therefore as a start, prospectsfor rural developmentpartnerships should not rule out well-established food crops in local demand such as maize, manioc, potatoesetc. Establishing partnerships requires organization at the different levels to permit collective negotiation. The emergenceof effective producer organizations and federations is criticalto the establishment of partnerships and will in practice bethe major challengeto establishing successful partnerships(Razafintsalama2003). The capacitiesof producer organizations are as yet very limited and they are very vulnerable to interference. Success will dependon producers gaining the confidence of operators. 130 Development of a partnership for the production of a stapleproduct would require attention to the following aspects: -- Organization of producers into production groups and a federative structure Organization of operators, brokers and others into stable groupings - Establishing an effective forum for negotiation - Defining a mechanism for determining prices and making payments - Financing, training, and equipping producersto meet quantity and quality targets - Establish provisions about the use of seeds (generally producerswill need to use the same - variety) Clear procedures for growing, stocking and treatment ofproduct - - Quality control and monitoring of yields and production methods Packing, collection and transport of product to operators. Achieving agreementon all these elements is a major undertakingand will require determination on the part of all members of the partnership. Major partnershipsare likely to require significant funding and technical assistance. Piloting projectswould provide valuable experience inthe use of partnerships. Facilitation of partnershipswill require appropriate policies and legislation to delimit the relationships betweenproducersand operatorsand to provide incentivesto operatorsto enter into partnerships. A more operational legal environment at the producer level (land rental contract) may provide to be important in the fostering of partnerships. A report preparedfor this review (Randrianjohary, 2003) has identified7 main observationsand 14recommendations. Main observations of the sector: 1. Intervention will be required at the local market level - in particular with cash crops-to guarantee a minimum producerrevenue level that allows farmers to recapture capital expenditures and to assure food needs. Specifically, better market information should be made available to villagers. There also may be a needto protect farmers from the impacts of globalization but at the same time assure that productsare competitive on the local markets. Government clarity with regardto pricingpolicy will be essential. 2. A subtle mechanism is neededthat will allow the integration o fcivil society, private sector, and governmenttechnical and extension support to partners. The State may needto finance at least part ofthe high short-term costs ofthis support to villagers. Village level organizations will require support from their conceptionuntil they are fully functional. As well, a mechanismwill be needed to better integrate agricultural production with agri-business. 3. There is a need to createan institutional mechanismthat can contribute to the establishment of sustainablepartnerships. 4. Various partnering alternativesshould be developedfor product market chains where there is obvious competitivenessand high production potential. Impacts ofprivatization ofthe cotton and sugar sectors should be examined in detail. Information should be exchangedamongcountries that have similar productivity so as to minimize local market perturbations. Promote market development for market chains that generateeither direct or indirect benefits with regardto food production. 131 I 5. Becausethere i s inadequateaccess to capital for the development ofthe private sector in Madagascar, a financing mechanism will be neededto facilitate capital access for partnerships, agro-industry and producers. For example ARD" proposed seeking international investors for the establishment of different financing institutions including: an agri-business developmentbank and a small to mid-size enterprise bank for the food industry amongothers 6. Promote local product preparationfacilities in various areas depending upon each region's competitive advantages. 7. Re-examinethe role of the government with regardto the economic management ofthe country. Specifically, 10 of the 46 enterprises that are to be privatized are inthe food production sector: SIRAMA Siramamy Malagasy, KAFEMA Kafe Malagasy, SEVIMA Societe d'exploitation de la viande a Madagascar, FEV Ferme d'Etat de Vohimasina, SORIMA, Socidte rizicole de Marovoay, SOMAPALM, Societe Malgachede Palmiersa Huile, SOAVOANIO Societe Sambava Voanio, BCL Bureau Central Laitier, FAMAMA FambolenaMahabibo de Mahajanga. and SIRANALA. The recommendations that come from this review respondto several questions and general areas of concern. The first question is what types ofpartnerships or mechanisms @orboth local and international markets) can rise above the constraints to higher sustainableproduction? The following results respondto this question. R1. Partnerships in the food sector that still require significant imports -among others rice, wheat, and cooking oils. R2. Partnershipsthat can alleviate or improve food security and nutrition. R3. Partnershipsconcerned with the exportation of non-traditional products with either high value added or products that require significant labor (and thus createjobs.) R4. Regional and international partnershipsthat include trading partnershipsand access to specific markets. The second question addressed is what are the enablingconditionsfor productivepartnerships? The following results respondto this question. R5. Promotepartnerships based upon the differential contexts and comparative advantages of each region. R6. Financihg will be essential to realize the potential for partnerships-especially partnerships that can transform individual actors into close businesspartnersthat have true negotiating power. R7. Promote partnershipsthat can facilitate additional investments in commercial agricultural market chains. R8.The appropriate partnershipmodel should follow the specific market situation. The following results are related to the development of an appropriate institutional landscape. Round Table organizedby USAID with the participationofARD, October 15,2002. 132 R9. Developto fulIcapacitythe Rural ObservationNetwork (Reseau d'observatoires Ruraux, ROR) to provide a channel for informationand to energize the institutional landscape. R10. Promote the "Tranomben'ny Tantsaha" (TT) or Agriculture Chamber ofCommerce to better represent agriculturalists. The following results reflect the role of thegovernment. R11. Concerningthe institutionalstructure, two areas requiresupport: producer services (Service d'Appui-Conseil, SAC) and management of litigation and conflicts. With regardto SAC, it is suggestedto enhance the legal stature ofthe GTDR to provideSAC including the improvement of GTDR functioning and technical capacity. With regardto litigation and conflicts, it is suggested to create an office of mediationto facilitate conflict resolution. R12.Concerningthe economic management of the State, there is astrongneedfor public investmentto facilitate partnerships andto keep continuedfocus on povertyreduction. Among others, the FER (Fonds d'Entretien Routier) andthe CRC (Comite de Reflexion sur la Competitivite) can helpregroupprivate sector concerns to better inform public sector investment decisions. R13.Concerningthe management of the public functions, there are many areas for improvement and a systein of performance-based payment could helpwith some problems. R14.A programfor the development of business partnershipsshouldbe established bythe government in the following market chains: "fruits andvegetables, essential oils, and natural products" and "rice: wheat, and cookingoils." 133 Appendix I RICE IMPORT TARIFF SIMULATION MODEL A simulation model built on a step by step accumulationof cost and taxes from the port in Tamatave to retail sales in Antananarivo allows to assess the effect of different import tariffs on the price of rice. Furthermore. by calculating the retail price in Antananativo back to the paddy farm gate paddy by subtractingtransport and operator costs, it is possibleto evaluate at what price levels imported rice starts competing in rural areas. Results ofthe model are presentedin Table A3-1. The model shows the relation betweenthe price of imported rice and domestic paddy in 1999 and 2000. For 2001, it providestwo figures: one virtual case with 100%compliance to import tariffs and VAT taxes and one realistic scenario with only 50% compliance. With the import tariff at 5% and VAT at 0%, the farm gate price of rice stood at FMG885 per kg in 1999. At this price, over 80% of all rice producing households had cost of production that were lower than FMG885/kg. The farm gate price of paddy was relatively favorable, despitethe low import tariff and absence of VAT, due to the relatively high world market price ofrice. Although world market prices decreased considerably in 2000, the farm gate price of paddy in fact increasedto FMG975/kg due to an increaseofthe import tariff to 15% and the introduction ofthe VAT at 20%. At FMG975/kg, about 90% of rice producing householdshaveproduction coststhat were lower than this figure. In 2001, world market prices reducedevenfurther. With import tariffs and VAT unchanged, this should have reducedthe farm gate price of paddy to about FMGSlYkg in case of full compliancewith these tariffs and taxes. At this price, about 75% of all rice producing households would still have been competitive. However, feedback from relevant stakeholders indicatesthat the effective import tariff and VAT ratewas at the most 50% and most probably only 30% in 2001. Assuming a 50% compliance rate for the existing import tariff and VAT levels, would in effect reducethe farm gate price of paddyto FMG687/kg. Given the sensitive relationship between farm gate price and production costs over the FMG700-900 range, it implies that only 49% of all rice producing householdscould produceat this price, thereby explaining the distressingsignals that were receivedfrom the rural areas concerning rice prices in late 2001. Basedon January 2002 world market price, the model indicatesthe following price levels at the retail as well as the farm gate level in relation to simulatedincreases ofthe import tariff to respectively 20%, 25% and 30%: 0 Assuming a 30% import tax and 20% VAT, paddy price would range around FMGl,000-1,050/kg, the price of rice would be FMG2,336/kg in Tana. Around 95% of all rice producing householdswould be competitive. 0 Assuming a 25% import tax and 20% VAT, the price of paddy would range around FMG950-1000/kg at the farm gate; the retail price in Tana would be FMG2.145/kg. Over 87% of all rice producinghouseholds would be competitive 0 Assuming a 20% import tax and 20% VAT, the price of paddy would range from FMG 900-9501kg; the rice price in Tanawould be FMG 1,982kg. Over 82% o f all rice producing householdswould be competitive at this price. This simulation shows that an increase in rice import tariffs over the analyzedrange would not have a dramatic effect on the retail price of rice. However, at the same time, it shows that an increasewould not significantly increasethe percentageof competitive rice producing households. Given the sensitivity of paddy farm gate price and competitivenessover the FMG600-900 range, the challenge is to effectuate an import tariff that keepsthe farm gate price 134 c at the upper level of this range. This requires inthe first place to put the conditions in place for an effective compliance of the official import tariffs and tax rates. Similarly, it requires close monitoring o f tlie world market price in order to adjust the import tariff accordingly in case o f a significant decrease so as to avoid an abrupt negative impact on the farm gate price o f paddy in the sensitive range. Madagascar commitments with WTO will have to be considered in this respect. CAF Main tvpes of rice imnortcd liSS/ton Riz Brisures FOB 155 135 122,5 122,5 122,5 122,5 122,5 Riz semi blanchi FOB 192 153 130 130 130 130 130 Riz 100%blanchi FOB 240 190 165 165 165 165 165 Coiit CAF addit (US$) 28 28 28 28 28 28 28 Taux de change US$ 6800 6791 6480 6480 6480 6400 6400 EQuivprix en FMG/kp avant taxeq Fmg/kg Riz Brisures 1244 1107 975 975 975 963 963 Riz semi blanchi 1496 1229 1024 1 024 1024 1011 1011 Riz 100%blanchi 1 822 1480 1251 1251 1251 1235 1235 % vol en riz semi-blanch Frais de dechagement Transp Frais de chargement 30 30 30 30 30 30 30 ' O r t Coiit de transuort 130 140 140 140 140 140 140 Marge operateur 3% 3% 3% 3% 3% Cout de transport 120 I20 120 120 120 120 120 Margecumulee comm 12% 12% 12% 12% 12% 12% 12% Equ Prix achat Riz local 181.5 I964 1698 1485 1950 1782 1638 66% Coiit de transformation 80 80 80 80 80 80 80 Coiit transport 75 75 75 75 75 75 75 Autres coiit (collecte ) 70 70 70 70 70 70 70 Margecollecte 9yo 9% 9% 9% 9% 9% 9% 135 Appendix IIPoverty and Biodiversity in Madagascar (WDR 2003) Poverty and biodiversity in Madagascar used for even lower productivity uses, such as (from World Development Report 2003, cattle, or it is abandoned. In drier parts of the Box 8.3) country, grazing and fuelwood extraction spur forest degradation. So, while Madagascar's biodiversity is among the richest 115,000 square kilometers o f forest have been and most unusual in the world, an asset difficult lost since 1960, the area under cultivation for to value in monetary terms but with great staple crops has expanded by only 15,000 square potential to support ecotourism and perhaps bio- kilometers. prospecting industries. Of its 12,000 plant Forest destruction has not only failed to species, 85 percent are found only on yield new productive land; it has degraded the Madagascar. Its 32 endemic lemur species are productivity o f existing farmlands and an attraction for ecotourists. Alkaloids extracted infrastructure. Denuded hillsides are easily from its rosy periwinkle plant from the basis for eroded: 130,000 hectares of irrigated land have some o f the most effective cancer treatment sustained damage or are threatened by sediment. drugs, achieving a 90 percent remission rate Sediment also clogs hydropower facilities and against childhood leukemia. Yet over the past threatens freshwater and marine ecosystems. 40 years, Madagascarhas liquidated about half Madagascar hopes to alleviate poverty its forests, which contain the overwhelming and reduce pressure on its biodiversity by majority of its biodiversity assets, without boosting agricultural productivity. Improving realizing offsetting gains in other assets. The roads in agriculturally productive areas may country has fallen deeper into poverty, with its increasefarmer revenues, reduce fertilizer prices, GDP per capita falling from $383 (in 1995 promote intensification, and absorb labor- dollars) in 1960 to $246 today. In 1997, 16 reducing incentives for farmers to migrate to the percent o f children died before age five. forest frontier. In addition, transferring property What happened? AgricuItural and management rights of natural resources to productivity stagnated while population tripled. local communities is generating incentives for Madagascar's people depend heavily 011rice and more sustainable use and conservation tillage, a few other staple crops. In 1960 average rice that better protect natural resources and that have productivity was 1.8 tons a hectare-about the the capacity to improve profitability. Expansion same as Indonesia, and much more than the 1 ton on the tiny industrial sector may also relieve a hectare average in Mali. By 2000 productivity pressure on the land. had doubled in Mali and more than doubled in In the medium to long run, Indonesia, but it was almost unchanged in Madagascar's unique natural assets may provide Madagascar. Static productivity-despite a the basis for a lucrative tourist industry based on substantial increase in irrigated rice area-reflects ecotourisme and resorts. The country may also in part the implosion of the nation's road be able to benefit from global markets for network. which fell from 55,000 kilometers in biodiversity and carbon sequestration services-if 1960 to 33.000 in 2000. It reflects also a low these markets develop on a large scale. and declining rate of fertilizer use: on11 4 kilograms per hectare. against a Sub-Saharan WRI 2000. of 96. * Meanwhile, population grew from 5.4 average of 12 and a developing country average Source: World Bank staff. Forest area, million to 15.5 million. cultivated area, yields from FAOSTAT The combination of an expanding database; child morality from Gwatkin population and stagnant productivity generated (2000) pressures for agricultural expansion through forest conversion. Small farmers expanded slash-and-burn cultivation of rice into forest lands officially belonging to the state. The practice is attractive to farmers because of its low labor and input requirements and relatively attractive yields in the first two years. But yields rapidly decline to less than half a ton per hectare after a year or two. Subsequently, the land is 136